United States
Environmental Protectio
LhI K ^Agency
EPA 430-R-25-003
Inventory of
U.S. Greenhouse Gas
Emissions and Sinks
1990-2023
-------
HOW TO OBTAIN COPIES
You can download this document from https://www.epa.gov/ghgemissions/inventory-us-
greenhouse-gas-emissions-and-sinks. All data tables in this report, inclusive of data for interim
years not shown in tables, will be made available in CSV format within 4-6 weeks following
publication of the final report on EPA's website.
RECOMMENDED CITATION
EPA (2025). Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023. U.S.
Environmental Protection Agency, EPA 430-R-25-003.
FOR FURTHER INFORMATION
Contact GHGInventory@epa.gov.
For more information regarding greenhouse gas emissions, see the EPA web site at
https://www.epa.gov/ghgemissions.
-------
Preface
The United States Environmental Protection Agency (EPA) prepares the official U.S. Inventory of
Greenhouse Gas Emissions and Sinks on an annual basis. In an effort to engage the public and
researchers across the country, the EPA has instituted an annual public review and comment
process for this document. The availability of the draft document in the docket ID no. EPA-HO-OAR-
2024-05914 was announced via Federal Register Notice FRL-9448-04-QAR. Comments received
during and after closure of the public review period are posted to the same docket and will be
considered for the next edition of this annual report. Responses to comments are typically posted
to EPA'swebsite within 2-4 weeks following publication of the final report.
-------
Table of Contents
LIST OF TABLES, FIGURES, BOXES AND EQUATIONS IV
Tables iv
Figures xx
Boxes xxiii
Equations xxiv
EXECUTIVE SUMMARY ES-1
ES.1 Background Information ES-2
ES.2 Summary of Trends in U.S. Greenhouse Gas Emissions and Sinks ES-3
ES.3 Overview of Sector Emissions and Trends ES-16
ES.4 Other Information ES-22
1 INTRODUCTION 1-1
1.1 Greenhouse Gases 1-2
1.2 Global Warming Potentials 1-9
1.3 National Inventory Arrangements 1-11
1.4 Inventory Preparation Process 1-12
1.5 Methodology and Data Sources 1-17
1.6 Key Categories 1-19
1.7 Quality Assurance and Quality Control 1-25
1.8 Uncertainty Analysis 1-28
1.9 Completeness 1-33
2 TRENDS IN GREENHOUSE GAS EMISSIONS AND REMOVALS 2-1
2.1 Overview of U.S. Greenhouse Gas Emissions and Sinks Trends 2-1
2.2 Emissions and Sinks by Economic Sector 2-34
2.3 Precursor Greenhouse Gas Emissions 2-49
3 ENERGY 3-2
3.1 Fossil Fuel Combustion (Source Category 1 A) 3-8
3.2 Carbon Emitted from Non-Energy Uses of Fossil Fuels (Source Category 1 A) 3-57
3.3 Incineration of Waste (Source Category 1A) 3-67
3.4 Coal Mining (Source Category 1 B1a) 3-71
3.5 Abandoned Underground Coal Mines (Source Category 1 B1 a) 3-79
Table of Contents i
-------
3.6 Petroleum Systems (Source Category 1 B2a) 3-85
3.7 Natural Gas Systems (Source Category 1 B2b) 3-101
3.8 Abandoned Oil and Gas Wells (Source Categories 1 B2a and 1 B2b) 3-120
3.9 C02Transport, Injection, and Geological Storage (Source Category 1C) 3-125
3.10 International Bunker Fuels (Source Category 1: Memo Items) 3-132
3.11 Biomass and Biofuels Consumption (Source Category 1 A) 3-138
3.12 Energy Sources of Precursor Greenhouse Gases 3-143
4 INDUSTRIAL PROCESSES AND PRODUCT USE 4-2
4.1 Cement Production (Source Category 2A1) 4-11
4.2 Lime Production (Source Category 2A2) 4-16
4.3 Glass Production (Source Category 2A3) 4-23
4.4 Other Process Uses of Carbonates (Source Category 2A4) 4-28
4.5 Ammonia Production (Source Category 2B1) 4-36
4.6 Urea Consumption for Non-Agricultural Purposes (Source Category 2B10) 4-42
4.7 Nitric Acid Production (Source Category 2B2) 4-46
4.8 Adipic Acid Production (Source Category 2B3) 4-51
4.9 Caprolactam, Glyoxal and Glyoxylic Acid Production (Source Category 2B4) 4-56
4.10 Carbide Production and Consumption (Source Category 2B5 & 2B10) 4-60
4.11 Titanium Dioxide Production (Source Category 2B6) 4-65
4.12 Soda Ash Production (Source Category 2B7) 4-68
4.13 Petrochemical Production (Source Category 2B8) 4-72
4.14 HCFC-22 Production (Source Category 2B9a) 4-83
4.15 Production of Fluorochemicals Other Than HCFC-22 (Source Category 2B9b) 4-86
4.16 Non-EOR Carbon Dioxide Utilization (Source Category 2H2 and 2H3) 4-98
4.17 Phosphoric Acid Production (Source Category 2B10) 4-104
4.18 Iron and Steel Production (Source Category 2C1) and Metallurgical Coke Production 4-109
4.19 Ferroalloy Production (Source Category 2C2) 4-123
4.20 Aluminum Production (Source Category 2C3) 4-128
4.21 Magnesium Production (Source Category 2C4) 4-136
4.22 Lead Production (Source Category 2C5) 4-143
4.23 Zinc Production (Source Category 2C6) 4-147
4.24 Electronics Industry (Source Category 2E) 4-153
4.25 Substitution of Ozone Depleting Substances (Source Category 2F) 4-175
4.26 Electrical Equipment (Source Category 2G1) 4-182
4.27 SF6 and PFCs from Other Product Use (Source Category 2G.2) 4-194
4.28 Nitrous Oxide from Product Uses (Source Category 2G3) 4-201
4.29 Industrial Processes and Product Use Sources of Precursor Gases 4-205
5 AGRICULTURE 5-2
5.1 Enteric Fermentation (Source Category 3A) 5-5
5.2 Manure Management (Source Category 3B) 5-14
ii Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
5.3 Rice Cultivation (Source Category 3C) 5-26
5.4 Agricultural Soil Management (Source Category 3D) 5-34
5.5 Liming (Source Category 3G) 5-56
5.6 Urea Fertilization (Source Category 3H) 5-60
5.7 Field Burning of Agricultural Residues (Source Category 3F) 5-63
6 LAND USE, LAND USE CHANGE AND FORESTRY 6-2
6.1 Representation of the U.S. Land Base 6-10
6.2 Forest Land Remaining Forest Land (Source Category 4A1) 6-29
6.3 Land Converted to Forest Land (Source Category 4A2) 6-58
6.4 Cropland Remaining Cropland (Source Category 4B1) 6-67
6.5 Land Converted to Cropland (Source Category 4B2) 6-85
6.6 Grassland Remaining Grassland (Source Category 4C1) 6-96
6.7 Land Converted to Grassland (Source Category 4C2) 6-108
6.8 Wetlands Remaining Wetlands (Source Category 4D1) 6-120
6.9 Land Converted to Wetlands (Source Category 4D2) 6-171
6.10 Settlements Remaining Settlements (Source Category 4E1) 6-196
6.11 Land Converted to Settlements (Source Category 4E2) 6-221
6.12 Other Land Remaining Other Land (Source Category4F1) 6-231
6.13 Land Converted to Other Land (Source Category4F2) 6-231
7 WASTE 7-2
7.1 Landfills (Source Category 5A1) 7-5
7.2 Wastewater Treatment and Discharge (Source Category 5D) 7-22
7.3 Composting (Source Category 5B1) 7-37
7.4 Anaerobic Digestion at Biogas Facilities (Source Category 5B2) 7-42
7.5 Waste Incineration (Source Category 5C1) 7-49
7.6 Waste Sources of Precursor Greenhouse Gases 7-49
8 OTHER 8-1
9 RECALCULATIONS AND IMPROVEMENTS 9-1
Key Recalculations and Improvements for 1990-2023 Inventory 9-3
10 REFERENCES AND ABBREVIATIONS 10-1
Executive Summary 10-1
Introduction 10-2
Trends 10-4
Energy 10-5
Industrial Processes and Product Use 10-32
Agriculture 10-68
Land Use, Land-Use Change and Forestry 10-94
Waste 10-143
Recalculations and Improvements 10-159
Abbreviations 10-160
Table of Contents iii
-------
List of Tables, Figures, Boxes and
Equations
Tables
Table ES-1: Global Warming Potentials (100-Year Time Horizon) Used in this Report ES-3
Table ES-2: Recent Trends in U.S. Greenhouse Gas Emissions and Sinks (MMT C02 Eq.) ES-4
Table ES-3: Recent Trends in U.S. Greenhouse Gas Emissions and Sinks by Inventory Sector
(MMT C02 Eq.) ES-16
Table ES-4: U.S. Greenhouse Gas Emissions and Removals (Net Flux) from Land Use, Land-
Use Change, and Forestry (MMT C02 Eq.) ES-20
Table ES-5: U.S. Greenhouse Gas Emissions Allocated to Economic Sectors (MMT C02 Eq.) ES-22
Table ES-6: U.S. Greenhouse Gas Emissions with Electricity-Related Emissions Distributed
by Economic Sector (MMT C02 Eq.) ES-24
Table ES-7: Recent Trends in Various U.S. Data (Index 1990 = 100) ES-26
Table 1-1: Global Atmospheric Concentration, Rate of Concentration Change, and
Atmospheric Lifetime of Selected Greenhouse Gases 1-4
Table 1-2: Global Warming Potentials and Atmospheric Lifetimes (Years) Used in this Report 1-9
Table 1 -3: Comparison of 100-Year GWP values 1-11
Table 1 -4: Summary of Key Categories for the United States (1990 and 2023) by Sector 1-21
Table 1-5: Estimated Overall Inventory Quantitative Uncertainty for 1990 (MMT C02 Eq. and
Percent) 1-30
Table 1-6: Estimated Overall Inventory Quantitative Uncertainty for 2023 (MMT C02 Eq. and
Percent) 1-31
Table 1-7: Quantitative Assessment of Trend Uncertainty (MMT C02 Eq. and Percent) 1-32
Table 1 -8: Inventory Sector Descriptions 1 -33
Table 1-9: List of Annexes 1-34
Table 2-1: Recent Trends in U.S. Greenhouse Gas Emissions and Sinks by Gas (MMT C02 q.) 2-4
Table 2-2: Recent Trends in U.S. Greenhouse Gas Emissions and Sinks by Gas (kt) 2-7
Table 2-3: Recent Trends in U.S. Greenhouse Gas Emissions and Sinks by Inventory
Sector/Category (MMT C02 Eq.) 2-11
Table 2-4: Emissions from Energy by Gas (MMT C02 Eq.) 2-14
Table 2-5: C02 Emissions from Fossil Fuel Combustion by End-Use Sector (MMT C02 Eq.) 2-17
Table 2-6: Emissions from Industrial Processes and Product Use (MMT C02 Eq.) 2-23
Table 2-7: Emissions from Agriculture (MMT C02 Eq.) 2-27
Table 2-8: U.S. Greenhouse Gas Emissions and Removals (Net Flux) from Land Use, Land-
Use Change, and Forestry (MMT C02 Eq.) 2-30
Table 2-9: Emissions from Waste (MMT C02 Eq.) 2-33
Table 2-10: U.S. Greenhouse Gas Emissions Allocated to Economic Sectors (MMT C02 Eq.
and Percent of Total in 2023) 2-35
Table 2-11: Electric Power-Related Greenhouse Gas Emissions (MMT C02 Eq.) 2-40
Table 2-12: U.S. Greenhouse Gas Emissions by Economic Sector and Gas with Electricity-
Related Emissions Distributed (MMT C02 Eq.) and Percent of Total in 2023 2-42
Table 2-13: Transportation-Related Greenhouse Gas Emissions (MMT C02 Eq.) 2-45
Table 2-14: Recent Trends in Various U.S. Data (Index 1990 = 100) 2-48
iv Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Table 2-15: Emissions of NOx, CO, NMVOCs, NH3, and S02 (kt) 2-50
Table 3-1: C02, CH4, and N20 Emissions from Energy (MMT C02 Eq.) 3-5
Table 3-2: C02, CH4, and N20 Emissions from Energy (kt) 3-6
Table 3-3: C02, CH4, and N20 Emissions from Fossil Fuel Combustion (MMT C02 Eq.) 3-8
Table 3-4: C02, CH4, and N20 Emissions from Fossil Fuel Combustion (kt) 3-8
Table 3-5: C02 Emissions from Fossil Fuel Combustion by FuelType and Sector (MMT C02
Eq.) 3-9
Table 3-6: Annual Change in C02 Emissions and Total 2023 C02 Emissions from Fossil Fuel
Combustion for Selected Fuels and Sectors (MMT C02 Eq. and Percent) 3-10
Table 3-7: C02 Emissions from Stationary Fossil Fuel Combustion (MMT C02 Eq.) 3-14
Table 3-8: CH4 Emissions from Stationary Combustion (MMT C02 Eq.) 3-14
Table 3-9: N20 Emissions from Stationary Combustion (MMT C02 Eq.) 3-15
Table 3-10: C02, CH4, and N20 Emissions from Fossil Fuel Combustion by Sector (MMT C02
Eq.) 3-16
Table 3-11: C02, CH4, and N20 Emissions from Fossil Fuel Combustion by End-Use Sector
with Electricity Emissions Distributed (MMT C02 Eq.) 3-17
Table 3-12: Electric Power Generation by Fuel Type (Percent) 3-19
Table 3-13: C02 Emissions from Fossil Fuel Combustion in Transportation End-Use Sector
(MMT C02 Eq.) 3-33
Table 3-14: CH4 Emissions from Mobile Combustion (MMT C02 Eq.) 3-35
Table 3-15: N20 Emissions from Mobile Combustion (MMT C02 Eq.) 3-36
Table 3-16: Carbon Intensity from Direct Fossil Fuel Combustion by Sector (MMT C02
Eq./QBtu) 3-42
Table 3-17: U.S. Energy Consumption and Energy-Related C02 Emissions Per Capita and Per
Dollar GDP 3-43
Table 3-18: Approach 2 Quantitative Uncertainty Estimates for C02 Emissions from Energy-
Related Fossil Fuel Combustion by FuelType and Sector (MMT C02 Eq. and
Percent) 3-45
Table 3-19: Comparison of Electric Power Sector Emissions (MMT C02 Eq. and Percent) 3-46
Table 3-20: Comparison of Emissions Factors (MMT Carbon/QBtu) 3-47
Table 3-21: Approach 2 Quantitative Uncertainty Estimates for CH4 and N20 Emissions from
Energy-Related Stationary Combustion, Including Biomass (MMT C02 Eq. and
Percent) 3-51
Table 3-22: Approach 2 Quantitative Uncertainty Estimates for CH4 and N20 Emissions from
Mobile Sources (MMTC02 Eq. and Percent) 3-55
Table 3-23: C02 Emissions from Non-Energy Use Fossil Fuel Consumption (MMT C02 Eq.
and Percent C) 3-58
Table 3-24: Adjusted Consumption of Fossil Fuels for Non-Energy Uses (TBtu) 3-59
Table 3-25: 2023 Adjusted Non-Energy Use Fossil Fuel Consumption, Storage, and
Emissions 3-60
Table 3-26: Approach 2 Quantitative Uncertainty Estimates for C02 Emissions from Non-
Energy Uses of Fossil Fuels (MMT C02 Eq. and Percent) 3-63
Table 3-27: Approach 2 Quantitative Uncertainty Estimates for Storage Factors of Non-
Energy Uses of Fossil Fuels (Percent) 3-63
Table 3-28: C02, CH4, and N20 Emissions from the Combustion of Waste (MMT C02 Eq.) 3-68
Table 3-29: C02, CH4, and N20 Emissions from the Combustion of Waste (kt) 3-68
List of Tables, Figures, Boxes and Equations v
-------
Table 3-30: Municipal Solid Waste Combusted (Short Tons) 3-68
Table 3-31: Calculated Fossil C02 Content per Ton Waste Combusted (kgC02/ShortTon
Combusted) 3-69
Table 3-32: C02 Emissions from Combustion of Tires (MMT C02 Eq.) 3-69
Table 3-33: Approach 2 Quantitative Uncertainty Estimates for C02 and N20 from the
Incineration of Waste (MMT C02 Eq. and Percent) 3-70
Table 3-34: Coal Production (kt) 3-71
Table 3-35: CH4 Emissions from Coal Mining (MMT C02 Eq.) 3-72
Table 3-36: CH4 Emissions from Coal Mining (kt) 3-72
Table 3-37: C02 Emissions from Coal Mining (MMT C02 Eq.) 3-76
Table 3-38: C02 Emissions from Coal Mining (kt) 3-76
Table 3-39: Approach 2 Quantitative Uncertainty Estimates for CH4 and C02 Emissions from
Coal Mining (MMT C02 Eq. and Percent) 3-78
Table 3-40: CH4 Emissions from Abandoned Coal Mines (MMT C02 Eq.) 3-80
Table 3-41: CH4 Emissions from Abandoned Coal Mines (kt) 3-80
Table 3-42: Number of Gassy Abandoned Mines Present in U.S. Basins in 2023, Grouped by
Class According to Post-Abandonment State 3-83
Table 3-43: Approach 2 Quantitative Uncertainty Estimates for CH4 Emissions from
Abandoned Underground Coal Mines (MMT C02 Eq. and Percent) 3-85
Table 3-44: Total Greenhouse Gas Emissions (C02, CH4, and N20) from Petroleum Systems
(MMT C02 Eq.) 3-87
Table 3-45: CH4 Emissions from Petroleum Systems (MMTC02 Eq.) 3-88
Table 3-46: CH4 Emissions from Petroleum Systems (kt CH4) 3-88
Table 3-47: C02 Emissions from Petroleum Systems (MMT C02) 3-88
Table 3-48: C02 Emissions from Petroleum Systems (kt C02) 3-89
Table 3-49: N20 Emissions from Petroleum Systems (Metric Tons C02 Eq.) 3-89
Table 3-50: N20 Emissions from Petroleum Systems (Metric Tons N20) 3-89
Table 3-51: Approach 2 Quantitative Uncertainty Estimates for CH4 and C02 Emissions from
Petroleum Systems (MMT C02 Eq. and Percent) 3-93
Table 3-52: Recalculations of C02 in Petroleum Systems (MMT C02) 3-96
Table 3-53: Recalculations of CH4 in Petroleum Systems (MMT C02 Eq.) 3-96
Table 3-54: GOA Offshore Production Vent and Leak National CH4 Emissions (Metric Tons
CH4) 3-98
Table 3-55: GOA Offshore Production Vent and Leak National C02 Emissions (Metric Tons
C02) 3-98
Table 3-56: Pneumatic Controllers National CH4 Emissions (Metric Tons CH4) 3-98
Table 3-57: Chemical Injection Pump National CH4 Emissions (Metric Tons CH4) 3-99
Table 3-58: Produced Water National CH4 Emissions (Metric Tons CH4) 3-99
Table 3-59: Storage Tanks National C02 Emissions (kt C02) 3-99
Table 3-60: Miscellaneous Production Flaring National C02 Emissions (kt C02) 3-100
Table 3-61: Refining National CH4 Emissions (Metric Tons CH4) 3-100
Table 3-62: Refining National C02 Emissions (kt C02) 3-100
Table 3-63: Total Greenhouse Gas Emissions (CH4, C02, and N20) from Natural Gas Systems
(MMT C02 Eq.) 3-105
Table 3-64: CH4 Emissions from Natural Gas Systems (MMT C02 Eq.) 3-105
Table 3-65: CH4 Emissions from Natural Gas Systems (kt) 3-105
vi Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Table 3-66: C02 Emissions from Natural Gas Systems (MMT) 3-106
Table 3-67: C02 Emissions from Natural Gas Systems (kt) 3-106
Table 3-68: N20 Emissions from Natural Gas Systems (Metric Tons C02 Eq.) 3-106
Table 3-69: N20 Emissions from Natural Gas Systems (Metric Tons N20) 3-107
Table 3-70: Approach 2 Quantitative Uncertainty Estimates for CH4 and Non-combustion
C02 Emissions from Natural Gas Systems (MMT C02 Eq. and Percent) 3-110
Table 3-71: Recalculations of C02 in Natural Gas Systems (MMT C02) 3-113
Table 3-72: Recalculations of CH4 in Natural Gas Systems (MMT C02 Eq.) 3-113
Table 3-73: GOA Offshore Production Vent and Leak National CH4 Emissions (Metric Tons
CH4) 3-115
Table 3-74: GOA Offshore Production Vent and Leak National C02 Emissions (Metric Tons
C02) 3-115
Table 3-75: Chemical Injection Pumps National CH4 Emissions (Metric Tons CH4) 3-115
Table 3-76: Pneumatic Controllers National CH4 Emissions (Metric Tons CH4) 3-115
Table 3-77: Liquids Unloading National CH4 Emissions (Metric Tons CH4) 3-116
Table 3-78: Wellpad Equipment Leaks National CH4 Emissions (Metric Tons CH4) 3-116
Table 3-79: Produced Water National CH4 Emissions (Metric Tons CH4) 3-117
Table 3-80: Kimray Pumps National CH4 Emissions (Metric Tons CH4) 3-117
Table 3-81: G&B Storage Tanks National Emissions (Metric Tons CH4) 3-117
Table 3-82: G&B Station Blowdowns National Emissions (Metric Tons CH4) 3-117
Table 3-83: G&B Pneumatic Controllers National Emissions (Metric Tons CH4) 3-118
Table 3-84: G&B Dehydrator Vent National Emissions (Metric Tons CH4) 3-118
Table 3-85: Processing Segment AGR National C02 Emissions (kt C02) 3-118
Table 3-86: Processing Blowdowns National CH4 Emissions (Metric Tons CH4) 3-119
Table 3-87: LNG Export Terminals National C02 Emissions (kt C02) 3-119
Table 3-88: CH4 Emissions from Abandoned Oil and Gas Wells (MMT C02 Eq.) 3-121
Table 3-89: CH4 Emissions from Abandoned Oil and Gas Wells (kt) 3-121
Table 3-90: C02 Emissions from Abandoned Oil and Gas Wells (MMT C02) 3-121
Table 3-91: C02 Emissions from Abandoned Oil and Gas Wells (kt) 3-121
Table 3-92: Abandoned Oil Wells Activity Data, CH4 and C02 Emissions (kt) 3-122
Table 3-93: Abandoned Gas Wells Activity Data, CH4 and C02 Emissions (kt) 3-123
Table 3-94: Approach 2 Quantitative Uncertainty Estimates for CH4 and C02 Emissions from
Petroleum and Natural Gas Systems (MMT C02 Eq. and Percent) 3-124
Table 3-95: Emission from TIGS (kt C02) 3-125
Table 3-96: Allocation of Sequestered C02 for Inventory Adjustment (kt C02) 3-126
Table 3-97: Pipeline Mileage (Miles) 3-127
Table 3-98: Emissions from Injection and Storage (kt C02) 3-128
Table 3-99: Sequestered C02 (kt C02) 3-128
Table 3-100: Percentage of C02 (kt) Supplied to EOR from Different Sources 3-129
Table 3-101: Approach 2 Quantitative Uncertainty Estimates for C02 Emissions from TIGS
(ktC02 Eq. and Percent) 3-130
Table 3-102: C02, CH4, and N20 Emissions from International Bunker Fuels (MMT C02 Eq.) 3-133
Table 3-103: C02, CH4, and N20 Emissions from International Bunker Fuels (kt) 3-133
Table 3-104: Aviation Jet Fuel Consumption for International Transport (TBtu) 3-135
Table 3-105: Marine Fuel Consumption for International Transport (Million Gallons) 3-135
Table 3-106: C02 Emissions from Wood Consumption by End-Use Sector (MMT C02 Eq.) 3-138
List of Tables, Figures, Boxes and Equations vii
-------
Table 3-107: C02 Emissions from Wood Consumption by End-Use Sector (kt) 3-138
Table 3-108: C02 Emissions from Biogenic Components of MSW (MMT C02 Eq.) 3-139
Table 3-109: C02 Emissions from Biogenic Components of MSW (kt) 3-139
Table 3-110: C02 Emissions from Ethanol Consumption (MMT C02 Eq.) 3-139
Table 3-111: C02 Emissions from Ethanol Consumption (kt) 3-139
Table 3-112: C02 Emissions from Biodiesel Consumption (MMT C02 Eq.) 3-140
Table 3-113: C02 Emissions from Biodiesel Consumption (kt) 3-140
Table 3-114: Calculated Biogenic C02 Content per Ton Waste (kg C02/Short Ton
Combusted) 3-141
Table 3-115: Woody Biomass Consumption by Sector (Trillion Btu) 3-141
Table 3-116: Ethanol Consumption by Sector (Trillion Btu) 3-141
Table 3-117: Biodiesel Consumption by Sector (Trillion Btu) 3-141
Table 3-118: Approach 2 Quantitative Uncertainty Estimates for C02 Emissions from
Biomass and Biofuel Combustion (MMT C02 Eq. and Percent) 3-142
Table 3-119: NOx, CO, NMVOC, NH3, and S02 Emissions from Energy-Related Activities (kt) 3-144
Table 4-1: Emissions from Industrial Processes and Product Use (MMT C02 Eq.) 4-5
Table 4-2: Emissions from Industrial Processes and Product Use (kt) 4-7
Table 4-3: C02 Emissions from Cement Production (MMT C02 Eq.) 4-12
Table 4-4: C02 Emissions from Cement Production (kt C02) 4-12
Table 4-5: Clinker Production (kt) 4-14
Table 4-6: Approach 2 Quantitative Uncertainty Estimates for C02 Emissions from Cement
Production (MMTC02 Eq. and Percent) 4-14
Table 4-7: C02 Emissions from Lime Production (MMT C02 Eq.) 4-18
Table 4-8: Gross, Recovered, and Net C02 Emissions from Lime Production (kt C02) 4-18
Table 4-9: High-Calcium-and Dolomitic-Quicklime, High-Calcium-and Dolomitic-
Hydrated, and Dead-Burned-Dolomite Lime Production (kt) 4-19
Table 4-10: Adjusted Lime Production (kt) 4-20
Table 4-11: Approach 2 Quantitative Uncertainty Estimates for C02 Emissions from Lime
Production (MMTC02 Eq. and Percent) 4-21
Table 4-12: C02 Emissions from Glass Production (MMT C02 Eq.) 4-24
Table 4-13: C02 Emissions from Glass Production (kt C02) 4-24
Table 4-14: Limestone, Dolomite, Soda Ash, and Other Carbonates Used in Glass
Production (kt) and Average Annual Production Index for Glass and Glass
Product Manufacturing 4-26
Table 4-15: Approach 2 Quantitative Uncertainty Estimates for C02 Emissions from Glass
Production (MMTC02 Eq. and Percent) 4-27
Table 4-16: C02 Emissions from Other Process Uses of Carbonates (MMT C02 Eq.) 4-30
Table 4-17: C02 Emissions from Other Process Uses of Carbonates (kt C02) 4-30
Table 4-18: Limestone and Dolomite Consumption from Other Uses of Carbonates (kt) 4-31
Table 4-19: Limestone and Dolomite Consumption from Ceramics Production (kt) 4-32
Table 4-20: Other Uses of Soda Ash Consumption Not Associated with Glass Manufacturing
(kt) 4-33
Table 4-21: Magnesite and Limestone Consumption from Non-Metallurgical Magnesia
Production (kt) 4-33
Table 4-22: Approach 2 Quantitative Uncertainty Estimates for C02 Emissions from Other
Process Uses of Carbonates (MMT C02 Eq. and Percent) 4-35
viii Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Table 4-23: C02 Emissions from Ammonia Production (MMT C02 Eq.) 4-37
Table 4-24: C02 Emissions from Ammonia Production (kt C02) 4-38
Table 4-25: Total Ammonia Production, Total Urea Production, Recovered C02 Consumed
for Urea Production, and Sequestered C02 (kt) 4-40
Table 4-26: Approach 2 Quantitative Uncertainty Estimates for C02 Emissions from
Ammonia Production (MMT C02 Eq. and Percent) 4-41
Table 4-27: C02 Emissions from Urea Consumption for Non-Agricultural Purposes (MMT
C02 Eq.) 4-43
Table 4-28: C02 Emissions from Urea Consumption for Non-Agricultural Purposes (kt C02) 4-43
Table 4-29: Urea Production, Urea Applied as Fertilizer, Urea Imports, and Urea Exports (kt) 4-44
Table 4-30: Approach 2 Quantitative Uncertainty Estimates for C02 Emissions from Urea
Consumption for Non-Agricultural Purposes (MMT C02 Eq. and Percent) 4-45
Table 4-31: N20 Emissions from Nitric Acid Production (MMT C02 Eq.) 4-47
Table 4-32: N20 Emissions from Nitric Acid Production (kt N20) 4-47
Table 4-33: Nitric Acid Production (kt) 4-49
Table 4-34: Approach 2 Quantitative Uncertainty Estimates for N20 Emissions from Nitric
Acid Production (MMT C02 Eq. and Percent) 4-50
Table 4-35: N20 Emissions from Adipic Acid Production (MMT C02 Eq.) 4-52
Table 4-36: N20 Emissions from Adipic Acid Production (kt N20) 4-52
Table 4-37: Adipic Acid Production (kt) 4-54
Table 4-38: Approach 2 Quantitative Uncertainty Estimates for N20 Emissions from Adipic
Acid Production (MMT C02 Eq. and Percent) 4-55
Table 4-39: N20 Emissions from Caprolactam Production (MMT C02 Eq.) 4-57
Table 4-40: N20 Emissions from Caprolactam Production (kt N20) 4-57
Table 4-41: Caprolactam Production (kt) 4-58
Table 4-42: Approach 2 Quantitative Uncertainty Estimates for N20 Emissions from
Caprolactam, Glyoxaland Glyoxylic Acid Production (MMT C02 Eq. and Percent) 4-59
Table 4-43: C02 and CH4 Emissions from Silicon Carbide Production and Consumption
(MMT C02 Eq.) 4-61
Table 4-44: C02 and CH4 Emissions from Silicon Carbide Production and Consumption (kt) 4-61
Table 4-45: Production and Consumption of Silicon Carbide (Metric Tons) 4-63
Table 4-46: Approach 2 Quantitative Uncertainty Estimates for CH4 and C02 Emissions from
Silicon Carbide Production and Consumption (MMT C02 Eq. and Percent) 4-64
Table 4-47: C02 Emissions from Titanium Dioxide (MMT C02 Eq.) 4-65
Table 4-48: C02 Emissions from Titanium Dioxide (kt C02) 4-65
Table 4-49: Titanium Dioxide Production (kt) 4-67
Table 4-50: Approach 2 Quantitative Uncertainty Estimates for C02 Emissions from
Titanium Dioxide Production (MMT C02 Eq. and Percent) 4-68
Table 4-51: C02 Emissions from Soda Ash Production (MMT C02 Eq.) 4-70
Table 4-52: C02 Emissions from Soda Ash Production (kt C02) 4-70
Table 4-53: Trona Ore Used in Soda Ash Production (kt) 4-70
Table 4-54: Approach 2 Quantitative Uncertainty Estimates for C02 Emissions from Soda
Ash Production (MMT C02 Eq. and Percent) 4-71
Table 4-55: C02 and CH4 Emissions from Petrochemical Production (MMT C02 Eq.) 4-74
Table 4-56: C02 and CH4 Emissions from Petrochemical Production (kt) 4-74
Table 4-57: Production of Selected Petrochemicals (kt) 4-78
List of Tables, Figures, Boxes and Equations ix
-------
Table 4-58: Approach 2 Quantitative Uncertainty Estimates for CH4 Emissions from
Petrochemical Production and C02 Emissions from Petrochemical Production
(MMT C02 Eq. and Percent) 4-80
Table 4-59: HFC-23 Emissions from HCFC-22 Production (MMT C02 Eq.) 4-84
Table 4-60: HFC-23 Emissions from HCFC-22 Production (kt HFC-23) 4-84
Table 4-61: HCFC-22 Production (kt) 4-85
Table 4-62: Approach 2 Quantitative Uncertainty Estimates for HFC-23 Emissions from
HCFC-22 Production (MMT C02 Eq. and Percent) 4-85
Table 4-63: Emissions of HFCs, PFCs, SF6, and NF3 from Production of Fluorochemicals
Other Than HCFC-22 (MMTC02 Eq.) 4-89
Table 4-64: Emissions of HFCs, PFCs, SF6, and NF3 from Production of Fluorochemicals
Other Than HCFC-22 (Metric Tons) 4-90
Table 4-65: Emissions of Other Fluorinated GHGs from Production of Fluorochemicals
Other Than HCFC-22 (MMTCOz Eq.) 4-90
Table 4-66: Emissions of Other Fluorinated GHGs from Production of Fluorochemicals
Other Than HCFC-22 (Metric Tons) 4-91
Table 4-67: Production and Transformation of Fluorinated GHGs (kt)a 4-91
Table 4-68: Approach 1 Quantitative Uncertainty Estimates for HFC, PFC, SF6, and NF3 from
Production of Fluorochemicals other than HCFC-22 (MMT C02 Eq. and Percent) 4-96
Table 4-69: Net C02 Emissions from Non-EOR C02 Utilization (MMTC02 Eq.) 4-99
Table 4-70: Net C02 Emissions from Non-EOR C02 Utilization (kt C02) 4-99
Table 4-71: C02 Production (kt C02) and the Percent Used for Non-EOR Applications 4-102
Table 4-72: Approach 2 Quantitative Uncertainty Estimates for C02 Emissions from Non-
EOR C02 Utilization (MMTC02 Eq. and Percent) 4-103
Table 4-73: C02 Emissions from Phosphoric Acid Production (MMT C02 Eq.) 4-105
Table 4-74: C02 Emissions from Phosphoric Acid Production (kt C02) 4-105
Table 4-75: Phosphate Rock Domestic Consumption, Exports, and Imports (kt) 4-106
Table 4-76: Chemical Composition of Phosphate Rock (Percent by Weight) 4-107
Table 4-77: Approach 2 Quantitative Uncertainty Estimates for C02 Emissions from
Phosphoric Acid Production (MMT C02 Eq. and Percent) 4-108
Table 4-78: C02 Emissions from Metallurgical Coke Production (MMT C02 Eq.) 4-111
Table 4-79: C02 Emissions from Metallurgical Coke Production (kt C02) 4-111
Table 4-80: C02 Emissions from Iron and Steel Production (MMT C02 Eq.) 4-111
Table 4-81: C02 Emissions from Iron and Steel Production (kt C02) 4-112
Table 4-82: CH4 Emissions from Iron and Steel Production (MMT C02 Eq.) 4-112
Table 4-83: CH4 Emissions from Iron and Steel Production (kt CH4) 4-112
Table 4-84: Material Carbon Contents for Metallurgical Coke Production 4-114
Table 4-85: Production and Consumption Data for the Calculation of C02 Emissions from
Metallurgical Coke Production (Thousand Metric Tons) 4-115
Table 4-86: Production and Consumption Data for the Calculation of C02 Emissions from
Metallurgical Coke Production (Million ft3) 4-115
Table 4-87: Material Carbon Contents for Iron and Steel Production 4-116
Table 4-88: CH4 Emission Factors for Sinter and Pig Iron Production 4-117
Table 4-89: C02 Emission Factors for Sinter Production, Direct Reduced Iron Production,
and Pellet Production 4-117
x Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Table 4-90: Production and Consumption Data for the Calculation of C02 and CH4
Emissions from Iron and Steel Production (Thousand Metric Tons) 4-119
Table 4-91: Production and Consumption Data for the Calculation of C02 Emissions from
Iron and Steel Production (Million ft3 unless otherwise specified) 4-119
Table 4-92: Approach 2 Quantitative Uncertainty Estimates for C02 and CH4 Emissions from
Iron and Steel Production and Metallurgical Coke Production (MMT C02 Eq. and
Percent) 4-121
Table 4-93: Changes from Previous Inventory in C02 Emissions from Iron and Steel
Production (kt C02, % change) 4-122
Table 4-94: C02 and CH4 Emissions from Ferroalloy Production (MMT C02 Eq.) 4-124
Table 4-95: C02 and CH4 Emissions from Ferroalloy Production (kt) 4-124
Table 4-96: Production of Ferroalloys (Metric Tons) 4-126
Table 4-97: Approach 2 Quantitative Uncertainty Estimates for C02 Emissions from
Ferroalloy Production (MMT C02 Eq. and Percent) 4-127
Table 4-98: C02 Emissions from Aluminum Production (MMT C02 Eq.) 4-129
Table 4-99: C02 Emissions from Aluminum Production (kt C02) 4-129
Table 4-100: PFC Emissions from Aluminum Production (MMT C02 Eq.) 4-129
Table 4-101: PFC Emissions from Aluminum Production (kt) 4-130
Table 4-102: Summary of HVAE Emissions (MMT C02 Eq.) 4-133
Table 4-103: Summary of LVAE Emissions (MMT C02 Eq.) 4-134
Table 4-104: Production of Primary Aluminum (kt) 4-134
Table 4-105: Approach 2 Quantitative Uncertainty Estimates for C02 and PFC Emissions
from Aluminum Production (MMT C02 Eq. and Percent) 4-135
Table 4-106: SF6, HFC-134a, FK 5-1-12 and C02 Emissions from Magnesium Production
(MMT C02 Eq.) 4-137
Table 4-107: SF6, HFC-134a, FK 5-1-12 and C02 Emissions from Magnesium Production (kt) 4-137
Table 4-108: SF6 Emission Factors (kg SF6 per metric ton of magnesium) 4-140
Table 4-109: Approach 2 Quantitative Uncertainty Estimates for SF6, HFC-134a and C02
Emissions from Magnesium Production (MMT C02 Eq. and Percent) 4-142
Table 4-110: C02 Emissions from Lead Production (MMT C02 Eq.) 4-144
Table 4-111: C02 Emissions from Lead Production (kt C02) 4-144
Table 4-112: Lead Production (Metric Tons) 4-145
Table 4-113: Approach 2 Quantitative Uncertainty Estimates for C02 Emissions from Lead
Production (MMTC02 Eq. and Percent) 4-146
Table 4-114: C02 Emissions from Zinc Production (MMT C02 Eq.) 4-148
Table 4-115: C02 Emissions from Zinc Production (kt C02) 4-149
Table 4-116: Zinc Production (Metric Tons) 4-149
Table 4-117: Approach 2 Quantitative Uncertainty Estimates for C02 Emissions from Zinc
Production (MMTC02 Eq. and Percent) 4-153
Table 4-118: PFC, HFC, SF6, NF3, and N20 Emissions from Electronics Industry (MMT C02
Eq.) 4-157
Table 4-119: PFC, HFC, SF6, NF3, and N20 Emissions from Semiconductor Manufacture
(Metric Tons) 4-158
Table 4-120: F-HTF Emissions from Electronics Manufacture by Compound Group (kt C02
Eq.) 4-158
List of Tables, Figures, Boxes and Equations xi
-------
Table 4-121: Approach 2 Quantitative Uncertainty Estimates for HFC, PFC, SF6, NF3 and N20
Emissions from Electronics Manufacture (MMT C02 Eq. and Percent) 4-173
Table 4-122: Emissions of HFCs, PFCs, and C02 from ODS Substitutes (MMT C02 Eq.) 4-175
Table 4-123: Emissions of HFCs, PFCs, and C02 from ODS Substitution (MetricTons) 4-176
Table 4-124: Emissions of HFCs, PFCs, and C02 from ODS Substitutes by Sector (MMT C02
Eq.) 4-177
Table 4-125: Approach 2 Quantitative Uncertainty Estimates for HFC and PFC Emissions
from ODS Substitutes (MMT C02 Eq. and Percent) 4-180
Table 4-126: SF6 and CF4 Emissions from Electric Power Systems and Electrical Equipment
Manufacturers (MMT C02 Eq.) 4-183
Table 4-127: SF6 and CF4 Emissions from Electric Power Systems and Electrical Equipment
Manufacturers (kt) 4-184
Table 4-128: GHGRP-only Average Emission Rate (kg per mile) 4-187
Table 4-129: Categorization of Utilities and Timeseries for Application of Corresponding
Emission Estimation Methodologies 4-187
Table 4-130: California GHGRP and Voluntarily Reported SF6 Emissions Compared to
CARB's SF6 Emissions (MMT C02 Eq.) 4-188
Table 4-131: Approach 2 Quantitative Uncertainty Estimates for SF6 and CF4 Emissions from
Electrical Equipment (MMT C02 Eq. and Percent) 4-191
Table 4-132: SF6 and PFC Emissions from Other Product Use (MMT C02 Eq.) 4-195
Table 4-133: SF6andPFC Emissions from Other Product Use (kt) 4-195
Table 4-134: Approach 2 Quantitative Uncertainty Estimates for SF6 and PFC Emissions
from Other Product Use (MMT C02 Eq. and Percent) 4-200
Table 4-135: N20 Emissions from N20 Product Usage (MMT C02 Eq.) 4-202
Table 4-136: N20 Emissions from N20 Product Usage (kt N20) 4-202
Table 4-137: N20 Production (kt) 4-202
Table 4-138: Approach 2 Quantitative Uncertainty Estimates for N20 Emissions from N20
Product Usage (MMT C02 Eq. and Percent) 4-204
Table 4-139: NOx, CO, NMVOC, NH3, and S02 Emissions from Industrial Processes and
Product Use (kt) 4-205
Table 5-1: Emissions from Agriculture (MMT C02 Eq.) 5-3
Table 5-2: Emissions from Agriculture (kt) 5-3
Table 5-3: CH4 Emissions from Enteric Fermentation (MMT C02 Eq.) 5-6
Table 5-4: CH4 Emissions from Enteric Fermentation (kt CH4) 5-6
Table 5-5: Cattle Sub-Population Categories for 2023 Population Estimates 5-10
Table 5-6: Approach 2 Quantitative Uncertainty Estimates for CH4 Emissions from Enteric
Fermentation (MMT C02 Eq. and Percent) 5-12
Table 5-7: CH4 and N20 Emissions from Manure Management (MMTC02 Eq.) 5-16
Table 5-8: CH4 and N20 Emissions from Manure Management (kt) 5-17
Table 5-9: Approach 2 Quantitative Uncertainty Estimates for CH4 and N20 (Direct and
Indirect) Emissions from Manure Management (MMT C02 Eq. and Percent) 5-23
Table 5-10: IPCC (2006) Implied Emission Factor Default Values Compared with Calculated
Values for CH4 from Manure Management (kg/head/year) 5-24
Table 5-11: CH4 Emissions from Rice Cultivation (MMTC02 Eq.) 5-27
Table 5-12: CH4 Emissions from Rice Cultivation (kt CH4) 5-27
Table 5-13: Rice Area Harvested (1,000 Hectares) 5-30
xii Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Table 5-14: Average Ratooned Area as Percent of Primary Growth Area (Percent) 5-30
Table 5-15: Approach 2 Quantitative Uncertainty Estimates for CH4 Emissions from Rice
Cultivation (MMT C02 Eq. and Percent) 5-33
Table 5-16: N20 Emissions from Agricultural Soils (MMT C02 Eq.) 5-36
Table 5-17: N20 Emissions from Agricultural Soils (kt N20) 5-36
Table 5-18: Direct N20 Emissions from Agricultural Soils by Land Use Type and Nitrogen
Input Type (MMT C02 Eq.) 5-36
Table 5-19: Indirect N20 Emissions from Agricultural Soils (MMT C02 Eq.) 5-37
Table 5-20: Quantitative Uncertainty Estimates of N20 Emissions from Agricultural Soil
Management in 2023 (MMT C02 Eq. and Percent) 5-54
Table 5-21: Emissions from Liming (MMT C02 Eq.) 5-56
Table 5-22: Emissions from Liming (MMT C) 5-57
Table 5-23: Applied Minerals (MMT) 5-58
Table 5-24: Approach 2 Quantitative Uncertainty Estimates for C02 Emissions from Liming
(MMT C02 Eq. and Percent) 5-59
Table 5-25: C02 Emissions from Urea Fertilization (MMT C02 Eq.) 5-60
Table 5-26: C02 Emissions from Urea Fertilization (MMT C) 5-60
Table 5-27: Applied Urea (MMT) 5-61
Table 5-28: Quantitative Uncertainty Estimates for C02 Emissions from Urea Fertilization
(MMT C02 Eq. and Percent) 5-62
Table 5-29: CH4 and N20 Emissions from Field Burning of Agricultural Residues (MMT C02
Eq.) 5-63
Table 5-30: CH4, N20, CO, and NOx Emissions from Field Burning of Agricultural Residues
(kt) 5-64
Table 5-31: Agricultural Crop Production (kt of Product) 5-68
Table 5-32: U.S. Average Percent Crop Area Burned by Crop (Percent) 5-69
Table 5-33: Parameters for Estimating Emissions from Field Burning of Agricultural Residues 5-70
Table 5-34: Greenhouse Gas Emission Ratios and Conversion Factors 5-71
Table 5-35: Approach 2 Quantitative Uncertainty Estimates for CH4 and N20 Emissions from
Field Burning of Agricultural Residues (MMT C02 Eq. and Percent) 5-72
Table 6-1: Emissions and Removals (Net Flux) from Land Use, Land-Use Change, and
Forestry (MMT C02 Eq.) 6-5
Table 6-2: Emissions and Removals from Land Use, Land-Use Change, and Forestry by Gas
(MMT C02 Eq.) 6-7
Table 6-3: Emissions and Removals from Land Use, Land-Use Change, and Forestry by Gas
(kt) 6-8
Table 6-4: Managed and Unmanaged Land Area by Land-Use Categories for All 50 States
(Thousands of Hectares) 6-12
Table 6-5: Land Use and Land-Use Change for the U.S. Managed Land Base for All 50 States
(Thousands of Hectares) 6-13
Table 6-6: Data Sources Used to Determine Land Use and Land Area for the Conterminous
United States, Hawaii, and Alaska 6-19
Table 6-7: Total Land Area (Hectares) by Land Use Category for U.S. Territories 6-29
Table 6-8: Net C02 Flux from Forest Ecosystem Pools in Forest Land Remaining Forest Land
and Harvested Wood Pools (MMT C02 Eq.) 6-35
List of Tables, Figures, Boxes and Equations xiii
-------
Table 6-9: Net Carbon Flux from Forest Ecosystem Pools in Forest Land Remaining Forest
Land and Harvested Wood Pools (MMT C) 6-35
Table 6-10: Forest Area (1,000 ha) and Carbon Stocks in Forest Land Remaining Forest Land
and Harvested Wood Pools (MMT C) 6-36
Table 6-11: Estimates of C02 (MMT per Year) Emissions3 from Forest Fires in the
Conterminous 48 States, Hawaii, Puerto Rico, Guam, and Alaska 6-38
Table 6-12: Quantitative Uncertainty Estimates for Net C02 Flux from Forest Land
Remaining Forest Land: Changes in Forest Carbon Stocks (MMT C02 Eq. and
Percent) 6-44
Table 6-13: Recalculations of Forest Area (1,000 ha) and Carbon Stocks in Forest Land
Remaining Forest Land and Harvested Wood Pools (MMT C) 6-46
Table 6-14: Recalculations of Forest Area (1,000 ha) and Carbon Stocks in Forest Land
Remaining Forest Land (MMT C) in Interior Alaska 6-46
Table 6-15: Recalculations of Net Carbon Flux from Forest Ecosystem Pools in Forest Land
Remaining Forest Land and Harvested Wood Pools (MMT C) 6-47
Table 6-16: Recalculations of Net Carbon Flux from Forest Ecosystem Pools in Forest Land
Remaining Forest Land (MMT C) in Interior Alaska 6-47
Table 6-17: Non-C02 Emissions from Forest Fires (MMT C02 Eq.)a 6-49
Table 6-18: Non-C02 Emissions from Forest Fires (kt)a 6-49
Table 6-19: Quantitative Uncertainty Estimates of Non-C02 Emissions from Forest Fires
(MMT C02 Eq. and Percent)3 6-50
Table 6-20: N20 Fluxes from Soils in Forest Land Remaining Forest Land and Land
Converted to Forest Land (MMT C02 Eq. and kt N20) 6-51
Table 6-21: Quantitative Uncertainty Estimates of N20 Fluxes from Soils in Forest Land
Remaining Forest Land and Land Converted to Forest Land (MMT C02 Eq. and
Percent) 6-53
Table 6-22: Non-C02 Emissions from Drained Organic Forest Soilsa b (MMT C02 Eq.) 6-55
Table 6-23: Non-C02 Emissions from Drained Organic Forest Soilsa b (kt) 6-55
Table 6-24: States identified as having Drained Organic Soils, Area of Forest on Drained
Organic Soils, and Sampling Error 6-56
Table 6-25: Quantitative Uncertainty Estimates for Non-C02 Emissions on Drained Organic
Forest Soils (MMT C02 Eq. and Percent)3 6-57
Table 6-26: Net C02 Flux from Forest Carbon Pools in Land Converted to Forest Land by
Land Use Change Category (MMT C02 Eq.) 6-59
Table 6-27: Net Carbon Flux from Forest Carbon Pools in Land Converted to Forest Land by
Land Use Change Category (MMT C) 6-60
Table 6-28: Quantitative Uncertainty Estimates for Forest Carbon Pool Stock Changes (MMT
C02 Eq. per Year) in 2023 from Land Converted to Forest Land by Land Use
Change 6-64
Table 6-29: Recalculations of the Net Carbon Flux from Forest Carbon Pools in Land
Converted to Forest Land by Land Use Change Category (MMT C) 6-66
Table 6-30: Net C02 Flux from Live Biomass and Soil Carbon Stock Changes in Cropland
Remaining Cropland (MMT C02 Eq.) 6-69
Table 6-31: Net C02 Flux from Live Biomass and Soil Carbon Stock Changes in Cropland
Remaining Cropland (MMT C) 6-69
xiv Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Table 6-32: Thousands of Hectares of Land for Total Live Biomass Associated with
Subcategory Land-Use Conversions within Croplands 6-74
Table 6-33: Carbon stock factors for total live biomass associated with Cropland Remaining
Cropland 6-74
Table 6-34: Approach 2 Quantitative Uncertainty Estimates for Soil and Biomass Carbon
Stock Changes occurring within Cropland Remaining Cropland (MMT C02 Eq.
and Percent) 6-81
Table 6-35: Comparison of Managed Land Area in Cropland Remaining Cropland and Area
in the Current Cropland Remaining Cropland Inventory (Thousand Hectares) 6-83
Table 6-36: Net C02 Flux from Soil, Dead Organic Matter and Biomass Carbon Stock
Changes in Land Converted to Cropland by Land-Use Change Category (MMT
C02 Eq.) 6-86
Table 6-37: Net C02 Flux from Soil, Dead Organic Matter and Biomass Carbon Stock
Changes in Land Converted to Cropland (MMT C) 6-87
Table 6-38: Thousands of Hectares of Land for Total Live Biomass Associated with Land-Use
Conversions to Cropland 6-88
Table 6-39: Carbon Stock Change Factors for Total Live Biomass Associated with Land-Use
Conversions to Cropland 6-89
Table 6-40: Approach 2 Quantitative Uncertainty Estimates for Soil, Dead Organic Matter
and Biomass Carbon Stock Changes occurring within Land Converted to
Cropland (MMT C02 Eq. and Percent) 6-93
Table 6-41: Comparison of Managed Land Area in Land Converted to Cropland and the Area
in the current Land Converted to Cropland Inventory (Thousand Hectares) 6-95
Table 6-42: Net C02 Flux from Soil, Dead Organic Matter and Biomass Carbon Stock
Changes in Grassland Remaining Grassland (MMT C02 Eq.) 6-97
Table 6-43: Net C02 Flux from Soil, Dead Organic Matter and Biomass Carbon Stock
Changes in Grassland Remaining Grassland (MMT C) 6-97
Table 6-44: Approach 2 Quantitative Uncertainty Estimates for Carbon Stock Changes
Occurring Within Grassland Remaining Grassland (MMT C02 Eq. and Percent) 6-103
Table 6-45: Comparison of Managed Land Area in Grassland Remaining Grassland and the
Area in the current Grassland Remaining Grassland Inventory (Thousand
Hectares) 6-104
Table 6-46: CH4 and N20 Emissions from Biomass Burning in Grassland (MMT C02 Eq.) 6-106
Table 6-47: CH4, N20, CO, and NOx Emissions from Biomass Burning in Grassland (kt) 6-106
Table 6-48: Thousands of Grassland Hectares Burned Annually 6-106
Table 6-49: Uncertainty Estimates for Non-C02 Greenhouse Gas Emissions from Biomass
Burning in Grassland (MMT C02 Eq. and Percent) 6-107
Table 6-50: Net C02 Flux from Soil, Dead Organic Matter and Biomass Carbon Stock
Changes for Land Converted to Grassland (MMT C02 Eq.) 6-109
Table 6-51: Net C02 Flux from Soil, Dead Organic Matter and Biomass Carbon Stock
Changes for Land Converted to Grassland (MMT C) 6-110
Table 6-52: Thousands of Hectares of Land for Total Live Biomass Associated with Land-Use
Conversions to Grasslands 6-112
Table 6-53: Carbon stock factors for total live biomass associated with land use conversions
to grassland (non-woodland) 6-112
List of Tables, Figures, Boxes and Equations xv
-------
Table 6-54: Approach 2 Quantitative Uncertainty Estimates for Soil, Dead Organic Matter
and Biomass Carbon Stock Changes occurring within Land Converted to
Grassland (MMT C02 Eq. and Percent) 6-117
Table 6-55: Comparison of Managed Land Area in Land Converted to Grassland and Area in
the current Land Converted to Grassland Inventory (Thousand Hectares) 6-119
Table 6-56: Emissions from Peatlands Remaining Peatlands (MMT C02 Eq.) 6-122
Table 6-57: Emissions from Peatlands Remaining Peatlands (kt) 6-123
Table 6-58: Peat Production of Conterminous 48 States (kt) 6-124
Table 6-59: Peat Production of Alaska (Thousand Cubic Meters) 6-124
Table 6-60: Peat Production Area of Conterminous 48 States (Hectares) 6-125
Table 6-61: Peat Production Area of Alaska (Hectares) 6-125
Table 6-62: Peat Production (Hectares) 6-126
Table 6-63: Approach 2 Quantitative Uncertainty Estimates for C02, CH4, and N20
Emissions from Peatlands Remaining Peatlands (MMT C02 Eq. and Percent) 6-127
Table 6-64: Emissions and Removals from Coastal Wetlands Remaining Coastal Wetlands
(MMT C02 Eq.) 6-131
Table 6-65: Net C02 Flux from Carbon Stock Changes in Vegetated Coastal Wetlands
Remaining Vegetated Coastal Wetlands (MMT C02 Eq.) 6-133
Table 6-66: Net C02 Flux from Carbon Stock Changes in Vegetated Coastal Wetlands
Remaining Vegetated Coastal Wetlands (MMT C) 6-133
Table 6-67: CH4 Emissions from Vegetated Coastal Wetlands Remaining Vegetated Coastal
Wetlands (MMTC02 Eq. and ktCH4) 6-133
Table 6-68: Area of Vegetated Coastal Wetlands Remaining Vegetated Coastal Wetlands,
Vegetated Coastal Wetlands Converted to Unvegetated Open Water Coastal
Wetlands, and Unvegetated Open Water Coastal Wetlands Converted to
Vegetated Coastal Wetlands (hectares) 6-134
Table 6-69: Aboveground Biomass Carbon Stocks for Vegetated Coastal Wetlands (t C ha1) 6-134
Table 6-70: Root to Shoot Ratios for Vegetated Coastal Wetlands 6-134
Table 6-71: Annual Soil Carbon Accumulation Rates for Vegetated Coastal Wetlands (t C ha~
1 yr"1) 6-135
Table 6-72: IPCC Approach 1 Quantitative Uncertainty Estimates for Carbon Stock Changes
and CH4 Emissions occurring within Vegetated Coastal Wetlands Remaining
Vegetated Coastal Wetlands in 2023 (MMT C02 Eq. and Percent) 6-137
Table 6-73: Net C02 Flux from Carbon Stock Changes in Vegetated Coastal Wetlands
Converted to Unvegetated Open Water Coastal Wetlands (MMT C02 Eq.) 6-139
Table 6-74: Net C02 Flux from Carbon Stock Changes in Vegetated Coastal Wetlands
Converted to Unvegetated Open Water Coastal Wetlands (MMT C) 6-139
Table 6-75: Approach 1 Quantitative Uncertainty Estimates for C02 Flux Occurring within
Vegetated Coastal Wetlands Converted to Unvegetated Open Water Coastal
Wetlands in 2023 (MMT C02 Eq. and Percent) 6-142
Table 6-76: C02 Flux from Carbon Stock Changes from Unvegetated Open Water Coastal
Wetlands Converted to Vegetated Coastal Wetlands (MMT C02 Eq.) 6-144
Table 6-77: C02 Flux from Carbon Stock Changes from Unvegetated Open Water Coastal
Wetlands Converted to Vegetated Coastal Wetlands (MMT C) 6-144
xvi Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Table 6-78: Approach 1 Quantitative Uncertainty Estimates for Carbon Stock Changes
Occurring within Unvegetated Open Water Coastal Wetlands Converted to
Vegetated Coastal Wetlands in 2023 (MMT C02 Eq. and Percent) 6-147
Table 6-79: N20 Emissions from Aquaculture in Coastal Wetlands (MMT C02 Eq. and kt N20) 6-148
Table 6-80: Approach 1 Quantitative Uncertainty Estimates for N20 Emissions from
Aquaculture Production in Coastal Wetlands in 2023 (MMT C02 Eq. and Percent) 6-149
Table 6-81: CH4 Emissions from Flooded Land Remaining Flooded Land—Reservoirs (MMT
C02 Eq.) 6-151
Table 6-82: CH4 Emissions from Flooded Land Remaining Flooded Land—Reservoirs (kt
CH4) 6-152
Table 6-83: Surface and Downstream CH4 Emissions from Reservoirs in Flooded Land
Remaining Flooded Land in 2023 (kt CH4) 6-153
Table 6-84: IPCC (2019) Default CH4 Emission Factors for Surface Emission from Reservoirs
in Flooded Land Remaining Flooded Land 6-154
Table 6-85: National Totals of Reservoir Surface Area in Flooded Land Remaining Flooded
Land (millions of ha) 6-156
Table 6-86: State Breakdown of Reservoir Surface Area in Flooded Land Remaining Flooded
Land (millions of ha) 6-156
Table 6-87: Approach 2 Quantitative Uncertainty Estimates for CH4 Emissions from
Reservoirs in Flooded Land Remaining Flooded Land 6-158
Table 6-88: CH4 Emissions from Other Constructed Waterbodies in Flooded Land
Remaining Flooded Land (MMTC02 Eq.) 6-160
Table 6-89: CH4 Emissions from Other Constructed Waterbodies in Flooded Land
Remaining Flooded Land (kt CH4) 6-160
Table 6-90: CH4 Emissions from Other Constructed Waterbodies in Flooded Land
Remaining Flooded Land in 2022 (kt CH4) 6-161
Table 6-91: IPCC (2019) Default CH4 Emission Factors for Surface Emissions from Other
Constructed Waterbodies in Flooded Land Remaining Flooded Land 6-164
Table 6-92: National Surface Area Totals in Flooded Land Remaining Flooded Land - Other
Constructed Waterbodies (hectares) 6-165
Table 6-93: State Totals of Surface Area in Flooded Land Remaining Flooded Land— Canals
and Ditches (hectares) 6-167
Table 6-94: State Totals of Surface Area in Flooded Land Remaining Flooded Land—
Freshwater Ponds (hectares) 6-168
Table 6-95: Approach 2 Quantitative Uncertainty Estimates for CH4 Emissions from Other
Constructed Waterbodies in Flooded Land Remaining Flooded Land 6-170
Table 6-96: Net C02 Flux from Carbon Stock Changes in Land Converted to Vegetated
Coastal Wetlands (MMT C02 Eq.) 6-172
Table 6-97: Net C02 Flux from Carbon Stock Changes in Land Converted to Vegetated
Coastal Wetlands (MMT C) 6-172
Table 6-98: CH4 Emissions from Land Converted to Vegetated Coastal Wetlands (MMT C02
Eq. and kt CH4) 6-173
Table 6-99: Approach 1 Quantitative Uncertainty Estimates for Carbon Stock Changes
occurring within Land Converted to Vegetated Coastal Wetlands in 2023 (MMT
C02 Eq. and Percent) 6-176
List of Tables, Figures, Boxes and Equations xvii
-------
Table 6-100: CH4 Emissions from Land Converted to Flooded Land—Reservoirs (MMT C02
Eq.) 6-179
Table 6-101: CH4 Emissions from Land Converted to Flooded Land—Reservoirs (kt CH4) 6-179
Table 6-102: C02 Emissions from Land Converted to Flooded Land—Reservoirs (MMT C02) 6-179
Table 6-103: C02 Emissions from Land Converted to Flooded Land—Reservoirs (MMT C) 6-179
Table 6-104: Methane and C02 Emissions from Reservoirs in Land Converted to Flooded
Land in 2023 (kt CH4; kt C02) 6-180
Table 6-105: IPCC (2019) Default CH4 and C02 Emission Factors for Surface Emissions from
Reservoirs in Land Converted to Flooded Land 6-182
Table 6-106: National Totals of Reservoir Surface Area in Land Converted to Flooded Land
(thousands of ha) 6-184
Table 6-107: State Breakdown of Reservoir Surface Area in Land Converted to Flooded Land
(thousands of ha) 6-184
Table 6-108: Approach 2 Quantitative Uncertainty Estimates for CH4 and C02 Emissions
from Reservoirs in Land Converted to Flooded Land 6-186
Table 6-109: CH4 Emissions from Other Constructed Waterbodies in Land Converted to
Flooded Land (MMT C02 Eq.) 6-188
Table 6-110: CH4 Emissions from Other Constructed Waterbodies in Land Converted to
Flooded Land (ktCH4) 6-188
Table 6-111: C02 Emissions from Other Constructed Waterbodies in Land Converted to
Flooded Land (MMT C02 Eq.) 6-188
Table 6-112: C02 Emissions from Other Constructed Waterbodies in Land Converted to
Flooded Land (MMT C) 6-188
Table 6-113: CH4 and C02 Emissions from Other Constructed Waterbodies in Land
Converted to Flooded Land in 2023 (MT C02 Eq.) 6-188
Table 6-114: IPCC Default Methane and C02 Emission Factors for Other Constructed
Waterbodies in Land Converted to Flooded Land 6-191
Table 6-115: National Surface Area Totals of Other Constructed Waterbodies in Land
Converted to Flooded Land (hectares) 6-192
Table 6-116: State Surface Area Totals of Other Constructed Waterbodies in Land Converted
to Flooded Land (hectares) 6-193
Table 6-117: Approach 2 Quantitative Uncertainty Estimates for CH4 and C02 Emissions
from Other Constructed Waterbodies in Land Converted to Flooded Land 6-194
Table 6-118: Net C02 Flux from Soil C Stock Changes in Settlements Remaining
Settlements (MMTC02 Eq.) 6-197
Table 6-119: Net C02 Flux from Soil C Stock Changes in Settlements Remaining
Settlements (MMTC) 6-197
Table 6-120: Thousands of Hectares of Drained Organic Soils in Settlements Remaining
Settlements 6-197
Table 6-121: Uncertainty Estimates for C02 Emissions from Drained Organic Soils in
Settlements Remaining Settlements (MMT C02 Eq. and Percent) 6-198
Table 6-122: Area of Managed Land in Settlements Remaining Settlements that is not
included in the current Inventory (Thousand Hectares) 6-199
Table 6-123: Net Flux from Trees in Settlements Remaining Settlements (MMT C02 Eq. and
MMT C)a 6-200
xviii Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Table 6-124: Carbon Storage (kg C/m2 tree cover), Gross and Net Sequestration (kg C/m2
tree cover/year) and Tree Cover (percent) among Sampled U.S. Cities (see
Nowaketal. 2013) 6-203
Table 6-125: Estimated Annual Carbon Sequestration, Tree Cover, and Annual Carbon
Sequestration per Area of Tree Cover for settlement areas in the United States
by State and the District of Columbia (2023) 6-206
Table 6-126: Approach 2 Quantitative Uncertainty Estimates for Net C02 Flux from Changes
in Carbon Stocks in Settlement Trees (MMT C02 Eq. and Percent) 6-208
Table 6-127: Recalculations of the Settlement Tree Categories 6-208
Table 6-128: N20 Emissions from Soils in Settlements Remaining Settlements (MMT C02
Eq.) 6-210
Table 6-129: N20 Emissions from Soils in Settlements Remaining Settlements (kt N20) 6-210
Table 6-130: Quantitative Uncertainty Estimates of N20 Emissions from Soils in Settlements
Remaining Settlements (MMT C02 Eq. and Percent) 6-212
Table 6-131: Net Changes in Yard Trimmings and Food Scrap Carbon Stocks in Landfills
(MMT C02 Eq.) 6-214
Table 6-132: Net Changes in Yard Trimmings and Food Scrap Carbon Stocks in Landfills
(MMT C) 6-214
Table 6-133: Moisture Contents, Carbon Storage Factors (Proportions of Initial Carbon
Sequestered), Initial C Contents, and Decay Rates for Yard Trimmings and Food
Scraps in Landfills 6-218
Table 6-134: Carbon Stocks in Yard Trimmings and Food Scraps in Landfills (MMT C) 6-218
Table 6-135: Approach 2 Quantitative Uncertainty Estimates for C02 Flux from Yard
Trimmings and Food Scraps in Landfills (MMT C02 Eq. and Percent) 6-219
Table 6-136: Net C02 Flux from Soil, Dead Organic Matter and Biomass Carbon Stock
Changes for Land Converted to Settlements (MMT C02 Eq.) 6-222
Table 6-137: Net C02 Flux from Soil, Dead Organic Matter and Biomass Carbon Stock
Changes for Land Converted to Settlements (MMT C) 6-223
Table 6-138: Thousands of hectares of land for total live biomass associated with grasslands
(non-woodland) and croplands converted to settlements 6-224
Table 6-139: Approach 2 Quantitative Uncertainty Estimates for Soil, Dead Organic Matter
and Biomass Carbon Stock Changes occurring within Land Converted to
Settlements (MMTC02 Eq. and Percent) 6-228
Table 6-140: Area of Managed Land in Land Converted to Settlements that is not included in
the current Inventory (Thousand Hectares) 6-230
Table 7-1: Emissions from Waste (MMT C02 Eq.) 7-3
Table 7-2: Emissions from Waste (kt) 7-4
Table 7-3: CH4 Emissions from Landfills (MMT C02 Eq.) 7-8
Table 7-4: CH4 Emissions from Landfills (kt CH4) 7-9
Table 7-5: Approach 2 Quantitative Uncertainty Estimates for CH4 Emissions from Landfills
(MMT C02 Eq. and Percent) 7-16
Table 7-6: Materials Discarded in the Municipal Waste Stream by Waste Type from 1990 to
2018 (Percent) 7-21
Table 7-7: CH4 and N20 Emissions from Domestic and Industrial Wastewater Treatment
(MMT C02 Eq.) 7-25
Table 7-8: CH4 and N20 Emissions from Domestic and Industrial Wastewater Treatment (kt) 7-25
List of Tables, Figures, Boxes and Equations xix
-------
Table 7-9: Industrial Wastewater Treatment Systems with (CH4) Emissions 7-29
Table 7-10: Industrial Wastewater Treatment Systems with N20 Emissions 7-32
Table 7-11: Approach 2 Quantitative Uncertainty Estimates for 2023 Emissions from
Wastewater Treatment (MMT C02 Eq. and Percent) 7-34
Table 7-12: CH4 and N20 Emissions from Composting (MMT C02 Eq.) 7-39
Table 7-13: CH4 and N20 Emissions from Composting (kt) 7-39
Table 7-14: U.S. Waste Composted (kt) 7-40
Table 7-15: Tier 1 Quantitative Uncertainty Estimates for Emissions from Composting (MMT
C02 Eq. and Percent) 7-41
Table 7-16: CH4 Emissions from Anaerobic Digestion at Biogas Facilities (MT C02 Eq.) 7-43
Table 7-17: CH4 Emissions from Anaerobic Digestion at Biogas Facilities (kt CH4) 7-43
Table 7-18: Estimated U.S. Waste Digested (kt) 7-46
Table 7-19: Estimated Number of Stand-Alone AD Facilities Operating 7-46
Table 7-20: Approach 1 Quantitative Uncertainty Estimates for Emissions from Anaerobic
Digestion (MT C02 Eq. and Percent) 7-48
Table 7-21: Emissions of NOx, CO, NMVOC, NH3 and S02 from Waste (kt) 7-50
Table 9-1: Overall Impact of Recalculations by Gas Compared to 1990-2022 Inventory (MMT
C02 Eq.) 9-2
Table 9-2: Overall Impact of Recalculations by Sector Compared to 1990-2022 Inventory
(MMT C02 Eq.) 9-3
Table 9-3: Key Recalculations 9-4
Table 9-4: Revisions to U.S. Greenhouse Gas Emissions (MMT C02 Eq.) 9-5
Table 9-5: Revisions to U.S. Greenhouse Gas Emissions and Removals (Net Flux) from Land
Use, Land-Use Change, and Forestry (MMT C02 Eq.) 9-8
Figures
Figure ES-1: U.S. Greenhouse Gas Emissions and Sinks by Gas ES-4
Figure ES-2: Annual Percent Change in Net and Gross U.S. Greenhouse Gas Emissions and
Sinks Relative to the Previous Year ES-5
Figure ES-3: Impacts of Recalculations on Net Emissions ES-6
Figure ES-4: 2023 Total Gross U.S. Greenhouse Gas Emissions by Gas (Percentages based
on MMTCOz Eq.) ES-7
Figure ES-5: 2023 Sources and Sinks of Carbon Dioxide Emissions ES-8
Figure ES-6: 2023 End-Use Sector Emissions of Carbon Dioxide from Fossil Fuel
Combustion ES-9
Figure ES-7: Electric Power Generation and Emissions ES-11
Figure ES-8: 2023 Sources of Methane Emissions ES-12
Figure ES-9: 2023 Sources of Nitrous Oxide Emissions ES-13
Figure ES-10: 2023 Sources of Hydrofluorocarbon, Perfluorocarbon, Sulfur Hexafluoride,
and Nitrogen Trifluoride Emissions ES-15
Figure ES-11: U.S. Greenhouse Gas Emissions and Sinks by Inventory Sector ES-16
Figure ES-12: 2023 U.S. Energy Consumption by Energy Source (Percent) ES-18
Figure ES-13: U.S. Greenhouse Gas Emissions Allocated to Economic Sectors ES-22
xx Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Figure ES-14: U.S. Greenhouse Gas Emissions with Electricity-Related Emissions
Distributed to Economic Sectors ES-25
Figure ES-15: U.S. Greenhouse Gas Emissions Per Capita and Per Dollar of Gross Domestic
Product (GDP) ES-26
Figure ES-16: 2023 Key Categories (Approach 1 including LULUCF) ES-27
Figure 1 -1: Key Data Institutions and Data Sources by Sector 1-18
Figure 1-2: Example QC Processes from Inventory QA/QC Plan 1-26
Figure 2-1: U.S. Greenhouse Gas Emissions and Sinks by Gas 2-2
Figure 2-2: Annual Percentage Change in Net and Gross U.S. Greenhouse Gas Emissions
Relative to the Previous Year 2-2
Figure 2-3: 2023 Gross Total U.S. Greenhouse Gas Emissions by Gas (Percentages based on
MMT C02 Eq.) 2-3
Figure 2-4: U.S. Greenhouse Gas Emissions and Removals by Inventory Sector 2-11
Figure 2-5: Trends in Energy Sector Greenhouse Gas Sources 2-13
Figure 2-6: Trends in C02 Emissions from Fossil Fuel Combustion by End-Use Sector and
Fuel Type 2-18
Figure 2-7: Trends in End-Use Sector Emissions of C02from Fossil Fuel Combustion 2-19
Figure 2-8: Electric Power Generation (Billion kWh) and Emissions (MMT C02 Eq.) 2-20
Figure 2-9: Trends in Industrial Processes and Product Use Sector Greenhouse Gas Sources 2-23
Figure 2-10: Trends in Agriculture Sector Greenhouse Gas Sources 2-27
Figure 2-11: Trends in Emissions and Removals (Net C02 Flux) from Land Use, Land-Use
Change, and Forestry 2-30
Figure 2-12: Trends in Waste Sector Greenhouse Gas Sources 2-33
Figure 2-13: U.S. Greenhouse Gas Emissions Allocated to Economic Sectors 2-35
Figure 2-14: U.S. Greenhouse Gas Emissions with Electricity-Related Emissions Distributed
to Economic Sectors 2-41
Figure 2-15: Trends in Transportation-Related Greenhouse Gas Emissions 2-45
Figure 2-16: U.S. Greenhouse Gas Emissions Per Capita and Per Dollar of Gross Domestic
Product (GDP) 2-49
Figure 3-1: 2023 Energy Sector Greenhouse Gas Sources 3-3
Figure 3-2: Trends in Energy Sector Greenhouse Gas Sources 3-3
Figure 3-3: 2023 U.S. Fossil Carbon Flows (MMT C02 Eq.) 3-4
Figure 3-4: 2023 U.S. Energy Use by Energy Source 3-12
Figure 3-5: Annual U.S. Energy Use 3-12
Figure 3-6: 2023 C02 Emissions from Fossil Fuel Combustion by Sector and Fuel Type 3-13
Figure 3-7: Fuels Used in Electric Power Generation and Total Electric Power Sector C02
Emissions 3-20
Figure 3-8: Electric Power Retail Sales by End-Use Sector 3-21
Figure 3-9: Industrial Production Indices (Index 2017=100) 3-23
Figure 3-10: Fuels and Electricity Used in Industrial Sector, Industrial Output, and Total
Sector C02 Emissions (Including Electricity) 3-24
Figure 3-11: Fuels and Electricity Used in Residential and Commercial Sectors, Heating and
Cooling Degree Days, and Total Sector C02 Emissions (Including Electricity) 3-26
Figure 3-12: Annual Deviations from Normal Heating Degree Days for the United States
(1970-2023, Index Normal = 100) 3-27
List of Tables, Figures, Boxes and Equations xxi
-------
Figure 3-13: Annual Deviations from Normal Cooling Degree Days for the United States
(1970-2023, Index Normal = 100) 3-28
Figure 3-14: Fuels Used in Transportation Sector, On-road VMT, and Total Sector C02
Emissions 3-30
Figure 3-15: Sales-Weighted Fuel Economy of New Passenger Cars and Light-Duty Trucks,
1990-2023 3-32
Figure 3-16: Sales of New Passenger Cars and Light-Duty Trucks, 1990-2023 3-32
Figure 3-17: Mobile Source CH4 and N20 Emissions 3-35
Figure 3-18: Flow of C02 Capture and Sequestration 3-125
Figure 4-1: Industrial Processes and Product Use Sector Greenhouse Gas Sources 4-3
Figure 4-2: Trends in Industrial Processes and Product Use Sector Greenhouse Gas Sources 4-4
Figure 5-1: 2023 Agriculture Sector Greenhouse Gas Emission Sources 5-2
Figure 5-2: Trends in Agriculture Sector Greenhouse Gas Emission Sources 5-4
Figure 5-3: Annual CH4 Emissions from Rice Cultivation, 2020, Using the Tier 3 DayCent
Model 5-28
Figure 5-4: Sources and Pathways of Nitrogen that Result in N20 Emissions from
Agricultural Soil Management 5-35
Figure 5-5: Croplands, 2020 Annual Direct N20 Emissions Estimated Using the Tier 3
DayCent Model 5-38
Figure 5-6: Grasslands, 2020 Annual Direct N20 Emissions Estimated Using the Tier 3
DayCent Model 5-39
Figure 5-7: Croplands, 2020 Annual Indirect N20 Emissions from Volatilization Using the Tier
3 DayCent Model 5-40
Figure 5-8: Grasslands, 2020 Annual Indirect N20 Emissions from Volatilization Using the
Tier 3 DayCent Model 5-40
Figure 5-9: Croplands, 2020 Annual Indirect N20 Emissions from Leaching and Runoff Using
the Tier 3 DayCent Model 5-41
Figure 5-10: Grasslands, 2020 Annual Indirect N20 Emissions from Leaching and Runoff
Using the Tier 3 DayCent Model 5-41
Figure 6-1: 2023 LULUCF Chapter Greenhouse Gas Sources and Sinks 6-4
Figure 6-2: Trends in Emissions and Removals (Net C02 Flux) from Land Use, Land-Use
Change, and Forestry 6-4
Figure 6-3: Percent of Total Land Area for Each State in the General Land Use Categories for
2023 6-15
Figure 6-4: Area by Region for Forest Land Remaining Forest Land in the United States
(1990-2023) 6-33
Figure 6-5: Estimated Net Annual Changes in Carbon Stocks for All Carbon Pools in Forest
Land Remaining Forest Land in the United States (1990-2023) 6-37
Figure 6-6: Total Net Annual Soil Carbon Stock Changes for Mineral Soils under Agricultural
Management within States, 2020, Cropland Remaining Cropland 6-71
Figure 6-7: Total Net Annual Soil Carbon Stock Changes for Organic Soils under Agricultural
Management within States, 2020, Cropland Remaining Cropland 6-72
Figure 6-8: Total Net Annual Soil Carbon Stock Changes for Mineral Soils under Agricultural
Management within States, 2020, Grassland Remaining Grassland 6-98
Figure 6-9: Total Net Annual Soil Carbon Stock Changes for Organic Soils under Agricultural
Management within States, 2020, Grassland Remaining Grassland 6-99
xxii Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Figure 6-10: U.S. Reservoirs (black polygons) in the Flooded Land Remaining Flooded Land
Category in 2023 6-151
Figure 6-11: Total CH4 Emissions (Downstream + Surface) from Reservoirs in Flooded Land
Remaining Flooded Land in 2023 (kt CH4) 6-152
Figure 6-12: Selected Features from NWI that Meet Flooded Lands Criteria 6-156
Figure 6-13: 2023 CH4 Emissions from A) Ditches and Canals and B) Freshwater Ponds in
Flooded Land Remaining Flooded Land (kt CH4) 6-163
Figure 6-14: 2023 Surface Area of A) Ditches and Canals and B) Freshwater Ponds in
Flooded Land Remaining Flooded Land (hectares) 6-166
Figure 6-15: U.S. Reservoirs (black polygons) in the Land Converted to Flooded Land
Category in 2023 6-178
Figure 6-16: 2023 A) CH4 and B) C02 Emissions from U.S. Reservoirs in Land Converted to
Flooded Land 6-180
Figure 6-17: Selected Features from NWI that meet Flooded Lands Criteria 6-183
Figure 6-18: Number of Dams Built per Year from 1990 through 2023 6-184
Figure 6-19: 2023 A) CH4 and B) C02 Emissions from Other Constructed Waterbodies
(Freshwater Ponds) in Land Converted to Flooded Land (MT C02 Eq.) 6-190
Figure 6-20: Surface Area of Other Constructed Waterbodies in Land Converted to Flooded
Land (hectares) in 2023 6-192
Figure 7-1: 2023 Waste Sector Greenhouse Gas Sources 7-2
Figure 7-2: Trends in Waste Sector Greenhouse Gas Sources 7-3
Figure 7-3: Methodologies Used Across the Time Series to Compile the Emission Estimates
for MSW Landfills 7-11
Figure 7-4: Management of Municipal Solid Waste in the United States, 2018 7-20
Figure 7-5: MSW Management Trends from 1990 to 2018 7-21
Figure 7-6: Percent of Degradable Materials Diverted from Landfills from 1990 to 2018
(Percent) 7-22
Figure 9-1: Impacts of Recalculations on Net Emissions 9-2
Figure 9-2: Impacts from Recalculations to U.S. Greenhouse Gas Emissions and Sinks by
Sector 9-5
Boxes
Box ES-1: Relationship to the U.S. EPA's Greenhouse Gas Reporting Program ES-1
Box ES-2: Inventory Reporting Sectors ES-17
Box ES-3: Trends in Various U.S. Greenhouse Gas Emissions-Related Data ES-25
Box ES-4: Use of Ambient Measurements Systems for Validation of Emission Inventories ES-28
Box 1-1: Relationship to EPA's Greenhouse Gas Reporting Program 1-1
Box 1-2: The IPCC Sixth Assessment Report and Global Warming Potentials 1-10
Box 1-3: Examples of Verification Activities 1-26
Box 1 -4: Organization of Report 1 -33
Box 2-1: Methodology for Aggregating Emissions by Economic Sector 2-38
Box 2-2: Trends in Various U.S. Greenhouse Gas Emissions-Related Data 2-48
Box 3-1: Uses of EPA's Greenhouse Gas Reporting Program Energy Data 3-7
Box 3-2: Uses of Greenhouse Gas Reporting Program Data and Improvements in Reporting
Emissions from Industrial Sector Fossil Fuel Combustion 3-24
List of Tables, Figures, Boxes and Equations xxiii
-------
Box 3-3: Weather and Non-Fossil Energy Effects on C02 Emissions from Fossil Fuel
Combustion Trends 3-27
Box 3-4: Carbon Intensity of U.S. Energy Consumption 3-41
Box 3-5: Reporting of Lubricants, Waxes, and Asphalt and Road Oil Product Use in Energy
Sector 3-61
Box 4-1: Uses of EPA's Greenhouse Gas Reporting Program Energy Data 4-10
Box 5-1: Surrogate Data Method 5-31
Box 5-2: Tier 1 vs. Tier 3 Approach for Estimating N20 Emissions 5-43
Box 5-3: Data Splicing Method 5-44
Box 5-4: Comparison of the Tier 2 U.S. Inventory Approach and IPCC (2006) Default
Approach 5-57
Box 5-5: Comparison of Tier 2 U.S. Inventory Approach and IPCC (2006) Default Approach 5-67
Box 6-1: Preliminary Estimates of Land Use in U.S. Territories 6-28
Box 6-2: C02 Emissions from Forest Fires 6-37
Box 6-3: Surrogate Data Method 6-76
Box 6-4: Tier 3 Approach for Soil Carbon Stocks Compared to Tier 1 or 2 Approaches 6-77
Box 6-5: State-Level Case Studies for the Estimation of GHG Removals in Seagrasses 6-138
Box 7-1: Use of Greenhouse Gas Reporting Data in Waste Sector 7-5
Box 7-2: Description of a Modern, Managed Landfill in the United States 7-6
Box 7-3: Nationwide Municipal Solid Waste Data Sources 7-13
Box 7-4: Overview of U.S. Solid Waste Management Trends 7-20
Equations
Equation 1-1: Calculating C02 Equivalent Emissions 1-9
Equation 3-1: Estimating Fugitive C02 Emissions from Underground Mines 3-76
Equation 3-2: Estimating C02 Emissions from Drained Methane Flared or Catalytically
Oxidized 3-77
Equation 3-3: Decline Function to Estimate Venting Abandoned Mine Methane Emissions 3-81
Equation 3-4: Decline Function to Estimate Flooded Abandoned Mine Methane Emissions 3-82
Equation 4-1: 2006 IPCC Guidelines Tier 1 Emission Factor for Clinker (precursor to Equation
2.4) 4-13
Equation 4-2: 2006 IPCC Guidelines Tier 2 Emission Factor for Lime Production, High-
Calcium Lime (Equation 2.9) 4-18
Equation 4-3: 2006 IPCC Guidelines Tier 2 Emission Factor for Lime Production, Dolomitic
Lime (Equation 2.9) 4-18
Equation 4-4: 2006 IPCC Guidelines Tier 3: N20 Emissions From Nitric Acid Production
(Equation 3.6) 4-49
Equation 4-5: 2006 IPCC Guidelines Tier 2: N20 Emissions From Adipic Acid Production
(Equation 3.8) 4-53
Equation 4-6: 2006 IPCC Guidelines Tier 1: N20 Emissions From Caprolactam Production
(Equation 3.9) 4-58
Equation 4-7:2006 IPCC Guidelines Tier 1: Emissions from Carbide Production (Equation
3.11) 4-62
xxiv Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Equation 4-8: 2006IPCC Guidelines Tier 1: C02 Emissions from Titanium Production
(Equation 3.12) 4-66
Equation 4-9: C02 Emissions from Phosphoric Acid Production 4-105
Equation 4-10: C02 Emissions from Coke, Pig Iron, EAF Steel, and BOF Steel Production,
based on 2006 IPCC Guidelines Tier 2 Methodologies 4-113
Equation 4-11: 2006 IPCC Guidelines Tier 1: Emissions from Sinter, Direct Reduced Iron, and
Pellet Production (Equations 4.6, 4.7, and 4.8) 4-113
Equation 4-12: 2006 IPCC Guidelines Tier 1: C02 Emissions for Ferroalloy Production
(Equation 4.15) 4-124
Equation 4-13: 2006 IPCC Guidelines Tier 1: CH4 Emissions for Ferroalloy Production
(Equation 4.18) 4-125
Equation 4-14: CF4 Emissions Resulting from Low Voltage Anode Effects 4-133
Equation 4-15: 2006 IPCC Guidelines Tier 1: C02 Emissions From Lead Production (Equation
4.32 ) 4-144
Equation 4-16: 2006 IPCC Guidelines Tier 1: C02 Emissions from Zinc Production (Equation
4.33 ) 4-149
Equation 4-17: Waelz Kiln C02 Emission Factor for Zinc Produced 4-150
Equation 4-18: Waelz Kiln C02 Emission Factor for EAF Dust Consumed 4-150
Equation 4-19: Total Emissions from Electronics Industry 4-168
Equation 4-20: Total Emissions from Semiconductor Manufacturing 4-169
Equation 4-21: Total Emissions from MEMS Manufacturing 4-171
Equation 4-22: Total Emissions from PV Manufacturing 4-172
Equation 4-23: Estimation for SF6 Emissions from Electric Power Systems 4-184
Equation 4-24: Regression Equation for Estimating SF6 Emissions of Non-Reporting
Facilities in 1999 4-186
Equation 4-25: Regression Equation for Estimating SF6 Emissions of GHGRP-Only Reporters
in 2011 4-187
Equation 4-26: Total Emissions from Other Product Use 4-198
Equation 4-27: Total Emissions from Military Applications 4-198
Equation 4-28: Total Emissions from Scientific Applications 4-199
Equation 4-29: N20 Emissions from Product Use 4-202
Equation 5-1: Elemental C or N Released through Oxidation of Crop Residues 5-66
Equation 5-2: Emissions from Crop Residue Burning 5-67
Equation 5-3: Estimation of Greenhouse Gas Emissions from Fire 5-67
Equation 6-1: Net State Annual Carbon Sequestration 6-206
Equation 6-2: Total Carbon Stock for Yard Trimmings and Food Scraps in Landfills 6-217
Equation 6-3: Carbon StockAnnual Flux for Yard Trimmings and Food Scraps in Landfills 6-217
Equation 7-1: Landfill Methane Emissions 7-10
Equation 7-2: Net Methane Emissions from MSW Landfills 7-10
Equation 7-3: Net Methane Emissions from Industrial Waste Landfills 7-12
Equation 7-4: Greenhouse Gas Emission Calculation for Composting 7-39
Equation 7-5: Methane Emissions Calculation for Anaerobic Digestion 7-44
Equation 7-6: Methane Generation Calculation for Anaerobic Digestion 7-44
Equation 7-7: Weighted Average of Waste Processed 7-45
List of Tables, Figures, Boxes and Equations xxv
-------
Executive Summary
This Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023 (Inventory) identifies and
quantifies the anthropogenic1 emissions sources and removals (by sinks) of greenhouse gases in the
U.S. It is an essential tool for understanding the relative magnitude of different sources and sinks across
the U.S., as well as changes in these magnitudes over time. This Executive Summary provides the latest
information on U.S. anthropogenic greenhouse gas emissions and removals trends from 1990 through
2023. Throughout this report, emission and sink estimates are grouped into five reporting sectors (i.e.,
chapters): Energy; Industrial Processes and Product Use (IPPU); Agriculture; Land Use, Land-Use
Change, and Forestry (LULUCF); and Waste; and are calculated using methods that are consistent with
the 2006 Intergovernmental Panel on Climate Change (IPCC) Guidelines for National Greenhouse Gas
Inventories (2006 IPCC Guidelines) and, where appropriate, its supplements and refinements. The
structure of this report follows the common approach used by Parties to the United Nations Framework
Convention on Climate Change (UNFCCC). The presentation of emissions and removals provided in this
Inventory does not preclude alternative examinations (e.g., economic sectors). See Box ES-1 to
understand the relationship to other greenhouse gas data collected and reported by the U.S.
Environmental Protection Agency (U.S. EPA).
Box ES-1: Relationship to the U.S. EPA's Greenhouse Gas Reporting Program
The U.S. EPA also collects greenhouse gas data and other relevant information from individual facilities
and suppliers of certain fossil fuels and industrial gases through its Greenhouse Gas Reporting Program
(GHGRP), which is complementary to the U.S. Inventory.2 The GHGRP applies to direct greenhouse gas
emitters, fossil fuel suppliers, industrial greenhouse gas suppliers, and facilities that inject carbon
dioxide (C02) underground for sequestration or other reasons, and requires reporting by over 8,000
sources or suppliers in 46 industrial categories.3 Annual reporting is at the facility level, except for
certain suppliers of fossil fuels and industrial greenhouse gases. In general, the threshold for reporting is
25,000 metric tons or more of C02 Eq. per year. Facilities in most source categories subject to GHGRP
began reporting for the 2010 reporting year, while additional types of industrial operations began
reporting for reporting year 2011 (EPA, 2019). Methodologies used in the U.S. EPA's GHGRP are
consistent with the 2006 IPCC Guidelines. While the GHGRP does not provide full coverage of total
1 The term "anthropogenic," in this context, refers to greenhouse gas emissions and removals that are a direct result of
human activities or are the result of natural processes that have been affected by human activities (IPCC 2006).
2 On October 30, 2009, EPA promulgated a rule requiring annual reporting of greenhouse gas data from large greenhouse
gas emissions sources in the United States. Implementation of the rule, codified at 40 CFR Part 98, is referred to as EPA's
Greenhouse Gas Reporting Program (GHGRP).
3 See http://www.epa.gov/ghgreporting and http://ghgdata.epa.gov/ghgp/main.do.
Executive Summary ES-1
-------
annual U.S. greenhouse gas emissions and removals (e.g., the GHGRP excludes emissions from the
Agriculture and Land Use, Land-Use Change, and Forestry sectors), it is an important input to the
calculations of national-level emissions in this Inventory.
The GHGRP dataset provides not only annual emissions information, but also other annual information
such as activity data and emission factors that can improve and refine national emission estimates over
time. GHGRP data also allow the U.S. EPA to disaggregate national inventory estimates in new ways that
can highlight differences across regions and subcategories of emissions, along with enhancing the
application of quality assurance/quality control procedures and assessment of uncertainties. See Annex
9 for more information on specific uses of GHGRP data in the Inventory.
ES.1 Background Information
Greenhouse gases absorb infrared radiation, trapping heat in the atmosphere and making the planet
warmer. The most important greenhouse gases directly emitted by human activities include carbon
dioxide (C02), methane (CH4), nitrous oxide (N20), and several fluorine-containing halogenated
substances (hydrofluorocarbons [HFCs], perfluorocarbons [PFCs], sulfur hexafluoride [SF6] and
nitrogen trifluoride [NF3]). Although C02, CH4, and N20 occur naturally in the atmosphere, human
activities have changed their atmospheric concentrations. From the pre-industrial era (i.e., ending about
1750) to 2023, concentrations of these greenhouse gases have increased globally by 50.4,163.2, and
24.7 percent, respectively (IPCC 2013; NOAA/ESRL 2025a, 2025b, 2025c). This annual report estimates
the total national greenhouse gas emissions and removals associated with human activities across the
United States.
Global Warming Potentials
The IPCC developed the global warming potential (GWP) concept to compare the ability of a greenhouse
gas to trap heat in the atmosphere relative to another gas. A GWP is a quantified measure of the
warming impact of a unit of a specific greenhouse gas over a specific period of time relative to the same
unit of C02. The GWP of a greenhouse gas is defined as the ratio of the accumulated radiative forcing
within a specific time horizon caused by emitting 1 kilogram of the gas, relative to that of the reference
gas C02 (IPCC 2021); therefore, C02-equivalent emissions are provided in million metric tons of C02
equivalent (MMT C02 Eq.) for non-C02 greenhouse gases.4 5 All estimates are provided throughout the
main report in both C02 equivalents and unweighted units. Estimates for all gases in this Executive
Summary are presented in units of MMT C02 Eq. Emissions by gas in unweighted mass kilotons are also
provided in Chapter 2, Trends and individual sector chapters of this report.
Based on recent decisions under the UNFCCC6 in 2024, Parties began using 100-year GWP values from
the IPCC Fifth Assessment Report (AR5) for calculating C02-equivalents in their national greenhouse
4 Carbon comprises 12/44 of carbon dioxide byweight.
5 One million metric ton is equal to 1012 grams or one teragram.
6 See paragraphs 1 and 2 of the decision on common metrics adopted at the 27th UNFCCC Conference of Parties
(COP27), available online at https://unfccc.int/sites/default/files/resource/cp2022 10a01 E.pdf.
ES-2 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
gas inventories (IPCC 2013). This change reflects updated science and ensures that national
greenhouse gas inventories from all nations are comparable. A comparison of emission values with the
IPCC Sixth Assessment Report (AR6) (IPCC 2021) values can be found in Annex 6.1 of this report. The
100-year GWP values used in this report are listed below in Table ES-1.
Table ES-1: Global Warming Potentials (100-Year Time Horizon) Used in this Report
Gas GWP
C02 1
CH4a 28
N20 265
HFCs up to 12,400
PFCs up to 11,100
SF6 23,500
NFs 16,100
Other Fluorinated Gases See Annex 6
a The GWP of ChU includes the direct effects and those indirect effects due to the production of tropospheric ozone and
stratospheric water vapor. The indirect effect due to production of C02 is not included. See Annex 6 for additional information.
Source: IPCC (2013).
ES.2 Summary of Trends in U.S. Greenhouse
Gas Emissions and Sinks
In 2023, total gross U.S. greenhouse gas emissions were 6,197.3 MMT C02 Eq.Total gross U.S.
emissions, which exclude emissions and sinks from the LULUCF sector, decreased by 5.2 percent from
1990 to 2023, down from a high of 15.3 percent above 1990 levels in 2007. Gross emissions decreased
from 2022 to 2023 by 2.3 percent (146.8 MMT C02 Eq.), driven largely by a decrease in C02 emissions
from fossil fuel combustion. C02 emissions from fossil fuel combustion decreased by 3.0 percent in
2023 relative to 2022 and were 4.1 percent below 1990 emissions. Specifically, C02 emissions from coal
consumption decreased by 18.3 percent (164.1 MMT C02 Eq.) from 2022 to 2023. C02 emissions from
natural gas use increased by 1.0 percent (17.6 MMT C02 Eq.) and emissions from petroleum use
increased by 0.2 percent (3.1 MMTC02 Eq.) from 2022 to 2023. The decrease in coal use and associated
emissions from 2022 to 2023 is mainly due to reduced use in the electric power sector and is driving the
overall reduction. The increase in natural gas consumption and associated emissions in 2023 is
observed mostly in the electric power and industrial sectors, the increase in petroleum use is mainly in
the transportation sector.
Net emissions, including emissions and sinks from the LULUCF sector, were 5,257.4 MMT C02 Eq. in
2023. Overall, net emissions decreased by 3.3 percent from 2022 to 2023. Over the last 20 years, net
emissions decreased by nearly 20 percent. Trends in net emissions are illustrated in Table ES-2. Carbon
sequestration from the LULUCF sector offset 16.1 percent of total gross emissions in 2023.
Executive Summary ES-3
-------
Figure ES-1 and Figure ES-2 illustrate the overall trend in total U.S. emissions and sinks since 1990, by
gas and annual percent change relative to the previous year. Table ES-2 provides information on trends
in gross and net U.S. greenhouse gas emissions and sinks for 1990 through 2023. Unless otherwise
stated, all tables and figures provide total gross emissions and exclude the greenhouse gas fluxes from
the LULUCF sector. For an overview of the LULUCF sector, see Section ES-3.
Figure ES-1: U.S. Greenhouse Gas Emissions and Sinks by Gas
¦ Net Emissions (including Net CO2 Flux from LULUCF)
9,000
8,000
7,000
6,000
S 5'000
Oj
S 4,000
i—
I 3,000
2,000
1,000
0
-1,000
I HFCs, PFCs, SFe and NFb
I Nitrous Oxide
Methane
I Carbon Dioxide
I Net CO2 Flux from LULUCFa
Oi-irNrv-i^rLriicirvCOCTiOT-irvirO'g-Lriior^cocnoi-HrNm'srLnior^cocno-i-irNm
O^O^O^O^CTIO^O^O^O^O^OOOOOOOOOOt—It—I 1—I T—It—It—li—It—I ri tH (N (N (N (N
CTioiaiCTiaiaiCTicricncrioooooooooooooooooooooooo
r-It—It—It—It—It—It—It—It—It—IINfNCNININCN(NINCM(Nr\ICM(NININ(NINrM(NINOI(NIN(N
8 The term "flux" is used to describe the exchange of C02 to and from the atmosphere, with "net flux" being either positive or
negative depending on the overall balance. Removal and long-term storage of C02 from the atmosphere is also referred to as
"carbon sequestration."
Notes: Gas totals exclude CH4, and N20 greenhouse gas fluxes from the LULUCF sector. Net emissions values include both CH4
and N2O emissions and the net carbon flux from the LULUCF sector.
Table ES-2: Recent Trends in U.S. Greenhouse Gas Emissions and Sinks (MMT C02 Eq.)
Gas/Source
1990
2005
2019
2020
2021
2022
2023
Percent
Change
Since
1990
CO2
5,131.8
6,126.9
5,235.9
4,690.0
5,020.1
5,055.4
4,918.4
-4.2%
ChUfexcludes LULUCF sources)"
873.1
797.1
752.6
730.9
715.6
696.8
686.7
-21.4%
N2O (excludes LULUCF sources)"
407.8
424.8
416.4
391.4
398.4
387.5
387.0
-5.1%
HFCs
47.8
125.0
175.8
177.8
184.3
189.5
191.0
299.8%
PFCs
39.7
10.3
7.3
6.6
6.3
6.5
5.8
-85.5%
SFe
37.9
20.2
8.3
7.7
8.0
7.2
7.7
-79.6%
NF3
0.2
1.0
1.1
1.3
1.1
1.1
0.8
335.8%
Total Gross Emissions (Sources)a
6,538.3
7,505.3
6,597.4
6,005.7
6,333.8
6,344.1
6,197.3
-5.2%
ES-4 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
LULUCF Emissions b
59.1
71.8
63.2
82.6
81.0
68.6
60.6
2.6%
CH4
54.4
60.9
56.1
69.0
67.8
59.6
54.7
0.5%
N2O
4.7
10.9
7.0
13.7
13.1
9.0
5.9
26.7%
LULUCF Carbon Stock Change c
(1,096.9)
(1,040.7)
(982.6)
(1,034.2)
(1,043.8)
(973.9)
(1,000.5)
-8.8%
LULUCF Sector Net Totald
(1,037.9)
(968.9)
(919.4)
(951.6)
(962.9)
(905.3)
(939.9)
-9.4%
Net Emissions (Sources and
Sinks)
5,500.4
6,536.4
5,678.0
5,054.2
5,371.0
5,438.7
5,257.4
-4.4%
a Gross emissions totals do not include ChU and N20 emissions from Land Use, Land-Use Change, and Forestry (LULUCF).
LULUCF CH4 and N20 emissions are included in net emission totals.
b LULUCF emissions subtotal of CH4 and N20 are reported separately from gross emissions totals. LULUCF emissions include
the CH4 and N20 emissions reported for peatlands remaining peatlands, forest fires, drained organic soils, grassland fires, and
coastal wetlands remaining coastal wetlands; CH4 emissions from land converted to coastal wetlands, flooded land remaining
flooded land, and land converted to flooded land; and N20 emissions from forest soils and settlement soils.
c LULUCF carbon stock change is the net carbon stock change from the following categories: forest land remaining forest land,
land converted to forest land, cropland remaining cropland, land converted to cropland, grassland remaining grassland, land
converted to grassland, wetlands remaining wetlands, land converted to wetlands, settlements remaining settlements, and
land converted to settlements.
d The LULUCF sector net total is the net sum of all LULUCF ChUand N20 emissions to the atmosphere plus LULUCF net carbon
stock changes.
Notes: Total gross emissions are emissions presented without LULUCF. Net emissions are presented with LULUCF. Totals may
not sum due to independent rounding. Parentheses indicate negative values or sequestration.
Figure ES-2: Annual Percent Change in Net and Gross U.S. Greenhouse Gas Emissions
and Sinks Relative to the Previous Year
8%
6%
2 iiiiiiiill.il—ill -i.-il il II |L. I
n n "'i
-2%
-4%
-6%
-8% ¦ Change in Gross Emissions
¦ Change in Net Emissions
-10%
-12%
i-ici«jm'3-mvor-vOoa\o-i-irMco^'mvoi*-.ooo>o^-o^oooooooooooooooooooooooo
HHHHHHHHH(N(N(N(NMlNfMfMfN(NlN(N(\lN(N(NN(NINlNlNfMlN(N
Improvements and Recalculations Relative to the Previous
Inventory
Each year, some of the emission and removal estimates in the Inventory are recalculated and revised to
incorporate improved methods and/or data. The most common reason for recalculating U.S.
greenhouse gas emission estimates is to update recent historical data. Changes in historical data are
generally the result of changes in data supplied by other U.S. government agencies or organizations, as
Executive Summary ES-5
-------
they continue to make refinements and improvements. These improvements are implemented
consistently across the previous Inventory's time series, as necessary, (i.e., 1990 to 2022) to ensure that
the trend is accurate.
Collectively, all methodological changes and historical data updates made in the current Inventory
resulted in lower estimates of annual net emissions by an average of 56.0 MMT C02 Eq. (0.9 percent).
Figure ES-3: Impacts of Recalculations on Net Emissions
Below are categories with methodological and data-related recalculations resulting in an average
annual change of greater than 2.0 MMT C02 Eq. over the time series, in descending order.
• Forest land remaining forest land: changes in forest carbon stocks (C02)
• Land converted to settlements: changes in all ecosystem carbon stocks (C02)
• Substitution of ozone depleting substances (HFCs)
• Land converted to grassland: changes in all ecosystem carbon stocks (C02)
• Land converted to forest land: changes in forest carbon stocks (C02)
In addition, the current Inventory includes the following new categories that were not a part of the
previous Inventory that improve the completeness of the national estimates: C02 transport, injection
and storage; and perennial woody biomass carbon stock changes and biomass carbon stock changes
from croplands and lands converted to and from croplands (e.g., croplands converted to grasslands,
grasslands converted to croplands). This Inventory also now includes additional gases (NF3 and HFCs)
within the SF6 and PFCs from other product use category.
The results of all methodological changes and historical data updates and the inclusion of new sources
and sink estimates are summarized in the Recalculations and Improvements chapter (Chapter 9 of this
ES-6 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Inventory). For more detailed descriptions of each recalculation, including references for data, please
see the respective emission or sink category description(s) within the relevant chapter (Chapter 3,
Energy, Chapter 4, Industrial Processes and Product Use; Chapter 5, Agriculture; Chapter 6, Land Use,
Land-Use Change, and Forestry, and Chapter 7, Waste).
Emissions and Sinks Trends by Greenhouse Gas
Figure ES-4 illustrates the relative contribution of each gas to total gross U.S. emissions in 2023, in C02
equivalents (i.e., weighted by GWP). The primary greenhouse gas emitted by human activities in the
United States is C02, representing 79.4 percent of total greenhouse gas emissions. The largest source of
C02, and of overall greenhouse gas emissions, is fossil fuel combustion, primarily from transportation
and power generation. CH4 emissions account for 11.1 percent of emissions while N20 accounts for an
additional 6.2 percent of emissions. The major sources of methane include enteric fermentation
associated with domestic livestock, natural gas systems, and decomposition of wastes in landfills.
Agricultural soil management, wastewater treatment, stationary sources of fuel combustion, mobile
combustion and manure management are the major sources of N20 emissions. Emissions of
substitutes for ozone depleting substances are the primary contributor to aggregate HFC emissions.
PFC emissions are primarily attributable to fluorochemical production, electronics manufacturing, and
primary aluminum production. Electrical equipment systems account for most SF6 emissions. The
electronics industry and fluorochemical production are the only sources of NF3 emissions. U.S.
greenhouse gas emissions were partly offset by carbon sequestration in forests, trees in urban areas,
agricultural soils, landfilledyard trimmings and food scraps, and coastal wetlands, which together
offset 16.1 percent of total gross emissions in 2023 (as reflected in Figure ES-1). The following sections
describe in more detail each gas's contribution to total U.S. greenhouse gas emissions.
Figure ES-4: 2023 Total Gross U.S. Greenhouse Gas Emissions by Gas (Percentages
based on MMT C02 Eq.)
3.3%
HFCs, PFCs, SFe and NFs
Note: Emissions and sinks from the Land Use, Land-Use Change, and Forestry sector are excluded from the figure above.
Executive Summary ES-7
-------
Carbon Dioxide Emissions
Overall, gross C02 emissions have decreased by 4.2 percent since 1990 and decreased by 2.7 percent
since 2022, consistent with trends in fuel combustion emissions. In the United States, fossil fuel
combustion accounted for 92.7 percent of gross C02 emissions in 2023. Nationally, within fossil fuel
combustion, the transportation sector was the largest emitter of C02 in 2023, followed by electric power
generation. There are 28 additional sources of C02 emissions included in the Inventory (see Table 2-1),
including sources and sinks from the LULUCF sector. Changes in land use and forestry practices can
also lead to net C02 emissions (e.g., through conversion of forest land to agricultural or urban use) or to
a net sink for C02 (e.g., through net additions to forest biomass). See more on these LULUCF emissions
and removals or sinks in Table ES-4 and a summarization of C02 sources in Figure ES-5.
Figure ES-5: 2023 Sources and Sinks of Carbon Dioxide Emissions
Fossil Fuel Combustion
Non-Energy Use of Fuels
Iron and Steel Production
Cement Production
Natural Gas Systems
Petrochemical Production
Other Industrial Processes
Petroleum Systems
Incineration of Waste
Ammonia Production
Lime Production
Other Agriculture
Other Energy
Net Carbon Stock Change from LULUCF
MMT CO2 Eq.
Note: "Other Industrial Processes" includes emissions from aluminum production, carbide production and consumption, non-
EOR carbon dioxide utilization, ferroalloy production, glass production, lead production, magnesium production, other process
uses of carbonates, phosphoric acid production, substitution of ozone depleting substances, soda ash production, titanium
dioxide production, urea consumption for non-agricultural purposes, and zinc production. "Other Energy" includes emissions
from abandoned oiland gas wells; 002 transport, injection, and storage; and coal mining.
Fossil Fuel Combustion Trends
Historically, changes in emissions from fossil fuel combustion have been the driving factor affecting
overall U.S. emission trends. Important drivers include changes in demand for energy and a general
decline in the overall carbon intensity of fuels combusted for energy in recent years by non-transport
sectors of the economy. Between 1990 and 2023, C02 emissions from fossil fuel combustion decreased
by 4.1 percent; emissions decreased by 20.6 percent (1,184.8 MMT C02 Eq.) from 2005 level peak; and
from 2022 to 2023, these emissions decreased by 3.0 percent (143.4 MMT C02 Eq.).
The five major fuel-consuming economic sectors are transportation, electric power, industrial,
residential, and commercial. Carbon dioxide emissions are produced by the electric power sector as
fossil fuel is consumed to provide electricity to one of the other four economic sectors, or "end-use"
sectors. Greenhouse gas emissions from the commercial, residential, and industrial end-use sectors
ES-8 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
increase substantially when indirect emissions from electric power end-use are distributed, due to the
relatively large share of electricity use by buildings (e.g., heating, ventilation, and air conditioning;
lighting; and appliances) and use of electricity for powering industrial machinery.
Figure ES-6: 2023 End-Use Sector Emissions of Carbon Dioxide from Fossil Fuel
Combustion
2,000
1,500
1,781
I Direct Fossil Fuel Combustion
I Indirect Fossil Fuel Combustion
u
1,000
500
Transportation Industrial
Residential
Commercial U.S. Territories
• Transportation End-Use Sector. Transportation activities accounted for 39.1 percent of U.S. C02
emissions from fossil fuel combustion in 2023, with the largest contributors being light-duty
trucks (39.7 percent), followed by medium- and heavy-duty trucks (23.4 percent) and passenger
vehicles (16.6 percent). The overall trend from 1990 to 2023 shows that total transportation C02
emissions increased due largely to increased demand for travel, which was a result of a
confluence of factors including population growth, economic growth, urban sprawl, and low fuel
prices during the beginning of this period. While an increased demand for travel has led to
generally increasing C02 emissions since 1990, improvements in average new vehicle fuel
economy since 2005 have slowed the rate of increase of C02 emissions.
• Industrial End-Use Sector. Industrial C02 emissions, resulting both directly from the combustion
of fossil fuels7 and indirectly from the generation of electricity that is used by industry,
accounted for 26.3 percent of C02 emissions from fossil fuel combustion in 2023.
Approximately 66.2 percent of these emissions resulted from direct fossil fuel combustion to
produce steam and/or heat for industrial processes. The remaining emissions resulted from the
use of electricity for motors, electric furnaces, ovens, lighting, and other applications. Total
direct and indirect emissions from the industrial sector have declined by 23.4 percent since
1990. This decline is due to structural changes in the U.S. economy (i.e., shifts from a
manufacturing-based to a service-based economy), fuel switching, and efficiency
improvements. From 2022 to 2023, total energy use in the industrial sector decreased by 1.2
percent due largely to a decrease in total coal consumption.
• Residential and Commercial End-Use Sectors. The residential and commercial end-use sectors
accounted for 17.9 and 16.2 percent, respectively, of C02 emissions from fossil fuel combustion
7 This does not include fossil fuels used as feedstocks and reductants, which are reported under IPPU emissions.
Executive Summary ES-9
-------
in 2023 including indirect emissions from electricity. The residential and commercial sectors
relied heavily on electricity for meeting energy demands, with 62.3 and 67.0 percent,
respectively, of their emissions attributable to electricity use for building-related activities such
as lighting, heating, cooling, and operating appliances. The remaining emissions were due to the
consumption of natural gas and petroleum for heating and cooking. Total direct and indirect
emissions from the residential sector have decreased by 12.4 percent since 1990, and total
direct and indirect emissions from the commercial sector have decreased by 3.4 percent since
1990. From 2022 to 2023, a decrease in heating degree days (of 10.4 percent) decreased energy
demand for heating in the residential and commercial sectors; also, a 5.2 percent decrease in
cooling degree days compared to 2022 decreased demand for air conditioning in the residential
and commercial sectors.
• Electric Power Sector. Electricity generators used 29.9 percent of U.S. energy from fossil fuels
and emitted 31.0 percent of the C02 from fossil fuel combustion in 2023. Across the time series,
the type of energy source used to generate electricity, its carbon intensity, and the mix of electric
generation resources used to meet demand, are the main factors influencing emissions.8 Coal-
fired electric generation (in kilowatt-hours [kWh]) decreased from 54.1 percent of generation in
1990 to 16.6 percent in 2023.9 This corresponded with an increase in natural gas generation and
non-fossil fuel renewable energy generation, largely from wind and solar energy. Natural gas
generation (in kWh) represented 10.7 percent of electric power generation in 1990 and
increased over the 34-year period to represent 42.2 percent of electric power generation in
2023. Wind and solar generation (in kWh) represented 0.1 percent of electric power generation
in 1990 and increased over the 34-year period to represent 14.5 percent of electric power
generation in 2023. Between 2022 and 2023, coal electricity generation decreased by 17.9
percent, natural gas generation increased by 8.6 percent, and renewable energy generation
increased by 0.7 percent. While C02 emissions from fossil fuel combustion in the electric power
sector have decreased by 22.3 percent since 1990, the carbon intensity of the electric power
sector, in C02 Eq. per QBtu input, decreased by 31.3 percent during that same timeframe. This
trend is shown in Figure ES-7.
8 In line with IPCC guidelines, CO2 emissions from biomass combustion have been estimated separately from fossil fuel
CO2 emissions and are not included in the electric power sector totals and trends discussed in this section. Net carbon
fluxes from changes in biogenic carbon reservoirs are accounted for in the estimates for the LULUCF sector.
9 Values represent electricity net generation from the electric power sector. See Table 7.2b Electricity Net Generation:
Electric Power Sector of EIA (2024).
ES-10 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Figure ES-7: Electric Power Generation and Emissions
4,500
4,000
2- 3,500
3
c
J 3,000
0
'g 2,500
cD
c
Q)
ai 2,000
1
1,500
ai
1,000
500
Nuclear Generation (Billion kWh)
I Renewable Generation (Billion kWh)
Petroleum Generation (Billion kWh)
I Coal Generation (Billion kWh)
I Natural Gas Generation (Billion kWh)
¦ Total Emissions (MMT CO2 Eq.) [Right Axis]
3,500
3,000
2,500 uj
(N
o
u
2,000
1,500 w
1,000
500
OHrMfotLnvDNco^OHfMmtmvDNos^o
a^cricricrio^o^o^cricricrioooooooooo'-i
HHHrHHr4HHHHf\J(N(NrM(NlN(N(N(NlNNrM(N(N(N(NrMfM(NfN(NlN(NlN
Other significant C02 trends included:
• C02 emissions from natural gas and petroleum systems combined accounted for 1,2percent of
C02 emissions and 1.0 percent of total gross emissions in 2023. These emissions increased by
44.7 percent (18.8 MMT C02 Eq.) from 1990 to 2023. This increase is due primarily to increases
in the production segment, where flaring emissions from associated gas flaring, tanks, and
miscellaneous production flaring have increased over time.
• C02 emissions from iron, steel production and metallurgical coke production accounted for 0.9
percent of C02 and 0.7 percent of total gross emissions. Emissions decreased by 55.9 percent
(58.5 MMT C02 Eq.) from 1990 through 2023. This decrease was primarily due to restructuring of
the industry, technological improvements, and increased scrap steel utilization.
• Total carbon stock change (i.e., net C02 removals) in the LULUCF sector decreased by 8.8
percent between 1990 and 2023. This decrease was primarily due to a decrease in the rate of
net carbon accumulation in forest carbon stocks as well as an increase in emissions from land
converted to settlements. Disturbances on managed lands, particularly wildfire, are among the
major influences that affect the annual net carbon flux by altering the amount of carbon stored
in forest ecosystems.
Methane Emissions
CH4 is significantly more effective than C02 at trapping heat in the atmosphere by a factor of 28 over a
100-year time frame based on the IPCC Fifth Assessment Report estimate (IPCC 2013). Within the
Executive Summary ES-11
-------
United States, the main anthropogenic sources of CH4 include enteric fermentation from domestic
livestock, natural gas systems, landfills, domestic livestock manure management, flooded land, coal
mining, and petroleum systems, as shown in Figure ES-8.
Figure ES-8: 2023 Sources of Methane Emissions
Enteric Fermentation
Natural Gas Systems
Landfills
Manure Management
Flooded Land
Coal Mining
Petroleum Systems
Wastewater Treatment
Rice Cultivation
Other Energy
Other LULUCF
Stationary Combustion
Other Waste
Field Burning of Agricultural Residues
Other Industrial Processes
CFU as a Portion of
All Emissions
187
80 100 120
MMT CO2 Eq.
200
Note: "Other Energy" includes ChU emissions from abandoned oil and gas wells, abandoned underground coal mines,
incineration of waste, and mobile combustion. "Other Waste" includes ChUemissions from anaerobic digestion at biogas
facilities and composting. "Other Industrial Processes" includes ChU emissions from carbide production and consumption,
ferroalloy production, iron and steel production and metallurgical coke production, and petrochemical production. "Other
LULUCF" includes the ChU reported for peatlands remaining peatlands, forest fires, drained organic soils, grassland fires, coastal
wetlands remainingcoastalwetlands, and land converted to coastalwetlands.
Overall, CH4emissions in the United States in 2023, including LULUCF CH4 emissions, accounted for
741.3 MMT C02Eq., representing decreases of 20.1 percent (186.2 MMT C02 Eq.) since 1990 and 2.0
percent (15.1 MMT C02 Eq.) since 2022. Significant trends for the largest sources of anthropogenic CH4
emissions include the following.
• Enteric fermentation emissions increased from 1990 to 2023, largely due to increasing cattle
population. For example, emissions increased from 1990 to 1995 and then generally decreased
from 1996 to 2004, mainly due to fluctuations in beef cattle populations and increased
digestibility of feed for feed lot cattle. Emissions decreased again from 2008 to 2014 as beef
cattle populations again decreased. Emissions increased from 2014 to 2023, consistent with an
increase in beef cattle population over those same years. CH4 emissions from enteric
fermentation decreased by 2.9 percent (5.5 MMTC02 Eq.) from 2022 to 2023, however, largely
driven by a decrease in beef cattle populations.
• Natural gas systems were the second largest anthropogenic source category of CH4 emissions
in the United States in 2023, accounting for 21.9 percent of total CH4 emissions and 2.6 percent
oftotalgross emissions. Emissions have decreased by 26.0 percent (57.1 MMTC02 Eq.) since
1990, largely due to decreases in emissions from distribution, transmission, and storage.
11
CO2
CH4
N20
HFCs, PFCs, SFe and NF3
ES-12 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
• Landfills were the third largest anthropogenic source of CH4 emissions in the United States in
2023, accounting for 17.4 percent of total CH4 emissions and 1.9 percent of total gross
emissions and representing a decrease of 39.6 percent (78.3 MMT C02 Eq.) since 1990, with
small year-to-year increases. This downward trend in emissions coincided with increased
landfill gas collection and control systems, and a reduction of decomposable materials (i.e.,
paper and paperboard, food scraps, and yard trimmings) discarded in MSW landfills over the
time series.10
Nitrous Oxide Emissions
Nitrous oxide (N20) is produced by biological processes that occur in soil and water and by a variety of
anthropogenic activities in the agricultural, energy, industrial, and waste management fields. While total
N20 emissions are much lower than C02 emissions, N20 is 265 times more powerful than C02 at
trapping heat in the atmosphere over a 100-year timeframe (IPCC 2013). The main anthropogenic
activities producing N20 in the United States are agricultural soil management, wastewater treatment,
stationary fuel combustion, manure management, fuel combustion in motor vehicles, and nitric acid
production (see Figure ES-9).
Figure ES-9: 2023 Sources of Nitrous Oxide Emissions
Agricultural Soil Management
Wastewater Treatment
Stationary Combustion
Manure Management
Mobile Combustion
Nitric Acid Production
LULUCF Emissions
Other Industrial Processes
Composting
Adipic Acid Production
Other Energy
Field Burning of Agricultural Residues
0 5 10 15 20 25 30 35 40
MMT CO2 Eq.
Note: "Other Industrial Processes" includes N20 emissions from caprolactam, glyoxal, and glyoxylic acid production; the
electronics industry; and product uses. "Other Energy" includes N20 emissions from petroleum systems, natural gas systems,
and incineration of waste. Land Use, Land-Use Change, and Forestry emissions include N20 emissions reported for peatlands
remaining peatlands, forest fires, drained organic soils, grassland fires, coastalwetlands remaining coastal wetlands, forest
soils, and settlement soils.
10 Carbon dioxide emissions from landfills are not included specifically in summing waste sector totals. Net carbon fluxes
from changes in biogenic carbon reservoirs and decay of disposed wood products are accounted for in the estimates for
LULUCF.
Executive Summary ES-13
-------
Overall, N20 emissions in the United States in 2023, including LULUCF N20 emissions, accounted for
392.9 MMT C02 Eq., representing a decrease of 4.8 percent (19.6 MMT C02 Eq.) since 1990 and a
decrease of 0.9 percent (3.6 MMT C02 Eq.) since 2022. Significant trends for the largest sources of
anthropogenic N20 emissions include the following.
• Agricultural soils were the largest anthropogenic source of N20 emissions in 2023, accounting
for 75.4 percent of N20 emissions and 4.8 percent of total gross greenhouse gas emissions in
the United States. These emissions increased by 2.5 percent (7.2 MMT C02 Eq.) from 1990 to
2023 but fluctuated during that period due to annual variations in weather patterns, fertilizer
use, and crop production.
• Wastewater treatment, both domestic and industrial, was the second largest anthropogenic
source of N20 emissions in 2023, accounting for 5.3 percent of N20 emissions and 0.3 percent
of total gross greenhouse gas emissions in the United States in 2023. Emissions from
wastewater treatment increased by 41.0 percent (6.0 MMT C02 Eq.) since 1990 as a result of
growing U.S. population and protein consumption.
• Stationary combustion was the third largest source of anthropogenic N20 emissions in 2023,
accounting for 5.0 percent of N20 emissions and 0.3 percent of total gross U.S. greenhouse gas
emissions in 2023. Stationary combustion emissions peaked in 2007 and steadily declined until
2020. Emissions increased from 2021 to 2022 but decreased again in 2023. Stationary
combustion emissions have decreased by 12.4 percent (2.8 MMTC02 Eq.) since 1990.
Hydrofluorocarbon, Perfluorocarbon, Sulfur Hexafluoride, and Nitrogen
Trifluoride Emissions
HFCs are synthetic chemicals that are used as alternatives to ozone depleting substances (ODS), which
are being phased out under the Montreal Protocol and Clean Air Act Amendments of 1990. HFCs do not
deplete the stratospheric ozone layer and have been used as alternatives under the Montreal Protocol.
PFCs are emitted from the production of electronics and aluminum and also (in smaller quantities) from
their use as alternatives to ODS. SF6 is emitted from the manufacturing and use of electrical equipment
as well as the production of electronics and magnesium. NF3 is emitted from electronics production.
HFCs are also emitted during production of HCFC-22 and electronics (see Figure ES-10).
HFCs, PFCs, SF6, and NF3 are potent greenhouse gases. In addition to having very high GWPs, PFCs, SF6,
and NF3 have extremely long atmospheric lifetimes, resulting in their essentially irreversible
accumulation in the atmosphere once emitted. SF6 is the most potent greenhouse gas that the IPCC has
evaluated (IPCC 2021).
ES-14 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Figure ES-10: 2023 Sources of Hydrofluorocarbon, Perfluorocarbon, Sulfur
Hexafluoride, and Nitrogen Trifluoride Emissions
Substitution of Ozone Depleting Substances
Electrical Equipment
Fluorochemical Production
Electronics Industry
Magnesium Production and Processing
Other Product Manufacture and Use
Aluminum Production
189
HFCs, PFCs, SFo, and NF3 as a
Portion of All Emissions
3.3%
ICO2
ICH4
I N2O
I HFCs, PFCs, SFe and NFb
0
10 12
MMT CO2 Eq.
14
16
18
20
Some significant trends for the largest sources of U.S. HFC, PFC, SF6, and NF3 emissions include the
following.
• HFC and PFC emissions resulting from their use as substitutes for ODS (e.g.,
chlorofluorocarbons [CFCs]) are the largest share of fluorinated emissions (92.1 percent) in
2023 and have been consistently increasing from small amounts since 1990. This increase over
the time series was largely the result of efforts to phase out CFCs and other ODS in the United
States.
• SF6 emissions from electrical equipment decreased by 79.4 percent (19.6 MMT C02 Eq.) from
1990 to 2023. There are two factors contributing to this decrease: (1) a sharp increase in the
price of SF6 during the 1990s, and (2) a growing awareness of the environmental impact of SF6
emissions through programs such as EPA's SF6 Emission Reduction Partnership for Electric
Power Systems.
• HFC, PFC, SF6, and NF3 emissions from fluorochemical production decreased by 93.4 percent
(66.4 MMT C02 Eq.) from 1990 to 2023 due to a reduction in the HFC-23 emission rate from
HCFC-22 production (kg HFC-23 emitted/kg HCFC-22 produced), the imposition of emissions
controls at production facilities, and a decrease in SF6 production (due to the cessation of
production at the major SF6 production facility in 2010).
• PFC emissions from aluminum production decreased by 97.6 percent (18.8 MMT C02 Eq.) from
1990 to 2023, due to both industry emission reduction efforts and lower domestic aluminum
production.
• HFC, PFC, SF6, and NF3 emissions from use in electronics increased 27.2 percent (0.9 MMT C02
Eq.) from 1990 to 2023. Industrial growth, increasing chip complexity, and the adoption of
emissions reductions technologies contributed to the fluctuation in electronics industry
emissions.
Executive Summary ES-15
-------
ES.3 Overview of Sector Emissions and
Trends
In addition to understanding trends by GHG, this section provides an overview of the trends by inventory
sectors. Figure ES-11 illustrates that over the 34-year period of 1990 to 2023, total emissions from the
Energy and Waste sectors decreased by 6.2 percent (331.6 MMT C02 Eq.) and 29.7 percent (70.1 MMT
C02 Eq.) respectively. Emissions from the IPPU and Agriculture sectors grew by 4.6 percent (16.8 MMT
C02 Eq.), and 8.0 percent (43.9 MMT C02 Eq.), respectively. Over the same period, the overall net flux
from LULUCF (i.e., the net sum of all CH4 and N20 emissions to the atmosphere plus LULUCF net carbon
stock changes in units of MMT C02 Eq.) decreased by 9.4 percent (97.9 MMT C02 Eq.) and resulted in a
removal of 939.9 MMT C02 Eq. in 2023.
Figure ES-11: U.S. Greenhouse Gas Emissions and Sinks by Inventory Sector
¦ LULUCF (emissions) ¦ Agriculture
9,000 ¦ Waste ¦ Energy
¦ Industrial Processes and Product Use ¦ LULUCF (sinks)
8 000 — Emissions (including LULUCF Flux)
7,000 _
S 5,000
o
4,000
3,000
i-HfNj(n^-Lnv£>rssoocr>
cr>
Ot—irvjm'^-LnvDrvcocTio
oooooooooo
(NntuiyDNcoaio
oooooooooooooooooooooooo
Table ES-3: Recent Trends in U.S. Greenhouse Gas Emissions and Sinks by Inventory
Sector (MMT C02 Eq.)
Inventory Sector
1990
2005
2019
2020
2021
2022
2023
Percent
Change
Since
1990
Energy
5,381.9
6,356.2
5,420.9
4,860.2
5,170.1
5,196.2
5,050.4
-6.2%
Industrial Processes and Product Use
368.9
374.7
380.8
375.3
390.9
389.6
385.7
4.6%
Agriculture
551.5
582.5
620.8
600.4
605.8
593.3
595.4
8.0%
ES-16 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Waste
235.9
192.0
174.8
169.7
167.0
165.1
165.8
-29.7%
Total Gross Emissionsa (Sources)
6,538.31
7,505.3
6,597.4
6,005.7
6,333.8
6,344.1
6,197.3
-5.2%
LULUCF Sector Net Totalb
(1,037.9)
(968.9)
(919.4)
(951.6)
(962.9)
(905.3)
(939.9)
-9.4%
Net Emissions (Sources and Sinks)c
5,500.4
6,536.4
5,678.0
5,054.2
5,371.0
5,438.7
5,257.4
-4.4%
a Total emissions without LULUCF.
b The LULUCF sector net totalis the sum of all LULUCF ChUand N20 emissions to the atmosphere plus LULUCF net carbon stock
changes in units of MMT C02 Eq.
c Net emissions with LULUCF.
Notes: Totals may not sum due to independent rounding. Parentheses indicate negative values or sequestration.
Box ES-2: Inventory Reporting Sectors
Inventory reporting sectors are standardized to promote comparability for greenhouse gas inventories
across countries. These sectors include Energy (chapter 3), Industrial Processes and Product Use (IPPU)
(chapter 4), Agriculture (chapter 5), Land Use, Land-Use Change (chapter 6), and Forestry (LULUCF),
and Waste (chapter 7). These categories facilitate consistent reporting and comparison of greenhouse
gas emissions data.
In contrast, the economic sectors used in this report are aligned with more commonly used categories
within the United States, but which may differ from country to country. These sectors include
residential, commercial, industry, transportation, electric power, and agriculture. Emissions from the
electric power industry can be distributed to each economic sector to reflect their use of electricity. This
categorization helps to identify and analyze the sources of greenhouse gas emissions within the U.S.
economy.
Energy
Chapter 3, Energy, contains emissions of all greenhouse gases resulting from stationary and mobile
energy activities including fuel combustion and fugitive fuel emissions and the use of fossil fuels for
non-energy purposes. Energy-related activities, primarily fossil fuel combustion, accounted for the vast
majority of U.S. C02 emissions for the period of 1990 through 2023. Energy-related activities are also
responsible for CH4 and N20 emissions (39.6 percent and 9.3 percent of total U.S. emissions of each
gas, respectively).11 Overall, emission sources in the Energy sector account for a combined 81.5 percent
of total gross U.S. greenhouse gas emissions in 2023. Emissions from energy decreased by 2.8 percent
(145.8 MMTC02 Eq.) since 2022 and 6.2 percent (331.6 MMTC02 Eq.) since 1990.
In 2023, 82.6 percent of the energy used in the United States (in Btus) was produced through the
combustion of fossil fuels. The remaining 17.4 percent came from other energy sources, such as
hydropower, biomass, nuclear, wind, and solar energy (see Figure ES-12).
11 The contribution of energy non-CCb emissions is based on gross totals and excludes LULUCF methane (ChU) and nitrous
oxide (N2O) emissions. The contribution of energy-related CH4 and N2O including LULUCF non-C02 emissions is 36.3
percent and 7.9 percent respectively.
Executive Summary ES-17
-------
Figure ES-12: 2023 U.S. Energy Consumption by Energy Source (Percent)
Nuclear Electric Power
8.6%
Renewable E
Coal
8.7%
8.8%
Industrial Processes and Product Use
Chapter 4, Industrial Processes and Product Use, chapter contains greenhouse gas emissions
generated and emitted as the byproducts of non-energy-related industrial processes, which involve the
chemical or physical transformation of raw materials and can release waste gases such as C02, CH4,
N20, and fluorinated gases (e.g., HFC-23). These processes include iron and steel production and
metallurgical coke production, cement production, petrochemical production, ammonia production,
lime production, other process uses of carbonates (e.g., other uses of carbonates, other uses of soda
ash not associated with glass manufacturing, ceramics production, and non-metallurgical magnesia
production), nitric acid production, adipic acid production, urea consumption for non-agricultural
purposes, aluminum production, HCFC-22 production, other fluorochemical production, glass
production, soda ash production, ferroalloy production, titanium dioxide production, caprolactam
production, zinc production, phosphoric acid production, lead production, and silicon carbide
production and consumption. Most of these industries also emit C02 from fossil fuel combustion which,
in line with sectoral definitions for national inventory reporting, is included in the Energy sector.
Chapter 4 also contains emissions resulting from the release of HFCs, PFCs, SF6, and NF3 and other
human-made compounds used in industrial manufacturing processes and by end-consumers (e.g.,
residential and mobile air conditioning). These industries include electronics manufacturing, electric
power transmission and distribution, and magnesium metal production and processing. In addition,
N20 is used in and emitted by electronics industry and anesthetic and aerosol applications, PFCs and
SF6 are emitted in other product use, and C02 is consumed and emitted through various end-use
applications. In 2023, emissions resulting from the use of ODS substitutes (e.g., HFCs,
chlorofluorocarbons [CFCs]) by end-consumers was the largest source of IPPU emissions and
accounted for 49.0 percent of total IPPU emissions.
Of total U.S. C02, CH4, and N20 emissions, IPPU activities are responsible for 3.4, less than 0.5, and 3.8
percent respectively, as well as for all U.S. emissions of fluorinated gases including HFCs, PFCs, SF6 and
NF3. Overall, emission sources from the IPPU sector accounted for 6.2 percent of U.S. greenhouse gas
emissions in 2023. Between 1990 and 2023, IPPU emissions increased by 4.6 percent (16.8 MMTC02
ES-18 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Eq.), primarily due to growth in the use of HFCs as substitutes for ozone depleting substances. IPPU
emissions have decreased by 1.0 percent (3.8 MMT C02 Eq.) since 2022 largely due to decreased
production of fluorochemicals and decreased consumption of other process uses of carbonates.
Agriculture
Chapter 5, Agriculture, contains information on anthropogenic emissions from agricultural activities
(except fuel combustion, which is addressed in Chapter 3, Energy, and some agricultural C02, CH4, and
N20 fluxes, which are addressed in Chapter 6, Land Use, Land-Use Change, and Forestry).
Several agricultural activities contribute directly to emissions of greenhouse gases, including
agricultural soil management, enteric fermentation from domestic livestock production, livestock
manure management, rice cultivation, urea fertilization, liming, and field burning of agricultural
residues.
In 2023, agricultural activities were responsible for 9.6 percent of total gross U.S. greenhouse gas
emissions. Agriculture sector emissions increased by 2.1 MMT C02 Eq. (0.4 percent) since 2022 and
have increased by 43.9 MMT C02 Eq. (8.0 percent) since 1990, mostly from trends in enteric
fermentation and manure management. CH4, N20, and C02 are the greenhouse gases emitted by
agricultural activities. CH4 emissions from enteric fermentation and manure management represented
25.2 percent of total CH4 emissions from anthropogenic activities in 2023. Agricultural soil management
activities, such as application of synthetic and organic fertilizers, deposition of livestock manure, and
growing N-fixing plants, were the largest contributor to U.S. N20 emissions in 2023, accounting for 76.6
percent of total N20 emissions. C02 emissions from the application of crushed limestone and dolomite
(i.e., soil liming) and urea fertilization represented 0.2 percent of total C02 emissions from
anthropogenic activities.
Land Use, Land-Use Change, and Forestry
Chapter 6, Land Use, Land-Use Change, and Forestry, contains emissions and removals of C02 and
emissions of CH4 and N20 from managed lands in the United States. Consistent with the 2006IPCC
Guidelines, emissions and removals from managed lands are considered to be anthropogenic, while
emissions and removals from unmanaged lands are considered to be natural and therefore not included
in this report.12 The share of managed land in the United States is approximately 95 percent of total land
included in the Inventory,13 More information on the definition of managed land used in the Inventory is
provided in Chapter 6.
Overall, the Inventory results show that managed land is a net sink for C02 (i.e., carbon sequestration).
The primary drivers of fluxes on managed lands include forest management practices, tree planting in
urban areas, the management of agricultural soils, lands remaining and lands converted to reservoirs
and other constructed waterbodies, landfilling of yard trimmings and food scraps, and activities that
cause changes in carbon stocks in coastal wetlands. The main drivers for forest carbon sequestration
12 See http://www.ipcc-nggip.iges.or.ip/pub[ic/2006g[/pdf/4 Vo[ume4/V4 01 Ch1 lntroduction.pdf.
13 The current [and representation does not include [and in U.S. Territories, but there are planned improvements to include
these regions in future Inventories. U.S. Territories represent approximately 0.1 percent of the total land base for the
United States. See Box 6-2 in Chapter 6 of this report.
Executive Summary ES-19
-------
include forest growth and increasing forest area (i.e., afforestation), as well as a net accumulation of
carbon stocks in harvested wood pools. The net sequestration in settlements remaining settlements,
which occurs predominantly from urban forests (i.e., settlement trees) and landfilled yard trimmings and
food scraps, is a result of net tree growth and increased urban forest area, as well as long-term
accumulation of carbon from yard trimmings and food scraps in landfills.
The LULUCF sector in 2023 resulted in a net increase in carbon stocks (i.e., net C02 removals) of 1,000.5
C02 Eq.14 The removals of carbon offset 16.1 percent of total gross greenhouse gas emissions in 2023.
Emissions of CH4 and N20 from LULUCF activities in 2023 represented 1.2 percent of net greenhouse
gas emissions.15 Carbon dioxide removals from carbon stock changes are presented in Table ES-4 along
with CH4 and N20 emissions for LULUCF source categories.
Table ES-4: U.S. Greenhouse Gas Emissions and Removals (Net Flux) from Land Use,
Land-Use Change, and Forestry (MMT C02 Eq.)
Land-Use Category
1990
2005
2019
2020
2021
2022
2023
Forest Land Remaining Forest Land a
(1049.3)
(932.8)
(867.4)
(898.0)
(881.0)
(827.6)
(873.3)
Settlements Remaining Settlementsb
(109.1)
(115.2)
(131.4)
(131.7)
(132.1)
(132.1)
(131.7)
Land Converted to Forest Land 0
(103.6)
(103.6)
(103.9)
(103.8)
(103.8)
(103.8)
(103.8)
Cropland Remaining Cropland
1.0
(31.0)
(19.3)
(8.7)
(31.9)
(31.6)
(30.5)
Land Converted to Wetlandsd
6.8
1.9
0.7
0.7
0.7
0.7
0.6
Land Converted to Grasslande
35.6
21.9
20.9
24.1
19.9
20.9
20.9
Grassland Remaining Grasslandf
24.2
24.5
28.5
16.8
11.2
13.7
22.7
Land Converted to Cropland e
48.5
35.5
31.4
29.2
34.9
35.0
35.6
Wetlands Remaining Wetlandsd
38.5
40.9
39.7
39.7
39.7
39.7
39.7
Land Converted to Settlementse
69.5
89.0
81.4
80.3
79.7
79.8
79.8
LULUCF Carbon Stock Change g
(1,096.9)
(1,040.7)
(982.6)
(1,034.2)
(1,043.8)
(973.9)
(1,000.5)
LULUCF Emissions h
59.1
71.8
63.2
82.6
81.0
68.6
60.6
cm
54.4
60.9
56.1
69.0
67.8
59.6
54.7
N2O
4.7
10.9
7.0
13.7
13.1
9.0
5.9
LULUCF Sector Net Total1
(1,037.9)
(968.9)
(919.4)
(951.6)
(962.9)
(905.3)
(939.9)
a Includes the net changes to carbon stocks stored in all forest ecosystem pools and harvested wood products, emissions from
fires on both forest land remaining forest land and land converted to forest land, emissions from N fertilizer additions on both
forest land remainingforest land and land converted to forest land, and ChUand N20 emissions from drained organic soils on
both forest land remainingforest land and land converted to forest land.
b Estimates include N20 emissions from N fertilizer additions on both settlements remaining settlements and land converted to
settlements because it is not possible to separate the activity data at this time.
c Includes the net changes to carbon stocks stored in all forest ecosystem pools.
d Estimates include ChU emissions from flooded land remaining flooded land and land converted to flooded land.
14 LULUCF carbon stock change is the net C stock change from the following categories: forest land remainingforest land,
land converted to forest land, cropland remaining cropland, land converted to cropland, grassland remaining grassland,
land converted to grassland, wetlands remaining wetlands, land converted to wetlands, settlements remaining
settlements, and land converted to settlements.
15 LULUCF emissions include the CFU and N2O emissions reported for peatlands remaining peatlands, forest fires, drained
organic soils, grassland fires, and coastal wetlands remaining coastal wetlands; CH4 emissions from land converted to
coastal wetlands; and N2O emissions from forest soils and settlement soils.
ES-20 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
e Includes changes in mineral and organic soil carbon stocks for all land use conversions to cropland, grassland, and
settlements, respectively. Also includes aboveground/belowground biomass, dead wood, and litter carbon stock changes for
conversion of forest land to cropland, grassland, and settlements, respectively.
' Estimates include ChUand N20 emissions from fires on both grassland remaining grassland and land converted to grassland.
g LULUCF carbon stock change includes any carbon stock gains and losses from all land use and land use conversion categories.
h LULUCF emissions subtotal includes the ChU and N20 emissions reported for peatlands remaining peatlands, forest fires,
drained organic soils, grassland fires, and coastal wetlands remaining coastal wetlands; ChU emissions from land converted to
coastal wetlands, flooded land remaining flooded land, and land converted to flooded land; and N20 emissions from forest soils
and settlement soils. Emissions values are included in land-use category rows.
The LULUCF sector net total is the net sum of all LULUCF ChU and N20 emissions to the atmosphere plus LULUCF net carbon
stock changes in units of MMT C02 Eq.
Notes: Totals may not sum due to independent rounding.
Between 1990 and 2023, total carbon sequestration in the LULUCF sector decreased by 8.8 percent,
primarily due to a decrease in the rate of net carbon accumulation in forests and in cropland remaining
cropland, as well as an increase in C02 emissions from land converted to settlements. The overall net
flux from LULUCF (i.e., net sum of all CH4 and N20 emissions to the atmosphere plus LULUCF net
carbon stock changes in units of MMT C02 Eq.) resulted in a removal of 939.9 MMT C02 Eq. in 2023.
Flooded lands were the largest source of CH4 emissions from the LULUCF sector and the fifth largest
source of overall net CH4 emissions in 2023. Coastalwetlands remaining coastal wetlands were the
second largest source of CH4 emissions, followed by forest fires. Settlement soils were the largest
source of N20 emissions from the LULUCF sector in 2023.
Waste
Chapter 7, Waste, contains emissions from waste management activities (except the incineration of
waste, which is addressed in Chapter 3, Energy). Landfills were the largest source of anthropogenic
greenhouse gas emissions from waste management activities, accounting for 72.0 percent of total
greenhouse gas emissions from waste management activities, and 17.4 percent of total U.S. CH4
emissions.16Additionally, wastewater treatment accounted for 25.3 percent of total Waste sector
greenhouse gas emissions, 3.1 percent of U.S. CH4 emissions, and 5.4 percent of U.S. N20 emissions in
2023. Emissions of CH4 and N20 from commercial composting are also included in this chapter,
accounting for 1.6 percent (2.6 MMT C02 Eq.) and 1.1 percent (1.8 MMT C02 Eq.) of overall waste sector
emissions, respectively. Anaerobic digestion at biogas facilities generated CH4 emissions, accounting
for less than 0.05 percent of emissions from the Waste sector. Overall, emission sources in Chapter 7,
Waste, accounted for 2.7 percent of total gross U.S. greenhouse gas emissions in 2023. Waste sector
emissions decreased by 0.5 percent (0.7 MMT C02 Eq.) since 2022 and by 29.7 percent (70.1 MMT C02
Eq.) since 1990.
16 Landfills also store carbon, due to incomplete degradation of organic materials such as harvest wood products, yard
trimmings, and food scraps, as described in Chapter 6, Land Use, Land-Use Change, and Forestry. Also, the estimated
total methane emissions used to estimate contribution excludes methane emissions from the LULUCF sector.
Executive Summary ES-21
-------
ES.4 Other Information
Emissions and Sinks by Economic Sector
This report also characterizes emissions according to commonly used economic sector categories:
residential, commercial, industry, transportation, electric power, and agriculture.17 All emissions from
U.S. Territories are reported together as their own end-use sector in this characterization due to a lack of
specific consumption data for the individual end-use sectors. For more information on trends in the
Land Use, Land-Use Change, and Forestry sector, see Section 6.
Figure ES-13 shows the trend in emissions by economic sector from 1990 to 2023, and Table ES-
5summarizes emissions from each of these economic sectors.
Figure ES-13: U.S. Greenhouse Gas Emissions Allocated to Economic Sectors
Note: This figure excludes emissions and removals from Land Use, Land-Use Change, and Forestry and U.S. Territories.
Table ES-5: U.S. Greenhouse Gas Emissions Allocated to Economic Sectors (MMT C02
Eq.)
Economic Sectors
1990
2005
2019
2020
2021
2022
2023
Percent
Change
Since 1990
Transportation
1,520.8
1,971.8
1,874.2
1,625.3
1,805.5
1,804.0
1,822.5
19.8%
Electric Power
1,880.2
2,457.4
1,650.7
1,481.8
1,584.0
1,575.5
1,453.7
-22.7%
Industry
1,714.5
1,589.4
1,514.8
1,412.3
1,446.0
1,439.8
1,423.0
-17.0%
Agriculture
606.8
633.7
679.2
663.3
655.7
639.8
649.6
7.1%
17 The agriculture economic sector includes emissions from fossil fuel combustion and electricity use within the Agriculture
sector.
ES-22 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Commercial 447.0 422.1 469.2 442.3 448.6 469.0 455.1 1.8%
Residential 345.6 371.2 384.2 358.0 369.6 392.4 368.3 6.6%
U.S. Territories 23.4 59.7 25.1 22.6 24.4 23.7 25.1 6.9%
Total Gross Emissions (Sources) 6,538.3 7,505.3 6,597.4 6,005.7 6,333.8 6,344.1 6,197.3 -5.2%
LULUCFSector Net Total3 (1,037.9) (968.9) (919.4) (951.6) (962.9) (905.3) (939.9) -9.4%
Net Emissions (Sources and Sinks) 5,500.4 6,536.4 5,678.0 5,054.2 5,371.0 5,438.7 5,257.4 -4.4%
aThe Land Use, Land-Use Change, and Forestry (LULUCF) sector net total is the net sum of all LULUCF ChUand N20 emissions to
the atmosphere plus LULUCF net carbon stock changes.
Notes: Total gross emissions are presented without LULUCF. Total net emissions are presented with LULUCF. Totals may not sum
due to independent rounding. Parentheses indicate negative values or sequestration.
Using this categorization, emissions from transportation activities accounted for the largest portion
(29.4 percent) of total gross greenhouse gas emissions in 2023. Electric power accounted for the second
largest portion (23.5 percent) of greenhouse gas emissions in 2023, while emissions from industry
accounted for the third largest portion (23.0 percent). Emissions from industry have in general declined
over the past decade, due to a number of factors, including structural changes in the U.S. economy (i.e.,
shifts from a manufacturing-based to a service-based economy), fuel switching, and energy efficiency
improvements.
The remaining 24.2 percent of total gross greenhouse gas emissions were contributed by, in order of
magnitude, the agriculture, commercial, and residential sectors, plus emissions from U.S. Territories.
Activities related to agriculture accounted for 10.5 percent of emissions; unlike other economic sectors,
agricultural sector emissions were dominated by N20 emissions from agricultural soil management and
CH4 emissions from enteric fermentation. An increasing amount of carbon is stored in agricultural soils
each year, but this C02 sequestration is assigned to the LULUCF sector rather than the agriculture
economic sector. The commercial and residential sectors accounted for 7.3 percent and 5.9 percent of
emissions, respectively, and U.S. Territories accounted for 0.4 percent of emissions; emissions from
these sectors primarily consisted of C02 emissions from fossil fuel combustion. Carbon dioxide was
also emitted and sequestered by a variety of activities related to forest management practices, tree
planting in urban areas, the management of agricultural soils, landfilling of yard trimmings, and changes
in carbon stocks in coastalwetlands.
Electric power is ultimately used in the other economic sectors. Table ES-6 presents greenhouse gas
emissions from economic sectors with emissions related to electric power distributed into end-use
categories (i.e., emissions from electric power generation are allocated to the economic sectors in
which the electricity is used). To distribute electricity emissions among end-use sectors, emissions from
the source categories assigned to electric power were allocated to the residential, commercial,
industry, transportation, and agriculture economic sectors according to retail sales of electricity for
each end-use sector (EIA 2025).18 These source categories include C02from fossil fuel combustion and
the use of limestone and dolomite for flue gas desulfurization, C02 and N20 from incineration of waste,
CH4 and N20 from stationary sources, and SF6 from electrical equipment systems.
18 U.S. Territories consumption data that are obtained from EIA are only available at the aggregate level and cannot be
broken out by end-use sector. The distribution of emissions to each end-use sector for the 50 states does not apply to
Territories data.
Executive Summary ES-23
-------
When emissions from electric power use are distributed among these end-use sectors, transportation
and industrial account for the largest shares of U.S. greenhouse gas emissions (29.5 percent and 29.2
percent, respectively) in 2023. The commercial and residential sectors contributed the next largest
shares of total gross greenhouse gas emissions in 2023 (15.6 and 14.4 percent, respectively). Emissions
from the commercial and residential sectors increase substantially when emissions from electric power
use are included due to the relatively large share of electricity these sectors use for energy (e.g., lighting,
cooling, appliances). Table ES-6 shows the trends in these emissions by sector from 1990 to 2023.
Table ES-6: U.S. Greenhouse Gas Emissions with Electricity-Related Emissions
Distributed by Economic Sector (MMT C02 Eq.)
Economic Sectors
1990
2005
2019
2020
2021
2022
2023
Percent
Change
Since
1990
Transportation
1,523.9
1,976.6
1,878.5
1,628.9
1,809.5
1,808.6
1,827.7
19.9%
Industry
2,388.5
2,305.0
1,957.8
1,799.7
1,868.0
1,859.6
1,806.9
-24.4%
Commercial
1,002.5
1,244.3
1,037.1
936.9
981.8
1,007.5
965.1
-3.7%
Residential
957.9
1,247.7
984.4
919.9
958.7
975.0
891.1
-7.0%
Agriculture
641.9
672.0
714.4
697.7
691.4
669.6
681.5
6.2%
U.S. Territories
23.4
59.7
25.1
22.6
24.4
23.7
25.1
6.9%
Total Gross Emissions (Sources)
6,538.3
7,505.3
6,597.4
6,005.7
6,333.8
6,344.1
6,197.3
-5.2%
LULUCF Sector Net Totala
(1,037.9)
(968.9)
(919.4)
(951.6)
(962.9)
(905.3)
(939.9)
-9.4%
Net Emissions (Sources and Sinks)
5,500.4
6,536.4
5,678.0
5,054.2
5,371.0
5,438.7
5,257.4
-4.4%
aThe Land Use and Land-Use Change and Forestry (LULUCF) sector net totalis the net sum of all LULUCF ChUand N20
emissions to the atmosphere plus LULUCF net carbon stock changes.
Notes: Emissions from electric power are allocated based on aggregate electricity use in each end-use sector. Totals may not
sum due to independent rounding. Parentheses indicate negative values or sequestration.
ES-24 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Figure ES-14: U.S. Greenhouse Gas Emissions with Electricity-Related Emissions
Distributed to Economic Sectors
Note: This figure excludes emissions and removals from Land Use, Land-Use Change, and Forestry and U.S. Territories.
Box ES-3: Trends in Various U.S. Greenhouse Gas Emissions-Related Data
Total gross greenhouse gas emissions can be compared to other economic and social indices to
highlight changes over time. These comparisons include: (1) aggregate energy use, because energy-
related activities are the largest sources of emissions; (2) energy use per capita as a measure of
efficiency; (3) emissions per unit of total gross domestic product as a measure of national economic
activity; and (4) emissions per capita.
Table ES-7 provides data on various statistics related to U.S. greenhouse gas emissions normalized to
1990 as a baseline year. These values represent the relative change in each statistic since 1990.
Greenhouse gas emissions in the United States have declined at an average annual rate of 0.01 percent
since 1990, although changes from year to year have been significantly larger. This growth rate is slightly
slower than that for total energy use and fossil fuel consumption, and overall gross domestic product
(GDP), and national population (see Figure ES-15). The direction of these trends started to change after
2005, when greenhouse gas emissions, total energy use, and fossil fuel consumption began to peak.
Greenhouse gas emissions in the United States have decreased at an average annual rate of 1.0 percent
since 2005. Since 2005, GDP and national population, generally, continued to increase while energy has
decreased slightly noting 2020 was impacted by the COVID-19 pandemic.
Executive Summary ES-25
-------
Table ES-7: Recent Trends in Various U.S. Data (Index 1990 = 100)
Variable
1990
2005
2019
2020
2021
2022
2023
Avg. Annual
Growth Rate
Since 1990a
Avg. Annual
Growth Rate
Since 2005a
Greenhouse Gas Emissionsb
100
115
101
92
97
97
95
-0.1%
-1.0%
Energy Use0
100
119
117
107
113
115
113
0.4%
-0.2%
GDPd
100
159
206
202
214
219
225
2.5%
2.0%
Populatione
100
118
131
132
132
133
135
0.9%
0.8%
+ Absolute value does not exceed 0.05 percent.
a Average annual growth rate.
b Total gross GWP-weighted values.
c Energy content-weighted values (EIA 2025).
d GDP in chained 2017 dollars (BEA 2024).
e U.S. Census Bureau (2025).
Figure ES-15: U.S. Greenhouse Gas Emissions Per Capita and Per Dollar of Gross
Domestic Product (GDP)
Source: BEA (2024), U.S. Census Bureau (2025), and net emissions in this report.
Key Categories
Key categories are "inventory categories which individually, or as a group of categories (for which a
common method, emission factor and activity data are applied) are prioritized within the national
inventory system because their estimates have a significant influence on a country's total inventory of
greenhouse gases in terms of the absolute level, the trend, or the level of uncertainty in emissions or
removals" (IPCC 2006; IPCC 2019). A key category analysis identifies priority source or sink categories for
focusing efforts to improve overall Inventory quality. In addition, a qualitative review of key categories
and non-key categories can also help identify additional source and sink categories to consider for
improvement efforts, including reducing uncertainty.
ES-26 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Figure ES-16 presents the 2023 key categories identified by the Approach 1 level assessment, including
the LULUCF sector. A level assessment using Approach 1 identifies all source and sink categories that
cumulatively account for 95 percent of total (i.e., gross) emissions in a given year when assessed in
descending order of absolute magnitude.
Figure ES-16: 2023 Key Categories (Approach 1 including LULUCF)
Mobile Combustion: Road - CO2
Net Forest Land Remaining Forest Land: Forest C Pools - CO2
Stationary Combustion - Gas - Electricity Generation - CO2
Stationary Combustion - Coal - Electricity Generation - CO2
Stationary Combustion - Gas - Industrial - CO2
Direct Agricultural Soil Management - N2O
Stationary Combustion - Gas - Residential - CO2
Stationary Combustion - Oil - Industrial - CO2
Stationary Combustion - Gas - Commercial - CO2
Enteric Fermentation: Cattle - CFU
Mobile Combustion: Aviation - CO2
Natural Gas Systems - ChU
Emissions from Substitutes for Ozone Depleting Substances: Refrigeration and Air Conditioning - Several
Net Settlements Remaining Settlements: C Stocks in Trees and Soils - CO2
Non-Energy Use of Fuels - CO2
Net Land Converted to Forest Land - CO2
MSW Landfills - CH4
Net Forest Land Remaining Forest Land: HWP - CO2
Net Land Converted to Settlements - CO2
Mobile Combustion: Other - CO2
Stationary Combustion - Oil - Commercial - CO2
Stationary Combustion - Oil - Residential - CO2
Iron and Steel Production & Metallurgical Coke Production - CO2
Flooded Lands Remaining Flooded Lands - QU
Fugitive Emissions from Coal Mining - CH*
Cement Production - CO2
Petroleum Systems - CH4
Natural Gas Systems - CO2
Manure Management: Cattle - ChU
Stationary Combustion - Coal - Industrial - CO2
Net Land Converted to Cropland - CO2
Mobile Combustion: Marine - CO2
Mobile Combustion: Railways - CO2
Petrochemical Production - CO2
Net Cropland Remaining Cropland - CO2
Indirect Applied Nitrogen - N2O
Manure Management: Swine - CFU
Petroleum Systems - CO2
Net Grassland Remaining Grassland - CO2
Net Land Converted to Grassland - CO2
Domestic Wastewater Treatment - N2O
0 200 400 600 800 1,000 1,200 1,400
2023 Emissions (MMT CO2 Eq.)
Note: For a complete list of key categories and detailed discussion of the underlying key category analysis, see Annex 1. Bars
indicate key categories identified using Approach 1 level assessment, including the Land Use, Land-Use Change, and Forestry
(LULUCF) sector. The absolute values of net C02 emissions from LULUCF are presented in this figure but reported separately
from gross emissions totals. Refer to Table ES-4 for a breakout of emissions and removals for LULUCF by source/sink category.
For a complete list of key categories and more information regarding the overall key category analysis,
including approaches accounting for uncertainty and the influence of trends of individual source and
sink categories, see Chapter 1, Introduction, Section 1.5, and Annex 1.
Quality Assurance and Quality Control
The EPA seeks continuous improvements to the quality, transparency, ana usability of the Inventory. To
assist in these efforts, the United States implemented a U.S. Inventory QA/QC plan, which includes
Executive Summary ES-27
-------
expert and public reviews for the emissions estimates and this report, along with other verification
techniques, such as use of ambient measurements as described in Box ES-4.
Box ES-4: Use of Ambient Measurements Systems for Validation of Emission
Inventories
Several recent studies have estimated emissions at the national or regional level with estimated results
that sometimes differ from the results in this report and in previous reports. EPA has engaged with
researchers on how remote sensing, ambient measurement, and inverse modeling techniques for
estimating greenhouse gas emissions could assist in improving the understanding of inventory
estimates. In working with the research community to improve national greenhouse gas inventories, EPA
follows guidance from the IPCC on the use of measurements and modeling to validate emission
inventories.19 An area of particular interest in EPA's outreach efforts is how ambient measurement data
can be used to assess estimates or potentially be incorporated into the Inventory in a manner consistent
with this Inventory's transparency in calculation methodologies, and the ability of inverse modeling
techniques to attribute emissions and removals from remote sensing to anthropogenic sources, as
defined by the IPCC, versus natural sources and sinks.
The 2079 Refinement (IPCC 2019) Volume 1 General Guidance and Reporting, Chapter 6: Quality
Assurance, Quality Control and Verification, notes that emission estimates derived from atmospheric
concentration measurements can provide independent data sets as a basis for comparison with
inventory estimates. The 2019 Refinement provides guidance on conducting such comparisons (as
summarized in Table 6.2 of IPCC [2019] Volume 1, Chapter 6) and provides guidance on using such
comparisons to identify areas of improvement in national inventories (as summarized in Box 6.5 of IPCC
[2019] Volume 1, Chapter 6). Further, it identified fluorinated gases as particularly suitable for such
comparisons due to their limited natural sources, generally long atmospheric lifetimes, and well-
understood loss mechanisms, which makes it relatively more straightforward to model their emission
fluxes from observed mass quantities. Unlike emissions of C02, CH4, and N20, emissions of fluorinated
greenhouse gases are almost exclusively anthropogenic, meaning that the fluorinated greenhouse gas
emission sources included in this Inventory account for the majority of the total U.S. emissions of these
gases detectable in the atmosphere. This evaluation approach is also useful for gases and sources with
larger uncertainties in available bottom-up inventory methods and data, such as emissions of CH4,
which are primarily from uncertain biological (e.g., enteric fermentation) and fugitive (e.g., natural gas
production) activities.
In this Inventory, EPA includes the results from current and previous comparisons between fluorinated
gas emissions inferred from atmospheric measurements and fluorinated gas emissions estimated
based on bottom-up measurements and modeling. These comparisons, performed for HFCs and SF6
respectively, are described under the QA/QC and Verification discussions in Chapter 4, Section 4.25,
Substitution of Ozone Depleting Substances, and Section 4.26, Electrical Equipment.
Consistent with the 2019 Refinement, a key element to facilitate such comparisons is a spatially explicit
or gridded inventory as an input to inverse modeling. To improve the ability to compare methane
emissions from the national-level greenhouse gas inventory with observation-based estimates, a team
of researchers from U.S. EPA, SRON Netherlands Institute for Space Research, Harvard University, and
19 See http://www.ipcc-nggip.iges.or.ip/meeting/pdfiles/1003 Uncertaintv%20meeting report.pdf.
ES-28 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Lawrence Berkely National Laboratory and other coauthors developed a time series of anthropogenic
methane emissions maps with 0.1 °x 0.1° (10 kilometer (km) x 10 km) spatial resolution and monthly
temporal resolution for the contiguous United States.20 The gridded methane inventory is designed to be
consistent with the U.S. EPA's Inventory of U.S. Greenhouse Gas Emissions and Sinks estimates, which
presents national totals for different source types.21 The development of this gridded inventory is
consistent with the recommendations contained in two National Academies of Science reports
examining greenhouse gas emissions data (National Research Council 2010; National Academies of
Sciences, Engineering, and Medicine 2018).
Finally, in addition to the use of atmospheric concentration measurement data for comparison with
Inventory data, information from top-down studies is directly incorporated in the natural gas systems
calculations to quantify emissions from certain well blowout events.
Uncertainty Analysis of Emission and Sink Estimates
The 2006IPCC Guidelines (IPCC 2006), Volume 1, Chapter 3, describe the benefits of conducting an
uncertainty analysis, which include informing and prioritizing inventory improvements. This report
provides single estimates of uncertainty for all source and sink categories, and qualitative discussion of
specific factors affecting the uncertainty estimate. Some of the current estimates, such as those for C02
emissions from energy-related combustion activities, are considered to have low uncertainties. This is
because the amount of C02 emitted from energy-related combustion activities is directly related to the
amount of fuel consumed, the fraction of the fuel that is oxidized, and the carbon content of the fuel,
and for the United States, the uncertainties associated with estimating those factors are relatively small.
For some other categories of emissions and sinks, however, inherent variability or a lack of data
increases the uncertainty or systematic error associated with the estimates presented. Finally, an
analysis is conducted to assess uncertainties associated with the overall emissions, sinks, and trends
estimates. The overall uncertainty surrounding total net greenhouse gas emissions is estimated to be -6
to +5 percent in 1990 and -5 to +6 percent in 2023. When the LULUCF sector is excluded from the
analysis, the uncertainty is estimated to be -2 to +4 percent in 1990 and -2 to +4 percent in 2023.
20 See https://www.epa.gov/ghgemissions/us-gridded-methane-emissions.
21 See https://www.epa.gov/ghgemissions/inventorv-us-greenhouse-gas-emissions-and-sinks.
Executive Summary ES-29
-------
1 Introduction
This Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023 (Inventory) identifies and
quantifies the anthropogenic1 emissions sources and removals (by sinks) of greenhouse gases in
the U.S. It is an essential tool for understanding the relative magnitude of different sources and
sinks across the U.S., as well as changes in these magnitudes over time. This chapter provides an
overview of the annual Inventory compilation and review processes that contribute to the
Inventory's transparency, accuracy, consistency, and, including institutional arrangements,
methods, data sources, key categories, uncertainty, QA/QC, and improvement planning. Chapters
2 through 7 of this report are organized by the following inventory reporting sectors and their
respective source and sink categories: Energy; Industrial Processes and Product Use (IPPU);
Agriculture; Land Use, Land-Use Change, and Forestry (LULUCF); and Waste.
Box 1-1: Relationship to EPA's Greenhouse Gas Reporting Program
EPA also collects greenhouse gas emissions data from individual facilities and suppliers of certain fossil
fuels and industrial gases through its Greenhouse Gas Reporting Program (GHGRP).2 The GHGRP
applies to direct greenhouse gas emitters, fossil fuel suppliers, industrial greenhouse gas suppliers, and
facilities that inject carbon dioxide (C02) underground for sequestration or other reasons and requires
reporting by over 8,000 sources or suppliers in 46 industrial categories.3 Annual reporting is at the facility
level, except for certain suppliers of fossil fuels and industrial greenhouse gases. In general, the
threshold for reporting is 25,000 metric tons or more of C02 Eq. per year. Facilities in most source
categories4 subject to GHGRP began reporting for the 2010 reporting year while additional types of
industrial operations began reporting for the 2011 reporting year. Methodologies used in EPA's GHGRP
are consistent with the 2006IPCC Guidelines. While the GHGRP does not provide full coverage of total
annual U.S. greenhouse gas emissions and removals (e.g., the GHGRP excludes emissions from the
Agriculture and Land Use, Land-Use Change, and Forestry sectors), it is an important input to the
calculations of national-level emissions in this Inventory.
Data presented in this Inventory report and EPA's GHGRP are complementary. The GHGRP dataset
continues to be an important resource for the Inventory, providing not only annual emissions
1 The term "anthropogenic," in this context, refers to greenhouse gas emissions and removals that are a direct result of
human activities or are the result of natural processes that have been affected by human activities (IPCC 2006).
2 On October 30, 2009, EPA promulgated a rule requiring annual reporting of greenhouse gas data from large greenhouse
gas emissions sources in the United States. Implementation of the rule, codified at 40 CFR Part 98, is referred to as EPA's
Greenhouse Gas Reporting Program (GHGRP).
3 See http://www.epa.gov/ghgreporting and http://ghgdata.epa.gov/ghgp/main.do.
4 See https://www.ccdsupport.com/confluence/pages/viewpage.action?pageld=322699300
Introduction 1-1
-------
information, but also other annual information such as activity data and emission factors that can
improve and refine national emission estimates and trends overtime. GHGRPdata also allows EPA to
disaggregate national inventory estimates in new ways that can highlight differences across regions and
sub-categories of emissions, along with enhancing the application of QA/QC procedures and
assessment of uncertainties. EPA uses annual GHGRP data in several categories to improve the national
estimates presented in this Inventory, consistent with IPCC methodological guidance. See Annex 9 for
more information on specific uses of GHGRP data in the Inventory (e.g., natural gas systems).
Additionally, this Inventory presents the calculated emissions and removals in a given year for the
United States in a common manner that is in line with the inventory reporting guidelines for the
reporting of inventories under the UNFCCC. Use of consistent methods by countries to calculate
emissions and removals for their inventories helps to ensure that these reports are comparable.
The presentation of emissions and removals provided in this Inventory does not preclude
alternative examinations (e.g., economic sectors).
This report applies methods from the 2006 IPCC Guidelines for National Greenhouse Gas
Inventories (2006 IPCC Guidelines), the 2013 Supplement to the 2006 IPCC Guidelines for National
Greenhouse Gas Inventories: Wetlands (2013 Supplement), and the 2019 Refinement to the 2006
IPCC Guidelines for National Greenhouse Gas Inventories (2019 Refinement). The recently released
2079 Refinement clarifies and elaborates on the existing guidance in the 2006 IPCC Guidelines and
provides updates to default values of emission factors and other parameters based on updated
science. This report applies both the 2013 Supplement and updated guidance in the 2019
Refinement to improve accuracy and completeness of the Inventory. For more information on
specific uses, see Section 1.5, Methodology and Data Sources.
1.1 Greenhouse Gases
Although the Earth's atmosphere consists mainly of oxygen and nitrogen, neither plays a significant
role in enhancing the greenhouse effect because both are essentially transparent to terrestrial
radiation. The greenhouse effect is primarily a function of the concentration of water vapor, carbon
dioxide (C02), methane (CH4), nitrous oxide (N20), and other trace gases in the atmosphere that
absorb the terrestrial radiation leaving the surface of the Earth (IPCC 2021).
Naturally occurring greenhouse gases include water vapor, C02, CH4, N20, and ozone (03). Other
greenhouse gases—such as certain classes of halogenated substances that contain fluorine,
chlorine, or bromine (i.e., halons)—are, for the most part, solely a product of industrial activities.
Chlorofluorocarbons (CFCs) and hydrochlorofluorocarbons (HCFCs) are halocarbons that contain
chlorine, while halocarbons that contain bromine are referred to as bromofluorocarbons.
Stratospheric ozone depleting substances, such as CFCs, HCFCs, and halons, are covered under
the Montreal Protocol on Substances that Deplete the Ozone Layer. Countries typically do not
include these gases in national greenhouse gas inventories.5 Some other fluorine-containing
5 Emissions estimates of CFCs, HCFCs, halons and other ozone-depleting substances are included in this document for
informational purposes.
1-2 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
halogenated substances—hydrofluorocarbons (HFCs), perfluorocarbons (PFCs), sulfur
hexafluoride (SF6), and nitrogen trifluoride (NF3)—do not deplete stratospheric ozone but are potent
greenhouse gases and are included in this report.
Several substances influence the global radiation budget but are short-lived and therefore not well-
mixed, leading to spatially variable radiative forcing effects. The most important of these
substances are water vapor, aerosols, and tropospheric (ground level) ozone (03).
Tropospheric ozone is formed from chemical reactions in the atmosphere of precursor pollutants,
which include volatile organic compounds (VOCs), CH4, and nitrogen oxides (NOx), in the presence
of ultraviolet light (sunlight).
Aerosols are extremely small particles or liquid droplets suspended in the Earth's atmosphere that
are often composed of sulfur compounds, carbonaceous combustion products (e.g., black
carbon), crustal materials (e.g., dust) and other human-induced pollutants. They can affect the
absorptive characteristics of the atmosphere (e.g., scattering incoming sunlight away from the
Earth's surface, or, in the case of black carbon, absorbing sunlight) and can play a role in affecting
cloud formation and lifetime, as well as the radiative forcing of clouds and precipitation patterns.
In addition, carbon monoxide (CO), nitrogen dioxide (N02), and sulfur dioxide (S02) have direct
influences on radiative forcing, but those are generally smaller than their indirect influences
through their contributions to ozone, aerosols, and the lifetime of methane.
C02, CH4, and N20 are continuously emitted to and removed from the atmosphere by natural
processes on Earth. However, anthropogenic activities (e.g., fossil fuel combustion, cement
production, land-use, land-use change, and forestry, agriculture, or waste management) can cause
additional quantities of these and other greenhouse gases to be emitted or sequestered, thereby
changing their global average atmospheric concentrations. Natural activities, such as plant and
animal respiration and seasonal cycles of plant growth and decay, are examples of processes that
only cycle carbon or nitrogen between the atmosphere and organic biomass. Such processes,
except when directly or indirectly perturbed out of equilibrium by anthropogenic activities, generally
do not alter average atmospheric greenhouse gas concentrations over decadal timeframes.
Climatic changes resulting from anthropogenic activities, however, can have positive or negative
feed back effects on these natural systems. Atmospheric concentrations of these gases, along with
their rates of growth and atmospheric lifetimes, are presented in Table 1-1.
For further information on greenhouse gases, radiative forcing, and implications for climate change,
see the recent scientific assessment reports from the IPCC,6 the U.S. Global Change Research
Program (USGCRP),7 and the National Academies of Sciences, Engineering, and Medicine (NAS).8
6 See https://www.ipcc.ch/report/ar6/wg1/.
7 See https://nca2018.globalchange.gov/.
8 See https://www.nationalacademies.org/topics/climate.
Introduction 1-3
-------
Table 1-1: Global Atmospheric Concentration, Rate of Concentration Change, and
Atmospheric Lifetime of Selected Greenhouse Gases
Atmospheric Variable
CO2
CH4
N2O
SFe
cf4
Pre-industrial atmospheric
concentration
280 ppm
0.730 ppm
0.270 ppm
0.01 ppt
34.1 ppt
Atmospheric concentration
421.08 ppma'b
1.922 ppm0
0.337 ppmd
11.40 ppte
85.5 pptf
Rate of concentration change
2.35 ppm/yrg
8.82 ppb/yrgh
1.02 ppb/yrg
0.33 ppt/yrg
0.81 ppt/yrf
Atmospheric lifetime (years)
See footnote'
11.8'
109'
About 1,000k
50,000
a The atmospheric C02 concentration is the 2023 annual average at the Mauna Loa, HI station. Due to the eruption of the Mauna
Loa Volcano, measurements from Mauna Loa Observatory were suspended as of Nov. 29, 2022, and resumed in July 2023.
Observations starting from December 2022 to July 4, 2023, are from a site at the Mauna Kea Observatories, approximately 21
miles north of the Mauna Loa Observatory (NOAA/ESRL 2025a).
b The global atmospheric C02 concentration, computed using an average of sampling sites across the world, was 419.33 parts
per million (ppm) in 2023.
c The values presented are global 2023 annual average mole fractions (NOAA/ESRL 2025b).
d The values presented are global 2023 annual average mole fractions (NOAA/ESRL 2025c).
e The values presented are global 2023 annual average mole fractions expressed as parts per trillion (ppt) (NOAA/ESRL 2025d).
' The 2019 CF4 global mean atmospheric concentration is from the Advanced Global Atmospheric Gases Experiment (IPCC
2021).
g The rate of concentration change for C02 is an average of the rates from 2008 through 2023 and has fluctuated between 1.54 to
3.36 ppm per year over this period (NOAA/ESRL 2025a). The rate of concentration change for CH4. N2O, and SF6, is the average
rate of change between 2008 and 2024 (NOAA/ESRL 2025b; NOAA/ESRL 2025c; NOAA/ESRL 2025d). The rate of concentration
change for CF4is the average rate of change between 2011 and 2019 (IPCC 2021).
h The growth rate for atmospheric CH4 decreased from over 10 ppb/year in the 1980s to nearly zero in the early 2000s; recently,
the growth rate has been about 8.39 ppb/year (NOAA/ESRL 2025b).
For a given amount of C02 emitted, some fraction of the atmospheric increase in concentration is quickly absorbed by the
oceans and terrestrial vegetation, some fraction of the atmospheric increase will only slowly decrease over a number of years,
and a small portion of the increase will remain for many centuries or more.
' This table reports the "perturbation lifetime" for both ChUand N20, which takes into account the interactions between
emissions of the gas and its own atmospheric residence time.
k The lifetime for SF6 was revised from 3,200 years to about 1,000 years based on recent studies (IPCC 2021).
Sources: Pre-industrial atmospheric concentrations and atmospheric lifetimes for C02, ChUand N20 are from IPCC (2021), pre-
industrial atmospheric concentration for SF6 is from Rigby (2010), and pre-industrial atmospheric concentration for CF4is from
Meinshausen (2017).
A brief description of each greenhouse gas, its sources, and its role in the atmosphere follows.
• Water Vapor (H20): Water vapor is the largest contributor to the natural greenhouse effect.
Water vapor is fundamentally different from other greenhouse gases in that it can condense
and rain out when it reaches high concentrations, and the total amount of water vapor in the
atmosphere is in part a function of the Earth's temperature. While some human activities
such as evaporation from irrigated crops or power plant cooling release water vapor into the
air, these activities have been determined to have a negligible effect on global climate (IPCC
2021). The lifetime of water vapor in the troposphere is on the order of 10 days. Water vapor
can also contribute to cloud formation, and clouds can have both warming and cooling
effects by either trapping or reflecting heat. The relationship between warming and water
vapor concentrations causes a positive (or amplifying) feedback loop (i.e., temperature
increases cause greater atmospheric water vapor concentration which leads to further
increases in temperatures). Aircraft emissions can create contrails, which may also develop
into contrail-induced cirrus clouds, with complex regional and temporal net radiative
forcing effects that currently have a low level of scientific certainty (IPCC 2021).
1-4 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
• Carbon Dioxide (C02): In nature, carbon is cycled between various atmospheric, oceanic,
land biotic, marine biotic, and mineral reservoirs. The largest fluxes occur between the
atmosphere and terrestrial biota, and between the atmosphere and surface water of the
oceans. In the atmosphere, carbon predominantly exists in its oxidized form as C02.
Atmospheric C02 is part of this global carbon cycle, and therefore its fate is a complex
function of geochemical and biological processes. Carbon dioxide concentrations in the
atmosphere increased from approximately 280 parts per million by volume (ppm) in pre-
industrial times to 421.08 ppm in 2023, a 50.4 percent increase (IPCC 2021; NOAA/ESRL
2025a).9,10 The IPCC states that "Observed increases in well-mixed greenhouse gas (GHG)
concentrations since around 1750 are unequivocally caused by human activities" (IPCC
2021). The predominant source of anthropogenic C02 emissions is the combustion of fossil
fuels. Forest clearing, other biomass burning, and some non-energy production processes
(e.g., cement production) also emit notable quantities of C02.
• Methane (CH4): Methane is primarily produced through anaerobic decomposition of organic
matter in biological systems. Agricultural processes such as wetland rice cultivation,
enteric fermentation in animals, and the decomposition of animal wastes emit CH4, as does
the decomposition of municipal solid wastes and treatment of wastewater. Methane is also
emitted during the production and distribution of natural gas and petroleum and is released
as a byproduct of coal mining and incomplete fossil fuel combustion. Atmospheric
concentrations of CH4 have increased by about 163.3 percent since 1750, from a pre-
industrial value of about 730 parts per billion (ppb) to 1,922 ppb in 2023,11 although the rate
of increase decreased to near zero in the early 2000s and has recently increased again to
about 8.39 ppb/year. The IPCC has estimated that about half of the current CH4 flux to the
atmosphere (and the entirety of the increase in concentration) is anthropogenic, from
human activities such as agriculture, fossil fuel production and use, and waste disposal
(IPCC 2021).
Methane is primarily removed from the atmosphere through a reaction with the hydroxyl
radical (OH) and is ultimately converted to C02. Minor removal processes also include
reaction with chlorine in the marine boundary layer, soil sinks, and stratospheric reactions.
Increasing emissions of CH4 reduce the concentration of OH, creating a feedback loop that
increases the atmospheric lifetime of CH4 (IPCC 2021). Methane's reactions in the
atmosphere also lead to production of tropospheric ozone and stratospheric water vapor,
both of which also contribute to climate change. Tropospheric ozone also has negative
effects on human health and plant productivity.
• Nitrous Oxide (N20): Anthropogenic sources of N20 emissions include agricultural soils,
especially production of nitrogen-fixing crops and forages, the use of synthetic and manure
fertilizers, and manure deposition by livestock; fossil fuel combustion, especially from
mobile combustion; adipic (nylon) and nitric acid production; wastewater treatment and
waste incineration; and biomass burning. The atmospheric concentration of N20 has
9 The pre-industrial period is considered as the time preceding the year 1750 (IPCC 2021).
10 Carbon dioxide concentrations during the last 1,000 years of the pre-industrial period (i.e., 750 to 1750), a time of relative
climate stability, fluctuated by about+10 ppm around 280 ppm (IPCC 2021).
11 This value is the global 2023 annual average mole fraction (NOAA/ESRL 2025b).
Introduction 1-5
-------
increased by 24.8 percent since 1750, from a pre-industrial value of about 270 ppb to 337
ppb in 2023,12 a concentration that has not been exceeded in at least the last 800 thousand
years. Nitrous oxide is primarily removed from the atmosphere by the photolytic action of
sunlight in the stratosphere (IPCC 2021). Nitrous oxide also contributes to depletion of
stratospheric ozone.
• Ozone (03): Ozone is present in both the upper stratosphere,13 where it shields the Earth
from harmful levels of ultraviolet radiation, and at lower concentrations in the
troposphere,14 where it is the main component of anthropogenic photochemical "smog."
During the last two decades, emissions of anthropogenic chlorine and bromine-containing
halocarbons, such as CFCs, have depleted stratospheric ozone concentrations. This loss of
ozone in the stratosphere has resulted in negative radiative forcing, representing an indirect
effect of anthropogenic emissions of chlorine and bromine compounds (IPCC 2021). The
depletion of stratospheric ozone and its radiative forcing remained relatively unchanged
since 2000 for the last two decades and is starting to decline; recovery is expected to occur
shortly after the middle of the twenty-first century (WMO/UNEP 2018). Tropospheric ozone
is produced from complex chemical reactions of volatile organic compounds and CH4
mixing with NOx in the presence of sunlight. The tropospheric concentrations of ozone and
these other pollutants are short-lived and, therefore, spatially variable (IPCC 2021).
• Halocarbons, Sulfur Hexafluoride (SF6), and Nitrogen Trifluoride (NF3): Halocarbons are, for
the most part, man-made chemicals that have direct radiative forcing effects and could
also have an indirect effect. Halocarbons that contain chlorine (e.g., CFCs, HCFCs, methyl
chloroform, and carbon tetrachloride) and bromine (e.g., halons, methyl bromide, and
hydrobromofluorocarbons) result in stratospheric ozone depletion and are controlled under
the Montreal Protocol on Substances that Deplete the Ozone Layer. Although most CFCs
and HCFCs are potent global warming gases, their net radiative forcing effect on the
atmosphere is reduced because they cause stratospheric ozone depletion. The United
States phased out the production and importation of halons by 1994 and of CFCs by 1996.
A cap was placed on the production and importation of HCFCs in the United States,
beginning in 1996, and then followed by intermediate requirements and a complete phase-
out by the year 2030. Ozone depleting gases are reported in this Inventory under Annex 6.2
for informational purposes.
Hydrofluorocarbons (HFCs), perfluorocarbons (PFCs), sulfur hexafluoride (SF6), and
nitrogen trifluoride (NF3) are not ozone depleting substances. The most common HFCs are,
however, powerful greenhouse gases. Hydrofluorocarbons are primarily used as
replacements for ozone depleting substances but are also emitted as a byproduct of the
HCFC-22 (chlorodifluoromethane) manufacturing process. Other contributing sources to
12 This value is the global 2023 annual average (NOAA/ESRL 2025c).
13 The stratosphere is the layer from the troposphere up to roughly 50 kilometers. In the lower regions the temperature is
nearly constant but in the upper layer the temperature increases rapidly because of sunlight absorption by the ozone
layer. The ozone-layer is the part of the stratosphere from 19 kilometers up to 48 kilometers where the concentration of
ozone reaches up to 10 parts per million.
14 The troposphere is the layer from the ground up to 11 kilometers near the poles and up to 16 kilometers in equatorial
regions (i.e., the lowest layer of the atmosphere where people live). It contains roughly 80 percent of the mass of all gases
in the atmosphere and is the site for most weather processes, including most of the water vapor and clouds.
1-6 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
HFCemissions include the electronics industry and magnesium production and
processing. Perfluorocarbons, SF6, and NF3 are predominantly emitted from various
industrial processes including aluminum smelting, semiconductor manufacturing, electric
power transmission and distribution, and magnesium casting.
Precursor Greenhouse Gases
Precursor greenhouse gases are not direct greenhouse gases, but can indirectly impact Earth's
radiative balance, by altering the concentrations of other greenhouse gases (e.g., tropospheric
ozone) and atmospheric aerosol (e.g., particulate sulfate). A brief description of each precursor
greenhouse gas, its sources, and its role in the atmosphere follows. Precursors from sectors
included in this report are summarized and reported in Chapter 2.3 of the Inventory.
• Carbon Monoxide (CO): Carbon monoxide has an indirect radiative forcing effect by
elevating concentrations of CH4 and tropospheric ozone through chemical reactions with
other atmospheric constituents (e.g., the hydroxyl radical, OH) that would otherwise assist
in destroying CH4 and tropospheric ozone. Carbon monoxide is created when carbon-
containing fuels are burned incompletely. Through natural processes in the atmosphere, it
is eventually oxidized to C02(IPCC 2026). Carbon monoxide concentrations are both short-
lived in the atmosphere and spatially variable. National CO emissions are summarized and
reported in Chapter 2.3 of this report.
• Ammonia (NH3): The climate change effects of ammonia are indirect by contributing to
aerosol burden and aerosol nucleation. This process occurs through the reaction of NH3
with nitric and sulfuric acid (IPCC 2021). Ammonia also contributes to nitrogen deposition
which enhances N20 emissions from soils (IPCC 2021; IPCC 2006). Ammonia is emitted
during livestock management activities and crop production through the application of
mineral nitrogen fertilizers (IPCC 2021). For this reason, NH3 concentrations tend to peak
over large agricultural areas in the United States and have been increasing throughout the
U.S. in recent decades. National NH3 emissions are summarized and reported in Chapter
2.3 of this report.
• Nitrogen Oxides (NOx): The primary climate change effects of nitrogen oxides (i.e., NO and
N02) are indirect. Warming effects can occur due to reactions leading to the formation of
ozone in the troposphere, but cooling effects can occur due to the role of NOx as a
precursor to nitrate particles (i.e., aerosols) and due to destruction of stratospheric ozone
when emitted from very high-altitude aircraft.15 Additionally, NOx emissions are also likely to
decrease CH4 concentrations, thus having a negative radiative forcing effect (IPCC 2021).
Nitrogen oxides are created from lightning, soil microbial activity, biomass burning (both
natural and anthropogenic fires) fuel combustion, and, in the stratosphere, from the photo-
degradation of N20. Concentrations of NOx are both relatively short-lived in the atmosphere
and spatially variable. National NOx emissions are summarized and reported in Chapter 2.3
of this report.
15 NOx emissions injected higher in the stratosphere, primarily from fuel combustion emissions from high altitude
supersonic aircraft, can lead to stratospheric ozone depletion.
Introduction 1-7
-------
• Non-methane Volatile Organic Compounds (NMVOCs): Non-methane volatile organic
compounds include substances such as propane, butane, and ethane. These compounds
participate, along with NOx, in the formation of tropospheric ozone and other
photochemical oxidants. NMVOCs are emitted primarily from transportation and industrial
processes, as well as biomass burning and non-industrial consumption of organic solvents
(IPCC 2026). Concentrations of NMVOCs tend to be both short-lived in the atmosphere and
spatially variable (IPCC 2021). National NMVOCs are summarized and reported in Chapter
2.3 of this report.
Aerosols
Aerosols are extremely small particles or liquid droplets found in the atmosphere that are either
directly emitted into or are created through chemical reactions in the Earth's atmosphere. Aerosols
or their chemical precursors can be emitted by natural events such as dust storms, biogenic or
volcanic activity, or by anthropogenic processes such as transportation, coal combustion, cement
manufacturing, waste incineration, or biomass burning. Various categories of aerosols exist from
both natural and anthropogenic sources, such as soil dust, sea salt, biogenic aerosols, sulfates,
nitrates, volcanic aerosols, industrial dust, and carbonaceous16 aerosols (e.g., black carbon,
organic carbon). Aerosols can be removed from the atmosphere relatively rapidly by precipitation or
through more complex processes under dry conditions.
Aerosols affect radiative forcing differently than greenhouse gases. Their radiative effects occur
through direct and indirect mechanisms: directly by scattering and absorbing solar radiation (and to
a lesser extent scattering, absorption, and emission of terrestrial radiation); and indirectly by
increasing cloud droplets and ice crystals that modify the formation, precipitation efficiency, and
radiative properties of clouds (IPCC 2021). Despite advances in understanding of cloud-aerosol
interactions, the contribution of aerosols to radiative forcing are difficult to quantify because
aerosols generally have short atmospheric lifetimes, and have number concentrations, size
distributions, and compositions that vary regionally, spatially, and temporally (IPCC 2021).
The net effect of aerosols on the Earth's radiative forcing is believed to be negative (i.e., net cooling
effect on the climate). In fact, aerosols contributed a cooling influence of about 0.6 degrees
Celsius. Due to the high uncertainty in aerosol radiative forcing, estimates range from almost no net
global influence to as much as 1.4 degrees of cooling, offsetting a substantial portion of
greenhouse gas warming (IPCC 2021). Because aerosols remain in the atmosphere for only days to
weeks, their concentrations respond rapidly to changes in emissions.17 Not all aerosols have a
cooling effect. Current research suggests that another constituent of aerosols, black carbon, has a
positive radiative forcing by heating the Earth's atmosphere and causing surface warming when
deposited on ice and snow (IPCC 2021). Black carbon also influences cloud development, but the
direction and magnitude of this forcing is an area of active research.
18 Carbonaceous aerosols are aerosols that are comprised mainly of carbon and hydrogen. Those carbonaceous aerosols
with more hydrogen are classified as "organic carbon", and are generally reflective, while the aerosols that are nearly pure
carbon are classified as "black carbon" (also referred to as "soot") and can absorb light (IPCC 2021).
17 Volcanic activity can inject significant quantities of aerosol producing sulfur dioxide and other sulfur compounds into the
stratosphere, which can result in a longer lasting negative forcing effect (i.e., a few years) (IPCC 2021).
1-8 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
1.2 Global Warming Potentials
A global warming potential (GWP) is a quantified measure of the relative globally averaged radiative
forcing impacts of emissions of a particular greenhouse gas overtime (see Table 1-2). It is defined
as the accumulated radiative forcing within a specific time horizon caused by emitting 1 kilogram
(kg) of the gas, relative to that of the reference gas C02 (IPCC 2021). Direct radiative effects occur
when the gas itself absorbs radiation. Indirect radiative forcing occurs when chemical
transformations involving the original gas produce a gas or gases that are greenhouse gases, or
when a gas influences other radiatively important processes such as the atmospheric lifetimes of
other gases. The reference gas used is C02, and therefore GWP-weighted emissions are measured
in C02 equivalent (C02 Eq.).18 For example, the relationship between a kg of emissions of a gas and
a kg of C02 Eq. emissions can be expressed as follows and also adapted to other units (e.g. metric
tons, etc.):
Equation 1-1: Calculating C02 Equivalent Emissions
kg C02 Eq. = (kg emission of gas) x (GWP)
where,
kgC02Eq. = kilograms of C02 equivalent
kg = kilograms (equivalent to a thousand metric grams)
GWP = Global warming potential
GWP values allow for a comparison of the impacts of emissions and reductions of different gases.
According to the IPCC, GWPs typically have an uncertainty of ±40 percent.
Table 1-2: Global Warming Potentials and Atmospheric Lifetimes (Years) Used in this
Report
Gas Atmospheric Lifetime GWPa
CO2 See footnoteb 1
CH40 12.4 28
N2O 121 265
HFC-23 222 12,400
HFC-32 5.2 677
HFC-41d 2.8 116
HFC-125 28.2 3,170
HFC-134a 13.4 1,300
HFC-143a 47.1 4,800
HFC-152a 1.5 138
HFC-227ea 38.9 3,350
HFC-236fa 242 8,060
CF4 50,000 6,630
C2F6 10,000 11,100
18 Carbon comprises 12/44ths of carbon dioxide by weight.
Introduction 1-9
-------
C3F8 2,600 8,900
C-C4F8
3,200
9,540
SFe
3,200
23,500
NF3
500
16,100
Other Fluorinated Gases
See Annex 6
a 100-year time horizon.
b For a given amount of C02 emitted, some fraction of the atmospheric increase in concentration is quickly absorbed by the
oceans and terrestrial vegetation, some fraction of the atmospheric increase will only slowly decrease over a number of years,
and a small portion of the increase will remain for many centuries or more.
c The GWP of CH4 includes the direct effects and those indirect effects due to the production of tropospheric ozone and
stratospheric water vapor. The indirect effect due to the production of C02 is not included.
d See Table A-1 of 40 CFR Part 98
Source: IPCC (2013).
All estimates are provided throughout the report in both MMT C02 equivalents and unweighted
units. Consistent with greenhouse gas inventories from other countries, this report uses 100-year
GWP values from Table 8.A.1 in Appendix 8.A of the IPCC Fifth Assessment Report (AR5) for
calculating C02-equivalent emissions.
This reflects updated science and ensures that national greenhouse gas inventories reported by all
nations are comparable. A comparison of emission values with the previously used 100-year GWP
values from IPCC Fourth Assessment Report (AR4) (IPCC 2007), and the IPCC Sixth Assessment
Report (AR6) (IPCC 2021) values can be found in Annex 6.1 of this report. The 100-year GWP values
used in this report are listed in Table 1-2.
Greenhouse gases with atmospheric lifetimes longer than a couple of years (e.g., C02, CH4, N20,
HFCs, PFCs, SF6, NF3) tend to be evenly distributed throughout the atmosphere, and consequently
global average concentrations can be determined. The short-lived gases such as water vapor,
carbon monoxide, tropospheric ozone, ozone precursors (e.g., NOx, and NMVOCs), and
tropospheric aerosols (e.g., S02 products and carbonaceous particles), however, vary regionally,
and consequently it is difficult to quantify their global radiative forcing impacts.
Box 1-2: The IPCC Sixth Assessment Report and Global Warming Potentials
In 2021, the IPCC published its Sixth Assessment Report (AR6), which updated its comprehensive
scientific assessment of climate change. Within the AR6 report, the GWP values of gases were revised
relative to previous IPCC assessment reports, e.g., the IPCC Fifth Assessment Report (AR5) (IPCC 2014).
Although the AR5 GWP values are used throughout this report, it is straightforward to review the changes
to the GWP values and their impact on estimates of the total GWP-weighted emissions of the United
States. In the AR6, the IPCC used more recent estimates of the atmospheric lifetimes and radiative
efficiencies of some gases and updated background concentrations. The AR6 now includes climate-
carbon feedback effects for non-C02 gases, improving the consistency between treatment of C02 and
non-C02 gases. Indirect effects of gases on other atmospheric constituents (such as the effect of
methane on ozone) have also been updated to match more recent science.
Table 1-3 presents the new GWP values, relative to those presented in the AR5, using the 100-year time
horizon. All estimates provided throughout this report are also presented in unweighted units. For
informational purposes, emission estimates that use 100-year GWPs from other recent IPCC
Assessment Reports are presented in detail in Annex 6.1 of this report.
1-10 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Table 1-3: Comparison of 100-Year GWP values
100-Year GWP Values
Change Relative to AR5
Gas
AR5a
AR6b
AR6b
C02
1
1
NC
CIV
28
27
(1)
N2O
265
273
8
HFC-23
12,400
14,600
2,200
HFC-32
677
771
94
HFC-41
116
135
19
HFC-125
3,170
3,740
570
HFC-134a
1,300
1,530
230
HFC-143a
4,800
5,810
1,010
HFC-152a
138
164
26
HFC-227ea
3,350
3,600
250
HFC-236fa
8,060
8,690
630
LL
O
6,630
7,380
750
C2F6
11,100
12,400
1,300
C3F8
8,900
9,290
390
C-C4F8
9,540
10,200
660
SFe
23,500
24,300
800
NF3
16,100
17,400
1,300
NC (No Change)
aThe GWP values in this column reflect values used in this report from AR5 excluding climate-carbon feedbacks and the value
for fossil methane.
bThe GWP values in this column are from the AR6 report.
cThe GWP of CH4 includes the direct effects and those indirect effects due to the production of tropospheric ozone and
stratospheric water vapor. Including the indirect effect due to the production of C02 resulting from methane oxidation would lead
to an increase in AR5 methane GWP values by 2 for fossil methane and is not shown in this table.
Note: Parentheses indicate negative values.
Sources: IPCC (2021), IPCC(2013), IPCC (2007), IPCC (2001), IPCC (1996).
1.3 National Inventory Arrangements
The U.S. Environmental Protection Agency (EPA), in cooperation with other U.S. government
agencies, prepares and publishes the Inventory of U.S. Greenhouse Gas Emissions and Sinks. A
wide range of agencies and individuals are involved in supplying data to, planning methodological
approaches and improvements, reviewing, or preparing portions of the Inventory— including federal
and state government authorities, research and academic institutions, industry associations, and
private consultants.
Within EPA, the Office of Atmospheric Protection (OAP) is the lead office responsible for the
emission and removal calculations provided in the Inventory, as well as the completion of the
National Inventory Report including the CRTs. EPA's Office of Transportation and Air Quality (OTAQ)
and Office of Research and Development (ORD) are also involved in calculating emissions and
removals for the Inventory. EPA's OAP serves as the national inventory focal point for this report,
Introduction 1-11
-------
including responding to technical questions and comments on the U.S. Inventory. EPA staff
coordinate the annual methodological choice, activity data collection, emission and removal
calculations, uncertainty assessment, QA/QC processes, and improvement planning at the
individual source and sink category level. EPA's inventory coordinator leads overall compilation and
publication of the entire Inventory and is responsible for the synthesis of information along with the
consistent application of cross-cutting IPCC good practice across the Inventory.
Several other government agencies contribute to the collection and analysis of the underlying
activity data used in the Inventory calculations and to the calculation of estimates integrated in the
report. These agencies include the U.S. Department of Agriculture (USDA), National Oceanic and
Atmospheric Administration (NOAA), the U.S. Geological Survey (USGS), the Federal Highway
Administration (FHWA), the Department of Transportation (DOT), the Bureau of Transportation
Statistics (BTS), the Department of Commerce (DOC), and the Federal Aviation Administration
(FAA). Academic and research centers also provide activity data and calculations to EPA, as well as
individual companies participating in voluntary outreach efforts with EPA. Other U.S. agencies also
provide official data for use in the Inventory such as the U.S. Department of Energy's and the U.S.
Department of Defense. For more information on the source data provided by U.S. government
agencies and external groups, see the Methodology and Data Sources section of this chapter. EPA
engages with agencies regularly on data needs and improvements to ensure sufficient activity
collection for annual compilation of estimates.
1.4 Inventory Preparation Process
This section describes EPA's approach to preparing the annual U.S. Inventory, which includes both
this report and associated tables (e.g., Common Reporting Tables or "CRTs"). The inventory
coordinator at EPA, with support from the cross-cutting compilation staff, is responsible for
coordinating aggregation of all emission and removal estimates, conducting the overall uncertainty
analysis of Inventory emissions and trends overtime, and ensuring consistency and quality
throughout this report and associated tables. Emission and removal calculations, including
associated uncertainty analysis for individual sources and/or sink categories, are the responsibility
of individual source and sink category leads who are most familiar with each category, underlying
data, and the unique national circumstances relevant to its emissions or removals profile. Using
IPCC methodological decision trees and suggested good practice guidance, the individual leads
determine the most appropriate methodology and collect the relevant activity data to use in the
calculations, based upon their expertise, as well as coordinate with researchers and expert
consultants familiar with the sources and sinks. Each year, the coordinator oversees the multi-
stage process for collecting information from each category lead to compile all information and
data for the Inventory.
1-12 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Methods Selection, Data Collection, and Estimate
Development
EPA and other agency source and sink category leads coordinate the collection of input data (e.g.,
activity data and other information) and, as necessary, evaluate or develop the estimation
methodology for the individual source and/or sink categories. Unless there are improvements ready
to incorporate (e.g., methodological refinements), the methodology for the previous year is applied
to the current year of the Inventory, and inventory analysts collect any new data or update data that
have changed from the previous year. If estimates for a new source or sink category are being
developed for the first time, or if the methodology is changing for an existing category, then the
category lead will develop and implement the new or refined methodology, gather the appropriate
activity data and other information for the entire time series, and conduct any further category-
specific review with involvement of relevant experts from industry, government, and universities
(see Chapter 9 and Box ES-3 on EPA's approach to recalculations). Once the methodology is in
place and the data are collected, the category leads calculate emission and removal estimates.
The leads then update or create the relevant report text and accompanying annexes for the
Inventory. Category leads are also responsible for completing the relevant sectoral background
tables of the CRTs, conducting quality control (QC) checks, preparing relevant category materials
for quality assurance (QA), or expert reviews, category-level uncertainty assessments, and
reviewing data for publication in EPA's GHG Data Explorer.19
In the Inventory, the treatment of confidential business information (CBI) is based on EPA internal
guidelines, as well as regulations20 applicable to the data used. EPA has specific procedures in
place to safeguard CBI during the inventory compilation process. When information derived from
CBI data is used for development of inventory calculations, EPA procedures ensure that these
confidential data are sufficiently aggregated to protect confidentiality while still providing useful
information for analysis. For example, within the Energy and Industrial Processes and Product Use
(IPPU) sectors, EPA has used aggregated facility-level data from the Greenhouse Gas Reporting
Program (GHGRP) to develop, inform, and/or QA U.S. emission estimates. In 2014, EPA's GHGRP,
with industry engagement, compiled criteria that would be used for aggregating its confidential
data to shield the underlying CBI from public disclosure.21 In the Inventory, EPA is publishing only
data values that meet the GHGRP aggregation criteria.22 Specific uses of aggregated facility-level
data are described in the respective methodological sections within those chapters. In addition,
EPA uses historical data reported voluntarily to EPA through various voluntary initiatives with U.S.
industry (e.g., EPA Voluntary Aluminum Industrial Partnership [VAIP]) and follows guidelines
established under the voluntary programs for managing CBI.
19 See https://cfpub.epa.gov/ghgdata/inventorvexplorer/.
20 40 CFR Part 2, Subpart B titled "Confidentiality of Business Information" which is the regulation establishing rules
governing handling of data entitled to confidentiality treatment. See https://www.ecfr.gov/cgi-
hin/textidx?SID=a764?35c9eadf9afe05fe04c07a?8939&mc=true&node=sp40.1 .?.b&rgn=div6
21 Federal Register Notice on "Greenhouse Gas Reporting Program: Publication of Aggregated Greenhouse Gas Data." See
pp. 79 and 110 of notice at https://www.gpo.gov/fdsvs/pkg/FR-2014-06-09/pdf/2014-13425.pdf.
22 U.S. EPA Greenhouse Gas Reporting Program. Developments on Publication of Aggregated Greenhouse Gas Data,
November 25, 2014. See http://www.epa.gov/ghgreporting/confidential-business-information-ghg-reporting.
Introduction 1-13
-------
Data Compilation and Archiving
The inventory coordinator at EPA, with support from the data manager, collects the source and sink
data for each category and estimates and aggregates the emission and removal estimates into a
summary data file that links the individual source and sink category data files together. This
summary data file contains all of the essential data in one central location, in formats commonly
used in the Inventory document. In addition to the data from each category, other national trend
and related data are also gathered in the summary sheet for use in the Executive Summary,
Introduction, and Trends chapters of the Inventory report. Trend analysis necessitates gathering
supplemental data to understand and explain trends, including annual economic activity (e.g.,
gross domestic product), population, and the annual use of electricity, energy, fossil and non-fossil
fuels.
Similarly, analysis of the key categories in the current Inventory and analysis of recalculations
impacts compared to the previous inventory cycle are completed in a separate data file based on
output from the summary data file. The uncertainty estimates for each source and sink category are
also aggregated into uncertainty summary data files that are used to conduct the overall Inventory
uncertainty analysis (see Section 1.8).
A cloud-based management system, maintained within EPA's IT infrastructure by the inventory
coordinator, provides a platform for facilitating collaboration on inventory preparation during each
compilation phase and the efficient storage and archiving of electronic document and data files
each annual cycle. Previous final published inventories are also maintained on a report archive
page on EPA's Greenhouse Gas Emissions website.23
National Inventory Report Preparation
This report is compiled from the sections developed by each source or sink category lead using a
standard template to ensure comparable information across inventory categories within each
sector. The inventory coordinator at EPA, with support from the document manager, collects the
source and sink categories' report chapters and methodological annexes. In addition, the inventory
coordinator prepares a brief overview of each chapter that summarizes the emissions and
removals from all sources and sinks discussed in the chapters. The Executive Summary,
Introduction, Trends in Greenhouse Gas Emissions and Removals, and Recalculations and
Improvements chapters are also drafted at this time to reflect the trends and impact from
improvements for the time series of the current Inventory. Finally, the uncertainty analysis and key
category analysis are compiled and updated in the report as part of final analysis steps. Throughout
the report, text boxes are also created to provide additional documentation (e.g., definitions)
and/or to examine the data aggregated in different ways than as presented in the main text of the
report.
Commonly-used international tables, or CRTs are another format for transparently publishing
Inventory data. The tables are compiled from individual time series input data sheets completed by
each category lead, which contain emissions and/or removals and activity data, estimates,
23 See https://www.epa.gov/ghgemissions/us-greenhouse-gas-inventorv-report-archive.
1-14 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
methodological and completeness notations, and associated explanations. The inventory
coordinator and cross-cutting compilation staff import the U.S. category and subcategory
background data into the online reporting platform to export CRTs, assuring consistency and
completeness across all sectoral background tables. The summary reports for emissions and
removals, methods, and emission factors used; the summary tables indicating completeness of
estimates; the recalculation tables; and the emission and removal trends tables are automatically
compiled by the online reporting software and reviewed by the inventory coordinator with support
from the cross-cutting compilation staff. Internal automated quality checks within the software, as
well as checks by the cross-cutting and category leads, are completed for the entire time series of
tables.
Quality Assurance and Quality Control, and Uncertainty
Analysis
Quality assurance (QA) during inventory preparation and compilation consists of a two-stage review
process that includes an expert review and public review. During the first stage, a 30-day expert
review, the first draft of updated sectoral chapters, including a guidance memo and charge
questions is sent to technical experts who are not directly involved in preparing estimates. This
stage is intended to provide an objective review of the methodological approaches and data
sources used, especially for sources and sinks which have experienced any changes since the
previous inventory. The expert review follows good practices from EPA's Peer Review handbook.24
Expert reviewers include other federal agency staff, researchers, industry experts, and others who
have technical knowledge of the data, industry, and methods. EPA reviews and updates expert
participation and outreach on an annual basis prior to each expert review cycle. Experts are
identified in various ways; for example, many reach out to EPA with technical feedback and are
added to the expert reviewer list. Reviewers are also identified through direct outreach by inventory
staff based on expertise. Currently, EPA's expert list includes nearly 300 experts across all sectors.
The second review stage following expert review is a public review that consists of a second draft of
the document, including cross-cutting synthesis chapters, being released for review to the U.S.
public through a notice in the U.S. Federal Register. The entire draft inventory is typically published
on the EPA website and a federal eDocket. Public comments are submitted and tracked using this
publicly accessible eDocket. Similar to expert review, some public comments received may require
further discussion with commenters or other experts and/or additional research. Feedback from all
QA/QC processes that contribute to improving inventory quality overtime are described within
each category's Planned Improvement section. See also the Improvement Planning section that
follows in this chapter.
EPA also publishes responses to comments received during both expert and public reviews along
with the publication of the final report on its website.25 While all phases significantly contribute to
improving inventory quality, the public review phase is also essential for promoting the openness of
24 See https://www.epa.gov/osa/peer-review-handbook-4th-edition-2015.
25 See https://www.epa.gov/ghgemissions/inventorv-us-greenhouse-gas-emissions-and-sinks-1990-2022
Introduction 1-15
-------
the inventory development process and the transparency of the inventory methods and underlying
input data sources.
Quality control is implemented by staff preparing inventory estimates and —is applied at every
stage of inventory development and document preparation. Further information about QA/QC
practices and planning can be found in Section 1.7 and Annex 8. More information on uncertainty
analysis is found in Section 1.8.
Publication
After the final revisions to incorporate any comments from the expert review and public review
periods, EPA prepares the final Inventory, which includes this report, and the accompanying CRTs.
EPA publishes the final Inventory report on EPA's website.26On EPA's website, users can also
visualize and download the current time-series estimates from the GHG Inventory Data Explorer
Tool,27and users can also download more detailed data presented in tables within the report and
report annex in CSV format. Concurrently, EPA makes the official U.S. inventory data available using
commonly-used international tables.
Improvement Planning
Each year, several emission and sink estimates in the inventory are recalculated and revised using
better methods and/or data, with the goal of improving inventory quality and reducing uncertainties
and ensuring the transparency, accuracy, completeness, consistency, and overall usefulness of the
report. In this effort, the United States follows the 2006IPCC Guidelines (IPCC 2006) and its 2019
Refinement (IPCC 2019), which state,
"Both methodological changes and refinements over time are an essential part of improving
inventory quality It is good practice to change or refine methods when available data have changed;
the previously used method is not consistent with the IPCC guidelines for that category; a category
has become key; the previously used method is insufficient to reflect mitigation activities in a
transparent manner; the capacity for inventory preparation has increased; improved inventory
methods become available; and/or for correction of errors."
The EPA's OAP coordinates improvement planning across all sectors as well as cross-cutting
analyses. These plans are based on annual reviews and inputs from the technical teams leading the
compilation of source-level estimates. Planned improvements are also identified through
implementation of QC processes, the key category analysis, and the uncertainty analysis. The
inventory coordinator, with input from EPA source and sink category leads, maintains a log of all
planned improvements, by sector and cross-cutting, tracking the category significance, specific
category improvement, prioritization, anticipated time frame for implementation of each proposed
improvement, and implementation status. Improvements for significant or key categories are
usually prioritized unless the effort or resources required to implement that improvement are
disproportionate relative to improvements for other key categories.
26 See https://www.epa.gov/ghgemissions/inventorv-us-greenhouse-gas-emissions-and-sinks.
27 See https://cfpub.epa.gov/ghgdata/inventorvexplorer/.
1-16 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
1.5 Methodology and Data Sources
Emissions and removals of greenhouse gases from various source and sink categories have been
estimated using methodologies that are consistent with the 2006IPCC Guidelines for National
Greenhouse Gas Inventories (IPCC 2006) and its supplements and refinements. To a large extent,
this report makes use of published official economic and physical statistics for activity data along
with emission factors and other key parameters as inputs to the methods applied. Depending on
the category, activity data can include fuel consumption or deliveries, vehicle-miles traveled, raw
material processed, or commodity produced, etc. Emission factors are factors that relate
quantities of emissions to an activity. For more information on data sources see Figure 1-1, Box 1-1
on use of GHGRP data, and categories' methodology sections for more information on other data
sources. In addition to official statistics, the report utilizes findings from academic studies, trade
association surveys and statistical reports, along with expert judgment, consistent with the 2006
IPCC Guidelines (IPCC 2006).
The methodologies provided in the 2006 IPCC Guidelines (IPCC 2006) represent foundational
methodologies for a variety of source and sink categories, and many of these methodologies
continue to be improved and refined as new research and data become available. This report uses
those IPCC methodologies when applicable, and supplements them with refined guidance, other
available country-specific methodologies and data where possible (e.g., EPA's GHGRP). For
example, as noted earlier in this chapter, this report does apply recent supplements and
refinements to 2006 IPCC Guidelines (IPCC 2006) in estimating emissions and removals from coal
mining, wastewater treatment and discharge, low voltage anode effects (LVAE) during aluminum
production, drained organic soils, and management of wetlands, includingflooded lands. Choices
made regarding the methodologies and data sources used are provided in the Methodology and
Time-Series Consistency discussion of each category within each sectoral chapter of the report,
applying higher tiered methods when feasible, especially for key categories consistent with
methodological decision trees. Where additional detail is helpful and necessary to explain
methodologies and data sources used to estimate emissions, complete documentation is provided
in the annexes as indicated in the methodology sections of those respective source categories
(e.g., Annex 3.14 for forest land remaining forest land and land converted to forest land). Methods
used for key categories are summarized in Annex 1.
Introduction 1-17
-------
Figure 1-1: Key Data Institutions and Data Sources by Sector
Energy
Agriculture and LULUCF
IPPU
Waste
U.S. Energy Information
Administration
USDA U.S. Forest Service
Forest Inventory and Analysis
Program (FIA)
EPA Greenhouse Gas
Reporting Program (GHGRP)
EPA Greenhouse Gas
Reporting Program (GHGRP)
U.S. Departmentof
Commerce-Bureau of the
Census
USDA Natural Resource
Conservation Service (NRCS)
U.S. Geological Survey (USGS)
National Minerals Information
Center
EPA Office of Land and
Emergency Management
(OLEM)
U.S. Department of Defense -
Defense Logistics Agency
USDA National Agricultural
Statistics Service (NASS) and
Agricultural Research Service
(ARS)
American Chemistry Council
(ACC)
EPA Clean Watershed Needs
Survey (CWNS)
U.S. Departmentof Homeland
Security
EPA Office of Research and
Development (ORD)
American Iron and Steel
Institute (AISI)
American Housing Survey
U.S. Departmentof
Transportation - Federal
Highway Administration
U.S. Fish and Wildlife Service
U.S. InternationalTrade
Commission (USITC)
Data from research studies,
trade publications, and
industry associations
U.S. Department of U.S. Department of Air-Conditioning, Heating, and
Transportation - Federal Agriculture (USDA) Animal and Refrigeration Institute
Aviation Administration Plant Health Inspection
Service (APHIS)
U.S. Departmentof
Association of American Plant
Data from other U.S.
Transportation & Bureau of
Food Control Officials
government agencies,
Transportation Statistics
(AAPFCO)
research studies, trade
publications, and industry
association
U.S. Departmentof Labor- National Oceanic and UNEPTechnology and
Mine Safety and Health Atmospheric Administration Economic Assessment Panel
Administration (NOAA) (TEAP)
U.S. Department of Energy
and its National Laboratories
EPA Office of Land and
Emergency Management
(OLEM)
EPAAcid Rain Program
USDA Farm Service Agency
EPA MOVES Model
U.S. Geological Survey (USGS)
EPA Greenhouse Gas
Reporting Program (GHGRP)
U.S. Departmentof the
Interior (DOI) - Bureau of Land
Management (BLM)
U.S. Departmentof Labor-
Mine Safety and Health
Administration
EPA Office of Land and
Emergency Management
(OLEM)
American Association of
Railroads
Alaska Department of Natural
Resources
American Public
Transportation Association
U.S. Departmentof
Commerce - Bureau of the
Census
U.S. Departmentof Interior-
Bureau of Ocean Energy
Management
Data from research studies,
trade publications, and
industry associations
Federal Energy Regulatory
Commission
Data from research studies,
trade publications, and
industry associations
Note: This table is not an exhaustive list of all data sources.
1-18 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
1.6 Key Categories
Key categories are "inventory categories which individually, or as a group of categories (for which a
common method, emission factor and activity data are applied) are prioritized within the national
inventory system because their estimates have a significant influence on a country's total inventory
of greenhouse gases in terms of the absolute level, the trend, or the level of uncertainty in
emissions or removals. The term key category includes both source and sink categories" (IPCC
2006; IPCC 2019). A key category analysis identifies source or sink categories for focusing efforts to
improve overall inventory quality, including additional review when feasible.
The 2006 IPCC Guidelines (IPCC 2006) and its 2079 Refinement (IPCC 2019) define several
approaches, both quantitative and qualitative, to conduct a key category analysis and identify key
categories both in terms of absolute level and trend, along with consideration of uncertainty. This
report employs all approaches to identify key categories for the United States.
1. Approach 1: Identifies significant or key categories without considering uncertainty in its
calculations. The level assessment identifies all source and sink categories that
cumulatively account for 95 percent of total level, i.e., total emissions (gross) in a given year
when assessed in descending order of absolute magnitude. Trend analysis can identify
categories with trends that significantly influence overall trends by identifying all source and
sink categories that cumulatively account for 95 percent of the sum all the trend
assessments (e.g., percent change relative to national trend) when sorted in descending
order of absolute magnitude).
2. Approach 2: Incorporation of each category's uncertainty assessments in its calculations
can be used to identify any additional key categories not already identified from the
Approach 1 level and trend assessments by considering uncertainty. An Approach 2 level
assessment identifies all sources and sink categories that cumulatively account for 90
percent of the sum of all level assessments when sorted in descending order of magnitude.
Similarly, an Approach 2 trend analysis can identify categories whose trends contribute
significantly to overall trends weighing the relative trend difference with the category's
relative uncertainty assessment for 2023.
The level and trend analyses were performed twice, including, and excluding sources and sinks
from the land use, land-use change, and forestry (LULUCF) sector categories. For 2023, based on
the key category analysis, excluding the LULUCF sector, 33 categories accounted for 95 percent of
emissions (Table 1-4). Four categories account for 54 percent of emissions: C02from road
transport-related fuel combustion, C02 from gas-fired electricity generation, C02 from coal fired
electricity generation, and C02 from gas-fired industrial processes. When considering
uncertainties, additional categories such as emissions from substitutes for ozone depleting
substances in aerosols were also identified as a key category. In the trend analysis, 30 categories
were identified as key categories, and when considering uncertainties, 6 additional categories were
identified as key. The trend analysis shows that C02 emissions from coal-fired electricity
generation, C02 from road transport related combustion, and HFC and PFC emissions from
substitutes for ozone depleting substances in the refrigeration and air conditioning sector, and N20
Introduction 1-19
-------
from domestic wastewater treatment are also significant with respect to trends over the time
series.
When considering the contribution of the LULUCF sector, 41 categories accounted for 95 percent of
emissions and sinks, with the most significant category from LULUCF being net C02 emission from
forest land remaining forest land, forest carbon pools excluding harvested wood products. In the
trend analysis, 38 categories were identified as key, and when considering uncertainties, 4
additional categories were identified as key.
Finally, in addition to approaches described, a qualitative assessment of the source and sinks
categories was also considered to capture any additional key categories that were not identified
using the previously described quantitative approaches. For this Inventory, no additional categories
were identified using qualitative criteria recommend by IPCC, but EPA continues to review its
qualitative assessment on an annual basis. Find more information on the key category analysis,
including the approach to disaggregation of inventory estimates, see Annex 1 to this report.
1-20 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Table 1-4: Summary of Key Categories for the United States (1990 and 2023) by Sector
Approach 1
Approach 2
2023
Emissions
Level
Trend
Level
Trend
Level
Trend Level
Trend
Source/Sink Category Code
Greenhouse
Gas
Without
LULUCF
Without
LULUCF
With
LULUCF
With
LULUCF
Without
LULUCF
Without With
LULUCF LULUCF
With
LULUCF
(MMT C02
Eq.)
Energy
1.A.3.b Transportation: Road
C02
•
•
•
1,455.3
1 .A.1 Stationary Combustion - Gas -
C02
704.5
Electricity Generation
1 .A.1 Stationary Combustion - Coal -
Electricity Generation
C02
•
•
•
694.6
1 .A.2 Stationary Combustion - Gas -
Industrial
C02
•
•
•
514.8
1 .A.4.b Stationary Combustion - Gas -
Residential
C02
•
•
247.5
1 .A.2 Stationary Combustion - Oil - Industrial
C02
•
•
•
241.3
1 .A.4.a Stationary Combustion - Gas -
Commercial
C02
•
•
182.8
1 .A.3.a Transportation: Aviation
C02
179.7
1 .A.5 Non-Energy Use of Fuels
C02
•
107.1
1 .A.3.e Transportation: Other
C02
•
•
•
71.3
1 .A.4.a Stationary Combustion - Oil -
Commercial
C02
•
•
60.2
1 .A.4.b Stationary Combustion - Oil -
Residential
C02
•
•
•
59.6
1 .B.2 Natural Gas Systems
C02
37.7
1 .A.2 Stationary Combustion - Coal -
Industrial
C02
•
•
O
•
36.5
1 .A.3.d Transportation: Domestic Navigation
C02
34.6
1.A.3.c Transportation: Railways
C02
30.9
1 .B.2 Petroleum Systems
C02
•
•
•
•
•
•
•
23.3
1 .A.5 Stationary Combustion - Oil - U.S.
Territories
C02
O
O
17.9
1 .A.1 Stationary Combustion - Oil -
Electricity Generation
C02
o
•
o
•
O
•
•
14.7
1.A.5.b Transportation: Military
C02
•
•
4.6
Introduction 1-21
-------
Approach 1
Approach 2
2023
Emissions
Level
Trend Level
Trend
Level
Trend Level
Trend
Source/Sink Category Code
Greenhouse
Gas
Without
LULUCF
Without With
LULUCF LULUCF
With
LULUCF
Without
LULUCF
Without With
LULUCF LULUCF
With
LULUCF
(MMT C02
Eq.)
1 .A.4.a Stationary Combustion - Coal -
C02
•
•
1.1
Commercial
1 .A.4.b Stationary Combustion - Coal -
Residential
C02
•
0.0
1 .B.2 Natural Gas Systems
CH4
•
•
162.4
1 .B.1 Fugitive Emissions from Coal Mining
ch4
•
•
45.4
1 .B.2 Petroleum Systems
ch4
•
38.0
1 .B.2 Abandoned Oil and Gas Wells
ch4
8.5
1 .A.4.b Stationary Combustion - Residential
ch4
•
4.5
1 .A.1 Stationary Combustion - Coal -
Electricity Generation
n20
O
®
•
12.1
1.A.3.b Transportation: Road
N20
O
O
•
O
•
•
8.3
1 .A.2 Stationary Combustion - Industrial
N20
O
1.9
Industrial Processes and Product use
2.C.1 Iron and Steel Production &
C02
O
46.2
Metallurgical Coke Production
2.A.1 Cement Production
C02
40.6
2.B.8 Petrochemical Production
C02
•
30.5
2.B.3 Adipic Acid Production
N20
• •
1.2
2.F.1 Emissions from Substitutes for Ozone
Depleting Substances: Refrigeration and Air
HFCs, PFCs
•
• •
•
•
• •
•
154.7
conditioning
2.F.4 Emissions from Substitutes for Ozone
Depleting Substances: Aerosols
HFCs, PFCs
• •
•
• •
•
17.4
2.F.2 Emissions from Substitutes for Ozone
Depleting Substances: Foam Blowing Agents
HFCs, PFCs
• •
12.1
2.G Electrical Equipment
PFCs, SFs
O
O
•
•
•
5.1
2.B.9 Fluorochemical Production
PFCs, HFCs,
SFs, NF3
O
O
•
O
O
•
4.7
2.C.3 Aluminum Production
PFCs
O
0.5
1-22 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Approach 1
Approach 2
2023
Emissions
Level
Trend
Level
Trend
Level
Trend
Level
Trend
Source/Sink Category Code
Greenhouse
Gas
Without
LULUCF
Without
LULUCF
With
LULUCF
With
LULUCF
Without
LULUCF
Without
LULUCF
With
LULUCF
With
LULUCF
(MMT C02
Eq.)
Agriculture
3.A.1 Enteric Fermentation: Cattle
CH4
180.4
3.B.1 Manure Management: Cattle
ch4
•
•
•
•
•
•
•
•
37.6
3.B.2 Manure Management: Swine
ch4
•
•
24.3
3.C Rice Cultivation
ch4
18.7
3.D.1 Direct Agricultural Soil Management
ch4
266.8
3.D.2 Indirect Applied Nitrogen
n20
29.6
Waste
5.A MSW Landfills
ch4
•
•
•
•
100.6
5.A Industrial Landfills
ch4
•
•
•
18.9
5.D Domestic Wastewater Treatment
ch4
14.0
5.D Domestic Wastewater Treatment
n20
•
•
•
•
20.3
Land Use, Land Use Change, and Forestry
4.E.2 Net Land Converted to Settlements
co2
•
•
79.8
4.B.2 Net Land Converted to Cropland
co2
•
•
35.6
4.C.1 Net Grassland Remaining Grassland
co2
•
22.0
4.C.2 Net Land Converted to Grassland
co2
®
•
®
•
20.9
4.E.1 Net Settlements Remaining
Settlements: LYTFS
co2
O
•
O
•
-11.7
4.B.1 Net Cropland Remaining Cropland
co2
•
•
•
•
-30.5
4.A.1 Net Forest Land Remaining Forest
Land: HWP
co2
•
•
-90.9
4.A.2 Net Land Converted to Forest Land
co2
-103.8
4.E.1 Net Settlements Remaining
Settlements: C Stocks in Trees and Soils
co2
•
•
-122.6
4.A.1 Net Forest Land Remaining Forest
Land: Forest C Pools
co2
•
•
-789.1
4.D.1 Flooded Lands Remaining Flooded
Lands
ch4
45.8
Introduction 1-23
-------
Source/Sink Category Code
Greenhouse
Gas
Approach 1
Approach 2
2023
Emissions
(MMT C02
Eq.)
Level Trend Level Trend
Without Without With With
LULUCF LULUCF LULUCF LULUCF
Level Trend Level Trend
Without Without With With
LULUCF LULUCF LULUCF LULUCF
Subtotal Without LULUCF
Total Emissions Without LULUCF
Percent of Total Without LULUCF
6,025.4
6,197.3
97%
Subtotal With LULUCF
5,033.9
Total Emissions With LULUCF
5,257.4
Percent of Total With LULUCF
96%
NO (Not Occurring)
aSymbols correspond to theyear(s) in which a category is key: 1990 = O; 2023 = •. When both years are key, the open dot and filled dot are combined, for example, 1990 and 2023
= ®.
b Subtotal includes key categories from Level Approach 1 Without LULUCF, Level Approach 2 Without LULUCF, Trend Approach 1 Without LULUCF, and Trend Approach 2 Without
LULUCF.
c Subtotal includes key categories from Level Approach 1 With LULUCF, Level Approach 2 With LULUCF, Trend Approach 1 With LULUCF, and Trend Approach 2 With LULUCF.
Note: Parentheses indicate negative values (or sequestration).
1-24 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
1.7 Quality Assurance and Quality Control
As part of efforts to achieve its stated goals for inventory quality, transparency, and credibility, the
EPA has developed a quality assurance and quality control plan known as the Quality
Assurance/Quality Control and Uncertainty Management Plan for the U.S. Greenhouse Gas
Inventory (QA/QC plan). The QA/QC plan is designed to check, document, and improve the quality
of the Inventory overtime. It is also in alignment with good practice guidance laid out in the 2006
IPCC Guidelines.
The implementation of QA/QC and uncertainty analysis are guided by the QA/QC and inventory
coordinators, who help to maintain the QA/QC plan and the overall uncertainty analysis
procedures. The QA/QC coordinator works closely with the inventory coordinator and the category
leads to ensure consistent implementation of the QA/QC plan and uncertainty analysis across all
inventory categories. The QA/QC activities are integrated throughout the process of preparing the
Inventory. Quality control—in the form of both good practices (such as documentation procedures)
and checks on whether good practices and procedures are being followed—is applied at every
stage of inventory development and document preparation. Key elements of the EPA QA/QC plan
are summarized in Figure 1-2.
The QA/QC plan guides the process of ensuring inventory quality by defining data quality objectives,
describing data and methodology checks, developing processes governing peer review and public
comments, and developing guidance on conducting an analysis of the uncertainty surrounding the
emission and removal estimates. The QA/QC procedures also include feedback loops and provide
for corrective actions that are designed to improve the inventory estimates over time. Specific
improvements identified through QA/QC and uncertainty procedures are noted in the Planned
Improvements sections of respective categories.
Key elements of the QA/QC plan include:
• Procedures, Templates and Forms which standardize the implementation, documentation
and archival of QA/QC activities and related information throughout the whole Inventory
development process. They also guide the implementation of additional QC for more
significant sources. Uncertainty information templates facilitate compilation and review of
uncertainty inputs and the inventory uncertainty analysis.
• Quality Assurance (QA) processes for implementing annual expert and public reviews of the
Inventory. See Section 1.4 for more details on these QA processes.
• Quality Control (QC) procedures that apply General (Tier 1) and Category-specific (Tier
2)methodology or checks as recommended by 2006 IPCC Guidelines (IPCC 2006) to be
completed by staff compiling estimates. General checks are implemented on an annual
basis for each greenhouse gas emissions source or sink category included in this Inventory.
Where QC activities for a particular category go beyond the minimum general checks and
include category-specific checks (Tier 2) or include verification, further explanation is
provided within the respective source or sink category text.
Introduction 1-25
-------
• Communication of required and encouraged QA/QC practices, along with conveying
findings from implementation within the EPA, across federal agencies and departments,
research institutions, and organizations involved in supplying data or compiling estimates
for the Inventory. These communications include periodic trainings reviewing procedures
and lessons, or improvement areas from recent cycles.
• Record-keeping practices that ensure transparent documentation to track which
procedures have been followed, the results of the QA/QC, uncertainty analysis, and
feedback mechanisms for corrective action to improve the inventory estimates overtime.
Records also include tracking the application of more involved QA/QC procedures which
may take more than one cycle to fully implement as part of improvement planning,
especially for category-specific QC, prioritizing key categories (see Section 1.4).
These points and additional information about the QA/QC plan are further expanded upon in Annex
8.
Figure 1-2: Example QC Processes from Inventory QA/QC Plan
Data
Gathering
Check input data for
transcription errors,
units, and data integrity
Check data completeness
(geographic scope,
thresholds, etc.)
Identify input data
outliers forfurther
investigation
inspect automatic checks
for data quality
Identify modifications
that could provide
additional QA/QC checks
Document QC checks
implemented
Data
Documentation
Check citations of data
inputs, assumptions, and
workbook calculations
Check reference docket
for new citations
Inspect text for accuracy
and style consistency
Check that data or
methodology changes are
properly documented
Confirm access and
location of data and
calculation
Document QC checks
implemented
Calculating
Emissions
Check labels, units, and
conversion factors
Reproduce a sample of
calculations
Review completeness,
time series consistency
and trends
Confirm consistency of
key data and estimates
between report and
electronic estimation files
Review changes in data
and consistency with
IPCC methodology
Document QC checks
implemented
Cross-Cutting
Coordination
Confirm completeness of
data across sources and
sinks
Confirm use of
standardized data
summary templates
Inspect flags from
automated data checks
Ensure sums align across
database and document
Document QC checks
implemented
Identify improvements to
data management,
compilation and
documentation based on
QC findings
Box 1-3: Examples of Verification Activities
Consistent with IPCC guidance for national greenhouse gas inventories, verification activities include
comparisons with emission or removal estimates prepared by other bodies and comparisons with
estimates derived from fully independent assessments, e.g., atmospheric concentration
measurements. Verification activities provide information to improve inventories and are part of the
overall QA/QC system
Use of Lower Tier Methods. Complete a "top-down" reference approach for estimating C02 emissions
from fossil fuel combustion in addition to the "bottom-up" sectoral methodology for purposes of
verification is an IPCC good practice. This estimation method uses alternative methodologies and
different data sources than those contained in that section of the Energy chapter. The reference
1-26 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
approach estimates fossil fuel consumption by adjusting national aggregate fuel production data for
imports, exports, and stock changes rather than relying on end-user consumption surveys (see Annex 4
of this report). The reference approach assumes that once carbon-based fuels are brought into a
national economy, they are either saved in some way (e.g., stored in products, kept in fuel stocks, or left
unoxidized in ash) or combusted, and therefore the carbon in them is oxidized and released into the
atmosphere. Accounting for actual consumption of fuels at the sectoral or sub-national level is not
required.
Use of Ambient Measurements Systems for Validation of Emission Inventories. Several recent
studies have estimated emissions at the national or regional level with estimated results that
sometimes differ from EPA's estimate of emissions. EPA has engaged with researchers on how remote
sensing, ambient measurement, and inverse modeling techniques for estimating greenhouse gas
emissions could assist in improving the understanding of inventory estimates. In working with the
research community to improve national greenhouse gas inventories, EPA follows guidance from the
IPCC on the use of measurements and modeling to validate emission inventories.28 An area of particular
interest in EPA's outreach efforts is how ambient measurement data can be used to assess estimates or
potentially be incorporated into the Inventory in a manner consistent with this Inventory report's
transparency of its calculation methodologies, and the ability of inverse modeling to attribute emissions
and removals from remote sensing to anthropogenic sources, as defined by the IPCC for this report,
versus natural sources and sinks.
The 2019 Refinement to the IPCC 2006 Guidelines for National Greenhouse Gas Inventories (IPCC 2019)
Volume 1 General Guidance and Reporting, Chapter 6: Quality Assurance, Quality Control and
Verification notes that emission estimates derived from atmospheric concentration measurements can
provide independent data sets as a basis for comparison with inventory estimates. The 2019 Refinement
provides guidance on conducting such comparisons (as summarized in Table 6.2 of IPCC [2019] Volume
1, Chapter 6) and provides guidance on using such comparisons to identify areas of improvement in
national inventories (as summarized in Box 6.5 of IPCC [2019] Volume 1, Chapter 6). Further, it identified
fluorinated gases as particularly suitable for such comparisons due their limited natural sources, their
generally long atmospheric lifetimes, and well-understood loss mechanisms, which makes it relatively
more straightforward to model their emission fluxes from observed mass quantities. Unlike emissions of
C02, CH4, and N20, emissions of fluorinated greenhouse gases are almost exclusively anthropogenic,
meaning that the fluorinated greenhouse gas emission sources included in this Inventory account for
the majority of the total U.S. emissions of these gases detectable in the atmosphere. This evaluation
approach is also useful for gases and sources with larger uncertainties in available bottom-up inventory
methods and data, such as emissions of CH4, which are primarily from uncertain biological (e.g., enteric
fermentation) and fugitive (e.g., natural gas production) activities.
In this Inventory, EPA includes the results from current and previous comparisons between fluorinated
gas emissions inferred from atmospheric measurements and fluorinated gas emissions estimated
based on bottom-up measurements and modeling. These comparisons, performed for HFCs and SF6
respectively, are described under the QA/QC and Verification discussions in Chapter 4, Sections 4.25
28 See http://www.ipcc-nggip.iges.or.ip/meeting/pdfiles/1003 Uncertaintv%20meeting report.pdf.
Introduction 1-27
-------
Substitution of Ozone Depleting Substances and 4.26 Electrical Equipment in the IPPU chapter of this
report.
Consistent with the 2019 Refinement (IPCC 2019), a key element to facilitate such comparisons is a
spatially-explicit (or gridded) emissions inventory as an input to inverse modeling. To improve the ability
to compare methane emissions from the national-level greenhouse gas inventory with observation-
based emission estimates, a team of researchers from U.S. EPA, SRON Netherlands Institute for Space
Research, Harvard University, and Lawrence Berkely National Laboratory developed a time series of
annual anthropogenic methane emissions maps with 0.1 °x 0.1° (~10km x 10km) spatial resolution and
monthly temporal resolution for the contiguous United States.29 The gridded methane inventory is
designed to be consistent with the U.S. EPA Inventory of U.S. Greenhouse Gas Emissions and
Sinks estimates, which presents national totals for different source types.30 The development of this
gridded inventory is consistent with the recommendations contained in two National Academies of
Science reports examininggreenhousegas emissions data (National Research Council2010; National
Academies of Sciences, Engineering, and Medicine 2018).
Finally, in addition to use of atmospheric concentration measurement data for comparison with
Inventory data, information from top-down studies is directly incorporated in the Natural Gas Systems
calculations to quantify emissions from certain well blowout events.
1.8 Uncertainty Analysis
Emissions and removals calculated for the U.S. Inventory reflect best estimates for greenhouse gas
source and sink categories in the United States and are continuously revised and improved as new
information becomes available. Uncertainty assessment is an essential element of a complete and
transparent emissions inventory because it helps inform and prioritize inventory improvements. For
the U.S. Inventory, uncertainty analyses are conducted for each source and sink category as well as
for the uncertainties associated with the overall emission (current and base year) and trends
estimates. These analyses reflect the quantitative uncertainty in the emission (and removal)
estimates associated with uncertainties in their input parameters (e.g., activity data and EFs) and
serve to evaluate the relative contribution of individual input parameter uncertainties to the overall
Inventory, its trends, and each source and sink category.
The overall level and trend uncertainty estimates for total U.S. greenhouse gas emissions and sinks
was developed using the IPCC Approach 2 uncertainty estimation methodology (assuming a
Normal distribution for Approach 1 estimates), which employs a Monte Carlo stochastic simulation
technique. The IPCC provides good practice guidance on two approaches—Approach 1 and
Approach 2—to estimating uncertainty for both individual and combined source categories.
Approach 2 quantifies uncertainties based on a distribution of emissions (or removals), built-up
29 See https://www.epa.gov/ghgemissions/us-gridded-methane-emissions.
30 See https://www.epa.gov/ghgemissions/inventorv-us-greenhouse-gas-emissions-and-sinks.
1-28 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
from repeated calculations of emission estimation models and the underlying input parameters,
randomly selected according to their known distributions. Approach 2 methodology is applied to
each individual source and sink category wherever data and resources are permitted and is also
used to quantify the uncertainty in the overall Inventory and its Trends. Source and sink chapters in
this report provide additional details on the uncertainty analysis conducted for each source and
sink category. See Annex 7 of this report for further details on the U.S. process for estimating
uncertainty associated with the overall emission (base and current year) and trends estimates.
Consistent with IPCC (IPCC 2006), the United States has ongoing efforts to continue to improve the
overall Inventory uncertainty estimates presented in this report.
The United States has also implemented many improvements over the last several years to reduce
uncertainties across the source and sink categories and improve Inventory estimates. These
improvements result from the development of uncertainty assessments for sources previously not
estimated, new data sources that provide more accurate data and/or increased data coverage, as
well as methodological improvements. Following IPCC good practice, additional efforts to reduce
Inventory uncertainties can occur through efforts to incorporate excluded emission and sink
categories (see Annex 5), improve estimation methods, and collect more detailed, measured, and
representative data. Individual category chapters and Annex 7 both describe current ongoing and
planned Inventory and uncertainty analysis improvements. Consistent with IPCC (2006), the EPA
and collaborating organizations have ongoing efforts to continue to improve the category-specific
uncertainty estimates presented in this report, largely prioritized by considering improvements to
categories identified as significant by the Key Category Analysis.
Estimates of quantitative uncertainty for the total U.S. greenhouse gas emissions and sinks in 1990
(base year) and 2023 are shown in Table 1-5 and Table 1-6, respectively. The overall uncertainty
surrounding the total net emissions is estimated to be -6 to +5 percent in 1990 and -5 to +6 percent
in 2023. When the LULUCF sector is excluded from the analysis the uncertainty is estimated to be
-2 to +4 percent in 1990 and -2 to +4 percent in 2023.
Introduction 1-29
-------
Table 1-5: Estimated Overall Inventory Quantitative Uncertainty for 1990 (MMT C02 Eq.
and Percent)
1990 Emission
Uncertainty Range Relative to
Standard
Estimate
Greenhouse Gas Estimate"
Meanb
Deviation11
(MMT CO2 Eq.)
(%)
Lower
Upper
Lower
Upper
Gas
(MMT CO2 Eq.)
Boundc
Boundc
Bound
Bound
(MMT CO2 Eq.)
C02
5,131.8
5,024.0
5,365.2
-2%
5%
4,972.0
88.0
CH4d
873.1
767.4
914.6
-12%
5%
680.0
37.5
N2Od
407.8
348.7
509.8
-15%
25%
428.6
40.7
PFC, HFC, SF6, and NF3d
125.6
103.3
149.1
-18%
19%
220.7
12.2
Total Gross Emissions
6,538.3
6,384.5
6,787.1
-2%
4%
6,301.2
104.8
LULUCF Emissions6
59.1
55.9
63.3
-6%
7%
61.1
2.0
LULUCF Carbon Stock Change Fluxf
(1,096.9)
(1,386.3)
(949.2)
26%
-13%
(1,061.5)
111.7
LULUCF Sector Net Totals
(1,037.9)
(1,326.8)
(889.2)
28
-14%
(1,000.4)
111.7
Net Emissions (Sources and Sinks)
5,500.4
5,177.2
5,779.7
-6%
5%
(5,300.8)
154.6
aThe lower and upper bounds for emission estimates correspond to a 95 percent confidence interval, with the lower bound
corresponding to 2.5th percentile and the upper bound corresponding to 97.5th percentile.
b Mean value indicates the arithmetic average of the simulated emission estimates; standard deviation indicates the extent of
deviation of the simulated values from the mean.
cThe lower and upper bound emission estimates for the sub-source categories do not sum to total emissions because the low
and high estimates for total emissions were calculated separately through simulations.
dThe overall uncertainty estimates did not take into account the uncertainty in the GWP values for ChU, N20 and high GWP gases
used in the Inventory emission calculations for 1990.
e LULUCF emissions include the ChU and N20 emissions reported for peatlands remaining peatlands, forest fires, drained organic
soils, grassland fires, and coastal wetlands remaining coastal wetlands; ChU emissions from land converted to coastal
wetlands, land converted to flooded land, and flooded land remaining flooded land; and N20 emissions from forest soils and
settlement soils.
f LULUCF carbon stock change is the net C stock change from the following categories: forest land remainingforest land, land
converted to forest land, cropland remaining cropland, land converted to cropland, grassland remaininggrassland, land
converted to grassland, wetlands remaining wetlands, land converted to wetlands, settlements remaining settlements, and
land converted to settlements. Since the resulting flux is negative the signs of the resulting lower and upper bounds are
reversed.
gThe LULUCF sector net total is the net sum of allChU and N20 emissions to the atmosphere plus net carbon stock changes.
Notes: Total emissions (excluding emissions for which uncertainty was not quantified) are presented without LULUCF. Net
emissions are presented with LULUCF. Totals may not sum due to independent rounding. Parentheses indicate net
sequestration.
1-30 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Table 1-6: Estimated Overall Inventory Quantitative Uncertainty for 2023 (MMT C02 Eq.
and Percent)
2023 Emission
Uncertainty Range Relative to
Standard
Estimate
Greenhouse Gas Estimate"
Meanb
Deviation11
(MMT CO2 Eq.)
(%)
Lower
Upper
Lower
Upper
Gas
(MMT CO2 Eq.)
Boundc
Boundc
Bound
Bound
(MMT CO2 Eq.)
C02
4,918.4
4,818.4
5,128.7
-2%
4%
4,969.4
80.5
CH4d
686.7
638.6
762.4
-7%
11%
700.4
31.8
N2Od
387.0
328.8
493.6
-15%
28%
404.2
42.3
PFC, HFC, SF6, and NF3d
205.3
183.0
217.6
-11%
6%
199.7 8.9
Total Gross Emissions
6,197.3
6,088.8
6,466.6
-2%
4%
6,273.6
96.7
LULUCF Emissions6
60.6
57.5
65.0
-5%
7%
61.1
2.0
LULUCF Carbon Stock Change Fluxf
(1,000.5)
(1,257.8)
(862.8)
26%
-14%
(1,061.9)
101.6
LULUCF Sector Net Totals
(939.9)
(1,196.5)
(801.5)
27%
-15%
(1,000.8)
101.6
Net Emissions (Sources and Sinks)
5,257.4
5,002.9
5,549.1
-5%
6%
5,272.9
140.5
aThe lower and upper bounds for emission estimates correspond to a 95 percent confidence interval, with the lower bound
corresponding to 2.5th percentile and the upper bound corresponding to 97.5th percentile.
b Mean value indicates the arithmetic average of the simulated emission estimates; standard deviation indicates the extent of
deviation of the simulated values from the mean.
cThe lower and upper bound emission estimates for the sub-source categories do not sum to total emissions because the low
and high estimates for total emissions were calculated separately through simulations.
dThe overall uncertainty estimates did not take into account the uncertainty in the GWP values for ChU, N20 and high GWP gases
used in the Inventory emission calculations for 2023.
e LULUCF emissions include the ChU and N20 emissions reported for peatlands remaining peatlands, forest fires, drained organic
soils, grassland fires, and coastal wetlands remaining coastal wetlands; ChU emissions from land converted to coastal
wetlands, land converted to flooded land, and flooded land remaining flooded land; and N20 emissions from forest soils and
settlement soils.
f LULUCF carbon stock change is the net C stock change from the following categories: forest land remainingforest land, land
converted to forest land, cropland remaining cropland, land converted to cropland, grassland remaininggrassland, land
converted to grassland, wetlands remaining wetlands, land converted to wetlands, settlements remaining settlements, and
land converted to settlements. Since the resulting flux is negative the signs of the resulting lower and upper bounds are
reversed.
gThe LULUCF sector net total is the net sum of allChU and N20 emissions to the atmosphere plus net carbon stock changes.
Notes: Total emissions (excluding emissions for which uncertainty was not quantified) are presented without LULUCF. Net
emissions are presented with LULUCF. Totals may not sum due to independent rounding. Parentheses indicate net
sequestration.
The uncertainty for 2023 is similar to the uncertainty for 1990, though slightly lower. There have
been some improvements in significant categories which do not necessarily reduce uncertainties.
For example, use of higher tier methods may reveal a "more realistic acknowledgement of the
limitations of existing knowledge" of uncertainty, including reflecting additional complexity of a
system compared to a lower tier method (IPCC 2006). Methodological and data quality
improvements were also made for HFCs, PFCs, SF6 and NF3 this year but the uncertainties for these
sources also slightly increased. Use of extrapolation techniques to ensure time series consistency,
such as for direct N20 emissions from agricultural soil management, may also result in higher
uncertainties in the current year relative to the base year of the time series. Note, the methods and
data for fossil fuel combustion categories, the most significant source, has not changed
significantly and neither have uncertainties. It is also worth noting that some of the improvements
Introduction 1-31
-------
to shift to use of GHGRP have reduced uncertainties (e.g., CH4from landfills) but several have been
in less significant categories within the inventory (e.g., C02from ammonia production ). So, the
overall uncertainty for latest year reflects these offsetting effects and trends within the uncertainty
assessment.
In addition to the estimates of uncertainty associated with the current and base year estimates,
Table 1-7 presents the estimates of inventory trend uncertainty. The 2006IPCC Guidelines (IPCC
2006) defines trend as the difference in emissions between the base year (i.e., 1990) and the
current year (i.e., 2023) Inventory estimates. However, for purposes of understanding the concept of
trend uncertainty, the trend is defined in this Inventory as the percentage change in the gross
emissions (or net emissions) estimated for the current year, relative to the gross emission (or net
emissions) estimated for the base year. The uncertainty associated with this trend is referred to as
trend uncertainty and is reported as between -11 and 5 percent at the 95 percent confidence level
between 1990 and 2023. This indicates a range of approximately -11 percent below and 5 percent
above the trend estimate of -4 percent. See Annex 7 for trend uncertainty estimates for individual
source and sink categories by gas.
Table 1-7: Quantitative Assessment of Trend Uncertainty (MMT C02 Eq. and Percent)
(MMTCO2 Eq.)
Emissions
Trend Rangeb(%)
Base Year
2023
Trend
Lower
Upper
Gas
Emissions"
Emissions
(%)
Bound
Bound
C02
5,131.8
4,918.4
-4%
-8%
0%
CH4
873.1
686.7
-21%
-28%
-9%
N2O
407.8
387.0
-5%
-31%
49%
HFCs, PFCs,SF6, and NF3
125.6
205.3
64%
43%
119%
Total Gross Emissionsc
6,538.3
6,197.3
-5%
-9% 0%
LULUCF Emissions'1
59.1
60.6
3%
-6%
13%
LULUCF Carbon Stock Change Fluxe
(1,096.9)
(1,000.5)
-9%
-30%
19%
LULUCF Sector Net Total'
(1,037.9)
(939.9)
-9%
-32%
20%
Net Emissions (Sources and Sinks)c
5,500.4
5,257.4
-4%
-11%
5%
a Base Year is 1990 for all sources.
bThe trend range represents a 95 percent confidence interval for the emission trend, with the lower bound corresponding to 2.5th
percentile value and the upper bound corresponding to 97.5th percentile value.
c Totals exclude emissions for which uncertainty was not quantified.
d LULUCF emissions include the ChUand N20 emissions reported for peatlands remaining peatlands, forest fires, drained organic
soils, grassland fires, and coastal wetlands remaining coastal wetlands; ChU emissions from land converted to coastal
wetlands, land converted to flooded land, and flooded land remaining flooded land; and N20 emissions from forest soils and
settlement soils.
e LULUCF carbon stock change is the net C stock change from the following categories: forest land remainingforest land, land
converted to forest land, cropland remaining cropland, land converted to cropland, grassland remaininggrassland, land
converted to grassland, wetlands remaining wetlands, land converted to wetlands, settlements remaining settlements, and
land converted to settlements.
fThe LULUCF sector net totalis the net sum of allChU and N20 emissions to the atmosphere plus net carbon stock changes.
Notes: Totals may not sum due to independent rounding. Parentheses indicate net sequestration. Total emissions (excluding
emissions for which uncertainty was not quantified) are presented without LULUCF. Net emissions are presented with LULUCF.
1-32 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
1.9 Completeness
This report, along with its accompanying tables, serves as a thorough assessment of the
anthropogenic sources and sinks of greenhouse gas emissions for the United States for the time
series 1990 through 2023. This report is intended to be comprehensive and includes the vast
majority of emissions and removals identified as anthropogenic. In general, sources or sink
categories not accounted for in this Inventory are excluded because they are not occurring in the
United States, its territories, and tribal lands, or because data are unavailable to develop an
estimate and/or the categories were determined to be insignificant in terms of overall national
emissions.
The EPA is continually working to improve upon the understanding of such sources and sinks
currently not included and identify the data required to estimate any emissions and removals
currently excluded, focusing on categories that are anticipated to be significant. See Chapter 9 on
Improvements and Recalculations for more information on completeness and improvements
implemented this cycle. For a list of sources and sink categories not included and more information
on the significance of these categories, see Annex 5 and the respective category sections in each
sectoral chapter of this report.
Box 1-4: Organization of Report
The Inventory is organized according to the chapters listed below in Table 1-8, consistent with the
structure used by other countries.
Table 1-8: Inventory Sector Descriptions
Chapter (Inventory Sector) Activities Included
Energy Emissions of all greenhouse gases resulting from stationary and mobile energy
activities including fuel combustion and fugitive fuel emissions, and non-energy use of
fossil fuels.
Industrial Processes and
Emissions resulting from industrial processes and product use of greenhouse gases.
Product Use
Agriculture
Emissions from agricultural activities except fuel combustion, which is addressed
under Energy.
Land Use, Land-Use Change,
Emissions and removals of CO2, and emissions of CH4, and N2O from land use, land-
and Forestry
use change, and forestry.
Waste
Emissions from waste management activities.
Within each chapter, emissions are identified by the anthropogenic activity that is the source or sink of
the greenhouse gas emissions being estimated (e.g., coal mining). Overall, the following organizational
structure is consistently applied throughout this report:
Chapter/Reporting Sector: Overview of emissions and trends for each inventory reporting sector.
Source or Sink Category: Description of category pathway and emission/removal trends based on IPCC
methodologies.
Methodology and Time-Series Consistency: Description of analytical methods (e.g., from 2006 IPCC
Guidelines, or country-specific methods) employed to produce emission estimates and identification of
Introduction 1-33
-------
data references, primarily for activity data and emission factors, and a discussion of time-series
consistency
Uncertainty: A discussion and quantification of the uncertainty in emission estimates.
QA/QC and Verification: A discussion on steps taken to QA/QC and verify the emission estimates,
consistent with the U.S. QA/QC plan, and any key QC findings.
Recalculations Discussion: A discussion of any data or methodological changes that necessitate a
recalculation of previous years' emission estimates, and the impact of the recalculation on the emission
estimates, if applicable.
Planned Improvements: A discussion on any category-specific planned improvements, if applicable.
Special attention is given to co2 from fossil fuel combustion relative to other sources because of its
share of emissions and its dominant influence on emission trends. For example, each energy
consuming end-use sector (i.e., residential, commercial, industrial, and transportation), as well as the
electricity generation sector, is described individually. Additional information for certain source
categories and other topics is also provided in several methodological Annexes listed in Table 1-9.
Table 1-9: List of Annexes
ANNEX 1 Key Category Analysis
ANNEX 2 Methodology and Data for Estimating CO2 Emissions from Fossil Fuel Combustion
2.1.
Methodology for Estimating Emissions of CO2 from Fossil Fuel Combustion
2.2.
Methodology for Estimating the Carbon Content of Fossil Fuels
2.3.
Methodology for Estimating Carbon Emitted from Non-Energy Uses of Fossil Fuels
ANNEX 3 Methodological Descriptions for Additional Source or Sink Categories
3.1.
Methodology for Estimating Emissions of CH4, N2O, and Indirect Greenhouse Gases from Stationary
Combustion
3.2.
Methodology for Estimating Emissions of CH4, N2O, and Indirect Greenhouse Gases from Mobile Combustion
and Methodology for and Supplemental Information on Transportation-Related Greenhouse Gas Emissions
3.3.
Methodology for Estimating Emissions from Commercial Aircraft Jet Fuel Consumption
3.4.
Methodology for Estimating CH4 Emissions from Coal Mining
3.5.
Methodology for Estimating CH4 and CO2 Emissions from Petroleum Systems
3.6.
Methodology for Estimating CH4 Emissions from Natural Gas Systems
3.7.
Methodology for Estimating CO2 and N2O Emissions from Incineration of Waste
3.8.
Methodology for Estimating Emissions from International Bunker Fuels used by the U.S. Military
3.9.
Methodology for Estimating Emissions from Other Fluorochemical Production
3.10.
Methodology for Estimating HFC and PFC Emissions from Substitution of Ozone Depleting Substances
3.11.
Methodology for Estimating CH4 Emissions from Enteric Fermentation
3.12.
Methodology for Estimating CH4 and N2O Emissions from Manure Management
3.13.
Methodology for Estimating N2O Emissions, CH4 Emissions and Soil Organic C Stock Changes from Agricultural
Lands (Cropland and Grassland)
3.14.
Methodology for Estimating Net Carbon Stock Changes in Forest Land Remaining Forest Land and Land
Converted to Forest Land
3.15.
Methodology for Estimating CH4 Emissions from Landfills
3.16.
Methodology for Estimating CH4 and N2O Emissions from Wastewater Treatment and Discharge
1-34 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
ANNEX 4 IPCC Reference Approach for Estimating CO2 Emissions from Fossil Fuel Combustion
ANNEX 5 Assessment of the Sources and Sinks of Greenhouse Gas Emissions Not Included
ANNEX 6 Additional Information
6.1. Global Warming Potential Values
6.2. Ozone Depleting Substance Emissions
6.3. Greenhouse Gas Precursors: Mapping of NEI categories to the Inventory
6.4. Constants, Units, and Conversions
6.5. Chemical Formulas
ANNEX 7 Uncertainty
7.1. Overview
7.2. Methodology and Results
7.3. Reducing Uncertainty
7.4. Planned Improvements
7.5. Additional Information on Uncertainty Analyses by Source
ANNEX 8 QA/QC Procedures
8.1. Background
8.2. Purpose
8.3. Assessment Factors
8.4. Responses During the Review Process
ANNEX 9 Use of Greenhouse Gas Reporting Program (GHGRP) in Inventory
Introduction 1-35
-------
2 Trends in Greenhouse Gas
Emissions and Removals
2.1 Overview of U.S. Greenhouse Gas
Emissions and Sinks Trends
In 2023, total gross U.S. greenhouse gas emissions were 6,197.3 million metric tons of carbon dioxide
equivalent (MMT C02 Eq.). Total gross U.S. emissions, which exclude emissions and sinks from the land
use, land use change and forestry (LULUCF) sector, decreased by 5.2 percent from 1990 to 2023, down
from a high of 15.3 percent above 1990 levels in 2007. Gross emissions decreased from 2022 to 2023 by
2.3 percent (146.8 MMT C02 Eq.), driven largely by a decrease in C02 emissions from fossil fuel
combustion. C02 emissions from fossil fuel combustion decreased by 3.0 percent in 2023 relative to
2022 and were 4.1 percent below 1990 emissions. Specifically, C02 emissions from coal consumption
decreased by 18.3 percent (164.1 MMT C02 Eq.) from 2022 to 2023. C02 emissions from natural gas use
increased by 1.0 percent (17.6 MMTC02 Eq.) and emissions from petroleum use increased by 0.2
percent (3.1 MMT C02 Eq.) from 2022 to 2023. The decrease in coal use and associated emissions from
2022 to 2023 is mainly due to reduced use in the electric power sector and is driving the overall
reduction. The increase in natural gas consumption and associated emissions in 2023 is observed
mostly in the electric power and industrial sectors, the increase in petroleum use is mainly in the
transportation sector.
Net emissions, including emissions and sinks from the LULUCF sector were 5,257.4 MMT C02 Eq. in
2023. Overall, net emissions decreased by 3.3 percent from 2022 to 2023. Over the last 20 years, net
emissions decreased nearly 20 percent. Trends in net emissions are illustrated in Table 2-1. Carbon
sequestration in the LULUCF sector offset 16.1 percent of total gross emissions in 2023.
Figure 2-1 and Figure 2-2 illustrate the overall trend in total U.S. emissions and sinks since 1990, by gas
and by annual percentage changes relative to the previous year.
Trends in Greenhouse Gas Emissions and Removals 2-1
-------
Figure 2-1: U.S. Greenhouse Gas Emissions and Sinks by Gas
9,000
8,000
7,000
6,000
S 5'000
oj
8 4,000
I-
i 3,000
2,000
1,000
0
-1,000
I HFCs, PFCs, SFe and NFb
I Nitrous Oxide
Methane
I Carbon Dioxide
I Net CO2 Flux from LULUCFs
¦ Net Emissions (including Net CO2 Flux from LULUCF)
o h in m 5-
Ol Qt
Ln^r\cooio^Hr\iro
o^o^o^c^o^oooo
LO KO
o o
CTi Q1* Ch Ch CT> CTi CT> O O O O O O O
NCO^OHfNlMtLOlOMJDC^OrifNjrO
OOO
OOOOOOOOOOOOO
fN (N fN fN
OOOO
HHHHHHHHHHfN(NfMrNfNrMfNf\|(NfNrvJfNfNfNr\rvJfNjfMfNfNfNfNfNJfN
a The term "flux" is used to describe the exchange of C02 to and from the atmosphere, with net flux being either positive or
negative depending on the overall balance. Removal and long-term storage of C02 from the atmosphere is also referred to as
"carbon sequestration,"
Figure 2-2: Annual Percentage Change in Net and Gross U.S. Greenhouse Gas
Emissions and Sinks Relative to the Previous Year
6%
4%
2%
0%
-2%
-4%
-6%
-8%
-10%
-12%
mhhllLn„lll
hi Tii I || -¦
¦¦I
I Change in Gross Emissions
I Change in Net Emissions
^r-jrOTrir>\or->.o
CO1. CTiu*i^0*0*C
CTi On On On On On On On On
5
OOO
lNlNlNfMfMlNfMlNNlNl\rMlNfMfMfMfMlMlNrNjlNNfMfM
Emissions and Sinks by Gas
Figure 2-3 illustrates the relative contribution of each gas to total gross U.S. greenhouse gas emissions
in 2023, in C02 equivalents (i.e., weighted by global warming potential). The primary greenhouse gas
2-2 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
emitted by human activities in the United States is C02, representing 79.4 percent of total gross
greenhouse gas emissions. The largest source of C02—and of overall greenhouse gas emissions—is
fossil fuel combustion, primarilyfrom transportation and power generation. Methane (CH4) emissions
account for 11.1 percent of emissions. The major sources of methane include enteric fermentation
associated with domestic livestock, natural gas systems, and decomposition of waste in landfills. N20
accounts for an additional 6.2 percent of emissions. Agricultural soil management, wastewater
treatment, stationary sources of fuel combustion, and manure management are the major sources of
N20 emissions. Ozone depleting substance (ODS) substitute emissions were the primary contributor to
aggregate hydrofluorocarbon (HFC) emissions. Perfluorocarbon (PFC) emissions were attributable
primarily to fluorochemical production, electronics manufacturing, and primary aluminum production.
Electrical equipment accounted for most sulfur hexafluoride (SF6) emissions. The electronics industry
and fluorochemical production are the only sources of NF3 emissions.
Figure 2-3: 2023 Gross Total U.S. Greenhouse Gas Emissions by Gas (Percentages
based on MMT C02 Eq.)
3.3%
HFCs, PFCs, SFe and NF3
Note: Emissions and removals from LULUCF are excluded from the figure above.
From 1990 to 2023, total gross emissions of C02 decreased by 4.2 percent (213.4 MMT C02 Eq.), total
gross emissions of methane (CH4) decreased by 21.4 percent (186.5 MMT C02 Eq.), and total gross
emissions of nitrous oxide (N20) decreased by 5.1 percent (20.9 MMT C02 Eq.). During the same period,
emissions of fluorinated gases including HFCs, PFCs, SF6, and NF3 rose by 63.5 percent (79.7 MMT C02
Eq.). Rapidly growing emissions of HFCs drove this trend, overwhelming decreases in emissions of PFCs
and SF6. Emissions of HFCs, PFCs, SF6, and NF3 are emitted in smaller quantities but are significant
because many of them have extremely high global warming potentials (GWPs), and, in the cases of
PFCs, SF6, and NF3, very long atmospheric lifetimes. U.S. greenhouse gas emissions were partly offset
by carbon sequestration in managed forests, trees in urban areas, agricultural soils, landfilled yard
Trends in Greenhouse Gas Emissions and Removals 2-3
-------
trimmings, and coastal wetlands. These were estimated to offset 16.1 percent (1,000.5 MMT C02 Eq.) of
total gross emissions in 2023.
Table 2-1 provides information on trends in emissions and sinks from all U.S. anthropogenic sources and
sinks in weighted units of MMT C02 Eq., while unweighted gas emissions and sinks in kilotons (kt) are
provided in Table 2-2.
Table 2-1: Recent Trends in U.S. Greenhouse Gas Emissions and Sinks by Gas (MMT
C02 Eq.)
Gas/Source
1990
2005
2019
2020
2021
2022
2023
CO2
5,131.8
6,126.9
5,235.9
4,690.0
5,020.1
5,055.4
4,918.4
Fossil Fuel Combustion
4,752.2
5,744.1
4,852.6
4,342.3
4,654.6
4,702.8
4,559.4
Transportation
1,468.9
1,858.6
1,816.6
1,573.0
1,753.5
1,753.6
1,776.5
Electric Power Sector
1,820.0
2,400.1
1,606.7
1,439.6
1,540.9
1,531.7
1,414.2
Industrial
876.5
847.6
809.8
763.4
780.5
799.7
792.6
Residential
338.6
358.9
342.9
314.8
318.0
335.2
307.1
Commercial
228.3
227.1
251.7
229.3
237.5
259.2
244.2
U.S. Territories
20.0
51.9
24.8
22.3
24.1
23.5
24.9
Non-Energy Use of Fuels
99.1
125.0
106.5
97.9
111.7
101.7
107.1
Iron and Steel Production &
Metallurgical Coke Production
104.7
70.1
46.8
40.7
47.2
45.2
46.2
Cement Production
33.5
46.2
40.9
40.7
41.3
41.9
40.6
Natural Gas Systems
32.5
26.3
38.7
36.8
35.7
36.4
37.7
Petrochemical Production
20.1
26.9
28.5
27.9
30.7
28.8
30.5
Petroleum Systems
9.6
10.2
45.4
28.9
24.1
22.1
23.3
Incineration of Waste
12.9
13.3
12.9
12.9
12.5
12.5
12.4
Ammonia Production
14.4
10.2
12.4
12.3
11.5
11.9
12.2
Lime Production
11.7
14.6
12.1
11.3
11.9
12.2
11.5
Other Process Uses of
Carbonates
7.1
8.5
9.0
9.0
8.6
10.4
7.2
Urea Consumption for Non-
Agricultural Purposes
3.8
3.7
6.2
5.9
6.7
5.5
5.4
Liming
4.7
4.4
2.2
2.9
2.4
3.2
5.3
Urea Fertilization
2.4
3.5
4.9
5.0
5.1
5.2
5.3
Coal Mining
4.6
4.2
3.0
2.2
2.5
2.5
2.4
Non-EOR Carbon Dioxide
Utilization
1.5
1.4
2.4
2.8
2.9
2.8
2.1
Glass Production
2.3
2.4
1.9
1.9
2.0
2.0
1.8
Soda Ash Production
1.4
1.7
1.8
1.5
1.7
1.7
1.7
Ferroalloy Production
2.2
1.4
1.6
1.4
1.4
1.3
1.2
Aluminum Production
6.8
4.1
1.9
1.7
1.5
1.4
1.2
Titanium Dioxide Production
1.2
1.8
1.3
1.3
1.5
1.5
1.2
Zinc Production
0.6
1.0
1.0
1.0
1.0
0.9
0.9
Phosphoric Acid Production
1.5
1.3
0.9
0.9
0.9
0.8
0.9
Lead Production
0.5
0.6
0.5
0.5
0.5
0.5
0.5
2-4 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Gas/Source
1990
2005
2019
2020
2021
2022
2023
Carbide Production and
Consumption
0.2
0.2
0.2
0.2
0.2
0.2
0.2
CO2 Transport, Injection, and
Geological Storage
0.0
0.0
+
+
0.1
0.1
0.1
Abandoned Oil and Gas Wells
+
+
+
+
+
+
+
Substitution of Ozone Depleting
Substances
+
+
+
+
+
+
+
Magnesium Production and
Processing
0.1
+
+
+
+
+
+
Biomass and Biodiesel
Consumptions
237.9
245.4
332.0
294.7
302.0
304.4
300.5
International Bunker Fuelsb
103.6
113.3
113.6
69.6
80.2
98.2
96.2
CH4c
873.1
797.1
752.6
730.9
715.6
696.8
686.7
Enteric Fermentation
183.1
188.2
197.3
196.3
196.5
192.6
187.1
Natural Gas Systems
219.6
210.7
189.0
180.1
174.6
172.8
162.4
Landfills
197.8
147.7
128.2
122.6
120.7
118.7
119.5
Manure Management
39.1
55.0
66.7
66.9
66.4
64.7
65.1
Coal Mining
108.1
71.5
53.0
46.2
44.7
43.6
45.4
Petroleum Systems
50.0
48.4
50.8
50.6
45.1
36.3
38.0
Wastewater T reatment
22.7
22.7
21.1
21.0
20.7
20.9
21.1
Rice Cultivation
18.9
20.6
15.6
18.6
18.5
18.0
18.7
Stationary Combustion
9.7
8.8
9.8
7.9
7.9
8.7
8.8
Abandoned Oil and Gas Wells
7.8
8.2
8.5
8.5
8.6
8.5
8.5
Abandoned Underground Coal
Mines
8.1
7.4
6.6
6.5
6.2
6.1
6.1
Composting
0.4
2.1
2.5
2.6
2.6
2.6
2.6
Mobile Combustion
7.2
5.2
2.8
2.5
2.6
2.6
2.5
Field Burning of Agricultural
Residues
0.5
0.6
0.7
0.6
0.6
0.6
0.6
Anaerobic Digestion at Biogas
Facilities
+
+
+
+
+
+
+
Carbide Production and
Consumption
+
+
+
+
+
+
+
Ferroalloy Production
+
+
+
+
+
+
+
Iron and Steel Production &
Metallurgical Coke Production
+
+
+
+
+
+
+
Petrochemical Production
+
+
+
+
+
+
+
Incineration of Waste
+
+
+
+
+
+
+
International Bunker Fuelsb
0.2
0.1
0.1
0.1
0.1
0.1
0.1
N2Oc
407.8
424.8
416.4
391.4
398.4
387.5
387.0
Agricultural Soil Management
289.1
294.7
316.4
293.0
298.9
291.8
296.3
Wastewater T reatment
14.8
18.1
21.1
21.8
21.3
21.1
20.8
Stationary Combustion
22.3
30.5
22.1
20.5
22.0
22.6
19.6
Manure Management
13.4
15.2
16.8
16.9
17.1
17.0
16.8
Mobile Combustion
37.8
42.0
18.7
16.0
16.8
16.6
16.2
Nitric Acid Production
10.8
10.1
8.9
8.3
7.9
8.6
8.3
Trends in Greenhouse Gas Emissions and Removals 2-5
-------
Gas/Source
1990
2005
2019
2020
2021
2022
2023
N2O from Product Uses
3.8
3.8
3.8
3.8
3.8
3.8
3.8
Composting
0.3
1.5
1.8
1.8
1.8
1.8
1.8
Caprolactam, Glyoxal, and
Glyoxylic Acid Production
1.5
1.9
1.2
1.1
1.2
1.3
1.3
AdipicAcid Production
13.5
6.3
4.7
7.4
6.6
2.1
1.2
Incineration of Waste
0.4
0.3
0.4
0.3
0.4
0.3
0.3
Electronics Industry
+
0.1
0.2
0.3
0.3
0.3
0.3
Field Burning of Agricultural
Residues
0.2
0.2
0.2
0.2
0.2
0.2
0.2
Petroleum Systems
+
+
+
+
+
+
+
Natural Gas Systems
+
+
+
+
+
+
+
International Bunker Fuelsb
0.8
0.9
0.9
0.5
0.6
0.8
0.8
HFCs
47.8
125.0
175.8
177.8
184.3
189.5
191.0
Substitution of Ozone Depleting
Substancesd
0.3
102.7
169.7
173.7
179.9
184.8
189.0
Fluorochemical Production
47.3
22.2
5.8
3.8
4.0
4.3
1.7
Electronics Industry
0.2
0.2
0.3
0.3
0.4
0.3
0.3
Magnesium Production and
Processing
0.0
0.0
0.1
0.1
+
+
+
Other Product Manufacture and
Use
0.0
0.0
0.0
+
0.0
0.0
0.0
PFCs
39.7
10.3
7.3
6.6
6.3
6.5
5.8
Fluorochemical Production
17.7
4.1
3.0
2.5
2.6
2.8
2.7
Electronics Industry
2.5
3.0
2.7
2.6
2.7
2.8
2.4
Aluminum Production
19.3
3.1
1.4
1.4
0.9
0.8
0.5
Other Product Manufacture and
Use
0.1
0.1
0.1
0.1
0.1
0.1
0.2
Substitution of Ozone Depleting
Substances
0.0
+
+
+
+
+
+
Electrical Equipment
+
+
+
+
+
+
0.0
SFs
37.9
20.2
8.3
7.7
8.0
7.2
7.7
Electrical Equipment
24.6
11.8
6.0
5.5
5.5
4.9
5.1
Magnesium Production and
Processing
5.6
3.0
0.9
0.9
1.2
1.1
1.1
Other Product Manufacture and
Use
1.3
1.3
0.6
0.5
0.4
0.5
0.8
Electronics Industry
0.5
0.8
0.8
0.8
0.9
0.8
0.7
Fluorochemical Production
5.8
3.3
+
+
+
+
+
nf3
0.2
1.0
1.1
1.3
1.1
1.1
0.8
Electronics Industry
+
0.4
0.5
0.6
0.6
0.6
0.5
Fluorochemical Production
0.1
0.6
0.6
0.7
0.5
0.5
0.3
Other Product Manufacture and
Use
+
+
+
+
+
+
0.0
2-6 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Gas/Source
1990
2005
2019
2020
2021
2022
2023
Total Gross Emissions
(Sources)
6,538.3
7,505.3
6,597.4
6,005.7
6,333.8
6,344.1
6,197.3
LULUCF Emissionsc
59.1
71.8
63.2
82.6
81.0
68.6
60.6
CH4
54.4
60.9
56.1
69.0
67.8
59.6
54.7
N2O
4.7
10.9
7.0
13.7
13.1
9.0
5.9
LULUCF Carbon Stock Change6
(1,096.9)
(1,040.7)
(982.6)
(1,034.2)
(1,043.8)
(973.9)
(1,000.5)
LULUCF Sector Net Total'
(1,037.9)
(968.9)
(919.4)
(951.6)
(962.9)
(905.3)
(939.9)
Net Emissions (Sources and
Sinks)
5,500.4
6,536.4
5,678.0
5,054.2
5,371.0
5,438.7
5,257.4
+ Does not exceed 0.05 MMT C02 Eq.
a Emissions from biomass and biofuel consumption are not included specifically in Energy sector totals. Net carbon fluxes from
changes in biogenic carbon reservoirs are accounted for in the estimates for LULUCF.
b Emissions from international bunker fuels are not included in totals.
c LULUCF emissions of ChUand N20 are reported separately from gross emissions totals. LULUCF emissions include the ChU
and N20 emissions reported for peatlands remaining peatlands, forest fires, drained organic soils, grassland fires, and coastal
wetlands remaining coastal wetlands; CH4 emissions from land converted to coastal wetlands, flooded land remaining
flooded land, and land converted to flooded land; and N20 emissions from forest soils and settlement soils. Refer to Table 2-8
for a breakout of emissions and removals for LULUCF by gas and source category.
d Small amounts of PFC emissions from this source are included under HFCs due to confidential business information.
e LULUCF carbon stock change is the net carbon stock change from the following categories: forest land remaining forest land,
land converted to forest land, cropland remaining cropland, land converted to cropland, grassland remaining grassland, land
converted to grassland, wetlands remaining wetlands, land converted to wetlands, settlements remaining settlements, and
land converted to settlements. Refer to Table 2-8 for a breakout of emissions and removals for LULUCF by gas and source
category.
' The LULUCF sector net total is the net sum of all LULUCF CH4 and N20 emissions to the atmosphere plus LULUCF net carbon
stock changes.
Notes: Total (gross) emissions are presented without LULUCF. Net emissions are presented with LULUCF. Totals may not sum
due to independent rounding. Parentheses indicate negative values or sequestration.
Table 2-2: Recent Trends in U.S. Greenhouse Gas Emissions and Sinks by Gas (kt)
Gas/Source
1990
2005
2019
2020
2021
2022
2023
CO2
5,131,761
6,126,903
5,235,912
4,689,954
5,020,111
5,055,403
4,918,407
Fossil Fuel Combustion
4,752,234
5,744,138
4,852,647
4,342,309
4,654,629
4,702,769
4,559,379
Transportation
1,468,944
1,858,552
1,816,636
1,572,955
1,753,546
1,753,554
1,776,451
Electric Power Sector
1,819,951
2,400,057
1,606,721
1,439,566
1,540,933
1,531,678
1,414,177
Industrial
876,470
847,643
809,823
763,421
780,475
799,677
792,620
Residential
338,568
358,898
342,905
314,795
318,034
335,172
307,077
Commercial
228,293
227,130
251,749
229,264
237,528
259,182
244,161
U.S. Territories
20,010
51,857
24,813
22,308
24,114
23,506
24,893
Non-Energy Use of Fuels
99,104
124,988
106,487
97,881
111,718
101,697
107,069
Iron and Steel Production &
Metallurgical Coke Production
104,738
70,078
46,835
40,675
47,218
45,157
46,240
Cement Production
33,484
46,194
40,896
40,688
41,312
41,884
40,636
Natural Gas Systems
32,525
26,325
38,696
36,810
35,745
36,410
37,682
Petrochemical Production
20,075
26,882
28,483
27,926
30,656
28,788
30,540
Petroleum Systems
9,597
10,222
45,445
28,876
24,091
22,084
23,272
Incineration of Waste
12,900
13,254
12,948
12,921
12,476
12,484
12,425
Ammonia Production
14,404
10,234
12,388
12,335
11,458
11,945
12,211
Lime Production
11,700
14,552
12,112
11,299
11,870
12,208
11,548
Other Process Uses of Carbonates
7,103
8,472
8,973
9,012
8,583
10,383
7,163
Trends in Greenhouse Gas Emissions and Removals 2-7
-------
Gas/Source
1990
2005
2019
2020
2021
2022
2023
Urea Consumption for Non-
Agricultural Purposes
3,784
3,653
6,234
5,905
6,724
5,464
5,424
Liming
4,690
4,365
2,203
2,887
2,387
3,194
5,280
Urea Fertilization
2,417
3,504
4,950
5,031
5,105
5,193
5,258
Coal Mining
4,606
4,169
2,992
2,197
2,455
2,474
2,404
Non-EOR Carbon Dioxide Utilization
1,472
1,375
2,415
2,842
2,889
2,812
2,150
Glass Production
2,263
2,402
1,940
1,858
1,969
1,956
1,774
Soda Ash Production
1,431
1,655
1,792
1,461
1,714
1,704
1,723
Ferroalloy Production
2,152
1,392
1,598
1,377
1,426
1,327
1,245
Aluminum Production
6,831
4,142
1,880
1,748
1,541
1,446
1,237
Titanium Dioxide Production
1,195
1,755
1,340
1,340
1,541
1,541
1,233
Zinc Production
632
1,030
1,026
977
1,007
947
920
Phosphoric Acid Production
1,529
1,342
909
901
874
804
850
Lead Production
516
553
518
491
473
455
450
Carbide Production and
Consumption
243
213
175
154
172
210
183
CO2 Transport, Injection, and
Geological Storage
0
0
18
39
65
53
98
Abandoned Oil and Gas Wells
7
7
8
8
8
8
8
Substitution of Ozone Depleting
Substances
+
1
3
4
4
4
4
Magnesium Production and
Processing
129
4
2
3
3
3
2
Biomass and Biodiesel
Consumptiona
237,946
245,421
332,018
294,657
301,976
304,397
300,518
International Bunker Fuelsb
103,634
113,328
113,632
69,638
80,180
98,241
96,160
CH4c
31,183
28,468
26,877
26,102
25,558
24,884
24,524
Enteric Fermentation
6,539
6,722
7,045
7,010
7,017
6,878
6,683
Natural Gas Systems
7,842
7,525
6,751
6,431
6,236
6,173
5,802
Landfills
7,063
5,275
4,578
4,379
4,310
4,238
4,266
Manure Management
1,398
1,964
2,382
2,390
2,373
2,312
2,326
Coal Mining
3,860
2,552
1,892
1,648
1,595
1,558
1,623
Petroleum Systems
1,787
1,730
1,813
1,807
1,611
1,295
1,358
Wastewater T reatment
811
809
755
748
738
747
755
Rice Cultivation
677
735
558
664
661
642
667
Stationary Combustion
345
313
349
282
284
312
313
Abandoned Oil and Gas Wells
279
294
302
303
306
303
303
Abandoned Underground Coal
Mines
288
264
237
232
221
218
219
Composting
15
75
91
92
92
92
93
Mobile Combustion
258
187
101
90
91
92
91
Field Burning of Agricultural
Residues
19
23
23
22
22
22
22
Anaerobic Digestion at Biogas
Facilities
+
+
1
1
1
1
1
Carbide Production and
Consumption
1
+
+
+
+
+
+
2-8 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Gas/Source
1990
2005
2019
2020
2021
2022
2023
Ferroalloy Production
1
+
+
+
+
+
+
Iron and Steel Production &
Metallurgical Coke Production
1
1
+
+
+
+
+
Petrochemical Production
+
+
+
+
+
+
+
Incineration of Waste
+
+
+
+
+
+
+
International Bunker Fuelsb
7
5
4
3
3
3
3
N2Oc
1,539
1,603
1,571
1,477
1,503
1,462
1,460
Agricultural Soil Management
1,091
1,112
1,194
1,106
1,128
1,101
1,118
Wastewater T reatment
56
68
80
82
80
80
79
Stationary Combustion
84
115
84
77
83
85
74
Manure Management
50
57
63
64
65
64
63
Mobile Combustion
143
158
71
60
64
63
61
Nitric Acid Production
41
38
34
31
30
33
32
N2O from Product Uses
14
14
14
14
14
14
14
Composting
1
6
7
7
7
7
7
Caprolactam, Glyoxal, and Glyoxylic
Acid Production
6
7
5
4
5
5
5
AdipicAcid Production
51
24
18
28
25
8
4
Incineration of Waste
2
1
1
1
1
1
1
Electronics Industry
+
+
1
1
1
1
1
Field Burning of Agricultural
Residues
1
1
1
1
1
1
1
Petroleum Systems
+
+
+
+
+
+
+
Natural Gas Systems
+
+
+
+
+
+
+
International Bunker Fuelsb
3
3
3
2
2
3
3
HFCs
M
M
M
M
M
M
M
Substitution of Ozone Depleting
Substancesd
M
M
M
M
M
M
M
Fluorochemical Production
M
M
M
M
M
M
M
Electronics Industry
M
M
M
M
M
M
M
Magnesium Production and
Processing
0
0
+
+
+
+
+
Other Product Manufacture and Use
0
0
0
+
0
0
0
PFCs
M
M
M
M
M
M
M
Electronics Industry
M
M
M
M
M
M
M
Fluorochemical Production
M
M
M
M
M
M
M
Aluminum Production
M
M
M
M
M
M
M
SFe and PFCs from Other Product
Use
M
M
M
M
M
M
M
Substitution of Ozone Depleting
Substances
+
+
+
+
+
+
+
Electrical Equipment
+
+
+
+
+
+
0
SFs
1
1
+
+
+
+
+
Electrical Equipment
1
1
+
+
+
+
+
Magnesium Production and
Processing
+
+
+
+
+
+
+
Trends in Greenhouse Gas Emissions and Removals 2-9
-------
Gas/Source
1990
2005
2019
2020
2021
2022
2023
SFs and PFCs from Other Product
Use
+
+
+
+
+
+
+
Electronics Industry
+
+
+
+
+
+
+
Fluorochemical Production
+
+
+
+
+
+
+
nf3
+
+
+
+
+
+
+
Electronics Industry
+
+
+
+
+
+
+
Fluorochemical Production
+
+
+
+
+
+
+
Other Product Manufacture and Use
+
+
+
+
+
+
0
+ Does not exceed 0.5 kt.
M (Mixture of multiple gases)
a Emissions from biomass and biofuel consumption are not included specifically in Energy sector totals. Net carbon fluxes from
changes in biogenic carbon reservoirs are accounted for in the estimates for LULUCF.
b Emissions from international bunker fuels are not included in totals.
c LULUCF emissions of LULUCF ChU and N20 are reported separately from gross emissions totals. Refer to Table 2-8 for a
breakout of emissions and removals for LULUCF by gas and source category.
d Small amounts of PFC emissions from this source are included under HFCs due to confidential business information.
Notes: Totals by gas may not sum due to independent rounding. Parentheses indicate negative values or sequestration.
Emissions and Sinks by Inventory Sector
Emissions and removals of all gases can be summed from each source and sink category into a set of
five sectors defined by the national inventory reporting guidelines and methodological framework
provided by the Intergovernmental Panel on Climate Change (IPCC). Figure 2-4 and Table 2-3 illustrate
that over the 34-year period of 1990 to 2023, total emissions from the Energy and Waste sectors
decreased by 6.2 percent (331.6 MMTC02 Eq.) and 29.7 percent (70.1 MMTC02 Eq.), respectively
Emissions from Industrial Processes and Product Use and Agriculture grew by 4.6 percent (16.8 MMT
C02 Eq.) and 8.0 percent (43.9 MMT C02 Eq.), respectively Over the same period, total carbon
sequestration in the LULUCF sector decreased by 8.8 percent (96.4 MMT C02), and emissions from the
LULUCF sector increased by 2.6 percent (1.5 MMT C02 Eq.). The overall net flux from LULUCF (i.e., the
net sum of all CH4 and N20 emissions to the atmosphere plus LULUCF net carbon stock changes in
units of MMTC02 Eq.) decreased by 9.4 percent (97.9 MMT C02 Eq.) from 1990 levels.
2-10 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Figure 2-4: U.S. Greenhouse Gas Emissions and Removals by Inventory Sector
¦ LULUCF (emissions) ¦ Agriculture
9,000 ¦ Waste ¦ Energy
¦ Industrial Processes and Product Use ¦ LULUCF (sinks)
8 000 — ^et: Em'ss'ons (including LULUCF Flux)
Table 2-3: Recent Trends in U.S. Greenhouse Gas Emissions and Sinks by Inventory
Sector/Category (MMT C02 Eq.)
Inventory Sector/Category
1990
2005
2019
2020
2021
2022
2023
Energy
5,381.9
6,356.2
5,420.9
4,860.2
5,170.1
5,196.2
5,050.4
Fossil Fuel Combustion
4,752.2
5,744.1
4,852.6
4,342.3
4,654.6
4,702.8
4,559.4
Natural Gas Systems
252.1
237.0
227.7
216.9
210.4
209.3
200.1
Non-Energy Use of Fuels
99.1
125.0
106.5
97.9
111.7
101.7
107.1
Petroleum Systems
59.6
58.7
96.2
79.5
69.2
58.4
61.3
Coal Mining
112.7
75.6
56.0
48.3
47.1
46.1
47.8
Stationary Combustion
32.0
39.3
31.9
28.4
30.0
31.3
28.3
Mobile Combustion
45.0
47.2
21.5
18.5
19.4
19.2
18.8
Incineration of Waste
13.3
13.6
13.3
13.3
12.8
12.8
12.8
Abandoned Oil and Gas Wells
7.8
8.2
8.5
8.5
8.6
8.5
8.5
Abandoned Underground Coal Mines
8.1
7.4
6.6
6.5
6.2
6.1
6.1
CO2 Transport, Injection, and Geological Storage
0
0
+
+
0.1
0.1
0.1
Biomass and Biodiesel Consumption
237.9
245.4
332.0
294.7
302.0
304.4
300.5
International Bunker Fuels0
104.6
114.3
114.6
70.3
80.9
99.1
97.0
Industrial Processes and Product Use
368.9
374.7
380.8
375.3
390.9
389.6
385.7
Substitution of Ozone Depleting Substances
0.3
102.7
169.7
173.7
179.9
184.9
189.0
Iron and Steel Production & Metallurgical Coke
Production
104.8
70.1
46.8
40.7
47.2
45.2
46.2
Cement Production
33.5
46.2
40.9
40.7
41.3
41.9
40.6
Petrochemical Production
20.1
26.9
28.5
27.9
30.7
28.8
30.5
Ammonia Production
14.4
10.2
12.4
12.3
11.5
11.9
12.2
Trends in Greenhouse Gas Emissions and Removals 2-11
-------
Inventory Sector/Category
1990
2005
2019
2020
2021
2022
2023
Lime Production
11.7
14.6
12.1
11.3
11.9
12.2
11.5
Nitric Acid Production
10.8
10.1
8.9
8.3
7.9
8.6
8.3
Other Process Uses of Carbonates
7.1
8.5
9.0
9.0
8.6
10.4
7.2
Urea Consumption for Non-Agricultural Purposes
3.8
3.7
6.2
5.9
6.7
5.5
5.4
Electrical Equipment
24.6
11.8
6.0
5.5
5.5
4.9
5.1
Fluorochemical Production
71.0
30.0
9.3
7.0
7.1
7.6
4.7
Electronics Industry
3.3
4.5
4.5
4.5
4.9
4.8
4.2
N2O from Product Uses
3.8
3.8
3.8
3.8
3.8
3.8
3.8
Non-EOR Carbon Dioxide Utilization
1.5
1.4
2.4
2.8
2.9
2.8
2.1
Glass Production
2.3
2.4
1.9
1.9
2.0
2.0
1.8
Soda Ash Production
1.4
1.7
1.8
1.5
1.7
1.7
1.7
Aluminum Production
26.1
7.2
3.3
3.2
2.5
2.2
1.7
Caprolactam, Glyoxal, and Glyoxylic Acid
Production
1.5
1.9
1.2
1.1
1.2
1.3
1.3
Ferroalloy Production
2.2
1.4
1.6
1.4
1.4
1.3
1.3
Titanium Dioxide Production
1.2
1.8
1.3
1.3
1.5
1.5
1.2
AdipicAcid Production
13.5
6.3
4.7
7.4
6.6
2.1
1.2
Magnesium Production and Processing
5.7
3.0
1.0
0.9
1.2
1.1
1.1
Other Product Manufacture and Use
1.5
1.5
0.8
0.7
0.5
0.6
1.0
Zinc Production
0.6
1.0
1.0
1.0
1.0
0.9
0.9
Phosphoric Acid Production
1.5
1.3
0.9
0.9
0.9
0.8
0.9
Lead Production
0.5
0.6
0.5
0.5
0.5
0.5
0.5
Carbide Production and Consumption
0.3
0.2
0.2
0.2
0.2
0.2
0.2
Agriculture
551.5
582.5
620.8
600.4
605.8
593.3
595.4
Agricultural Soil Management
289.1
294.7
316.4
293.0
298.9
291.8
296.3
Enteric Fermentation
183.1
188.2
197.3
196.3
196.5
192.6
187.1
Manure Management
52.5
70.2
83.5
83.8
83.6
81.7
81.9
Rice Cultivation
18.9
20.6
15.6
18.6
18.5
18.0
18.7
Liming
4.7
4.4
2.2
2.9
2.4
3.2
5.3
Urea Fertilization
2.4
3.5
4.9
5.0
5.1
5.2
5.3
Field Burning of Agricultural Residues
0.7
0.8
0.9
0.8
0.8
0.8
0.8
Waste
235.9
192.0
174.8
169.7
167.0
165.1
165.8
Landfills
197.8
147.7
128.2
122.6
120.7
118.7
119.5
Wastewater T reatment
37.5
40.7
42.3
42.7
41.9
42.0
41.9
Composting
0.7
3.6
4.3
4.4
4.4
4.4
4.4
Anaerobic Digestion at Biogas Facilities
+
+
+
+
+
+
+
Total Gross Emissions6 (Sources)
6,538.3
7,505.3
6,597.4
6,005.7
6,333.8
6,344.1
6,197.3
LULUCF Sector Net Total'
(1,037.9)
(968.9)
(919.4)
(951.6)
(962.9)
(905.3)
(939.9)
Forest land
(1,152.9)
(1,036.3)
(971.3)
(1,001.9)
(984.9)
(931.5)
(977.1)
Cropland
49.6
4.4
12.1
20.5
3.0
3.5
5.0
Grassland
59.8
46.4
49.4
40.9
31.1
34.6
43.6
Wetlands
45.3
42.8
40.4
40.4
40.3
40.4
40.4
Settlements
(39.5)
(26.2)
(50.0)
(51.4)
(52.4)
(52.3)
(51.9)
Net Emissions (Sources and Sinks)g
5,500.4
6,536.4
5,678.0
5,054.2
5,371.0
5,438.7
5,257.4
+ Does not exceed 0.05 MMT C02 Eq.
2-12 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
a Includes ChU and N20 emissions from fuel combustion.
b Emissions from biomass and biofuel consumption are not included specifically in summing Energy sector totals. Net carbon
fluxes from changes in biogenic carbon reservoirs are accounted for in the estimates for LULUCF.
c Emissions from international bunker fuels are not included in totals.
d SF6 and PFCs from other product use category includes the use of HFCs and NF3.
e Total emissions without LULUCF.
f LULUCF emissions of ChUand N20 are reported separately from gross emissions totals. LULUCF emissions include the ChU
and N20 emissions reported for peatlands remaining peatlands, forest fires, drained organic soils, grassland fires, and coastal
wetlands remaining coastal wetlands; CH4 emissions from land converted to coastal wetlands, flooded land remaining
flooded land, and land converted to flooded land; and N20 emissions from forest soils and settlement soils. Refer to Table 2-8
for a breakout of emissions and removals for LULUCF by gas and source category.
g Net emissions with LULUCF.
Notes: Total (gross) emissions are presented without LULUCF. Net emissions are presented with LULUCF. Totals may not sum
due to independent rounding. Parentheses indicate negative values or sequestration.
Energy
Emissions from energy-related activities come from two main categories: 1) emissions associated with
fuel use (i.e., fossil fuel combustion, non-energy use of fossil fuels and waste incineration), and 2)
fugitive emissions mainly from coal, natural gas, and oil production. Energy emissions also include
some categories that are not added to Energy sector totals but are instead presented as memo items,
including international bunker fuels and biomass emissions. Energy-related activities, primarily fossil
fuel combustion, accounted for the vast majority of U.S. C02 emissions from 1990 through 2023. Fossil
fuel combustion is the largest source of energy-related emissions, with C02 being the primary gas
emitted (see Figure 2-5). Due to their relative importance, fossil fuel combustion-related C02 emissions
are considered in detail in the Energy chapter (see Chapter 3).
Figure 2-5: Trends in Energy Sector Greenhouse Gas Sources
8,000
7,000
CO2 Transport, Injection, and Geological Storage
Incineration of Waste
U.S Territories Fossil Fuel Combustion
I Non-Energy Use of Fuels
I Commerical Fossil Fuel Combustion
I Residential Fossil Fuel Combustion
I Fugitive Emissions
I Industrial Fossil Fuel Combustion
Transportation Fossil Fuel Combustion
I Electric Power Fossil Fuel Combustion
2; Lfi
6,000
J"L J
in ^
A S
3,000
2,000
1,000
0'-i«Nro^-Lnv£>rvoocT>Oi-irMro^-Lnv£>rs«.cocT>Oi-i(Nro^|-Lnv£>rs«.cocnoi-irs4ro
oio^o^cr>c^(Tio^cr>c^oooooooooO'—It—It—It—It—It—It—in-JCNirsirM
cnc^cnc^cnaicr>aicr>cnoooooooooooooooooooooooo
Trends in Greenhouse Gas Emissions and Removals 2-13
-------
In 2023, 82.6 percent of the energy used in the United States on a Btu basis was produced through the
combustion of fossil fuels. The remaining 17.4 percent came from other energy sources such as
hydropower, biomass, nuclear, wind, and solar energy. A discussion of specific trends related to C02
and other greenhouse gas emissions from energy use is presented here with more detail in the Energy
chapter. Energy-related activities are also responsible for CH4 and N20 emissions (39.6 percent and 9.3
percent of gross total U.S. emissions of each gas, respectively).1 Table 2-4 presents greenhouse gas
emissions from the Energy sector by source and gas.
Table 2-4: Emissions from Energy by Gas (MMT C02 Eq.)2
Gas/Source
1990
2005
2019
2020
2021
2022
2023
Percent
Change
Since
1990
CO2
4,911.0
5,923.1
5,059.2
4,521.0
4,841.2
4,878.0
4,742.3
-3%
Fossil Fuel Combustion
4,752.2
5,744.1
4,852.6
4,342.3
4,654.6
4,702.8
4,559.4
-4%
Transportation
1,468.9
1,858.6
1,816.6
1,573.0
1,753.5
1,753.6
1,776.5
21%
Electricity Generation
1,820.0
2,400.1
1,606.7
1,439.6
1,540.9
1,531.7
1,414.2
-22%
Industrial
876.5
847.6
809.8
763.4
780.5
799.7
792.6
-10%
Residential
338.6
358.9
342.9
314.8
318.0
335.2
307.1
-9%
Commercial
228.3
227.1
251.7
229.3
237.5
259.2
244.2
7%
U.S. Territories
20.0
51.9
24.8
22.3
24.1
23.5
24.9
24%
Non-Energy Use of Fuels
99.1
125.0
106.5
97.9
111.7
101.7
107.1
8%
Natural Gas Systems
32.5
26.3
38.7
36.8
35.7
36.4
37.7
16%
Petroleum Systems
9.6
10.2
45.4
28.9
24.1
22.1
23.3
142%
Incineration of Waste
12.9
13.3
12.9
12.9
12.5
12.5
12.4
-4%
Coal Mining
4.6
4.2
3.0
2.2
2.5
2.5
2.4
-48%
CO2 Transport, Injection, and
Geological Storage
0.0
0.0
+
+
0.1
0.1
0.1
100%
Abandoned Oil and Gas Wells
+
+
+
+
+
+
+
13%
Biomass-Wooda
215.2
206.9
216.7
189.5
191.5
194.3
187.7
-13%
International Bunker Fuels1
103.6
113.3
113.6
69.6
80.2
98.2
96.2
-7%
Biofuels-Ethanola
4.2
22.9
82.6
71.8
79.1
79.6
80.7
1,810%
Biofuels-BiodieseP
0.0
0.9
17.1
17.7
16.1
15.6
18.2
100%
Biofuels-MSWa
18.5
14.7
15.7
15.6
15.3
14.9
13.9
-25%
CH4
410.4
360.2
320.4
302.3
289.6
278.7
271.9
-34%
Natural Gas Systems
219.6
210.7
189.0
180.1
174.6
172.8
162.4
-26%
Coal Mining
108.1
71.5
53.0
46.2
44.7
43.6
45.4
-58%
Petroleum Systems
50.0
48.4
50.8
50.6
45.1
36.3
38.0
-24%
Stationary Combustion
9.7
8.8
9.8
7.9
7.9
8.7
8.8
-9%
Abandoned Oil and Gas Wells
7.8
8.2
8.5
8.5
8.6
8.5
8.5
9%
1 The contribution of energy non-CCb emissions is based on gross totals, so it excludes LULUCF methane (ChU) and
nitrous oxide (N2O) emissions. The contribution of energy-related CH4 and N2O including LULUCF non-CC>2 emissions, is
36.6 percent and 9.2 percent, respectively.
2 The full time series data is in CSV format available at https://www.epa.gov/ghgemissions/inventory-us-greenhouse-gas-
emissions-and-sinks.
2-14 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Gas/Source
1990
2005
2019
2020
2021
2022
2023
Percent
Change
Since
1990
Abandoned Underground Coal Mines
8.1
7.4
6.6
6.5
6.2
6.1
6.1
-24%
Mobile Combustion
7.2
5.2
2.8
2.5
2.6
2.6
2.5
-65%
Incineration of Waste
+
+
+
+
+
+
+
-19%
International Bunker Fuelsb
0.2
0.1
0.1
0.1
0.1
0.1
0.1
-55%
n2o
60.53
72.87
41.26
36.91
39.25
39.55
36.17
-40%
Stationary Combustion
22.3
30.5
22.1
20.5
22.0
22.6
19.6
-12%
Mobile Combustion
37.8
42.0
18.7
16.0
16.8
16.6
16.2
-57%
Incineration of Waste
0.4
0.3
0.4
0.3
0.4
0.3
0.3
-19%
Natural Gas Systems
+
+
+
+
+
+
+
57%
Petroleum Systems
+
+
+
+
+
+
+
73%
International Bunker Fuelsb
0.8
0.9
0.9
0.5
0.6
0.8
0.8
0%
Total
5,381.9
6,356.2
5,420.9
4,860.2
5,170.1
5,196.2
5,050.4
-6%
+ Does not exceed 0.05 MMT C02 Eq.
a Emissions from biomass and biofuel consumption are not included specifically in Energy sector totals. Net carbon fluxes from
changes in biogenic carbon reservoirs are accounted for in the estimates for LULUCF.
b Emissions from international bunker fuels are not included in totals. These values are presented for informational purposes
only, in line with the 2006IPCC Guidelines and the national inventory reporting guidelines.
Note: Totals may not sum due to independent rounding.
Fossil Fuel Combustion CO2 Emissions
As the largest contributor to U.S. greenhouse gas emissions, C02 from fossil fuel combustion has
accounted for approximately 73.6 percent of C02-equivalent total gross emissions on average across
the time series. Within the United States, fossil fuel combustion accounted for 92.7 percent of C02
emissions in 2023. Emissions from this source category include C02 associated with the combustion of
fossil fuels (coal, natural gas, and petroleum) for energy use. Fossil fuel combustion C02 emissions
decreased by 4.1 percent (192.9 MMT C02 Eq.) from 1990 to 2023 and were responsible for most of the
decrease in national emissions during this period. Similarly, C02 emissions from fossil fuel combustion
have decreased by 20.6 percent (1,184.8 MMTC02 Eq.) since 2005. From 2022 to 2023, these emissions
decreased by 3.0 percent (143.4 MMTC02 Eq.).
Historically, changes in emissions from fossil fuel combustion have been the main factor influencing
U.S. emission trends. Changes in C02 emissions from fossil fuel combustion since 1990 are affected by
many long-term and short-term factors, including population and economic growth, energy price
fluctuations and market trends, technological changes, carbon intensity of energy fuel choices, and
seasonal temperatures. On an annual basis, the overall consumption and mix of fossil fuels in the
United States fluctuates in response to changes in general economic conditions, overall energy prices,
the relative price of different fuels, weather, and the availability of non-fossil alternatives. For example,
coal consumption for electric power is influenced by factors such as the relative price of coal and
alternative sources, the ability to switch fuels, and longer-term trends in coal markets. Fossil fuel
combustion C02 emissions also depend on the type of fuel consumed or energy used and its carbon
intensity. Producing a unit of heat or electricity using natural gas instead of coal, for example, reduces
C02 emissions because of the lower carbon content of natural gas (see Table 3-12 in Chapter 3 for more
Trends in Greenhouse Gas Emissions and Removals 2-15
-------
detail on electricity generation by source and see Table A-21 in Annex 2.1 for more detail on the carbon
content coefficient of different fossil fuels).
Petroleum use is another major driver of C02 emissions from fossil fuel combustion, particularly in the
transportation sector, which has represented the largest source of C02 emissions from fossil fuel
combustion since 2017. Emissions from petroleum consumption for transportation (including bunker
fuels) increased by 1.2 percent from 2022 to 2023. Fuel economy of light-duty vehicles is an important
factor in transportation sector C02 emissions trends. The decline in new light-duty vehicle fuel economy
between 1990 and 2004 reflected the increasing market share of light-duty trucks, which grew from
about 29.6 percent of new vehicle sales in 1990 to 48.0 percent in 2004. Starting in 2005, average new
vehicle fuel economy began to increase while vehicle miles traveled (VMT) by light-duty vehicles grew
only modestly for much of the period, slowing the rate of increase of C02 emissions.
Trends in C02 emissions from fossil fuel combustion by end-use sector are presented in Table 2-5 and
Figure 2-6 based on the underlying U.S. energy consumer data collected by the U.S. Energy Information
Administration (EIA). Figure 2-7 further describes trends in direct and indirect C02 emissions from fossil
fuel combustion by end-use sector. Estimates of C02 emissions from fossil fuel combustion are
calculated from these EIA "end-use sectors" based on total fuel consumption and appropriate fuel
properties described below.3
• Transportation. ElA's fuel consumption data for the transportation sector consists of all vehicles
whose primary purpose is transporting people and/or goods from one physical location to
another.
• Electric Power. ElA's fuel consumption data for the electric power sector are composed of
electricity-only and combined-heat-and-power (CHP) plants within the North American Industry
Classification System (NAICS) 22 category whose primary business is to sell electricity, or
electricity and heat, to the public. (Non-utility power producers are included in this sector as
long as they meet the electric power sector definition.)
• Industry. EIA statistics for the industrial sector include fossil fuel consumption that occurs in the
fields of manufacturing, agriculture, mining, and construction. ElA's fuel consumption data for
the industrial sector consist of all facilities and equipment used for producing, processing, or
assembling goods. (EIA includes generators that produce electricity and/or useful thermal
output primarily to support on-site industrial activities in this sector.)
• Residential. ElA's fuel consumption data for the residential sector consist of living quarters for
private households.
• Commercial. ElA's fuel consumption data for the commercial sector consist of service-providing
facilities and equipment from private and public organizations and businesses. (EIA includes
generators that produce electricity and/or useful thermal output primarily to support the
activities at commercial establishments in this sector.)
3 Additional analysis and refinement of the EIA data is further explained in the Energy chapter of this report.
2-16 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Table 2-5: C02 Emissions from Fossil Fuel Combustion by End-Use Sector (MMT C02
Eq.)
End-Use Sector
1990
2005
2019
2020
2021
2022
2023
Transportation
1,472.0
1,863.3
1,820.9
1,576.4
1,757.5
1,758.0
1,781.5
Combustion
1,468.9
1,858.6
1,816.6
1,573.0
1,753.5
1,753.6
1,776.5
Electricity
3.0
4.7
4.2
3.5
4.0
4.5
5.0
Industrial
1,562.9
1,584.0
1,275.3
1,173.1
1,225.7
1,236.9
1,197.1
Combustion
876.5
847.6
809.8
763.4
780.5
799.7
792.6
Electricity
686.4
736.3
465.5
409.7
445.2
437.2
404.5
Residential
931.3
1,214.9
927.1
860.7
891.1
901.6
815.6
Combustion
338.6
358.9
342.9
314.8
318.0
335.2
307.1
Electricity
592.7
856.0
584.2
545.9
573.0
566.5
508.5
Commercial
766.0
1,030.1
804.5
709.7
756.2
782.7
740.3
Combustion
228.3
227.1
251.7
229.3
237.5
259.2
244.2
Electricity
537.7
803.0
552.7
480.5
518.7
523.5
496.1
U.S. Territoriesa
20.0
51.9
24.8
22.3
24.1
23.5
24.9
Total
4,752.2
5,744.1
4,852.6
4,342.3
4,654.6
4,702.8
4,559.4
Electric Power
1,820.0
2,400.1
1,606.7
1,439.6
1,540.9
1,531.7
1,414.2
a Fuel consumption by U.S. Territories (i.e., American Samoa, Guam, Puerto Rico, U.S. Virgin Islands, Wake Island, and other
outlying U.S. Pacific Islands) is included in this report.
Notes: Combustion-related emissions from electric power are allocated based on aggregate national electricity use by each
end-use sector. Totals may not sum due to independent rounding.
Trends in Greenhouse Gas Emissions and Removals 2-17
-------
Figure 2-6: Trends in C02 Emissions from Fossil Fuel Combustion by End-Use Sector
and Fuel Type
. 2,000
&
O
u
1,000
2,000
O
u
1,000
2,000
O
u
1,000
Coal ¦ Geothermal ¦ Natural Gas ¦ Petroleum
U.S. Territories
Commercial
Residential
1995 2002 2009 2016 2023
tr
LU
rsi
o
u
2,000
1,000
2,000
O
u
1,000
2,000
O
u
1,000
Industrial
Electric Power
Transportation
1995 2002 2009 2016 2023
Notes: Fossil fuel combustion for electric power also includes emissions of less than 0.5 MMT C02 Eq. from geothermal-based
generation. Although not technically a fossil fuel, geothermal energy-related C02 emissions are included for reporting purposes.
The source of C02 is non-condensable gases in subterranean heated water.
2-18 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Figure 2-7: Trends in End-Use Sector Emissions of C02 from Fossil Fuel Combustion
¦ Direct Fossil Fuel Combustion ¦ Indirect Fossil Fuel Combustion
2,000
iff l'500
(N
8 1,000
i-
z
z 500
0
2,000
iff 1,500
IN
8 1,000
I-
z
z 500
0
2,000
& 1,500
LU
IN
8 1,000
h
z
Z 500
0
Electric power was the second largest end-use emitter of C02 in 2023 (surpassed by transportation in
2017); electric power generators used 29.9 percent of U.S. energy from fossil fuels and emitted 31.0
percent of the C02 from fossil fuel combustion in 2023. C02 emissions from fossil fuel combustion in
the electric power sector decreased by 7.7 percent between 2022 and 2023 due to changes in the mix of
electric generation resources. Between 2022 and 2023 total electricity generation decreased by 1.1
percent, electricity generation from coal decreased by 18.8 percent, electricity generation from natural
gas increased by 7.4 percent, and electricity generation from renewables decreased by 0.4 percent.
Changes in electricity demand and the carbon intensity of fuels used for electric power generation have
a significant impact on C02 emissions. Carbon dioxide emissions from fossil fuel combustion from the
electric power sector have decreased by 22.3 percent since 1990, and the carbon intensity of the
electric power sector, in terms of C02 Eq. per QBtu input, has decreased by 21.8 percent during that
same timeframe. This trend is shown below in Figure 2-8. Overall C02 emissions from electric power
generation in 2023 decreased by 41 percent from a high in 2007 (see Figure 2-6, reflecting the continued
shift in the share of electric power generation from coal to natural gas and renewables since 2005.
Carbon dioxide emissions from coal combustion for electric power generation gradually increased
between 1990 and 2007, then began to decrease at a faster rate from 2008 to 2023. Carbon dioxide
U.S. Territories
Industrial
Commercial
Residential
2,000
& 1,500
8 1,000
500
0
2,000
& 1,500
8 1,000
500
0
Transportation
1995 2002 2009 2016 2023
1995 2002 2009 2016 2023
Trends in Greenhouse Gas Emissions and Removals 2-19
-------
emissions from natural gas combustion for electric power generation steadily increased between 1990
and 2023.
Figure 2-8: Electric Power Generation (Billion kWh) and Emissions (MMT C02 Eq.)
- Total Emissions (MMT CO2 Eq.) [Right Axis]
4,500
4,000
. 3,500
| 3,000
2,500
cD
c
OJ
(J
iD 2,000
o
S 1,500
cu
Nuclear Generation (Billion kWh)
I Renewable Generation (Billion kWh)
Petroleum Generation (Billion kWh)
I Coal Generation (Billion kWh)
I Natural Gas Generation (Billion kWh)
1,000
500
3,500
3,000
2,500 uj
(N
o
u
2,000
1,500 m
1,000
500
Electric power C02 emissions can also be allocated to the end-use sectors that use electricity, as
presented in Table 2-5. With electricity C02 emissions allocated to end-use sectors, the transportation
end-use sector represents the largest source of fossil fuel combustion emissions, accounting for
1,781.5 MMT C02 Eq. in 2023 or 39.1 percent of total C02 emissions from fossil fuel combustion, a 1.3
percent increase since 2022. The industrial end-use sector accounted for 26.3 percent of C02
emissions from fossil fuel combustion when including allocated electricity emissions, a decrease of 3.2
percent since 2022. From 2022 to 2023, total electricity use in the industrial sector decreased by 1.2
percent due to a decrease in total industrial production and manufacturing output. The residential and
commercial end-use sectors accounted for 17.9 and 16.2 percent, respectively, of C02 emissions from
fossil fuel combustion when including allocated electricity emissions. Both of these end-use sectors
were heavily reliant on electricity for meeting building-related energy needs, with electricity use for
lighting, heating, air conditioning, and operating appliances contributing 62.3 and 67.0 percent of
emissions from the residential and commercial end-use sectors, respectively. From 2022 to 2023, a
decrease in heating degree days (10.4 percent) decreased energy demand for heating in the residential
and commercial sectors; also, a 5.2 percent decrease in cooling degree days compared to 2022
decreased demand for air conditioning in the residential and commercial sectors. Total C02 emissions
from the residential and commercial end-use sectors when including allocated electricity emissions
decreased by 9.5 and 5.4 percent since 2022, respectively.
2-20 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Other Energy Sector Trends
Energy sector emissions decreased by 2.8 percent since 2022 and decreased by 6.2 percent since 1990.
Other notable trends in emissions from energy source categories (Figure 2-6 and Figure 2-7) over the 34-
year period from 1990 through 2023 included the following:
• Emissions (CH4, C02, and N20) from oil and gas systems decreased by 16.1 percent (50.3 MMT
C02 Eq.) since 1990 and decreased by 2.3 percent (6.2 MMT C02 Eq.) from 2022 to 2023. Natural
gas systems CH4 emissions have decreased by 26.0 percent (57.1 MMT C02 Eq.) since 1990,
due to a decrease in emissions from distribution, transmission and storage, and processing. The
decrease in distribution emissions is due mainly to reduced emissions from pipeline and
distribution station leaks, and the decrease in transmission and storage emissions is due mainly
to reduced compressor station emissions (including emissions from compressors and leaks).
Over the same time period (i.e., since 1990), methane emissions from the naturalgas
production segment decreased due to increased gathering and boosting emissions. Between
2022 and 2023, methane emissions from natural gas systems decreased 6.0 percent, due to a
decrease in emissions from production segment pneumatic controllers. Petroleum systems CH4
emissions decreased by 24.0 percent (12.0 MMT C02 Eq.) since 1990 and increased 4.9 percent
between 2022 and 2023. This increase is due primarily to decreases in emissions from offshore
platforms, tanks, and pneumatic controllers. Carbon dioxide emissions from natural gas and
petroleum systems increased by 44.7 percent (18.8 MMT C02) from 1990 to 2023 and increased
by4.2 percent between 2022 and 2023. This increase since 1990 is due primarily to increases in
the production segment, where emissions from associated gas flaring, tanks, and
miscellaneous production flaring have increased over time.
• Methane emissions from coal mining decreased by 58.0 percent (62.6 MMT C02 Eq.) from 1990
through 2023 primarily due to a decrease in the number of active mines and annual coal
production over the time period. Methane emissions from coal mining increased 4.1 percent
between 2022 and 2023. The number of mines increased, but overall coal production has
decreased.
• Nitrous oxide emissions from mobile combustion decreased by 57.0 percent (21.5 MMT C02
Eq.)from 1990 through 2023 and by 2.1 percent (0.4 MMTC02 Eq.) between 2022 and 2023,
primarily as a result of national vehicle criteria pollutant emissions standards and emission
control technologies for on-road vehicles.
• Nitrous oxide emissions from stationary combustion were the third largest source of
anthropogenic N20 emissions in 2023, accounting for 5.1 percent of N20 emissions and 0.4
percent of total gross U.S. greenhouse gas emissions in 2023. Stationary combustion emissions
peaked in 2007 and have steadily decreased since then.
• Carbon dioxide emissions from non-energy uses of fossil fuels increased by 8.0 percent (8.0
MMTC02 Eq.) from 1990 through 2023 and 5.3 percent (5.4 MMTC02 Eq.) between 2022 and
2023. Emissions from non-energy uses of fossil fuels were 107.1 MMT C02 Eq. in 2023, which
constituted 2.2 percent of total national C02 emissions, approximately the same proportion as
in 1990.
Trends in Greenhouse Gas Emissions and Removals 2-21
-------
• Carbon dioxide emissions from incineration of waste decreased by 3.7 percent (0.5 MMT C02
Eq.) from 1990 through 2023, as the volume of scrap tires and other fossil carbon-containing
materials in waste decreased.
Industrial Processes and Product Use
Greenhouse gases can be generated and emitted by industry in two different ways. First, they are
generated and emitted as the byproducts of many non-energy-related industrial activities. For example,
industrial processes can chemically or physically transform raw materials, which often release waste
gases such as C02, CH4, N20, and fluorinated gases (e.g., HFC-23). In the case of byproduct emissions,
the emissions are generated by an industrial process itself and are not directly a result of energy
consumed during the process.
Second, industrial manufacturing processes and use by end-consumers also release HFCs, PFCs, SF6,
and NF3and other man-made compounds. In addition to the use of HFCs and some PFCs as substitutes
for ODS, fluorinated compounds such as HFCs, PFCs, SF6, NF3, and others are also emitted through use
by a number of other industrial sources in the United States. These industries include the electronics
industry, electrical equipment, and magnesium metal production and processing. In addition, N20 is
used in and emitted by the electronics industry and anesthetic and aerosol applications, PFCs and SF6
are emitted from other product use, and C02 is consumed and emitted through various end-use
applications.
Emission sources in the Industrial Processes and Product Use (IPPU) sector accounted for 6.2 percent
of U.S. greenhouse gas emissions in 2023. Emissions from the IPPU sector increased by 4.6 percent
from 1990 to 2023. The use of HFCs as substitutes for ODS is the largest source of emissions in this
sector, contributing 49.0 percent of IPPU emissions in 2023 and driving growth since 1990. From 2022 to
2023, total emissions from IPPU decreased 1.0 percent. Emissions from ferroalloy production
decreased 6.2 percent, while emissions from electrical equipment and petrochemical production
increased 4.5 percent and 6.1 percent, respectively. Zinc production emissions decreased by 2.9
percent, and electronics industry emissions decreased by 12.5 percent. Figure 2-9 presents
greenhouse gas emissions from IPPU by source category.
2-22 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Figure 2-9: Trends in Industrial Processes and Product Use Sector Greenhouse Gas
Sources
o
u
500
450
400
350
300
250
200
150
100
50
0
I Electronics Industry
Other Product Manufacture and Use
I Mineral Industry
I Metal Industry
Chemical Industry
I Substitution of Ozone Depleting Substances
. CO f*-s
r* lo in
8 "
cr>
cn
*-• en
(N Ol f\J
3
CTi
ro
o
o-)
o
oo
fN CO • •
en
m m
m co
r-v
r-N oo cr>
co vo
ro m
SI 00
ft ^
m S § s
IC ro
o
cn
CTi
^-Hr\jroTj-Ln*jDrsvooo^O'-Hr\imTrijnvDr^coc^OT-HrNfn^LnvDis>.cocTiOi-—It—It—I--—It—I-!—li—li—I i—I i—I (~\J PsJ C\J cn
cr>cj^cr.cncna^cr>cncnoooooooooooooooooooooooo
!-!HT-lHHHHHH(NfN(NfN(N(NfM(NN(Nrsl(N(N(Nrslf\rMfMfNfM(NfN(NrM
Table 2-6: Emissions from industrial Processes and Product Use (MMT C02 Eq.)
1990
2005
2019
2020
2021
2022
2023
Percent
Change
Since
1990
C02
213.7
195.9
169.5
161.0
171.4
169.0
165.5
-22.5%
Iron and Steel Production &
Metallurgical Coke Production
104.7
70.1
46.8
40.7
47.2
45.2
46.2
-55.9%
Iron and Steel Production
99.1
66.2
43.8
38.3
44.0
42.2
43.3
-56.4%
Metallurgical Coke Production
5.6
3.9
3.0
2.3
3.2
3.0
3.0
-46.7%
Cement Production
33.5
46.2
40.9
40.7
41.3
41.9
40.6
21.4%
Petrochemical Production
20.1
26.9
28.5
27.9
30.7
28.8
30.5
52.1%
Ammonia Production
14.4
10.2
12.4
12.3
11.5
11.9
12.2
-15.2%
Lime Production
11.7
14.6
12.1
11.3
11.9
12.2
11.5
-1.3%
Other Process Uses of Carbonates
7.1
8.5
9.0
9.0
8.6
10.4
7.2
0.9%
Urea Consumption for Non-
Agricultural Purposes
3.8
3.7
6.2
5.9
6.7
5.5
5.4
43.3%
Non-EOR Carbon Dioxide
Utilization
1.5
1.4
2.4
2.8
2.9
2.8
2.1
46.0%
Glass Production
2.3
2.4
1.9
1.9
2.0
2.0
1.8
-21.6%
Soda Ash Production
1.4
1.7
1.8
1.5
1.7
1.7
1.7
20.4%
Ferroalloy Production
2.2
1.4
1.6
1.4
1.4
1.3
1.2
-42.1%
Trends iri Greenhouse Gas Emissions and Removals 2-23
-------
1990
2005
2019
2020
2021
2022
2023
Percent
Change
Since
1990
Aluminum Production
6.8
4.1
1.9
1.7
1.5
1.4
1.2
-81.9%
Titanium Dioxide Production
1.2
1.8
1.3
1.3
1.5
1.5
1.2
3.1%
Zinc Production
0.6
1.0
1.0
1.0
1.0
0.9
0.9
45.6%
Phosphoric Acid Production
1.5
1.3
0.9
0.9
0.9
0.8
0.9
-44.4%
Lead Production
0.5
0.6
0.5
0.5
0.5
0.5
0.5
-12.8%
Carbide Production and
Consumption
0.2
0.2
0.2
0.2
0.2
0.2
0.2
-24.7%
Substitution of Ozone Depleting
Substances®
+
+
+
+
+
+
+
30,504.1%
Magnesium Production and
Processing
0.1
+
+
+
+
+
+
-98.2%
CH4
0.1
+
+
+
+
+
+
-48.6%
Carbide Production and
Consumption
+
+
+
+
+
+
+
-38.5%
Ferroalloy Production
+
+
+
+
+
+
+
-48.6%
Iron and Steel Production &
Metallurgical Coke Production
+
+
+
+
+
+
+
-66.8%
Petrochemical Production
+
+
+
+
+
+
+
-11.9%
N20
29.6
22.2
18.7
20.8
19.7
16.1
14.9
-49.8%
Nitric Acid Production
10.8
10.1
8.9
8.3
7.9
8.6
8.3
-22.6%
N2O from Product Uses
3.8
3.8
3.8
3.8
3.8
3.8
3.8
-0.4%
Caprolactam, Glyoxal, and
Glyoxylic Acid Production
1.5
1.9
1.2
1.1
1.2
1.3
1.3
-10.5%
AdipicAcid Production
13.5
6.3
4.7
7.4
6.6
2.1
1.2
-91.5%
Electronics Industry
+
0.1
0.2
0.3
0.3
0.3
0.3
665.8%
HFCs
47.8
125.0
175.8
177.8
184.3
189.5
191.0
299.8%
Substitution of Ozone Depleting
Substances®
0.3
102.7
169.7
173.7
179.9
184.8
189.0 74,673.9%
Fluorochemical Production
47.3
22.2
5.8
3.8
4.0
4.3
1.7
-96.5%
Electronics Industry
0.2
0.2
0.3
0.3
0.4
0.3
0.3
54.8%
Magnesium Production and
Processing
0.0
0.0
0.1
0.1
+
+
+
100.0%
Other Product Manufacture and
Use
0.0
0.0
0.0
+
0.0
0.0
0.0
0.0%
PFCs
39.7
10.3
7.3
6.6
6.3
6.5
5.8
-85.5%
Fluorochemical Production
17.7
4.1
3.0
2.5
2.6
2.8
2.7
-84.8%
Electronics Industry
2.5
3.0
2.7
2.6
2.7
2.8
2.4
-4.3%
Aluminum Production
19.3
3.1
1.4
1.4
0.9
0.8
0.5
-97.6%
Other Product Manufacture and
Use
0.1
0.1
0.1
0.1
0.1
0.1
0.2
22.5%
Substitution of Ozone Depleting
Substances
0.0
+
+
+
+
+
+
100.0%
Electrical Equipment
+
+
+
+
+
+
0.0
-100.0%
2-24 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
1990
2005
2019
2020
2021
2022
2023
Percent
Change
Since
1990
SF6
37.9
20.2
8.3
7.7
8.0
7.2
7.7
-79.6%
Electrical Equipment
24.6
11.8
6.0
5.5
5.5
4.9
5.1
-79.4%
Magnesium Production and
Processing
5.6
3.0
0.9
0.9
1.2
1.1
1.1
-80.0%
Other Product Manufacture and
Use
1.3
1.3
0.6
0.5
0.4
0.5
0.8
-38.3%
Electronics Industry
0.5
0.8
0.8
0.8
0.9
0.8
0.7
40.6%
Fluorochemical Production
5.8
3.3
+
+
+
+
+
-100.0%
NF3
0.2
1.0
1.1
1.3
1.1
1.1
0.8
335.8%
Fluorochemical Production
0.1
0.6
0.6
0.7
0.5
0.5
0.3
117.6%
Electronics Industry
+
0.4
0.5
0.6
0.6
0.6
0.5
1,034.7%
Other Product Manufacture and
Use
+
+
+
+
+
+
+
0.0%
Total
368.9
374.7
380.8
375.3
390.9
389.6
385.7
4.6%
+ Does not exceed 0.05 MMT C02 Eq.
a Small amounts of PFC emissions from this source are included under HFCs due to confidential business information.
Note: Totals may not sum due to independent rounding.
IPPU sector emissions decreased 1.0 percent since 2022 but have increased 4.6 percent since 1990.
Some significant trends in U.S. emissions from IPPU source categories over the 34-year period from
1990 through 2023 included the following:
• HFC and PFC emissions resulting from the substitution of ODS (e.g., chlorofluorocarbons
[CFCs]) increased from small amounts in 1990 to 189.0 MMT C02 Eq. in 2023 (an increase of
74,680 percent).
• Combined C02 and CH4 emissions from iron and steel production and metallurgical coke
production increased by 2.4 percent from 2022 to 2023 to 46.2 MMT C02 Eq. and declined by
55.9 percent (58.5 MMT C02 Eq.) from 1990 through 2023, due to restructuring of the industry.
The trend in the United States has been a shift toward fewer integrated steel mills and more
electric arc furnaces (EAFs). EAFs use scrap steel as their main input and generally have lower
on-site emissions.
• Carbon dioxide emissions from petrochemical production increased by 52.1 percent between
1990 and 2023, from 20.1 MMTC02 Eq. to 30.5 MMTC02 Eq. The increase in emissions is largely
driven by the production of ethylene more than doubling over that period.
• Carbon dioxide emissions from ammonia production have decreased by 15.2 percent (2.2 MMT
C02 Eq.) since 1990. Ammonia production relies on natural gas as both a feedstock and a fuel,
and as such, market fluctuations and volatility in natural gas prices affect the production of
ammonia from year to year. Emissions from ammonia production have increased since 2016,
due to the addition of new ammonia production facilities and new production units at existing
facilities. Agricultural demands continue to drive demand for nitrogen fertilizers and the need for
new ammonia production capacity.
Trends in Greenhouse Gas Emissions and Removals 2-25
-------
• Carbon dioxide emissions from cement production increased by 21.4 percent (7.2 MMT C02 Eq.)
from 1990 through 2023. Emissions rose from 1990 through 2006 and then fell until 2009, due to
a decrease in demand for construction materials during the economic recession. Since 2010,
C02 emissions from cement production have risen by 29.2 percent.
• HFC, PFC, SF6, and NF3 emissions from fluorochemical production decreased by 93.4 percent
(66.4 MMT C02 Eq.) from 1990 to 2023 due to a reduction in the HFC-23 emission rate from
HCFC-22 production (kg HFC-23 emitted/kg HCFC-22 produced), the imposition of emissions
controls at production facilities, and a decrease in SF6 production due to the cessation of
production at the major SF6 production facility in 2010.
• PFC emissions from aluminum production decreased by 97.6 percent (18.8 MMT C02 Eq.) from
1990 to 2023, due to both industry emission reduction efforts and lower domestic aluminum
production.
• SF6 emissions from electrical equipment decreased by 79.4 percent (19.6 MMT C02 Eq.) from
1990 to 2023 due to a sharp increase in the price of SF6 during the 1990s and industry emission
reduction efforts.
• HFC, PFC, SF6, and NF3 emissions from use in electronics increased 27.2 percent (0.9 MMT C02
Eq.) from 1990 to 2023. Industrial growth, increasing chip complexity, and the adoption of
emissions reductions technologies contributed to the fluctuation in electronics industry
emissions.
Agriculture
Agricultural activities contribute directly to emissions of greenhouse gases through a variety of
processes, including the following source categories: enteric fermentation in domestic livestock,
livestock manure management, rice cultivation, agricultural soil management, liming, urea fertilization,
and field burning of agricultural residues. Methane and N20 are the primary greenhouse gases emitted
by agricultural activities, with small amounts of C02 also emitted.4 Carbon stock changes from
agricultural soils are included in the LULUCF sector.
In 2023, agricultural activities were responsible for emissions of 595.4 MMT C02 Eq., or 9.6 percent of
total U.S. greenhouse gas emissions. Agricultural soil management activities, such as the application of
synthetic and organic fertilizers, deposition of livestock manure, and growing N-fixing plants, were the
largest contributors to agricultural-related emissions (49.8 percent) and were the largest source of U.S.
N20 emissions in 2023, accounting for 76.6 percent. Methane emissions from enteric fermentation and
manure management represented 27.2 percent and 9.5 percent of total CH4 emissions from
anthropogenic activities, respectively, in 2023. Enteric fermentation is the largest anthropogenic source
of CH4 emissions, while manure management is the fourth largest anthropogenic source of CH4 and N20
emissions. Carbon dioxide emissions from the application of crushed limestone and dolomite (i.e., soil
liming) and urea fertilization represented 0.2 percent of total C02 emissions from anthropogenic
activities. Liming and urea fertilization are the only sources of C02 emissions reported in the Agriculture
sector. All other C02 emissions and removals (e.g., carbon stock changes from the management of
4 The contribution of agriculture non-CC>2 emissions is based on gross totals and excludes LULUCF methane (CFU) and
nitrous oxide (N2O) emissions. The contribution of agriculture CH4 and N2O including LULUCF non-C02 emissions, is
40.5 percent and 48.3 percent, respectively.
2-26 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
croplands) are included in the LULUCF sector. Figure 2-10 and Table 2-7 illustrate agricultural
greenhouse gas emissions by source and gas.
Figure 2-10: Trends in Agriculture Sector Greenhouse Gas Sources
I Field Burning of Agricultural Residues
I Urea Fertilization
Liming
I Rice Cultivation
I Manure Management
I Enteric Fermentation
I Agricultural Soil Management
<3- <3-
LO LO
00
r-N r-N
VD LO
m t
00 CO CO
LO LO LO
rvj
o
Table 2-7: Emissions from Agriculture (MMT C02 Eq.)
Gas/Source
1990
2005
2019
2020
2021
2022
2023
Percent
Change
Since
1990
C02
7.1
7.9
7.2
7.9
7.5
8.4
10.5
48.3%
Liming
4.7
4.4
2.2
2.9
2.4
3.2
5.3
12.6%
Urea Fertilization
2.4
3.5
4.9
5.0
5.1
5.2
5.3
117.5%
CH4
241.7
264.4
280.2
282.4
282.0
275.9
271.6
12.3%
Enteric Fermentation
183.1
188.2
197.3
196.3
196.5
192.6
187.1
2.2%
Manure Management
39.1
55.0
66.7
66.9
66.4
64.7
65.1
66.4%
Rice Cultivation
18.9
20.6
15.6
18.6
18.5
18.0
18.7
-1.4%
Field Burning of Agricultural Residues
0.5
0.6
0.7
0.6
0.6
0.6
0.6
14.8%
N2O
302.7
310.2
333.4
310.1
316.3
309.0
313.3
3.5%
Agricultural Soil Management
289.1
294.7
316.4
293.0
298.9
291.8
296.3
2.5%
Manure Management
13.4
15.2
16.8
16.9
17.1
17.0
16.8
25.5%
Field Burning of Agricultural Residues
0.2
0.2
0.2
0.2
0.2
0.2
0.2
15.8%
Total
551.5
582.5
620.8
600.4
605.8
593.3
595.4
8.0%
Note: Totals may not sum due to independent rounding.
Trends in Greenhouse Gas Emissions and Removals 2-27
-------
Agriculture sector emissions increased by 0.4 percent since 2022 and increased by 8.0 percent since
1990. Some significant trends in U.S. emissions from Agriculture source categories (Figure 2-10) over the
34-year time series from 1990 through 2023 included the following:
• Annual N20 emissions from agricultural soils fluctuated between 1990 and 2023, and overall
emissions were 2.5 percent (7.2 MMT C02 Eq.) higher in 2023 than in 1990. Year-to-year
fluctuations are largely a reflection of annual variation in weather patterns, synthetic fertilizer
use, and crop production.
• Enteric fermentation emissions increased from 1990 to 2023, largely due to increasing cattle
population. For example, emissions increased from 1990 to 1995 and then generally decreased
from 1996 to 2004, mainly due to fluctuations in beef cattle populations and increased
digestibility of feed for feed lot cattle. Emissions decreased again from 2008 to 2014 as beef
cattle populations again decreased. Emissions increased from 2014 to 2023, consistent with an
increase in beef cattle population over those same years. CH4 emissions from enteric
fermentation decreased by 2.9 percent (5.5 MMTC02 Eq.) from 2022 to 2023, however, largely
driven by a decrease in beef cattle populations.
• Emissions from manure management decreased by 56.0 percent between 1990 and 2023. This
includes an increase of 66.4 percent (26.0 MMT C02 Eq.) for CH4 and an increase of 25.5 percent
(3.4 MMT C02 Eq.) for N20. The majority of the increase observed in CH4 emissions resulted from
swine and dairy cattle manure, where emissions increased by 39.8 and 109.1 percent,
respectively, from 1990 to 2023. From 2022 to 2023, CH4 emissions from manure management
increase by 0.6 percent, mainly due to minor shifts in the animal populations and the resultant
effects on manure management system allocations.
• Liming emissions increased by 65.3 percent relative to 2022 and increased by 12.6 percent (0.6
MMT C02Eq.) relative to 1990, while urea fertilization emissions increased by 1.2 percent
relative to 2022 and 117.5 percent (2.8 MMT C02 Eq.) relative to 1990.
2-28 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Land Use, Land-Use Change, and Forestry
When humans alter the terrestrial biosphere through land use, changes in land use, and land
management practices, they also influence the carbon stock fluxes on these lands and cause emissions
of CH4 and N20. Overall, managed land is a net sink for C02 (i.e. carbon sequestration) in the United
States. The primary driver of fluxes on managed lands is from management of forest lands, but also
includes trees in settlements (i.e., urban areas), afforestation, conversion of forest lands to settlements
and croplands, the management of croplands and grasslands, flooded lands, and the landfillingofyard
trimmings and food scraps. The main drivers for net forest sequestration include net forest growth,
increasing forest area, and a net accumulation of carbon stocks in harvested wood pools. The net
sequestration in settlements remaining settlements is driven primarily by carbon stock gains in urban
forests (i.e., settlement trees) through net tree growth and increased urban area, as well as long-term
accumulation of carbon in landfills from additions of yard trimmings and food scraps.
The LULUCF sector in 2023 resulted in a net increase in carbon stocks (i.e., net C02 removals) of 1,000.5
MMT C02 Eq. (Table 2-8).5 This represents an offset of 16.1 percent of total (i.e., gross) greenhouse gas
emissions in 2023. Emissions of CH4 and N20 from LULUCF activities in 2023 were 60.6 MMT C02 Eq.
and represented 1.2 percent of net greenhouse gas emissions.6 Between 1990 and 2023, total net
carbon sequestration in the LULUCF sector decreased by 8.8 percent, primarily due to a decrease in the
rate of net carbon accumulation in forests and cropland remaining cropland, as well as an increase in
C02 emissions from land converted to settlements. Total flux, accounting for both removals and
emissions, was a net removal of 939.9 MMT C02 Eq., a 3.8 percent increase in removals from 2022.
Flooded land remaining flooded land was the largest source of CH4 emissions from LULUCF and the fifth
largest source overall of net CH4 emissions in 2023, totaling 45.8 MMT C02 Eq. (1,635.9 kt of CH4).
Coastal wetland remaining coastal wetland was the second largest source of CH4 emissions from
LULUCF in 2023, totaling 4.3 MMT C02 Eq. (154.6 kt of CH4). Settlement soils were the largest source of
N20 emissions from LULUCF in 2023, totaling 2.5 MMT C02 Eq. (9.6 kt of N20). Figure 2-11 and Table 2-8
illustrate LULUCF emissions and removals by land-use category and gas.
5 LULUCF carbon stock change is the net carbon stock change from the following categories: forest land remaining forest
land, land converted to forest land, cropland remaining cropland, land converted to cropland, grassland remaining
grassland, land converted to grassland, wetlands remaining wetlands, land converted to wetlands, settlements
remaining settlements, and land converted to settlements.
6 LULUCF emissions include the CFU and N2O emissions reported for peatlands remaining peatlands, forest fires, drained
organic soils, grassland fires, and coastal wetlands remaining coastal wetlands; CH4 emissions from flooded land
remaining flooded land, land converted to flooded land, and land converted to coastal wetlands; and N2O emissions
from forest soils and settlement soils.
Trends in Greenhouse Gas Emissions and Removals 2-29
-------
Figure 2-11: Trends in Emissions and Removals (Net Flux) from Land Use, Land-Use
Change, and Forestry
¦ Forest Land Remaining Forest Land ¦ Settlements Remaining Settlements ¦ Land Converted to Settlements
¦ Land Converted to Forest Land Cropland Remaining Cropland ¦ Wetlands Remaining Wetlands
600 ¦ Land Converted to Cropland Grassland Remaining Grassland ™ Net Emissions (Sources and Sinks)
Land Converted to Wetlands Land Converted to Grassland
400
co Ln co
n t y)
O O o
vo u: m
8 S ®
s
200
5 Ln *-
- cn
r-i cr>
co r-t k m oi
fN PO LD
iS" -200
o
u
-400
HI II
-600
-800
-1,000
-1,200
OT-Hr\jm<3-Ln\£>r-»ooCT>OT-irvjro^rLnv£>r^co<3-Lnvor^coCT'>0'-Hrvjpi-i
cricr>criCT»o^c7vcr>o^cr>o^ooooooooooi-H->-i-^Hi-ii-i^-i'-i'-Hi-i'-ifMrMrvjrNi
o^cr.chcr.oooooooooooooooooooooooo
HH^HHHHHHHf\|(NfNfNNfNlN(NfNNfNfN(NfNfN(NfSfN(N(NOJMrNN
Table 2-8: U.S. Greenhouse Gas Emissions and Removals (Net Flux) from Land Use,
Land-Use Change, and Forestry (MMT C02 Eq.)
Land-Use Category
1990
2005
2019
2020
2021
2022
2023
Percent
Change
Since
1990
Forest Land Remaining Forest Land
(1,049.3)
(932.8)
(867.4)
(898.0)
(881.0)
(827.6)
(873.3)
-16.8%
Changes in Forest Carbon Stocks0
(1,054.9)
(950.0)
(877.1)
(926.5)
(907.9)
(842.4)
(880.0)
-16.6%
Non-C02 Emissions from Forest Firesb
5.4
16.7
9.3
28.0
26.4
14.3
6.2
14.8%
N2O Emissions from Forest Soils0
0.1
0.4
0.4
0.4
0.4
0.4
0.4
455.1%
Non-C02 Emissions from Drained
Organic Soilsd
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.0%
Land Converted to Forest Land
(103.6)
(103.6)
(103.9)
(103.8)
(103.8)
(103.8)
(103.8)
0.2%
Changes in Forest Carbon Stocks®
(103.6)
(103.6)
(103.9)
(103.8)
(103.8)
(103.8)
(103.8)
0.2%
Cropland Remaining Cropland
1.0
(31.0)
(19.3)
(8.7)
(31.9)
(31.6)
(30.5)
-3,036.6%
Changes in Mineral and Organic Soil
Carbon Stocks
1.0
(31.0)
(19.3)
(8.7)
(31.9)
(31.6)
(30.5)
-3,036.6%
2-30 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Land-Use Category
1990
2005
2019
2020
2021
2022
2023
Percent
Change
Since
1990
Land Converted to Cropland
48.5
35.5
31.4
29.2
34.9
35.0
35.6
-26.7%
Changes in all Ecosystem Carbon
Stocks'
48.5
35.5
31.4
29.2
34.9
35.0
35.6
-26.7%
Grassland Remaining Grassland
24.2
24.5
28.5
16.8
11.2
13.7
22.7
-6.1%
Changes in Mineral and Organic Soil
Carbon Stocks
24.0
23.7
28.2
15.8
10.2
13.1
22.0
-8.2%
Non-C02 Emissions from Grassland
Fires8
0.2
0.8
0.3
1.1
0.9
0.6
0.7
227.6%
Land Converted to Grassland
35.6
21.9
20.9
24.1
19.9
20.9
20.9
-41.3%
Changes in all Ecosystem Carbon
Stocks'
35.6
21.9
20.9
24.1
19.9
20.9
20.9
-41.3%
Wetlands Remaining Wetlands
38.5
40.9
39.7
39.7
39.7
39.7
39.7
3.3%
Changes in Organic Soil Carbon Stocks
in Peatlands
1.1
1.1
0.6
0.6
0.5
0.6
0.6
-42.7%
Non-C02 Emissions from Peatlands
Remaining Peatlands
+
+
+
+
+
+
+
-44.4%
Changes in Biomass, DOM, and Soil
Carbon Stocks in Coastal Wetlands
(10.8)
(10.1)
(11.1)
(11.1)
(11.1)
(11.1)
(11.1)
2.9%
ChU Emissions from Coastal Wetlands
Remaining Coastal Wetlands
4.2
4.2
4.3
4.3
4.3
4.3
4.3
3.6%
N2O Emissions from Coastal Wetlands
Remaining Coastal Wetlands
0.1
0.2
0.1
0.1
0.1
0.1
0.1
14.9%
CH4 Emissions from Flooded Land
Remaining Flooded Land
43.9
45.5
45.8
45.8
45.8
45.8
45.8
4.2%
Land Converted to Wetlands
6.8
1.9
0.7
0.7
0.7
0.7
0.6
-90.6%
Changes in Biomass, DOM, and Soil
Carbon Stocks in Land Converted to
Coastal Wetlands
0.5
0.5
{+)
{+)
{+)
{+)
{+)
-94.8%
CH4 Emissions from Land Converted to
Coastal Wetlands
0.3
0.3
0.2
0.2
0.2
0.2
0.2
-42.1%
Changes in Land Converted to Flooded
Land
3.4
0.7
0.3
0.3
0.3
0.3
0.3
-92.1%
CH4 Emissions from Land Converted to
Flooded Land
2.7
0.5
0.2
0.2
0.2
0.2
0.2
-93.0%
Settlements Remaining Settlements
(109.1)
(115.2)
(131.4)
(131.7)
(132.1)
(132.1)
(131.7)
20.8%
Changes in Organic Soil Carbon Stocks
9.9
10.1
14.6
15.1
15.6
16.0
16.4
66.2%
Changes in Settlement Tree Carbon
Stocks
(96.5)
(117.0)
(135.4)
(136.6)
(137.6)
(138.4)
(139.0)
44.0%
N2O Emissions from Settlement Soilsh
2.1
3.1
2.5
2.5
2.5
2.5
2.5
23.1%
Changes in Yard Trimming and Food
Scrap Carbon Stocks in Landfills
(24.5)
(11.4)
(13.1)
(12.8)
(12.5)
(12.3)
(11.7)
-52.2%
Land Converted to Settlements
69.5
89.0
81.4
80.3
79.7
79.8
79.8
14.9%
Changes in all Ecosystem Carbon
Stocks'
69.5
89.0
81.4
80.3
79.7
79.8
79.8
14.9%
Trends in Greenhouse Gas Emissions and Removals 2-31
-------
Land-Use Category
1990
2005
2019
2020
2021
2022
2023
Percent
Change
Since
1990
LULUCF Emissions1
59.1
71.8
63.2
82.6
81.0
68.6
60.6
2.6%
CH4
54.4
60.9
56.1
69.0
67.8
59.6
54.7
0.5%
n2o
4.7
10.9
7.0
13.7
13.1
9.0
5.9
26.7%
LULUCF Carbon Stock Change1
(1,096.9)
(1,040.7)
(982.6)
(1,034.2)
(1,043.8)
(973.9)
(1,000.5)
-8.8%
LULUCF Sector Net Total"
(1,037.9)
(968.9)
(919.4)
(951.6)
(962.9)
(905.3)
(939.9)
-9.4%
+ Absolute value does not exceed 0.05 MMT C02 Eq.
a Includes the net changes to carbon stocks stored in all forest ecosystem pools (estimates include carbon stock changes from
drained organic soils from both forest land remaining forest land and land converted to forest land) and harvested wood
products.
b Estimates include emissions from fires on both forest land remaining forest land and land converted to forest land.
c Estimates include emissions from N fertilizer additions on both forest land remaining forest land and land converted to forest
land.
d Estimates include ChUand N20 emissions from drained organic soils on both forest land remaining forest land and land
converted to forest land. Carbon stock changes from drained organic soils are included with the forest land remaining forest
land forest ecosystem pools.
e Includes the net changes to carbon stocks stored in all forest ecosystem pools.
f Includes changes in mineral and organic soil carbon stocks for all land use conversions to cropland, grassland, and
settlements. Also includes aboveground/belowground biomass, dead wood, and litter carbon stock changes for conversion of
forest land to cropland, grassland, and settlements.
g Estimates include ChUand N20 emissions from fires on both grassland remaining grassland and land converted to grassland.
h Estimates include N20 emissions from N fertilizer additions on both settlements remaining settlements and land converted to
settlements because it is not possible to separate the activity data at this time.
LULUCF carbon stock change includes any carbon stock gains and losses from all land use and land-use conversion categories.
' LULUCF emissions subtotal includes the ChU and N20 emissions reported for peatlands remaining peatlands, forest fires,
drained organic soils, grassland fires, and coastal wetlands remaining coastal wetlands; ChU emissions from flooded land
remainingflooded land, and land converted to flooded land, and land converted to coastal wetlands; and N20 emissions from
forest soils and settlement soils. Emissions values are included in land-use category rows.
kThe LULUCF sector net totalis the net sum of all LULUCF ChUand N20 emissions to the atmosphere plus LULUCF net carbon
stock changes in units of MMT C02 Eq.
Notes: Totals may not sum due to independent rounding. Parentheses indicate net sequestration.
Overall CH4 and N20 emissions from LULUCF decreased by 11.6 percent from 2022 and increased by
2.6 percent since 1990, while total net sequestration decreased by 8.8 percent since 1990 and
increased 2.7 percent from 2022. Other trends from 1990 to 2023 in fluxes from LULUCF categories
(Figure 2-11) over the 34-year period included the following:
• Annual carbon sequestration by forest land (i.e., annual carbon stock accumulation in the five
ecosystem carbon pools and harvested wood products for forest land remaining forest land and
land converted to forest land) has decreased by 15.1 percent since 1990. This is primarily due to
decreased carbon stock gains in land converted to forest land and the harvested wood products
pools within forest land remaining forest land.
• Annual carbon sequestration from settlements remaining settlements (which includes organic
soils, settlement trees, and landfilled yard trimmings and food scraps) has increased by 20.8
percent over the period from 1990 to 2023. This is primarily due to an increase in urbanized land
area in the United States with tree growth.
• Annual emissions from land converted to settlements increased by 14.9 percent from 1990 to
2023 due primarily to carbon stock losses from forest land converted to settlements and
mineral soils carbon stocks from grassland converted to settlements.
2-32 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Waste
Waste management and treatment activities are sources of CH4 and N20 emissions (see Figure 2-12 and
Table 2-9). Overall, emission sources accounted for in the Waste chapter generated 165.8 MMT C02Eq.,
or 2.7 percent of total U.S. greenhouse gas emissions in 2023. In 2023, landfills were the largest source
of waste emissions, accounting for 72.0 percent of waste-related emissions. Landfills are also the third-
largest source of U.S. anthropogenic CH4 emissions, generating 119.5 MMT C02 Eq. and accounting for
17.4 percent of total U.S. CH4 emissions in 2023.7 Additionally, wastewater treatment generated
emissions of 41.9 MMT C02 Eq. and accounted for 25.3 percent of waste emissions, 3.1 percent of U.S.
CH4 emissions, and 5.4 percent of U.S. N20 emissions in 2023. Emissions of CH4 and N20 from
composting are also accounted for in this chapter, generating emissions of 2.6 MMT C02 Eq. and 1.8
MMT C02 Eq., accounting for 1.6 and 1.1 percent of Waste sector emissions, respectively. Anaerobic
digestion at biogas facilities generated CH4 emissions of less than 0.05 MMT C02 Eq., accounting for
less than 0.05 percent of emissions from the Waste sector.
Figure 2-12: Trends in Waste Sector Greenhouse Gas Sources
CTi O CO Gi
m n J n n ^ q
01 fN fN rM rsi m ^
rsi
™ °
o
i-H T
t; Ln
rsi en 2 ci
S °
en
cm ro
00 CO
fS rn fN
,-h ^ r-»
o
r-»
S ^ m lO
2 VO vo
50
0
OHNntullDNOOOlO-la>CT»CT»C^CT»
-------
Composting
0.4
2.1
2.5
2.6
2.6
2.6
2.6
507.7%
Anaerobic Digestion at Biogas Facilities
+
+
+
+
+
+
+
1,427.9%
N20
15.1
19.5
22.9
23.6
23.1
22.9
22.6
50.3%
Wastewater T reatment
14.8
18.1
21.1
21.8
21.3
21.1
20.8
41.0%
Composting
0.3
1.5
1.8
1.8
1.8
1.8
1.8
507.7%
Total
235.9
192.0
174.8
169.7
167.0
165.1
165.8
-29.7%
+ Does not exceed 0.05 MMT C02 Eq.
Note: Totals may not sum due to independent rounding.
Waste sector emissions increased by 0.5 percent since 2022 and decreased by 29.7 percent since 1990.
Some notable trends in U.S. emissions from Waste source categories (Figure 2-12) over the 34-year
period from 1990 through 2023 included the following:
• Net CH4 emissions from landfills decreased by 78.3 MMT C02 Eq. (39.6 percent), with small
increases occurring in interim years. This downward trend in emissions coincided with
increased landfill gas collection and control systems, and a reduction of decomposable
materials (i.e., paper and paperboard, food scraps, and yard trimmings) discarded in municipal
solid waste (MSW) landfills over the time series.
• CH4and N20 emissions from wastewater treatment decreased by 6.9 percent (1.6 MMT C02 Eq.)
and increased by 41.0 percent (6.0 MMT C02 Eq.), respectively. Methane emissions from
domestic wastewater treatment have decreased since 1999 due to decreasing percentages of
wastewater being treated in anaerobic systems, including reduced use of on-site septic systems
and central anaerobic treatment systems. N20 emissions from wastewater treatment processes
gradually increased across the time series as a result of increasing U.S. population and protein
consumption.
• Combined CH4 and N20 emissions from commercial composting have increased by 507.7
percent (3.7 MMT C02 Eq.) since 1990. The growth in composting since the 1990s is largely due
to growing legislation by state and local governments discouraging the disposal of yard
trimmings and food waste in landfills and increased collection of yard trimmings.
2.2 Emissions and Sinks by Economic Sector
This report also characterizes gross emissions according to commonly used economic sector
categories: residential, commercial, industry, transportation, electric power, and agriculture. All
emissions from U.S. Territories are reported together as their own end-use sector in this characterization
due to a lack of specific consumption data for the individual end-use sectors. See Box 2-1 for more
information on how economic sectors are defined. For more information on trends in the LULUCF
sector, see Section 2.1.
Using this categorization, transportation activities accounted for the largest portion (29.4 percent) of
total U.S. greenhouse gas emissions in 2023. Emissions from electric power accounted for the second
largest portion (23.5 percent), while emissions from industry accounted for the third-largest portion
(23.0 percent) of total U.S. greenhouse gas emissions in 2023. Emissions from industry have in general
declined over the past decade due to a number of factors, including structural changes in the U.S.
2-34 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
economy (i.e., shifts from a manufacturing-based to a service-based economy), fuel switching, and
efficiency improvements.
The remaining 24.2 percent of U.S. greenhouse gas emissions were contributed by, in order of
magnitude, the agriculture, commercial, and residential sectors, plus emissions from U.S. Territories.
Activities related to agriculture accounted for 10.5 percent of emissions; unlike other economic sectors,
agricultural sector emissions were dominated by N20 emissions from agricultural soil management and
CH4 emissions from enteric fermentation, rather than C02 from fossil fuel combustion. An increasing
amount of carbon is stored in agricultural soils each year, but this carbon sequestration is assigned to
the LULUCF sector rather than the agriculture economic sector. The commercial and residential sectors
accounted for roughly 7.3 percent and 5.9 percent of greenhouse gas emissions, respectively, and U.S.
Territories accounted for 0.4 percent of emissions; emissions from these sectors primarily consisted of
C02 emissions from fossil fuel combustion. Carbon dioxide was also emitted and sequestered (in the
form of carbon) by a variety of activities related to forest management practices, tree planting in urban
areas, the management of agricultural soils, landfilling of yard trimmings, and changes in carbon stocks
in coastal wetlands. Table 2-10 presents a detailed breakdown of emissions from each of these
economic sectors by source category, as they are defined in this report. Figure 2-13 shows the trend in
emissions by sector from 1990 to 2023.
Figure 2-13: U.S. Greenhouse Gas Emissions Allocated to Economic Sectors
Note: Emissions and removals from Land Use, Land-Use Change, and Forestry are excluded from the figure above. Excludes U.S.
Territories.
Table 2-10: U.S. Greenhouse Gas Emissions Allocated to Economic Sectors (MMT CO2 Eq. and
Percent of Total in 2023)
Sector/Source
1990
2005
2019
2020
2021
2022
2023
Percent of
Total
Emissions"
Transportation
1,520.8
1,971.8
1,874.2
1,625.3
1,805.5
1,804.0
1,822.5
29.4%
CO2 from Fossil Fuel Combustion
1,468.9
1,858.6
1,816.6
1,573.0
1,753.5
1,753.6
1,776.5
28.7%
Trends in Greenhouse Gas Emissions and Removals 2-35
-------
Sector/Source
1990
2005
2019
2020
2021
2022
2023
Percent of
Total
Emissions"
Substitution of Ozone Depleting
Substances
0.0
63.1
34.0
32.5
31.2
29.6
27.9
0.4%
Mobile Combustionb
40.0
40.0
14.7
12.0
12.7
12.2
11.6
0.2%
Non-Energy Use of Fuels
11.8
10.2
8.8
7.8
8.0
8.7
6.6
0.1%
Electric Power Industry
1,880.2
2,457.4
1,650.7
1,481.8
1,584.0
1,575.5
1,453.7
23.5%
CO2 from Fossil Fuel Combustion
1,820.0
2,400.1
1,606.7
1,439.6
1,540.9
1,531.7
1,414.2
22.8%
Stationary Combustionb
18.7
27.7
20.2
18.9
20.4
20.9
18.1
0.3%
Incineration of Waste
13.3
13.6
13.3
13.3
12.8
12.8
12.8
0.2%
Electrical Equipment
24.6
11.8
6.0
5.5
5.5
4.9
5.1
0.1%
Other Process Uses of Carbonates
3.6
4.2
4.5
4.5
4.3
5.2
3.6
0.1%
Industry
1,714.5
1,589.4
1,514.8
1,412.3
1,446.0
1,439.8
1,423.0
23.0%
CO2 from Fossil Fuel Combustion
822.6
798.1
752.8
701.9
731.9
754.5
739.7
11.9%
Natural Gas Systems
252.1
237.0
227.7
216.9
210.4
209.3
200.1
3.2%
Non-Energy Use of Fuels
83.9
107.2
97.4
89.9
103.5
92.9
100.4
1.6%
Petroleum Systems
59.6
58.7
96.2
79.5
69.2
58.4
61.3
1.0%
Coal Mining
112.7
75.6
56.0
48.3
47.1
46.1
47.8
0.8%
Iron and Steel Production
104.8
70.1
46.8
40.7
47.2
45.2
46.2
0.7%
Cement Production
33.5
46.2
40.9
40.7
41.3
41.9
40.6
0.7%
Substitution of Ozone Depleting
Substances
+
8.0
33.1
33.9
32.2
33.4
35.1
0.6%
Petrochemical Production
20.1
26.9
28.5
27.9
30.7
28.8
30.5
0.5%
Landfills (Industrial)
12.2
16.1
18.8
18.9
18.9
18.9
18.9
0.3%
Ammonia Production
14.4
10.2
12.4
12.3
11.5
11.9
12.2
0.2%
Lime Production
11.7
14.6
12.1
11.3
11.9
12.2
11.5
0.2%
Abandoned Oil and Gas Wells
7.8
8.2
8.5
8.5
8.6
8.5
8.5
0.1%
Nitric Acid Production
10.8
10.1
8.9
8.3
7.9
8.6
8.3
0.1%
Wastewater T reatment
6.6
7.1
7.6
7.6
7.7
7.7
7.6
0.1%
Abandoned Underground Coal Mines
8.1
7.4
6.6
6.5
6.2
6.1
6.1
0.1%
Mobile Combustion
3.6
5.6
5.6
5.3
5.5
5.8
5.9
0.1%
Urea Consumption for Non-
Agricultural Purposes
3.8
3.7
6.2
5.9
6.7
5.5
5.4
0.1%
Fluorochemical Production
71.0
30.0
9.3
7.0
7.1
7.6
4.7
0.1%
Electronics Industry
3.3
4.5
4.5
4.5
4.9
4.8
4.2
0.1%
N2O from Product Uses
3.8
3.8
3.8
3.8
3.8
3.8
3.8
0.1%
Other Process Uses of Carbonates
3.6
4.2
4.5
4.5
4.3
5.2
3.6
0.1%
Stationary Combustion
4.8
4.5
3.8
3.6
3.6
3.6
3.3
0.1%
Non-EOR Carbon Dioxide Utilization
1.5
1.4
2.4
2.8
2.9
2.8
2.1
0.0%
Glass Production
2.3
2.4
1.9
1.9
2.0
2.0
1.8
0.0%
Soda Ash Production
1.4
1.7
1.8
1.5
1.7
1.7
1.7
0.0%
Aluminum Production
26.1
7.2
3.3
3.2
2.5
2.2
1.7
0.0%
Caprolactam, Glyoxal, and Glyoxylic
Acid Production
1.5
1.9
1.2
1.1
1.2
1.3
1.3
0.0%
Ferroalloy Production
2.2
1.4
1.6
1.4
1.4
1.3
1.3
0.0%
2-36 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Sector/Source
1990
2005
2019
2020
2021
2022
2023
Percent of
Total
Emissions"
Titanium Dioxide Production
1.2
1.8
1.3
1.3
1.5
1.5
1.2
0.0%
AdipicAcid Production
13.5
6.3
4.7
7.4
6.6
2.1
1.2
0.0%
Magnesium Production and
Processing
5.7
3.0
1.0
0.9
1.2
1.1
1.1
0.0%
Other Product Manufacture and Use
1.5
1.5
0.8
0.7
0.5
0.6
1.0
0.0%
Zinc Production
0.6
1.0
1.0
1.0
1.0
0.9
0.9
0.0%
Phosphoric Acid Production
1.5
1.3
0.9
0.9
0.9
0.8
0.9
0.0%
Lead Production
0.5
0.6
0.5
0.5
0.5
0.5
0.5
0.0%
Carbide Production and
Consumption
0.3
0.2
0.2
0.2
0.2
0.2
0.2
0.0%
CO2 Transport, Injection, and
Geological Storage
0.0
0.0
+
+
0.1
0.1
0.1
0.0%
Agriculture
606.8
633.7
679.2
663.3
655.7
639.8
649.6
10.5%
N2O from Agricultural Soil
Management
289.1
294.7
316.4
293.0
298.9
291.8
296.3
4.8%
Enteric Fermentation
183.1
188.2
197.3
196.3
196.5
192.6
187.1
3.0%
Manure Management
52.5
70.2
83.5
83.8
83.6
81.7
81.9
1.3%
CO2 from Fossil Fuel Combustion
53.9
49.6
57.1
61.6
48.6
45.2
52.9
0.9%
Rice Cultivation
18.9
20.6
15.6
18.6
18.5
18.0
18.7
0.3%
Liming
4.7
4.4
2.2
2.9
2.4
3.2
5.3
0.1%
Urea Fertilization
2.4
3.5
4.9
5.0
5.1
5.2
5.3
0.1%
Mobile Combustionb
1.4
1.6
1.2
1.2
1.2
1.2
1.225
0.0%
Field Burning of Agricultural Residues
0.7
0.8
0.9
0.8
0.8
0.8
0.8
0.0%
Stationary Combustionb
0.1
+
0.1
0.1
0.1
0.1
0.1
0.0%
Commercial
447.0
422.1
469.2
442.3
448.6
469.0
455.1
7.3%
CO2 from Fossil Fuel Combustion
228.3
227.1
251.7
229.3
237.5
259.2
244.2
3.9%
Landfills (Municipal)
185.5
131.6
109.4
103.7
101.8
99.8
100.6
1.6%
Substitution of Ozone Depleting
Substances
+
24.7
67.4
68.3
69.1
69.6
70.0
1.1%
Wastewater T reatment
30.9
33.6
34.7
35.1
34.3
34.3
34.3
0.6%
Composting
0.7
3.6
4.3
4.4
4.4
4.4
4.4
0.1%
Stationary Combustionb
1.5
1.5
1.6
1.5
1.6
1.7
1.6
0.0%
Anaerobic Digestion at Biogas
Facilities
+
+
+
+
+
+
+
0.0%
Residential
345.6
371.2
384.2
358.0
369.6
392.4
368.3
5.9%
CO2 from Fossil Fuel Combustion
338.6
358.9
342.9
314.8
318.0
335.2
307.1
5.0%
Substitution of Ozone Depleting
Substances
0.2
7.0
35.1
39.0
47.3
52.2
56.1
0.9%
Stationary Combustionb
6.8
5.3
6.2
4.2
4.2
5.0
5.2
0.1%
U.S. Territories
23.4
59.7
25.1
22.6
24.4
23.7
25.1
0.4%
CO2 from Fossil Fuel Combustion
20.0
51.9
24.8
22.3
24.1
23.5
24.9
0.4%
Non-Energy Use of Fuels
3.4
7-6
0.2
0.2
0.2
0.1
0.1
0.0%
Stationary Combustionb
0.1
0.2
0.1
0.1
0.1
0.1
0.1
0.0%
Trends in Greenhouse Gas Emissions and Removals 2-37
-------
Percent of
Total
Sector/Source
1990
2005
2019
2020
2021
2022
2023
Emissions"
Total Gross Emissions (Sources)
6,538.3
7,505.3
6,597.4
6,005.7
6,333.8
6,344.1
6,197.3
100.0%
LULUCF Sector Net Totalc
(1,037.9)
(968.9)
(919.4)
(951.6)
(962.9)
(905.3)
(939.9)
-15.2%
Net Emissions (Sources and Sinks)
5,500.4
6,536.4
5,678.0
5,054.2
5,371.0
5,438.7
5,257.4
84.8%
+ Does not exceed 0.05 MMT C02 Eq. or 0.05 percent.
a Percent of total (gross) emissions excluding emissions from LULUCF for 2023.
b Includes ChU and N20 emissions from fuel combustion.
cThe LULUCF sector net total is the net sum of all LULUCF ChUand N20 emissions to the atmosphere plus LULUCF net carbon
stock changes.
Notes: Total gross emissions presented are without LULUCF. Total net emissions are presented with LULUCF. Totals may not sum
due to independent rounding. Parentheses indicate negative values or sequestration.
Box 2-1: Methodology for Aggregating Emissions by Economic Sector
This report also characterizes emissions according to following economic sector categories. Discussing
greenhouse gas emissions relevant to U.S.-specific economic sectors improves communication of the
report's findings.
The electric power economic sector includes C02,CH4and N20 emissions from the combustion of fossil
fuels that are included in the EIA electric power sector. Carbon dioxide, CH4, and N20 emissions from
waste incineration are included in the electric power economic sector, as the majority of MSW is
combusted in plants that produce electricity. The electric power economic sector also includes SF6
from electrical equipment, and a portion of C02 from other process uses of carbonates (from pollution
control equipment installed in electric power plants).
The transportation economic sector includes C02 emissions from the combustion of fossil fuels that
are included in the EIA transportation fuel-consuming sector. Additional analyses and refinement of the
EIA data are further explained in the Energy chapter of this report. Emissions of CH4 and N20 from
mobile combustion are also apportioned to the transportation economic sector based on the EIA
transportation fuel-consuming sector. Emissions of ODS substitutes are apportioned to the
transportation economic sector based on emissions from refrigerated transport and motor vehicle air-
conditioning systems. Finally, C02 emissions from non-energy uses of fossil fuels identified as
lubricants for transportation vehicles are included in the transportation economic sector.
The industry economic sector includes C02 emissions from the combustion of fossil fuels that are
included in the EIA industrial fuel-consuming sector, minus the agricultural use of fuel explained below.
The CH4 and N20 emissions from stationary and mobile combustion are also apportioned to the
industry economic sector based on the EIA industrial fuel-consuming sector, minus emissions
apportioned to the agriculture economic sector. Emissions of ODS substitutes are apportioned based
on their specific end-uses within the source category, with most emissions falling within the industry
economic sector. Finally, CH4 emissions from industrial landfills and CH4 and N20 from industrial
wastewater treatment are included in the industry economic sector.
Additionally, all process-related emissions from sources with methods considered within the IPCC
IPPU sector are apportioned to the industry economic sector. This includes the process-related
emissions (i.e., emissions from the actual process to make the material, not from fuels to power the
plant) from activities such as cement production, iron and steel production and metallurgical coke
production, and ammonia production. Additionally, fugitive emissions from energy production sources,
2-38 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
such as natural gas systems, coal mining, and petroleum systems are included in the industry
economic sector. A portion of C02 from other process uses of carbonates (from pollution control
equipment installed in large industrial facilities) is also included in the industry economic sector.
Finally, all remaining C02 emissions from non-energy uses of fossil fuels are assumed to be industrial in
nature (besides the lubricants for transportation vehicles specified above) and are attributed to the
industry economic sector.
The agriculture economic sector includes C02 emissions from the combustion of fossil fuels that are
based on supplementary sources of agriculture fuel use data, because EIA includes agriculture
equipment in the industrial fuel-consuming sector. Agriculture fuel use estimates are obtained from
U.S. Department of Agriculture survey data, in combination with EIA Fuel Oil and Kerosene Sales (FOKS)
data (EIA 2020). Agricultural operations are based on annual energy expense data from the Agricultural
Resource Management Survey (ARMS) conducted by the National Agricultural Statistics Service (NASS)
of the USDA. NASS collects information on farm production expenditures including expenditures on
diesel fuel, gasoline, LP gas, natural gas, and electricity use on the farm with the annual ARMS. A USDA
publication (USDA/NASS 2024) shows national totals, as well as selected States and ARMS production
regions. These supplementary data are subtracted from the industrial fuel use reported by EIA to obtain
agriculture fuel use. Carbon dioxide emissions from fossil fuel combustion, and CH4 and N20 emissions
from stationary and mobile combustion, are then apportioned to the agriculture economic sector based
on agricultural fuel use.
The other IPCC Agriculture emission source categories apportioned to the agriculture economic sector
include N20 emissions from agricultural soils, CH4 from enteric fermentation, CH4 and N20 from
manure management, CH4 from rice cultivation, C02 emissions from liming and urea application, and
CH4 and N20 from field burning of agricultural residues.
The residential economic sector includes C02 emissions from the combustion of fossil fuels that are
included in the EIA residential fuel-consuming sector. Stationary combustion emissions of CH4 and N20
are also based on the EIA residential fuel-consuming sector. Emissions of ODS substitutes are
apportioned to the residential economic sector based on emissions from residential air-conditioning
systems. N20 emissions from the application of fertilizers to developed land (termed "settlements" by
the IPCC) are also included in the residential economic sector.
The commercial economic sector includes C02 emissions from the combustion of fossil fuels that are
included in the EIA commercial fuel-consuming sector. Emissions of CH4 and N20 from mobile
combustion are also apportioned to the commercial economic sector based on the EIA commercial
fuel-consuming sector. Emissions of ODS substitutes are apportioned to the commercial economic
sector based on emissions from commercial refrigeration/air-conditioning systems. Public works
sources, including direct CH4 from municipal landfills, CH4 from anaerobic digestion at biogas facilities,
CH4 and N20 from domestic wastewater treatment, and composting, are also included in the
commercial economic sector.
Emissions with Electricity Distributed to Economic Sectors
It is also useful to view greenhouse gas emissions from economic sectors with emissions related to
electric power distributed into end-use categories (i.e., emissions from the electric power sector are
allocated to the economic end-use sectors in which the electricity is used). For example, greenhouse
Trends in Greenhouse Gas Emissions and Removals 2-39
-------
gas emissions from some economic sectors, e.g., commercial, residential and industry, increase
substantially when indirect emissions from electricity end-use are included, due to the relatively large
share of electricity use by buildings (75 percent of the electricity generated in the United States for
heating, ventilation, and air conditioning; lighting; and appliances, etc.)8 and use of electricity for
powering industrial machinery.
The generation, transmission, and distribution of electricity directly accounted for 23.5 percent of total
U.S. greenhouse gas emissions in 2023. Electric power-related emissions decreased by 22.7 percent
since 1990 mainly due to fuel switching. From 2022 to 2023, electric power-related emissions
decreased by 7.7 percent. Between 2022 and 2023, the consumption of natural gas for electric power
generation increased by 6.9 percent, while consumption of petroleum and coal decreased by 28.3
percent and 18.4 percent, respectively. Electric power-related emissions in 2023 are still lower than pre-
pandemic 2019 levels. Table 2-11 provides a detailed summary of emissions from electric power-related
activities.
From 2022 to 2023, electricity sales to the residential end-use sector and commercial end-use sector
decreased by 3.9 percent and increased 1.2 percent, respectively. Electricity sales to the industrial
sectors decreased by 1.1 percent. Overall, from 2022 to 2023, the amount of electricity retail sales (in
kWh) decreased by 1.3 percent.
Table 2-11: Electric Power-Related Greenhouse Gas Emissions (MMT C02 Eq.)
Gas/Fuel Type or Source
19901
2005
2019
2020
2021
2022
2023
C02
1,836.4
2,417.5
1,624.2
1,457.0
1,557.7
1,549.4
1,430.2
Fossil Fuel Combustion
1,820.0
2,400.1
1,606.7
1,439.6
1,540.9
1,531.7
1,414.2
Coal
1,546.5
1,982.8
973.5
788.2
910.1
851.5
694.6
Natural Gas
175.4
318.9
616.6
634.8
612.8
659.3
704.5
Petroleum
97.5
98.0
16.2
16.2
17.7
20.5
14.7
Geothermat
0.5
0.5
0.4
0.4
0.4
0.4
0.4
Incineration of Waste
12.9
13.3
12.9
12.9
12.5
12.5
12.4
Other Process Uses of Carbonates
3.6
4.2
4.5
4.5
4.3
5.2
3.6
CH4
0.5
1.0
1.4
1.4
1.4
1.5
1.5
Stationary Sources®
0.5
1.0
1.4
1.4
1.4
1.5
1.5
Incineration of Waste
+
+
+
+
+
+
+
N2O
18.6
27.1
19.1
17.9
19.4
19.8
16.9
Stationary Sources®
18.2
26.7
18.8
17.5
19.0
19.4
16.6
Incineration of Waste
0.4
0.3
0.4
0.3
0.4
0.3
0.3
SF6
24.6
11.8
6.0
5.5
5.5
4.9
5.1
Electrical Equipment
24.6
11.8
6.0
5.5
5.5
4.9
5.1
CF4
+
+
+
+
+
+
0.0
Electrical Equipment
+
+
+
+
+
+
0.0
Total
1,880.21
2,457.4
1,650.7
1,481.8
1,584.0
1,575.5
1,453.7
+ Does not exceed 0.05 MMT C02 Eq.
a Includes only stationary combustion emissions related to the generation of electricity.
Note: Totals may not sum due to independent rounding.
8 See https://www.nrel.gov/news/features/2023/nrel-researchers-reveal-how-buildings-across-the-united-states-do-
and-could-use-energv.html.
2-40 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
To distribute electricity emissions among economic end-use sectors, emissions from the source
categories assigned to the electric power sector were allocated to the residential, commercial, industry,
transportation, and agriculture economic sectors according to each economic sector's share of retail
sales of electricity (EIA 2020; USDA/NASS 2024). These source categories include C02 from fossil fuel
combustion, CH4and N2Ofrom stationary combustion, incineration of waste, other process uses of
carbonates, and SF6 from electrical equipment. Note that only 50 percent of the emissions from other
process uses of carbonates were associated with electric power and distributed as described; the
remaining emissions from other process uses of carbonates were attributed to the industry economic
end-use sector.9
When emissions from electricity use are distributed to these economic end-use sectors, 2023
emissions from transportation account for the largest share of total U.S. greenhouse gas emissions
(29.5 percent), followed closely by emissions from industrial activities (29.2 percent). The relative share
of emissions from the commercial and residential sectors also increased substantially when emissions
from electricity are included (15.6 and 14.4 percent, respectively). In all economic end-use sectors
except agriculture, C02 accounts for more than 77.3 percent of greenhouse gas emissions, primarily
from the combustion of fossil fuels. Table 2-12 presents a detailed breakdown of emissions from each of
these economic sectors, with emissions from electric power distributed to them. Figure 2-14 shows the
trend in these emissions by sector from 1990 to 2023.
Figure 2-14: U.S. Greenhouse Gas Emissions with Electricity-Related Emissions
Distributed to Economic Sectors
Note: Emissions and removals from Land Use, Land-Use Change, and Forestry are excluded from the figure above. Excludes U.S.
Territories.
Table 2-12: U.S. Greenhouse Gas Emissions by Economic Sector and Gas with
Electricity-Related Emissions Distributed (MMT C02 Eq.) and Percent of Total in 2023
9 Emissions were not distributed to U.S. Territories, since the electric power sector only includes emissions related to the
generation of electricity in the 50 states and the District of Columbia.
Trends in Greenhouse Gas Emissions and Removals 2-41
-------
Emissions by Gas
1990
2005
2019
2020
2021
2022
2023
Percent3
Transportation
1,523.9
1,976.6
1,878.5
1,628.9
1,809.5
1,808.6
1,827.7
29.5%
Direct Emissions
1,520.8
1,971.8
1,874.2
1,625.3
1,805.5
1,804.0
1,822.5
29.4%
C02
1,480.8
1,868.7
1,825.5
1,580.7
1,761.6
1,762.3
1,783.0
28.8%
CH4
6.4
4.0
1.6
1.4
1.5
1.5
1.4
0.0%
N2O
33.6
35.9
13.0
10.6
11.2
10.7
10.2
0.2%
HFCsb
0.0
63.1
34.0
32.5
31.2
29.6
27.86
0.4%
Electricity-Related
3.1
4.8
4.3
3.6
4.1
4.6
5.1
0.1%
CO2
3.1
4.8
4.3
3.5
4.0
4.5
5.1
0.1%
CH4
+
+
+
+
+
+
+
0.0%
N2O
+
0.1
0.1
+
+
0.1
0.1
0.0%
SFe
+
+
+
+
+
+
+
0.0%
Industry
2,388.5
2,305.0
1,957.8
1,799.7
1,868.0
1,859.6
1,806.9
29.2%
Direct Emissions
1,714.5
1,589.4
1,514.8
1,412.3
1,446.0
1,439.8
1,423.0
23.0%
CO2
1,163.4
1,137.7
1,102.4
1,016.2
1,064.9
1,072.2
1,065.5
17.2%
CH4
414.8
371.9
336.5
320.4
307.8
296.0
289.0
4.7%
N2O
35.7
29.8
26.0
27.6
26.8
23.4
22.2
0.4%
HFCs, PFCs,SFe
100.7
50.0
49.9
48.2
46.5
48.1
46.3
0.7%
and NF3
Electricity-Related
674.0
715.6
443.0
387.4
422.0
419.8
383.9
6.2%
CO2
658.3
704.0
435.9
380.9
415.0
412.8
377.7
6.1%
cm
0.2
0.3
0.4
0.4
0.4
0.4
0.4
0.0%
N2O
6.7
7.9
5.1
4.7
5.2
5.3
4.5
0.1%
SFe
8.8
3.4
1.6
1.4
1.5
1.3
1.3
0.0%
Residential
957.9
1,247.7
984.4
919.9
958.7
975.0
891.1
14.4%
Direct Emissions
345.6
371.2
384.2
358.0
369.6
392.4
368.3
5.9%
CO2
338.6
358.9
342.9
314.8
318.0
335.2
307.1
5.0%
CH4
5.9
4.5
5.3
3.6
3.6
4.3
4.5
0.1%
N2O
0.9
0.8
0.8
0.6
0.6
0.7
0.7
0.0%
SFe
0.2
7.0
35.1
39.0
47.3
52.2
56.1
0.9%
Electricity-Related
612.4
876.5
600.2
561.9
589.1
582.7
522.7
8.4%
CO2
598.1
862.2
590.6
552.5
579.3
573.0
514.3
8.3%
CH4
0.2
0.3
0.5
0.5
0.5
0.5
0.5
0.0%
N2O
6.1
9.7
7.0
6.8
7.2
7.3
6.1
0.1%
SFe
8.0
4.2
2.2
2.1
2.1
1.8
1.8
0.0%
Commercial
1,002.5
1,244.3
1,037.1
936.9
981.8
1,007.5
965.1
15.6%
Direct Emissions
447.0
422.1
469.2
442.3
448.6
469.0
455.1
7.3%
CO2
228.3
227.1
251.7
229.3
237.5
259.2
244.2
3.9%
ch4
203.6
150.9
127.3
121.4
119.1
117.4
118.5
1.9%
N2O
15.1
19.4
22.8
23.4
22.9
22.8
22.5
0.4%
HFCs
+
24.7
67.4
68.3
69.1
69.6
70.0
1.1%
Electricity-Related
555.5
822.2
567.9
494.6
533.2
538.5
510.0
8.2%
CO2
542.6
808.9
558.7
486.3
524.3
529.6
501.8
8.1%
CH4
0.1
0.3
0.5
0.5
0.5
0.5
0.5
0.0%
2-42 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Emissions by Gas
1990
2005
2019
2020
2021
2022
2023
Percent3
n2o
5.5
9.1
6.6
6.0
6.5
6.8
5.9
0.1%
SFe
7.3
4.0
2.1
1.8
1.9
1.7
1.8
0.0%
Agriculture
641.9
672.0
714.4
697.7
691.4
669.6
681.5
11.0%
Direct Emissions
606.8
633.7
679.2
663.3
655.7
639.8
649.6
10.5%
C02
61.0
57.4
64.2
69.5
56.1
53.6
63.5
1.0%
cm
241.9
264.6
280.4
282.6
282.2
276.1
271.7
4.4%
N2O
303.9
311.6
334.5
311.2
317.4
310.1
314.4
5.1%
Electricity-Related
35.2
38.3
35.2
34.4
35.7
29.9
31.9
0.5%
CO2
34.3
37.7
34.7
33.8
35.1
29.4
31.4
0.5%
CH4
+
+
+
+
+
+
+
0.0%
N2O
0.3
0.4
0.4
0.4
0.4
0.4
0.4
0.0%
SFe
0.5
0.2
0.1
0.1
0.1
0.1
0.1
0.0%
U.S. Territories
23.4
59.7
25.1
22.6
24.4
23.7
25.1
0.4%
Total Gross
Emissions (Sources)
6,538.3
7,505.3
6,597.4
6,005.7
6,333.8
6,344.1
6,197.3
100.0%
LULUCF Sector Net
Total0
(1,037.9)
(968.9)
(919.4)
(951.6)
(962.9)
(905.3)
(939.9)
-15.2%
Net Emissions
(Sources and Sinks)
5,500.4
6,536.4
5,678.0
5,054.2
5,371.0
5,438.7
5,257.4
84.8%
+ Does not exceed 0.05 MMT C02 Eq. or 0.05 percent.
a Percent of total (gross) emissions excluding emissions from LULUCF for the year 2023.
b Includes primarily HFC-134a.
cThe LULUCF sector net total is the net sum of all LULUCF ChUand N20 emissions to the atmosphere plus LULUCF net carbon
stock changes.
Notes: Total gross emissions are presented without LULUCF. Net emissions are presented with LULUCF. Emissions from electric
power are allocated based on aggregate electricity use in each end-use sector. Totals may not sum due to independent
rounding.
Transportation
When electricity-related emissions are distributed to economic end-use sectors, transportation
activities accounted for 29.5 percent of U.S. greenhouse gas emissions in 2023. The largest sources of
transportation greenhouse gas emissions in 2023 were light-duty trucks, which include sport utility
vehicles, pickup trucks, and minivans (39.7 percent); medium- and heavy-duty trucks (23.4 percent);
passenger cars (16.6 percent); commercial aircraft (7.2 percent); pipelines (3.9 percent); ships and
boats (2.4 percent); and other aircraft (2.8 percent), and rail (1.9 percent). These figures include direct
C02, CH4, and N20 emissions from fossil fuel combustion used in transportation, indirect emissions
from electricity use, and emissions from non-energy use (i.e., lubricants) used in transportation, as well
as HFC emissions from mobile air conditioners and refrigerated transport allocated to these vehicle
types.
From 1990 to 2023, total transportation emissions from fossil fuel combustion increased by 20.9
percent due, in large part, to increased demand for travel. From 2022 to 2023, emissions increased by
1.3 percent. VMT by light-duty vehicles (passenger cars and light-duty trucks) increased by 48.6 percent
from 1990 to 2023 as a result of a confluence of factors including population growth, economic growth,
urban sprawl, and periods of low fuel prices. The rise in transportation-related C02 emissions,
combined with an increase in HFCs from close to zero emissions in 1990 to 27.9 MMT C02 Eq. in 2023,
Trends in Greenhouse Gas Emissions and Removals 2-43
-------
led to an increase in overall greenhouse gas emissions from transportation activities of 19.9 percent
from 1990 to 2023.
The decline in new light-duty vehicle fuel economy between 1990 and 2004 reflected the increasing
market share of light-duty trucks, which grew from approximately 29.6 percent of new vehicle sales in
1990 to 48.0 percent in 2004. Starting in 2005, average new vehicle fuel economy began to increase
while light-duty vehicle VMT grew only modestly between 2005 and 2013. Light-duty vehicle VMT grew by
less than one percent or declined each year between 2005 and 2013, then grew at a faster rate until
2016 (2.6 percent from 2014 to 2015, and 2.5 percent from 2015 to 2016). Between 2016 and 2022 the
rate of light-duty vehicle VMT growth slowed to one percent or less each year. From 2022 to 2023, light-
duty vehicle VMT increased by 2.3 percent. Average new vehicle fuel economy has increased almost
every year since 2005, while light-duty truck market share decreased to 33.0 percent in 2009 and has
since varied from year to year between 33.0 and 63.1 percent. Light-duty truck market share was about
62.5 percent of new passenger vehicle sales in model year 2023 (EPA 2023).
Table 2-13 provides a detailed summary of greenhouse gas emissions from transportation-related
activities with electricity-related emissions included in the totals. Historically, the majority of electricity
use in the transportation sector was for rail transport. However, more recently there has been increased
electricity use in on-road electric and plug-in hybrid electric vehicles. Despite this increase, almost all of
the energy used for transportation was supplied by petroleum-based products, with more than half
related to gasoline consumption in automobiles and other highway vehicles. Other fuel uses, especially
diesel fuel for freight trucks and jet fuel for aircraft, accounted for the remainder. Indirect emissions
from electricity are less than 1 percent of direct emissions in the transportation sector. For a more
detailed breakout of emissions by fuel type by vehicle see Table A-93 in Annex 3.
2-44 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Figure 2-15: Trends in Transportation-Related Greenhouse Gas Emissions
2,800 Lubricants
Motorcycles
2,600 ¦ Buses
¦ Pipelines
2,400 ¦ Rail
! Ships and Boats
I Aircraft
I Medium- and Heavy-Duty Trucks
Light-Duty Trucks
I Passenger Cars
2,200
2,000
1,800
m 1,600
° 1,400
i-
S
s 1,200
1,000
800
600
400
200
0
S n n
lo r-v
LT)
^ rvj
r-- v£>
CM vjQ
vD -
vO
ro co
_r
co
LO
CO
LO
CM • •
cti cn
N N vD N
co \£> r-v rv.
ro cr> cn ^
¦T t-T
oo
^ co cn
i—1 ^ r~-> r-v
r-s^m^Tcoco
ojcnnco^co00
CO ^ t LO N N
h" ^ r- ^ _r J-
O CT>
1-1 o
CO CO
CO
rsj
co
Table 2-13: Transportation-Related Greenhouse Gas Emissions (MMT C02 Eq.)
Gas/Vehicle Type
1990
2005
2019
2020
2021
2022
2023
Passenger Cars
658.7
576.1
372.7
309.6
325.4
307.3
303.3
C02
632.6
529.1
361.3
299.8
316.3
298.9
296.0
CH4
3.8
1.8
0.3
0.2
0.2
0.2
0.2
N2O
22.3
16.4
2.7
2.0
1,9
1.6
1.4
HFCs
0.0
28.8
8.4
7.6
7.0
6.6
5.705
Light-Duty Trucks
289.7
653.8
729.8
636.9
705.8
704.5
726.2
CO2
280.0
606.4
708.3
618.0
688.0
688.8
711.9
CH4
1.4
1.3
0.6
0.5
0.5
0.5
0.5
N2O
8.2
15.8
5.5
4.3
4.2
3.8
3.4
HFCs
0.0
30.2
15.4
14.2
13.0
11.4
10.4
Medium- and Heavy-Duty Trucks
236.6
390.8
413.8
400.3
434.6
430.2
427.5
CO2
235.1
386.0
405.2
391.3
425.1
420.6
418.1
CH4
0.5
0.2
0.1
0.1
0.1
0.1
0.1
N2O
1.0
1.5
2.7
2.8
3.1
3.2
3.2
HFCs
0.0
3.2
5.8
6.1
6.3
6.3
6.1
Buses
13.1
18.0
24.8
20.2
22.3
24.4
24.4
CO2
12.9
17.6
24.2
19.7
21.8
23.8
23.8
Trends iri Greenhouse Gas Emissions and Removals 2-45
-------
Gas/Vehicle Type
1990
2005
2019
2020
2021
2022
2023
cm
0.1
+
+
+
+
+
+
n2o
0.1
0.1
0.2
0.1
0.2
0.2
0.2
HFCs
0.0
0.2
0.4
0.4
0.4
0.4
0.4
Motorcycles
3.3
4.8
6.8
6.2
7.0
8.3
8.7
CO2
3.3
4.7
6.7
6.1
6.9
8.2
8.6
cm
+
+
+
+
+
+
+
N2O
+
+
0.1
0.1
0.1
0.1
0.1
Commercial Aircraft3
110.8
133.8
137.8
92.0
120.0
130.8
130.8
CO2
109.9
132.7
136.7
91.3
119.0
129.7
129.7
CH4
0.0
0.0
0.0
0.0
0.0
0.0
0.0
N2O
0.9
1.1
1.1
0.7
1.0
1.1
1.1
Other Aircraft"
78.0
59.5
45.6
31.0
35.5
37.0
51.1
CO2
77.3
59.0
45.2
30.7
35.1
36.7
50.7
cm
0.1
0.1
*
+
+
+
+
N2O
0.6
0.5
0.4
0.2
0.3
0.3
0.4
Ships and Boats°
47.0
45.5
40.0
32.2
50.7
49.9
43.8
CO2
46.3
44.3
35.5
27.5
45.4
44.4
38.0
CH4
0.4
0.5
0.4
0.4
0.5
0.5
0.5
N2O
0.2
0.2
0.2
0.1
0.3
0.3
0.2
HFCs
0.0
0.5
3.9
4.2
4.5
4.8
5.1
Rail
39.0
51.4
39.7
34.2
35.5
35.6
33.9
CO2
38.5
50.8
39.1
33.7
34.9
35.0
33.4
cm
0.1
0.1
0.1
0.1
0.1
0.1
0.1
N2O
0.3
0.4
0.3
0.3
0.3
0.3
0.3
HFCs
0.0
0.1
0.1
0.1
0.1
0.1
0.115
Other Emissions from Electric Powerd
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Pipelines6
36.0
32.8
58.5
58.5
64.9
72.0
71.3
CO2
36.0
32.8
58.5
58.5
64.9
72.0
71.3
Lubricants
11.8
10.2
8.8
7.8
8.0
8.7
6.6
CO2
11.8
10.2
8.8
7.8
8.0
8.7
6.6
Total Transportation
1,523.9
1,976.6
1,878.5
1,628.9
1,809.5
1,808.6
1,827.7
International Bunker Fuels'
54.7
44.6
26.2
22.7
22.7
25.3
23.5
Ethanol CC>2g
4.1
21.6
78.7
68.1
75.4
75.0
76.4
Biodiesel CC>2g
0.0
0.9
17.1
17.7
16.1
15.6
18.2
+ Does not exceed 0.05 MMT C02 Eq.
a Consists of emissions from jet fuel consumed by domestic operations of commercial aircraft (no bunkers).
b Consists of emissions from jet fuel and aviation gasoline consumption by general aviation and military aircraft.
c Fluctuations in emission estimates are associated with fluctuations in reported fuel consumption and may reflect issues with
data sources.
d Other emissions from electric power are a result of waste incineration (as the majority of MSWis combusted in "trash-to-
steam" electric power plants), electrical equipment, and a portion of other process uses of carbonates (from pollution control
equipment installed in electric power plants).
e C02 estimates reflect natural gas used to power pipelines, but not electricity. While the operation of pipelines produces CH4
and N2O, these emissions are not directly attributed to pipelines in the Inventory.
1 Emissions from International Bunker Fuels include emissions from both civilian and military activities; these emissions are not
included in the transportation totals.
2-46 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
g Ethanol and biodiesel C02 estimates are presented for informational purposes only. See Section 3.11 and the estimates in
LULUCF (see Chapter 6), in line with IPCC methodological guidance, for more information on ethanol and biodiesel.
Notes: Passenger cars and light-duty trucks include vehicles typically used for personal travel and less than 8,500 lbs; medium-
and heavy-duty trucks include vehicles larger than 8,500 lbs. HFC emissions primarily reflect HFC-134a. Totals may not sum
due to independent rounding.
Industry
The industry economic sector includes C02 emissions from fossil fuel combustion from all
manufacturing facilities, in aggregate, and with the distribution of electricity-related emissions (e.g.,
powering industrial machinery), accounted for 29.2 percent of U.S. greenhouse gas emissions in 2023.
This end-use sector also includes emissions that are produced as a byproduct of the non-energy-related
industrial process activities. Various activities produce these non-energy-related emissions, including
CH4 emissions from petroleum and natural gas systems, fugitive CH4 and C02 emissions from coal
mining, byproduct C02 emissions from cement production, and HFC, PFC, SF6, and NF3 byproduct
emissions from the electronics industry.
Since 1990, industry sector emissions have declined by 24.4 percent. The decline has occurred both in
direct emissions and indirect emissions associated with electricity use. Structural changes within the
U.S. economy that led to shifts in industrial output away from energy-intensive manufacturing products
to less energy-intensive products (e.g., shifts from producing steel to computer equipment) have had a
significant effect on industrial emissions.
Commercial
The commercial end-use sector, including electricity-related emissions, accounted for 15.6 percent of
U.S. greenhouse gas emissions in 2023. Like the residential sector it is heavily reliant on electricity for
meeting energy needs, with electricity use for building-related activities like lighting, heating, air
conditioning, and operating appliances. The remaining emissions were largely due to the direct
consumption of natural gas and petroleum products, primarily for heating and cooking needs. Energy-
related emissions from the commercial sector have generally been increasing since 1990, and annual
variations are often correlated with short-term fluctuations in energy use caused by weather conditions,
rather than prevailing economic conditions. Decreases in energy-related emissions in the commercial
sector in recent years can be largely attributed to an overall reduction in energy use driven by a
reduction in heating degree days and increases in energy efficiency.
Municipal landfills and wastewater treatment are included in the commercial sector, with landfill
emissions decreasing since 1990 and wastewater treatment emissions increasing slightly.
Residential
The residential end-use sector, including electricity-related emissions, accounted for 14.4 percent of
U.S. greenhouse gas emissions in 2023. This sector is heavily reliant on electricity for meeting energy
needs, with electricity use for building-related activities like lighting, heating, air conditioning, and
operating appliances. The remaining emissions were largely due to the direct consumption of natural
gas and petroleum products, primarily for heating and cooking needs. Emissions from the residential
sector have generally been increasing since 1990, and annual variations are often correlated with short-
term fluctuations in energy use caused by weather conditions, rather than prevailing economic
conditions. In the long term, the residential sector is also affected by population growth, migration
Trends in Greenhouse Gas Emissions and Removals 2-47
-------
trends toward warmer areas, and changes in housing and building attributes (e.g., larger sizes and
improved insulation). A shift toward energy-efficient products and more stringent energy efficiency
standards for household equipment has also contributed to recent trends in energy demand in
households.
Agriculture
The agriculture end-use sector accounted for 11.0 percent of U.S. greenhouse gas emissions in 2023
when electricity-related emissions are distributed, and includes a variety of processes, including enteric
fermentation in domestic livestock, livestock manure management, and agricultural soil management.
In 2023, agricultural soil management was the largest source of N20 emissions, and enteric
fermentation was the largest source of CH4 emissions in the United States. This sector also includes
small amounts of C02 emissions from fossil fuel combustion by motorized farm equipment such as
tractors. Indirect emissions from electricity use in agricultural activities (e.g., powering buildings and
equipment) are about 5 percent of direct emissions.
Box 2-2: Trends in Various U.S. Greenhouse Gas Emissions-Related Data
Total (gross) greenhouse gas emissions can be compared to other economic and social indices to
highlight changes over time. These comparisons include: (1) aggregate energy use, because energy-
related activities are the largest sources of emissions; (2) energy use per capita as a measure of
efficiency; (3) emissions per unit of total gross domestic product as a measure of national economic
activity; and (4) emissions per capita.
Table 2-14 provides data on various statistics related to U.S. greenhouse gas emissions normalized to
1990 as a baseline year. These values represent the relative change in each statistic since 1990.
Greenhouse gas emissions in the United States have decreased at an average annual rate of 0.1 percent
since 1990, although changes from year to year have been significantly larger. This growth rate is slightly
slower than that for total energy use, overall gross domestic product (GDP) and national population (see
Table 2-14 and Figure 2-16). The direction of these trends started to change after 2005, when
greenhouse gas emissions, total energy use and associated fossil fuel consumption began to peak.
Greenhouse gas emissions in the United States have decreased at an average annual rate of 1.0 percent
since 2005. Since 2005, GDP, and national population, generally continued to increase, and energy use
has decreased slightly noting 2020 was impacted by the COVID-19 pandemic.
Table 2-14: Recent Trends in Various U.S. Data (Index 1990 = 100)
Variable
1990
2005
2019
2020
2021
2022
2023
Avg. Annual
Change
Since 1990a
Avg. Annual
Change
Since 2005a
Greenhouse Gas Emissionsb
100
115
101
92
97
97
95
-0.1%
-1.0%
Energy Use0
100
119
117
107
113
115
113
0.4%
-0.2%
GDPd
100
159
206
202
214
219
225
2.5%
2.0%
Population®
100
118
131
132
132
133
135
0.9%
0.8%
a Average annual growth rate.
b Gross total GWP-weighted values.
c Energy-content-weighted values (EIA 2025).
d GDP in chained 2017 dollars (BEA 2024).
e U.S. Census Bureau (2025).
2-48 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Figure 2-16: U.S. Greenhouse Gas Emissions Per Capita and Per Dollar of Gross
Domestic Product (GDP)
Source: BEA (2024), U.S. Census Bureau (2025), and net estimates in this report.
220
200
180
160
| 140
« 120
-S 100
C
80
60
40
20
0
Real GDP
Population
Energy Use
Emissions
Energy Use per Capita
Emissions per Capita
Emissions per GDP
o^Hrsjn^-LnvDi^coc^o^-trvjro^-Lr)«X)r^cocriO^HfMn^-LnvDr^a3aio^rNjro
oioiffiffiffiffioioicncnooooooooooHHH^H^^H^H(N(SfNi(N
o-.o-.mo-.mo-.crvo-iCT*cTvoooooooooooooooooooooooo
HHrHHrHrHrHrHrHrHrM(NrvlrMrv|fNfM(NrMfN(NfNrMrMfMrMrMfN|fNrMfNlrMfM(N
2.3 Precursor Greenhouse Gas Emissions
This section summarizes emissions of compounds that are precursors to greenhouse gases, which
include carbon monoxide (CO), nitrogen oxides (NOx), non-methane volatile organic compounds
(NMVOCs), ammonia (NH3), and sulfur dioxide (S02). These gases are not direct greenhouse gases, but
can indirectly impact Earth's radiative balance, by altering the concentrations of other greenhouse gases
(e.g., tropospheric ozone) and atmospheric aerosol (e.g., particulate sulfate). Carbon monoxide is
produced when carbon-containing fuels are combusted incompletely in energy, transportation, and
industrial processes, and is also emitted from practices such as agricultural burning and waste disposal
and treatment. Anthropogenic sources of nitrogen oxides (i.e., NO and N02) are primarily fossil fuel
combustion (for energy, transportation, industrial process) and agricultural burning. Anthropogenic
sources of NMVOCs, which include hundreds of organic compounds that participate in atmospheric
chemical reactions (propane, butane, xylene, toluene, ethane, and many others)—are emitted primarily
from transportation, industrial processes, oil and natural gas production, waste practices, agricultural
burning, and non-industrial consumption of organic solvents. Primary sources of ammonia (NH3) are
livestock waste and fertilizer application, and additional contributions come from industrial processes
and on-road vehicles. In the United States, S02 is primarily emitted from coal combustion for electric
power generation and the metals industry.
As noted above and summarized in Chapter 6 of IPCC (2021), these compounds can have important
indirect effects on Earth's radiative balance. For example, reactions between NMVOCs and NOx in the
presence of sunlight lead to formation of tropospheric ozone, a greenhouse gas. Concentrations of
NMVOCs, NOx, and CO can also impact the abundance and lifetime of primary greenhouse gases. This
largely occurs by altering the atmospheric concentrations of the hydroxyl radical (OH), which is the main
Trends in Greenhouse Gas Emissions and Removals 2-49
-------
sink for atmospheric CH4. For example, NO* emissions can lead to increases in 03 concentrations and
subsequent OH production, which will increase the amount of OH molecules that are available to
destroy CH4. In contrast, NMVOCs and CO can both react directly with OH, leadingto lower OH
concentrations, a longer atmospheric lifetime of CH4, and a decrease in C02 production (i.e., CO+OH->
C02). Changes in atmospheric CH4 can also feedback on background concentrations of tropospheric
03. Other indirect impacts include the formation of sulfate and nitrate aerosol from emissions of NOx
and S02, both of which have a net negative impact on radiative forcing.
Since 1970, the United States has published triennial estimates of emissions of CO, NOx, NMVOCs,
NH3, and S02 (EPA 2024), which are regulated under the Clean Air Act. Emissions of each of these
precursor greenhouse gases has decreased significantly since 1990 as a result of implementation of
Clean Air Act programs, as well as technological improvements.10 Precursor emission estimates for this
report for 1990 through 2023 were obtained from data published on EPA's National Emissions Inventory
(NEI) Air Pollutants Emissions Trends Data website (EPA 2024). For Table 2-15, NEI-reported emissions of
CO, NOx, S02, NH3, and NMVOCs are recategorized from NEI Emissions Inventory System (EIS) source
categories to those more closely aligned with reporting sectors and categories in this report, based on
the crosswalk detailed in Annex 6.3. Table 2-15 shows that fuel combustion accounts for the majority of
emissions of these precursors. Industrial processes—such as the manufacture of chemical and allied
products, metals processing, and industrial uses of solvents—are also significant sources of CO, NOx,
and NMVOCs. Precursor emissions from Agriculture and LULUCF categories are estimated separately
and therefore are not taken from EPA (2024).
Table 2-15: Emissions of NOx, CO, NMVOCs, NH3, and S02 (kt)
Gas/Activity
1990
2005
2019
2020
2021
2022
2023
NOx
22,898
19,903
7,841
7,141
7,178
6,879
6,531
Energy
21,966
18,863
7,048
6,237
6,300
6,103
5,823
IPPU
774
672
440
391
402
390
389
Agriculture
21
159
192
180
163
156
148
LULUCF
53
158
89
257
237
137
80
Waste
84
51
73
76
76
75
74
CO
133,263
84,345
47,186
53,144
52,559
46,803
41,849
Energy
124,712
64,455
30,349
28,430
28,820
28,004
27,475
IPPU
4,096
1,701
1,011
852
899
885
882
Agriculture
407
7,393
9,431
8,665
8,257
8,853
8,638
LULUCF
3,069
9,618
5,214
13,855
13,241
7,721
3,513
Waste
979
1,178
1,181
1,342
1,343
1,340
1,340
NMVOCs
20,975
14,372
10,893
10,891
10,999
10,858
10,772
Energy
13,067
8,694
5,444
5,305
5,561
5,403
5,342
IPPU
6,982
3,668
2,996
3,364
3,505
3,403
3,403
Agriculture
57
1,858
2,297
2,048
1,761
1,881
1,856
LULUCF
NA
NA
NA
NA
NA
NA
NA
Waste
870
152
156
173
172
171
171
10 More information is available online at: https://www.epa.gov/clean-air-act-overview/progress-cleaning-air-and-
improving-peoples-health and https://gispub.epa.gov/neireport/2017/.
2-50 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Gas/Activity
1990
2005
2019
2020
2021
2022
2023
NH3
511
2,763
2,935
3,021
3,190
2,839
2,834
Energy
229
219
179
180
267
271
268
IPPU
193
117
65
57
56
56
56
Agriculture
16
2,408
2,672
2,700
2,783
2,429
2,427
LULUCF
NA
NA
NA
NA
NA
NA
NA
Waste
73
18
19
84
84
83
83
S02
20,924
13,174
1,759
1,546
1,706
1,718
1,524
Energy
19,398
12,312
1,344
1,173
1,315
1,341
1,151
IPPU
1,490
776
309
265
273
261
260
Agriculture
+
65
83
75
85
84
82
LULUCF
NA
NA
NA
NA
NA
NA
NA
Waste
36
20
23
33
32
31
31
+ Does not exceed 0.5 kt.
NA (Not Available)
Note: Totals by gas may not sum due to independent rounding.
Source: (EPA 2024) except for estimates from forest fires, grassland fires, and field burning of agricultural residues. Emission
categories from EPA (2024) are aggregated into sectors and categories reported under Table ES-3.
Trends in Greenhouse Gas Emissions and Removals 2-51
-------
Energy
¦ ¦III
-------
3 Energy
Energy-related activities were the primary sources of U.S. anthropogenic greenhouse gas emissions,
accounting for 81.5 percent of total gross greenhouse gas emissions on a carbon dioxide (C02)
equivalent basis in 2023.1 This included 96.4, 39.6, and 9.3 percent of the nation's C02, methane (CH4),
and nitrous oxide (N20) emissions, respectively.2 Energy-related C02 emissions alone constituted 76.5
percent of total gross U.S. greenhouse gas emissions from all sources on a C02-equivalent basis, while
the non-C02 emissions from energy-related activities represented a much smaller portion of total gross
national emissions (5.0 percent collectively).
Emissions from fossil fuel combustion contribute the vast majority of energy-related emissions, with
C02 being the primary gas emitted (see Figure 3-1 and Figure 3-2). Globally, approximately 33,809
million metric tons (MMT) of C02 were added to the atmosphere through the combustion of fossil fuels
in 2023, of which the United States accounted for approximately 13 percent.3 Due to their relative
importance over time (see Figure 3-2), fossil fuel combustion-related C02 emissions are considered in
more detail than other energy-related emissions in this report (see Figure 3-3).
Fossil fuel combustion also emits CH4 and N20. Stationary combustion of fossil fuels was the third
largest source of N20 emissions in the United States and mobile fossil fuel combustion was the fifth
largest source. Energy-related activities other than fuel combustion, such as the production,
transmission, storage, and distribution of fossil fuels, also emit greenhouse gases. These emissions
consist primarily of fugitive CH4 emissions from natural gas systems, coal mining, and petroleum
systems.
1 Estimates are presented in units of million metric tons of carbon dioxide equivalent (MMT CO2 Eq.), which weight each
gas by its global warming potential, or GWP, value. See section on global warming potentials in the Executive Summary.
2 The contribution of energy non-CCb emissions is based on gross totals so excludes LULUCF methane (CH4) and nitrous
oxide (N2O) emissions. The contribution of energy-related methane (CH4) and (N2O) including LULUCF non-CCb
emissions, is 37.1 percent and 9.8 percent respectively.
3 Global CO2 emissions from fossil fuel combustion were taken from International Energy Agency Global energy-related
CO2 emissions, 2023. Available at: https://iea.blob.core.windows.net/assets/33e2badc-b839-4cl8-84ce-
f6387b3c008f/C02Emissionsin2023.pdf flEA 2023).
3-2 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Figure 3-1: 2023 Energy Sector Greenhouse Gas Sources
CO2 Emissions from Fossil Fuel Combustion
Natural Gas Systems
Non-Energy Use of Fuels
Petroleum Systems
Coal Mining
Non-CCh Emissions from Stationary Combustion
Non-C02 Emissions from Mobile Combustion
Incineration of Waste
Abandoned Oil and Gas Wells
Abandoned Underground Coal Mines
CO2 Transport, Injection, and Geological Storage
4,559
< 0.5
50
100
150
MMT CO2 Eq.
200
250
300
Figure 3-2: Trends in Energy Sector Greenhouse Gas Sources
8,000
7,000
6,000
5,000
4,000
3,000
2,000
1,000
CO2 Transport, Injection, and Geological Storage
Incineration of Waste
U.S Territories Fossil Fuel Combustion
I Non-Energy Use of Fuels
I Commerical Fossil Fuel Combustion
I Residential Fossil Fuel Combustion
1 Fugitive Emissions
i Industrial Fossil Fuel Combustion
Transportation Fossil Fuel Combustion
I Electric Power Fossil Fuel Combustion
in
f in ^
0"> cr>
s
OHiNni-Ln^DNKiffi
oicnoioioioioio^cnoi
(MfM(N(M(N(NfMfM(MfNNrvl(N(N(M(MNNNNfNNfvJ
Energy 3-3
-------
Figure 3-3: 2023 U.S. Fossil Carbon Flows (MMT C02 Eq.)
International
Table 3-1 summarizes emissions from the Energy sector in units of MMT C02 Eq., while unweighted gas
emissions in kilotons (kt) are provided in Table 3-2. Overall, emissions due to energy-related activities
were 5,050.4 MMT C02 Eq. in 2023,4 a decrease of 6.2 percent since 1990 and a decrease of 2.8 percent
since 2022. Trends are driven by a number of factors including a shift from coal to natural gas and
renewables in the electric power sector.
4 This Inventory report presents CO2 equivalent values based on the IPCC Fifth Assessment Report (AR5) GWP values. See
Chapter 1, Introduction for more information.
3-4 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Table 3-1: C02, CH4, and N20 Emissions from Energy (MMT C02 Eq.)
Gas/Source
1990
2005
2019
2020
2021
2022
2023
C02
4,911.0
5,923.1
5,059.2
4,521.0
4,841.2
4,878.0
4,742.3
Fossil Fuel Combustion
4,752.2
5,744.1
4,852.6
4,342.3
4,654.6
4,702.8
4,559.4
Transportation
1,468.9
1,858.6
1,816.6
1,573.0
1,753.5
1,753.6
1,776.5
Electricity Generation
1,820.0
2,400.1
1,606.7
1,439.6
1,540.9
1,531.7
1,414.2
Industrial
876.5
847.6
809.8
763.4
780.5
799.7
792.6
Residential
338.6
358.9
342.9
314.8
318.0
335.2
307.1
Commercial
228.3
227.1
251.7
229.3
237.5
259.2
244.2
U.S. Territories
20.0
51.9
24.8
22.3
24.1
23.5
24.9
Non-Energy Use of Fuels
99.1
125.0
106.5
97.9
111.7
101.7
107.1
Natural Gas Systems
32.5
26.3
38.7
36.8
35.7
36.4
37.7
Petroleum Systems
9.6
10.2
45.4
28.9
24.1
22.1
23.3
Incineration of Waste
12.9
13.3
12.9
12.9
12.5
12.5
12.4
Coal Mining
4.6
4.2
3.0
2.2
2.5
2.5
2.4
CO2 Transport, Injection, and
Geological and Storage
0.0
0.0
+
+
0.1
0.1
0.1
Abandoned Oil and Gas Wells
+
+
+
+
+
+
+
Biomass-Wooda
215.2
206.9
216.7
189.5
191.5
194.3
187.7
International Bunker Fuelsb
103.6
113.3
113.6
69.6
80.2
98.2
96.2
Biofuels-EthanoP
4.2
22.9
82.6
71.8
79.1
79.6
80.7
Biofuels-BiodieseP
0.0
0.9
17.1
17.7
16.1
15.6
18.2
Biomass-MSW
18.5
14.7
15.7
15.6
15.3
14.9
13.9
CH4
410.4
360.2
320.4
302.3
289.6
278.7
271.9
Natural Gas Systems
219.6
210.7
189.0
180.1
174.6
172.8
162.4
Coal Mining
108.1
71.5
53.0
46.2
44.7
43.6
45.4
Petroleum Systems
50.0
48.4
50.8
50.6
45.1
36.3
38.0
Stationary Combustion
9.7
8.8
9.8
7.9
7.9
8.7
8.8
Abandoned Oil and Gas Wells
7.8
8.2
8.5
8.5
8.6
8.5
8.5
Abandoned Underground Coal Mines 8.1
7.4
6.6
6.5
6.2
6.1
6.1
Mobile Combustion
7.2
5.2
2.8
2.5
2.6
2.6
2.5
Incineration of Waste
+
+
+
+
+
+
+
International Bunker Fuelsb
0.2
0.1
0.1
0.1
0.1
0.1
0.1
N20
60.5
72.9
41.3
36.9
39.3
39.6
36.2
Stationary Combustion
22.3
30.5
22.1
20.5
22.0
22.6
19.6
Mobile Combustion
37.8
42.0
18.7
16.0
16.8
16.6
16.2
Incineration of Waste
0.4
0.3
0.4
0.3
0.4
0.3
0.3
Petroleum Systems
+
+
+
+
+
+
+
Natural Gas Systems
+
+
+
+
+
+
+
International Bunker Fuelsb
0.8
0.9
0.9
0.5
0.6
0.8
0.8
Total
5,381.9
6,356.2
5,420.9
4,860.2
5,170.1
5,196.2
5,050.4
+ Does not exceed 0.05 MMT C02 Eq.
a Emissions from biomass and biofuel consumption are not included specifically in summing energy sector totals. Net carbon
fluxes from changes in biogenic carbon reservoirs are accounted for in the estimates for LULUCF.
b Emissions from international bunker fuels are not included in totals. These values are presented for informational purposes
only, in line with the 2006IPCC Guidelines.
Note: Totals may not sum due to independent rounding.
Energy 3-5
-------
Table 3-2: C02, CH4, and N20 Emissions from Energy (kt)
Gas/Source
1990
2005
2019
2020
2021
2022
2023
C02
4,910,974
5,923,104
5,059,240
4,521,041
4,841,186
4,877,978
4,742,336
Fossil Fuel Combustion
4,752,234
5,744,138
4,852,647
4,342,309
4,654,629
4,702,769
4,559,379
Non-Energy Use of Fuels
99,104
124,988
106,487
97,881
111,718
101,697
107,069
Natural Gas Systems
32,525
26,325
38,696
36,810
35,745
36,410
37,682
Petroleum Systems
9,597
10,222
45,445
28,876
24,091
22,084
23,272
Incineration of Waste
12,900
13,254
12,948
12,921
12,476
12,484
12,425
Coal Mining
4,606
4,169
2,992
2,197
2,455
2,474
2,404
CO2 Transport, Injection,
and Geological and Storage
0
0
18
39
65
53
98
Abandoned Oil and Gas
Wells
7
7
8
8
8
8
8
Biomass-Wooda
215,186
206,901
216,652
189,516
191,471
194,318
187,690
International Bunker Fuelsb
103,634
113,328
113,632
69,638
80,180
98,241
96,160
Biofuels-EthanoP
4,227
22,943
82,578
71,848
79,064
79,593
80,708
Biofuels-BiodieseP
0
856
17,080
17,678
16,112
15,622
18,185
Biomass-MSW
18,534
14,722
15,709
15,614
15,329
14,864
13,936
CH4
14,659
12,864
11,443
10,795
10,344
9,952
9,709
Natural Gas Systems
7,842
7,525
6,751
6,431
6,236
6,173
5,802
Coal Mining
3,860
2,552
1,892
1,648
1,595
1,558
1,623
Petroleum Systems
1,787
1,730
1,813
1,807
1,611
1,295
1,358
Stationary Combustion
345
313
349
282
284
312
313
Abandoned Oil and Gas
Wells
279
294
302
303
306
303
303
Abandoned Underground
Coal Mines
288
264
237
232
221
218
219
Mobile Combustion
258
187
101
90
91
92
91
Incineration of Waste
+
+
+
+
+
+
+
International Bunker Fuelsb
7
5
4
3
3
3
3
N20
228
275
156
139
148
149
136
Stationary Combustion
84
115
84
77
83
85
74
Mobile Combustion
143
158
71
60
64
63
61
Incineration of Waste
2
1
1
1
1
1
1
Petroleum Systems
+
+
+
+
+
+
+
Natural Gas Systems
+
+
+
+
+
+
+
International Bunker Fuelsb
3
3
3
2
2
3
3
+ Does not exceed 0.5 kt.
a Emissions from biomass and biofuel consumption are not included specifically in summing Energy sector totals. Net carbon
fluxes from changes in biogenic carbon reservoirs are accounted for in the estimates for LULUCF.
b Emissions from international bunker fuels are not included in totals. These values are presented for informational purposes
only, in line with the 2006IPCC Guidelines.
Note: Totals by gas may not sum due to independent rounding.
3-6 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Emissions estimates reported in the Energy chapter from fossil fuel combustion and fugitive sources
include those from all 50 states, including Hawaii and Alaska, and the District of Columbia. Emissions
are also included from tribal lands and U.S. Territories to the extent they are known to occur (e.g., coal
mining does not occur in U.S. Territories). For some sources there is a lack of detailed information on
U.S. Territories including some non-C02 emissions from biomass combustion. As part of continuous
improvement efforts, EPA reviews this on an ongoing basis to ensure emission sources are included
across all geographic areas including U.S. Territories if they are occurring. See Annex 5 for more
information on EPA's assessment of the sources not included in this Inventory.
Each year, some emission and sink estimates in the Inventory are recalculated and revised with
improved methods and/or data. In general, recalculations are made to the U.S. greenhouse gas
emission estimates either to incorporate new methodologies or, most commonly, to update recent
historical data. These improvements are implemented consistently across the previous Inventory's time
series (i.e., 1990 to 2022) to ensure that the trend is accurate. Key updates in this year's Inventory
include new data on the activity of battery and plug-in hybrid electric vehicles, the incorporation of
MOVES5 output data to replace MOVES3, updated values for natural gas and petroleum consumed by
all sectors and U.S. Territories for the years 2020 through 2022, updated electricity statistics which
affected commercial sector wood consumption for the years 2014 through 2022, updates for offshore
production sources in Gulf of America federal and state waters, and revisions to GHGRP data
submissions. The impact of these recalculations averaged an increase of 6.4 MMT C02 Eq. (0.1 percent)
per year across the time series. For more information on specific methodological updates, please see
the Recalculations Discussion section for each category in this chapter.
Box 3-1: Uses of EPA's Greenhouse Gas Reporting Program Energy Data
EPA's Greenhouse Gas Reporting Program (GHGRP) dataset continues to be an important resource for
the Inventory, providing not only annual emissions information, but also other annual information, such
as activity data and emission factors that can improve and refine national emission estimates and
trends over time. GHGRP data also allow EPA to disaggregate national inventory estimates in new ways
that can highlight differences across regions and sub-categories of emissions, along with enhancing
application of QA/QC procedures and assessment of uncertainties.
EPA uses annual GHGRP data in a number of Energy sector categories to improve the national estimates
presented in this Inventory consistent with IPCC guidelines (see Box 3-2 of this chapter, and Sections
3.3 Incineration of Waste, 3.4 Coal Mining, 3.6 Petroleum Systems, 3.7 Natural Gas Systems 3.9 C02
Transport, Injection and Storage, and 3.11 Biomass and Biofuels Consumption ). Methodologies used in
EPA's GHGRP are consistent with IPCC guidelines, including higher tier methods. Under EPA's GHGRP,
facilities collect detailed information specific to their operations according to detailed measurement
standards. It should be noted that the definitions and provisions for reporting fuel types in EPA's GHGRP
may differ from those used for national inventory reporting. The Inventory report is a comprehensive
accounting of all emissions from fuel types identified in the IPCC guidelines and provides a separate
reporting of emissions from biomass.
In addition to using GHGRP data to estimate emissions (Sections 3.3 Incineration of Waste, 3.4 Coal
Mining, 3.6 Petroleum Systems, 3.7 Natural Gas Systems, 3.9 C02 Transport, Injection and Storage, and
3.11 Biomass and Biofuels Consumption), EPA also uses the GHGRP fuel consumption activity data in
the Energy sector to disaggregate industrial end-use sector emissions in the category of C02 emissions
from fossil fuel combustion, for use in reporting emissions in common data tables (see Box 3-2). The
Energy 3-7
-------
industrial end-use sector activity data collected for the Inventory (EIA 2024) represent aggregated data
for the industrial end-use sector. EPA's GHGRP collects industrial fuel consumption activity data by
individual categories within the industrial end-use sector. Therefore, GHGRP data are used to provide a
more detailed breakout of total emissions in the industrial end-use sector within that source category.
As indicated in the respective Planned Improvements sections for source categories in this chapter, EPA
continues to examine the uses of facility-level GHGRP data to improve the national estimates presented
in this Inventory. See Annex 9 for more information on use of EPA's GHGRP in the Inventory.
3.1 Fossil Fuel Combustion (Source
Category 1A)
Emissions from the combustion of fossil fuels for energy include the greenhouse gases C02, CH4, and
N20. Given that C02 is the primary gas emitted from fossil fuel combustion and represents the largest
share of U.S. total emissions, C02 emissions from fossil fuel combustion are discussed at the beginning
of this section. An overview of CH4 and N20 emissions from the combustion of fuels in stationary
sources is then presented, followed by fossil fuel combustion emissions for all three gases by sector:
electric power, industrial, residential, commercial, U.S. Territories, and transportation.
Methodologies for estimating C02 emissions from fossil fuel combustion differ from the estimation of
CH4 and N20 emissions from stationary combustion and mobile combustion. Thus, three separate
descriptions of methodologies, uncertainties, recalculations, and planned improvements are provided
at the end of this section. Total C02, CH4, and N20 emissions from fossil fuel combustion are presented
in Table 3-3 and Table 3-4.
Table 3-3: C02, CH4, and N2Q Emissions from Fossil Fuel Combustion (MMT C02 Eq.)
Gas
1990
2005
2019
2020
2021
2022
2023
C02
4,752.2
5,744.1
4,852.6
4,342.3
4,654.6
4,702.8
4,559.4
cm
16.9
14.0
12.6
10.4
10.5
11.3
11.3
n2o
60.1 |
72.51
40.9
36.6
38.9
39.2
35.8
Total
4,829.2
5,830.6
4,906.1
4,389.3
4,704.0
4,753.2
4,606.5
Note: Totals may not sum due to independent rounding.
Table 3-4:
C02, CH4, and N20 Emissions from Fossil Fuel Combustion (kt)
Gas
1990| 2005 2019
2020
2021
2022
2023
CO2
4,752,2341
5,744,1381
4,852,647
4,342,309
4,654,629
4,702,769
4,559,379
CH4
603
501
449
372
375
404
404
N2O
227
274
154
138
146
148
135
3-8 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
C02 from Fossil Fuel Combustion
Carbon dioxide is the primary gas emitted from fossil fuel combustion and represents the largest share
of U.S. total greenhouse gas emissions. Carbon dioxide emissions from fossil fuel combustion are
presented in Table 3-5. In 2023, C02 emissions from fossil fuel combustion decreased by 3.1 percent
relative to the previous year and were 4.1 percent below emissions in 1990 (as shown in Table 3-6). The
decrease in C02 emissions from fossil fuel combustion was a result of a 1.9 percent decrease in fossil
fuel energy use. Carbon dioxide emissions from coal consumption decreased by 18.3 percent (164.1
MMT C02 Eq.) from 2022 to 2023. While carbon dioxide emissions from natural gas use increased by 1.0
percent (17.6 MMT C02 Eq.) and emissions from petroleum use increased by 0.2 percent (3.1 MMT C02
Eq.) from 2022 to 2023. The increase in natural gas consumption and associated emissions in 2023 is
observed mostly in the electric power and industrial sectors, the increase in petroleum use is mainly in
the transportation sector, while the coal decrease is mainly due to reduced use in the electric power
sector. In 2023, C02 emissions from fossil fuel combustion were 4,559.4 MMT C02 Eq. (see Table 3-5).5
Table 3-5: C02 Emissions from Fossil Fuel Combustion by Fuel Type and Sector (MMT
CQ2Eq.)
Fuel/Sector
1990
2005
2019
2020
2021
2022
2023
Coal
1,719.8
2,113.7
1,028.1
835.6
957.4
898.8
734.7
Residential
3.0
0.8
NO
NO
NO
NO
NO
Commercial
12.0
9.3
1.6
1.4
1.4
1.4
1.1
Industrial
157.8
117.8
49.4
43.0
43.0
43.0
36.5
Transportation
NO
NO
NO
NO
NO
NO
NO
Electric Power
1546.5
1982.8
973.5
788.2
910.1
851.5
694.6
U.S. Territories
0.5
3.0
3.6
3.1
2.9
2.9
2.5
Natural Gas
998.6
1,166.2
1,649.3
1,617.2
1,622.1
1,708.2
1,725.8
Residential
237.8
262.2
275.5
256.4
258.6
272.0
247.5
Commercial
142.0
162.9
192.9
173.5
180.4
192.3
182.8
Industrial
407.4
387.8
501.5
491.1
501.2
509.5
514.8
Transportation
36.0
33.1
58.9
58.8
65.2
72.3
71.7
Electric Power
175.4
318.9
616.6
634.8
612.8
659.3
704.5
U.S. Territories
NO
1.3
3.8
2.6
3.9
2.7
4.5
Petroleum
2,033.3
2,463.8
2,174.9
1,889.1
2,074.8
2,095.3
2,098.5
Residential
97.8
95.9
67.4
58.4
59.4
63.2
59.6
Commercial
74.3
54.9
57.2
54.4
55.7
65.4
60.2
Industrial
311.2
342.0
258.9
229.3
236.3
247.1
241.3
Transportation
1,432.9
1,825.5
1,757.7
1,514.2
1,688.4
1,681.2
1,704.7
Electric Power
97.5
98.0
16.2
16.2
17.7
20.5
14.7
U.S. Territories
19.5
47.6
17.5
16.6
17.3
17.8
17.9
Geothermal3
0.5
0.5
0.4
0.4
0.4
0.4
0.4
Electric Power
0.5
0.5
0.4
0.4
0.4
0.4
0.4
5 An additional discussion of fossil fuel emission trends is presented in the Trends in U.S. Greenhouse Gas Emissions
chapter.
Energy 3-9
-------
Fuel/Sector
1990 |
2005
2019
2020
2021
2022
2023
Total
4,752.2
5,744.1
4,852.6
4,342.3
4,654.6
4,702.8
4,559.4
NO (Not Occurring)
a Although not technically a fossil fuel, geothermal energy-related C02 emissions are included for reporting purposes. The source
of C02 is non-condensable gases in subterranean heated water.
Note: Totals may not sum due to independent rounding.
Trends in C02 emissions from fossil fuel combustion are influenced by many long-term and short-term
factors. On a year-to-year basis, the overall demand for fossil fuels in the United States and other
countries generally fluctuates in response to changes in general economic conditions, energy prices,
weather, and the availability of non-fossil alternatives. For example, in a year with increased
consumption of goods and services, low fuel prices, severe summer and winter weather conditions,
nuclear plant closures, and lower precipitation feeding hydroelectric dams, there would likely be
proportionally greater fossil fuel consumption than a year with poor economic performance, high fuel
prices, mild temperatures, and increased output from nuclear and hydroelectric plants.
Longer-term changes in energy usage patterns, however, tend to be more a function of aggregate
societal trends that affect the scale of energy use (e.g., population, number of cars, size of houses, and
number of houses), the efficiency with which energy is used in equipment (e.g., cars, HVAC systems,
power plants, steel mills, and light bulbs), and social planning and consumer behavior (e.g., walking,
bicycling, or telecommuting to work instead of driving).
Carbon dioxide emissions also depend on the source of energy and its carbon intensity. The amount of
carbon in fuels varies significantly by fuel type. For example, coal contains the highest amount of carbon
per unit of useful energy. Petroleum has roughly 75 percent of the carbon per unit of energy as coal, and
natural gas has only about 55 percent.6 Table 3-6 shows annual changes in emissions during the last five
years for coal, petroleum, and natural gas in selected sectors.
Table 3-6: Annual Change in C02 Emissions and Total 2023 C02 Emissions from Fossil
Fuel Combustion for Selected Fuels and Sectors (MMT C02 Eq. and Percent)
Sector Fuel Type
2019 to 2020
2020 to 2021
2021 to 2022
2022 to 2023
Total 2023
Transportation Petroleum
-243.5 -13.9%
174.2 11.5%
-7.1 -0.4%
23.5 1.4%
1,704.7
Electric Power Coal
-185.4 -19.0%
121.9 15.5%
-58.6 -6.4%
-156.9 -18.4%
694.6
Electric Power Natural Gas
18.2 3.0%
-22.1 -3.5%
46.5 7.6%
45.2 6.9%
704.5
Industrial Natural Gas
-10.4 -2.1%
10.0 2.0%
8.3 1.7%
5.3 1.0%
514.8
Residential Natural Gas
-19.1 -6.9%
2.3 0.9%
13.3 5.2%
-24.5 -9.0%
247.5
Commercial Natural Gas
-19.5 -10.1%
6.9 4.0%
12.0 6.6%
-9.5 -5.0%
182.8
Transportation All Fuels"
-243.7 -13.4%
180.6 11.5%
0.0 0.0%
22.9 1.3%
1,776.5
Electric Power All Fuels3
-167.2 -10.4%
101.4 7.0%
-9.3 -0.6%
-117.5 -7.7%
1,414.2
Industrial AllFuelsa
-46.4 -5.7%
17.1 2.2%
19.2 2.5%
-7.1 -0.9%
792.6
Residential All Fuelsa
-28.1 -8.2%
3.2 1.0%
17.1 5.4%
-28.1 -8.4%
307.1
Commercial AllFuelsa
-22.5 -8.9%
8.3 3.6%
21.7 9.1%
-15.0 -5.8%
244.2
AllSectorsa b All Fuels0
-510.3 -10.5%
312.3 7.2%
48.1 1.0%
-143.4 -3.1%
4,559.4
a Includes sector and fuel combinations not shown in this table.
b Includes U.S. Territories.
6 Based on national aggregate carbon content of all coal, natural gas, and petroleum fuels combusted in the United States.
See Annex 2.2 for more details on fuel carbon contents.
3-10 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Note: Totals may not sum due to independent rounding.
As shown in Table 3-6, recent trends in C02 emissions from fossil fuel combustion show a 10.5 percent
decrease from 2019 to 2020, a 7.2 percent increase from 2020 to 2021, a 1.0 percent increase from
2021 to 2022, and a 3.1 percent decrease from 2022 to 2023. These changes contributed to an overall
6.0 percent decrease in C02 emissions from fossil fuel combustion from 2019 to 2023.
Recent trends in C02 emissions from fossil fuel combustion are largely driven by the electric power
sector, which until 2017 has accounted for the largest portion of these emissions. The types of fuels
consumed to produce electricity have changed in recent years. Electric power sector consumption of
natural gas primarily increased due to increased production capacity as natural gas-fired plants
replaced coal-fired plants and increased electricity demand related to heating and cooling needs (EIA
2018; EIA 2024c). Total net electric power generation from all fossil and non-fossil sources decreased by
2.9 percent from 2019 to 2020, increased by 2.7 percent from 2020 to 2021, increased by 3.0 percent
from 2021 to 2022, and decreased by 1.1 percent from 2022 to 2023 (EIA 2025a). Carbon dioxide
emissions from the electric power sector decreased from 2022 to 2023 by 7.7 percent due to increased
production and use of natural gas and decreased production and use of coal for electric power
generation. Carbon dioxide emissions from coal consumption for electric power generation decreased
by 28.7 percent overall since 2019, including an 18.4 percent decrease from 2022 to 2023.
Petroleum use in the transportation sector is another major driver of emissions, representing the largest
source of C02 emissions from fossil fuel combustion in 2023. Emissions from petroleum consumption
for transportation have decreased by 3.0 percent since 2019, even as there was a less than 0.05 percent
increase in VMT over the same time period. As of 2017, the transportation sector is the largest source of
national C02 emissions - whereas in prior years, electric power was the largest source sector.
In the United States, 82.6 percent of the energy used in 2023 was produced through the combustion of
fossil fuels such as petroleum, natural gas, and coal (see Figure 3-4 and Figure 3-5). Specifically,
petroleum supplied the largest share of domestic energy demands, accounting for 37.8 percent of total
U.S. energy used in 2023. Natural gas and coal followed in order of fossil fuel energy demand
significance, accounting for approximately 35.9 percent and 8.7 percent of total U.S. energy used,
respectively. Petroleum was consumed primarily in the transportation end-use sector and the majority
of coal was used in the electric power sector. Natural gas was broadly consumed in all end-use sectors
except transportation (see Figure 3-6) (EIA 2025a). The remaining portion of energy used in 2023 was
supplied by nuclear electric power (8.6 percent) and by a variety of renewable energy sources (8.89
percent), primarily wind energy, hydroelectric power, solar, geothermal and biomass (EIA 2025a).7
7 Renewable energy, as defined in ElA's energy statistics, includes the following energy sources: hydroelectric power,
geothermal energy, biomass, solar energy, and wind energy.
Energy 3-11
-------
Figure 3-4: 2023 U.S. Energy Use by Energy Source
Nuclear Electric Power
8.6%
Note: Totals may not sum due to independent rounding.
Figure 3-5: Annual U.S. Energy Use
120
£ 100
f-v.
co
Ox
o
rH
fN
m
s
in
vD
r-*
CO
&
O
fN
m
in
vD
I-*.
CO
CTv
O
,-1
fN
cn
o\
cn
Q\
0*
o
CTv
O*
©
o
O
o
o
O
o
O
©
*—•
rH
rH
rH
rH
rH
rH
fN
fN
fN
fN
a>
a*
CTv
&
Ov a*
CTv
Ov
a*
o
o
o
o
o
o
o
o
O
o
o
o
O
O
o
o
o
o
o
O
O
O
O
o
rH
T"H
rH rH
H
*"H
H
rsi
fN
cn
OJ
fN
(N
fN
fN
fN
fN
fN
(N
fN
fN
fN
fN
fN
fN
fN
fN
fN
fN
fN
fN
3-12 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Figure 3-6: 2023 C02 Emissions from Fossil Fuel Combustion by Sector and Fuel Type
2,500
2,000
Relative Contribution by Fuel Type
<0.05%
(Geothermal)
2 1,000
¦ Natural Gas
¦ Geothermal
¦ Petroleum
¦ Coal
1,776
500
244
307
0
25
U.S. Territories Commercial
Residential
Industrial
Electricity Transportation
Generation
aAlthough not technically a fossil fuel, geothermal energy-related C02 emissions are included for reporting purposes. The source
of C02 is non-condensable gases in subterranean heated water.
Fossil fuels are generally combusted for the purpose of producing energy for useful heat and work.
During the combustion process, the carbon stored in the fuels is oxidized and emitted as C02 and
smaller amounts of other gases, including CH4, carbon monoxide (CO), and non-methane volatile
organic compounds (NMVOCs).8 These other carbon-containing non-C02 gases are emitted as a
byproduct of incomplete fuel combustion, but are, for the most part, eventually oxidized to C02 in the
atmosphere. Therefore, as per IPCC guidelines, it is assumed that all of the carbon in fossil fuels used to
produce energy is eventually converted to atmospheric C02.
The carbon intensity of the electric power sector is impacted by the amount of non-fossil energy sources
of electricity. The utilization (i.e., capacity factors)9 of nuclear power plants in 2023 remained high at 92
percent. In 2023, nuclear power represented 19 percent of total electricity generation. Since 1990, the
wind and solar power sectors have shown strong growth and have become relatively important sources
of electricity. Between 1990 and 2023, renewable energy generation (in kWh) from solar and wind energy
have increased from 0.1 percent in 1990 to 14 percent of total electricity generation in 2023, which
helped drive the decrease in the carbon intensity of the electricity supply in the United States.
Stationary Combustion
The direct combustion of fuels by stationary sources in the electric power, industrial, commercial, and
residential sectors represent the greatest share of U.S. greenhouse gas emissions. Table 3-7 presents
C02 emissions from fossil fuel combustion by stationary sources. The C02 emitted is closely linked to
the type of fuel being combusted in each sector (see the Methodology section of C02 from Fossil Fuel
Combustion). In addition to C02 emissions, CH4 and N20 are emitted from fossil fuel combustion as
8 See the sections entitled Stationary Combustion and Mobile Combustion in this chapter for information on non-C02 gas
emissions from fossil fuel combustion.
9 The capacity factor equals actual generation divided by maximum potential generation based on net summer capacity.
Net summer capacity is defined as "The maximum output that generating equipment can supply to system load, as
demonstrated by a multi-hour test, at the time of summer peak demand (period of June 1 through September 30)" (EIA
2024g). Data for both the generation and net summer capacity are from EIA (2024a).
Energy 3-13
-------
well. Table 3-8 and Table 3-9 present CH4 and N20 emissions from the combustion of fuels in stationary
sources. The CH4 and N20 emissions are linked to the type of fuel being combusted as well as the
combustion technology (see the Methodology section for CH4 and N20 from Stationary Combustion).
Table 3-7: C02 Emissions from Stationary Fossil Fuel Combustion (MMT C02 Eq.)
Sector/Fuel Type
1990
2005
2019
2020
2021
2022
2023
Electric Power
1,820.0
2,400.1
1,606.7
1,439.6
1,540.9
1,531.7
1,414.2
Coal
1546.5
1982.8
973.5
788.2
910.1
851.5
694.6
Natural Gas
175.4
318.9
616.6
634.8
612.8
659.3
704.5
Fuel Oil
97.5
98.0
16.2
16.2
17.7
20.5
14.7
Geothermal
0.5
0.5
0.4
0.4
0.4
0.4
0.4
Industrial
876.5
847.6
809.8
763.4
780.5
799.7
792.6
Coal
157.8
117.8
49.4
43.0
43.0
43.0
36.5
Natural Gas
407.4
387.8
501.5
491.1
501.2
509.5
514.8
Fuel Oil
311.2
342.0
258.9
229.3
236.3
247.1
241.3
Residential
338.6
358.9
342.9
314.8
318.0
335.2
307.1
Coal
3.0
0.8
NO
NO
NO
NO
NO
Natural Gas
237.8
262.2
275.5
256.4
258.6
272.0
247.5
Fuel Oil
97.8
95.9
67.4
58.4
59.4
63.2
59.6
Commercial
228.3
227.1
251.7
229.3
237.5
259.2
244.2
Coal
12.0
9.3
1.6
1.4
1.4
1.4
1.1
Natural Gas
142.0
162.9
192.9
173.5
180.4
192.3
182.8
Fuel Oil
74.3
54.9
57.2
54.4
55.7
65.4
60.2
U.S. Territories
20.0
51.9
24.8
22.3
24.1
23.5
24.9
Coal
0.5
3.0
3.6
3.1
2.9
2.9
2.5
Natural Gas
NO
1.3
3.8
2.6
3.9
2.7
4.5
Fuel Oil
19.5
47.6
17.5
16.6
17.3
17.8
17.9
Total
3,283.3
3,885.6
3,036.0
2,769.4
2,901.1
2,949.2
2,782.9
NO (Not Occurring)
Note: Totals may not sum due to independent rounding.
Table 3-8: CH4 Emissions from Stationary Combustion (MMT C02 Eq.)
Sector/Fuel Type
1990
2005
2019
2020
2021
2022
2023
Electric Power
0.5
1.0
1.4
1.4
1.4
1.5
1.5
Coal
0.4
0.4
0.2
0.2
0.2
0.2
0.2
Fuel Oil
0.0
0.0
+
+
+
+
+
Natural gas
0.1
0.5
1.2
1.2
1.2
1.3
1.3
Wood
0.0
0.0
+
+
+
+
+
Industrial
2.1
1.9
1.7
1.6
1.6
1.6
1.5
Coal
0.5
0.3
0.1
0.1
0.1
0.1
0.1
Fuel Oil
0.2
0.2
0.2
0.1
0.1
0.2
0.1
Natural gas
0.2
0.2
0.3
0.2
0.3
0.3
0.3
Wood
1.2
1.2
1.1
1.1
1.1
1.0
1.0
3-14 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Sector/Fuel Type
1990
2005
2019
2020
2021
2022
2023
Commercial
1.2
1.2
1.3
1.2
1.3
1.3
1.3
Coal
0.0
0.0
+
+
+
+
+
Fuel Oil
0.3
0.2
0.2
0.2
0.2
0.3
0.2
Natural gas
0.4
0.4
0.5
0.4
0.5
0.5
0.5
Wood
0.5
0.6
0.6
0.6
0.6
0.6
0.6
Residential
5.9
4.5
5.3
3.6
3.6
4.3
4.5
Coal
0.3
0.1
NO
NO
NO
NO
NO
Fuel Oil
0.4
0.4
0.3
0.2
0.2
0.3
0.2
Natural Gas
0.6
0.7
0.7
0.6
0.6
0.7
0.6
Wood
4.6
3.4
4.4
2.8
2.7
3.4
3.6
U.S. Territories
0.0
0.1
+
+
+
+
+
Coal
0.0
0.0
+
+
+
+
+
Fuel Oil
0.0
0.1
+
+
+
+
+
Natural Gas
NO
0.0
+
+
+
+
+
Wood
NE
NE
NE
NE
NE
NE
NE
Total
9.7
8.8
9.8
7.9
7.9
8.7
8.8
+ Does not exceed 0.05 MMT C02 Eq.
NO (Not Occurring)
NE (Not Estimated)
Note: Totals may not sum due to independent rounding.
Table 3-9: N20 Emissions from Stationary Combustion (MMT C02 Eq.)
Sector/Fuel Type
1990
2005
2019
2020
2021
2022
2023
Electric Power
18.2
26.7
18.8
17.5
19.0
19.4
16.6
Coal
17.9
24.9
14.8
13.5
15.1
15.2
12.1
Fuel Oil
0.1
0.1
+
+
+
+
+
Natural Gas
0.3
1.7
3.9
4.0
3.9
4.2
4.4
Wood
+
+
+
+
+
+
+
Industrial
2.8
2.6
2.2
2.0
2.1
2.0
1.9
Coal
0.7
0.5
0.2
0.2
0.2
0.2
0.2
Fuel Oil
0.5
0.5
0.3
0.3
0.3
0.3
0.3
Natural Gas
0.2
0.2
0.2
0.2
0.2
0.2
0.2
Wood
1.5
1.5
1.4
1.4
1.4
1.3
1.2
Commercial
0.3
0.3
0.3
0.3
0.3
0.3
0.3
Coal
+
+
+
+
+
+
+
Fuel Oil
0.2
0.1
0.1
0.1
0.1
0.1
0.1
Natural Gas
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Wood
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Residential
0.9
0.8
0.8
0.6
0.6
0.7
0.7
Coal
+
+
NO
NO
NO
NO
NO
Fuel Oil
0.2
0.2
0.2
0.1
0.1
0.1
0.1
Natural Gas
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Wood
0.6
0.4
0.5
0.3
0.3
0.4
0.5
Energy 3-15
-------
Sector/Fuel Type
1990
2005
2019
2020
2021
2022
2023
U.S. Territories
+
0.1
0.1
0.1
0.1
0.1
0.1
Coal
+
+
+
+
+
+
+
Fuel Oil
+
0.1
+
+
+
+
+
Natural Gas
NO
+
+
+
+
+
+
Wood
NE
NE
NE
NE
NE
NE
NE
Total
22.3
30.5
22.1
20.5
22.0
22.6
19.6
+ Does not exceed 0.05 MMT C02 Eq.
NO (Not Occurring)
NE (Not Estimated)
Note: Totals may not sum due to independent rounding.
Fossil Fuel Combustion Emissions by Sector
Table 3-10 provides an overview of the C02, CH4, and N20 emissions from fossil fuel combustion by
sector, including transportation, electric power, industrial, residential, commercial, and U.S. Territories.
Table 3-10: C02, CH4, and N20 Emissions from Fossil Fuel Combustion by Sector (MMT
C02 Eq.)
End-Use Sector
1990
2005
2019
2020
2021
2022
2023
Transportation
1,513.9
1,905.8
1,838.2
1,591.5
1,772.9
1,772.7
1,795.2
CO2
1,468.9
1,858.6
1,816.6
1,573.0
1,753.5
1,753.6
1,776.5
cm
7.2
5.3
2.8
2.5
2.5
2.6
2.5
N2O
37.8
41.9
18.7
16.0
16.8
16.5
16.2
Electric Power
1,838.7
2,427.8
1,626.9
1,458.5
1,561.3
1,552.6
1,432.3
CO2
1,820.0
2,400.1
1,606.7
1,439.6
1,540.9
1,531.7
1,414.2
CH4
0.5
1.0
1.4
1.4
1.4
1.5
1.5
N2O
18.2
26.7
18.8
17.5
19.0
19.4
16.6
Industrial
881.3
852.2
813.7
767.1
784.2
803.3
796.0
CO2
876.5
847.6
809.8
763.4
780.5
799.7
792.6
CH4
2.1
1.9
1.7
1.6
1.6
1.6
1.5
N2O
2.8
2.6
2.2
2.0
2.1
2.0
1.9
Residential
345.4
364.2
349.1
319.0
322.3
340.2
312.2
CO2
338.6
358.9
342.9
314.8
318.0
335.2
307.1
CH4
5.9
4.5
5.3
3.6
3.6
4.3
4.5
N2O
0.9
0.8
0.8
0.6
0.6
0.7
0.7
Commercial
229.8
228.6
253.4
230.8
239.1
260.8
245.7
CO2
228.3
227.1
251.7
229.3
237.5
259.2
244.2
CH4
1.2
1.2
1.3
1.2
1.3
1.3
1.3
N2O
0.3
0.3
0.3
0.3
0.3
0.3
0.3
U.S. Territories3
20.1
52.1
24.9
22.4
24.2
23.6
25.0
Total
4,829.2
5,830.6
4,906.1
4,389.3
4,704.0
4,753.2
4,606.5
a U.S. Territories are not apportioned by sector, and emissions shown in the table are total greenhouse gas emissions from all fuel
combustion sources.
Note: Totals may not sum due to independent rounding.
3-16 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Other than the greenhouse gases C02, CH4, and N20, gases emitted from stationary combustion include
the greenhouse gas precursors nitrogen oxides (NOx), CO, NMVOCs, and sulfur dioxide (S02). Methane
and N20 emissions from stationary combustion sources depend upon fuel characteristics and the size
and vintage of combustion device, along with combustion technology, pollution control equipment,
ambient environmental conditions, and operation and maintenance practices. Nitrous oxide emissions
from stationary combustion are closely related to air-fuel mixes and combustion temperatures, as well
as the characteristics of any pollution control equipment that is employed. Methane emissions from
stationary combustion are primarily a function of the CH4 content of the fuel and combustion efficiency.
Mobile combustion also produces emissions of CH4, N20, and greenhouse gas precursors including NOx,
CO, and NMVOCs. As with stationary combustion, N20 and NOx emissions from mobile combustion are
closely related to fuel characteristics, air-fuel mixes, combustion temperatures, and the use of pollution
control equipment. Nitrous oxide from mobile sources, in particular, can be formed by the catalytic
processes used to control NOx, CO, and hydrocarbon emissions. Carbon monoxide emissions from mobile
combustion are significantly affected by combustion efficiency and the presence of post-combustion
emission controls. Carbon monoxide emissions are highest when air-fuel mixtures have less oxygen than
required for complete combustion. These emissions occur especially in vehicle idle, low speed, and cold
start conditions. Methane and NMVOC emissions from motor vehicles are a function of the CH4 content of
the motor fuel, the amount of hydrocarbons passing uncombusted through the engine, and any post-
combustion control of hydrocarbon emissions (such as catalytic converters).
An alternative method of presenting combustion emissions is to allocate emissions associated with
electric power to the sectors in which it is used. Four end-use sectors are defined: transportation,
industrial, residential, and commercial. In Table 3-11 below, electric power emissions have been
distributed to each end-use sector based upon the sector's share of national electricity use, with the
exception of CH4 and N20 from transportation electricity use.10 Emissions from U.S. Territories are also
calculated separately due to a lack of end-use-specific consumption data.11 This method assumes that
emissions from combustion sources are distributed across the four end-use sectors based on the ratio
of electricity use in that sector. The results of this alternative method are presented in Table 3-11.
Table 3-11: C02, CH4, and N20 Emissions from Fossil Fuel Combustion by End-Use
Sector with Electricity Emissions Distributed (MMT C02 Eq.)
End-Use Sector
1990
2005
2019
2020
2021
2022
2023
Transportation
1,517.0
1,910.5
1,842.4
1,594.9
1,776.8
1,777.2
1,800.2
CO2
1,472.0
1,863.3
1,820.9
1,576.4
1,757.5
1,758.0
1,781.5
cm
7.2
5.3
2.8
2.5
2.5
2.6
2.5
n2o
37.8
41.9
18.7
16.0
16.8
16.5
16.2
Industrial
1,574.8
1,597.0
1,285.1
1,182.2
1,235.3
1,246.5
1,205.7
CO2
1,562.9
1,584.0
1,275.3
1,173.3
1,225.7
1,236.9
1,197.1
CH4
2.2
2.2
2.1
2.0
2.0
2.0
1.9
N2O
9.7
10.8
7.6
7.1
7.6
7.6
6.7
10 Separate calculations are performed for transportation-related ChU and N2O. The methodology used to calculate these
emissions is discussed in the Mobile Combustion section.
n U.S. Territories (including American Samoa, Guam, Puerto Rico, U.S. Virgin Islands, Wake Island, and other outlying U.S.
Pacific Islands) consumption data obtained from EIA are only available at the aggregate level and cannot be broken out
by end-use sector. The distribution of emissions to each end-use sector for the 50 states does not apply to territories
data.
Energy 3-17
-------
End-Use Sector
1990 2005 2019 2020 2021 2022 2023
Residential 944.2 1,230.1 940.7 872.1 902.9 914.4 827.3
C02 931.3 1,214.9 927.1 860.7 891.1 901.6 815.6
CH4 6.01 4.91 5.8 4.2 4.2 4.9 5.0
N2O 6.9 10.3 7.7 7.3 7.7 7.9 6.7
Commercial 773.1 1,040.9 813.1 717.5 764.6 791.5 748.2
CO2 766.0 1,030.1 804.5 709.7 756.2 782.7 740.3
CH4 1.31 1.51 1.8 1.7 1.7 1.8 1.8
N2O 5.7 9.3 6.8 6.2 6.7 7.0 6.1
U.S. Territories3 20.1 52.1 24.9 22.4 24.2 23.6 25.0
Total 4,829.2 5,830.6 4,906.1 4,389.3 4,704.0 4,753.2 4,606.5
a U.S. Territories are not apportioned by sector, and emissions are total greenhouse gas emissions from all fuel combustion
sources.
Notes: Totals may not sum due to independent rounding. Emissions from fossil fuel combustion by electric power are allocated
based on aggregate national electricity use by each end-use sector.
Electric Power Sector
The process of generating electricity is the largest stationary source of C02 emissions in the United
States, representing 28.8 percent of total C02 emissions from all C02 emissions sources across the
United States and 31.0 percent of C02 emissions from fossil fuel combustion in 2023. Methane and N20
accounted for a small portion of total greenhouse gas emissions from electric power generation,
representing 0.1 percent and 1.2 percent, respectively. Methane and N20 from electric power
represented 13.4 and 46.5 percent of total CH4 and N20 emissions from fossil fuel combustion in 2023,
respectively.
For the underlying energy data used in this chapter, the Energy Information Administration (EIA) places
electric power generation into three functional categories: the electric power sector, the commercial
sector, and the industrial sector. The energy use and emissions associated with the electric power
sector are included here. As defined by EIA, the electric power sector consists of electric utilities and
independent power producers whose primary business is the production of electricity. This includes
both regulated utilities and non-utilities (e.g., independent power producers, qualifying co-generators,
and other small power producers). Energy use and emissions associated with electric generation in the
commercial and industrial sectors is reported in those other sectors where the producer of the power
indicates that its primary business is something other than the production of electricity.12
Total greenhouse gas emissions from the electric power sector have decreased by 22.1 percent since
1990. From 1990 to 2007, electric power sector emissions increased by 33 percent, driven by a
significant increase in electricity demand (39 percent) while the carbon intensity of electricity generated
showed a modest decline (2.1 percent). From 2008 to 2023, as electricity demand increased by 3.8
percent, electric power sector emissions decreased by 40 percent, driven by a significant drop (29
percent) in the carbon intensity of electricity generated. Overall, the carbon intensity of the electric
power sector, in terms of C02 Eq. per QBtu, decreased by 31 percent from 1990 to 2023 with additional
trends detailed in Box 3-4. This trend is shown in Figure 3-7. This recent decarbonization of the electric
power sector is a result of several key drivers.
12 Utilities primarily generate power for the U.S. electric grid for sale to retail customers. Non-utilities typically generate
electricity for sale on the wholesale electricity market (e.g., to utilities for distribution and resale to retail customers).
Where electricity generation occurs outside the ElA-defined electric power sector, it is typically for the entity's own use.
3-18 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Coal-fired electric generation (in kilowatt-hours [kWh]) decreased from 54 percent of generation in 1990
to 17 percent in 2023.13 This corresponded with an increase in natural gas generation and renewable
energy generation, largely from wind and solar energy. Natural gas generation (in kWh) represented 11
percent of electric power generation in 1990 and increased over the 34-year period to represent 42
percent of electric power sector generation in 2023 (see Table 3-12). Natural gas has a much lower
carbon content than coal and is generated in power plants that are generally more efficient in terms of
kWh produced per Btu of fuel combusted, which has led to lower emissions as natural gas replaces
coal-powered electricity generation. Natural gas and coal used in the United States in 2023 had an
average carbon content of 14.43 MMT C/Qbtu and 26.15 MMT C/Qbtu respectively.
Table 3-12: Electric Power Generation by Fuel Type (Percent)
Fuel Type
1990
2005
2019
2020
2021
2022
2023
Coal
54.1%
51.1%
24.2%
19.9%
22.6%
20.3%
16.6%
Natural Gas
10.7%
17.5%
37.3%
39.5%
37.3%
38.8%
42.2%
Nuclear
19.9%
20.0%
20.4%
20.5%
19.7%
18.9%
19.2%
Renewables
11.3%
8.3%
17.6%
19.5%
19.8%
21.4%
21.5%
Petroleum
4.1%
3.0%
0.4%
0.4%
0.5%
0.5%
0.4%
Other Gases®
+%
0.1%
0.1%
0.1%
0.1%
0.1%
0.1%
Net Electricity Generation (Billion kWh)b
2,905
3,902
3,966
3,851
3,955
4,076
4,031
+ Does not exceed 0.05 percent.
a Other gases include blastfurnace gas, propane gas, and other manufactured and waste gases derived from fossil fuels.
b Represents net electricity generation from the electric power sector. Excludes net electricity generation from commercial and
industrial combined-heat-and-power and electricity-only plants. Does not include electricity generation from purchased
steam as the fuel used to generate the steam cannot be determined.
In 2023, C02 emissions from the electric power sector decreased by 7.7 percent relative to 2022. This
decrease in C02 emissions was primarily driven by a decrease in coal consumed to produce electricity
in the electric power sector. Consumption of coal for electric power decreased by 18.4 percent while
consumption of natural gas increased 6.9 percent from 2022 to 2023, leading to an overall decrease in
emissions. There has also been a rapid increase in renewable electricity generation in the electric power
sector in recent years and electricity generation from renewable sources remained relatively flat from
2022 to 2023 (see Table 3-12). A decrease in coal-fired electricity generation and increases in natural gas
and renewable energy sources for electricity generation contributed to a decoupling of emissions trends
from electric power generation trends starting around 2005 (EIA 2024g) (see Figure 3-7).
The shift from coal to natural gas comes from a variety of factors including the increase in natural gas
generation, particularly between 2005 and 2020 and the relative prices of using coal vs. natural gas for
electricity generation (EIA 2022a; EIA 2022b). The aging coal fleet and coal plant retirements also
contributes to why this trend is continuing (EIA 2025b). From 2022 to 2023, coal consumption
decreased by 18 percent while natural gas consumption increased by 7 percent.
Also, in 2023 the Petra Nova project sequestered 359,840 metric tons of C02 from a coal fired power
plant. These emissions have been netted out of the results shown in this chapter for electric power
sector coal C02 emissions. More information on C02 transport, injection, and geologic sequestration
can be found in Section 3.9.
13 Values represent electricity net generation from the electric power sector (EIA 2024a).
Energy 3-19
-------
Renewable energy generation (in kWh) from wind and solar energy increased from 0.1 percent of total
generation in 1990 to 5 percent in 2015 and increased at a faster pace to 15 percent of total generation
in 2023. The decrease in carbon intensity occurred even as total electricity retail sales increased 43
percent, from 2,713 billion kWh in 1990 to 3,874 billion kWh in 2023.
Figure 3-7: Fuels Used in Electric Power Generation and Total Electric Power Sector
C02 Emissions
Electricity was used primarily in the residential, commercial, and industrial end-use sectors for lighting,
heating, electric motors, appliances, electronics, and air conditioning (see Figure 3-8). Note that
transportation is an end-use sector as well but is not shown in Figure 3-8 due to the sector's relatively
low percentage of electricity use. The Transportation Sector and Mobile Combustion section provides a
break-out of C02 emissions from electricity use in the transportation end-use sector.
3-20 inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Figure 3-8: Electric Power Retail Sales by End-Use Sector
In 2023, electricity sales to the residential and commercial end-use sectors, as presented in Figure 3-8,
decreased by 3.9 percent and increased 1.2 percent relative to 2022, respectively. Electricity sales to
the industrial sector in 2023 decreased by approximately 1.1 percent relative to 2022. The sections
below describe end-use sector energy use in more detail. Overall, in 2023, the amount of electricity
retail sales (in kWh) decreased by 1.3 percent relative to 2022.
Industrial Sector
Industrial sector C02, CH4, and N20 emissions accounted for 17, 13, and 5 percent of C02, CH4, and
N20 emissions from fossil fuel combustion, respectively, in 2023. Carbon dioxide, CH4, and N20
emissions resulted from the direct consumption of fossil fuels for steam and process heat production.
The industrial end-use sector, per the underlying energy use data from EIA, includes activities such as
manufacturing, construction, mining, and agriculture. The largest of these activities in terms of energy
use is manufacturing, of which six industries—petroleum refineries, chemicals, paper, primary metals,
food, and nonmetallic mineral products—represent the majority of the energy use (EIA 2025a; EIA
2009b).
There are many dynamics that impact emissions from the industrial sector including economic activity,
changes in the make-up of the industrial sector, changes in the emissions intensity of industrial
processes, and weather-related impacts on heating and cooling of industrial buildings.14Structural
changes within the U.S. economy that lead to shifts in industrial output away from energy-intensive
manufacturing products to less energy-intensive products (e.g., from steel to computer equipment)
have had a significant effect on industrial emissions.
From 2022 to 2023, total industrial production and manufacturing output increased by 0.2 percent (FRB
2023). Over this period, output increased slightly across production indices for Food, Nonmetallic
Mineral Products, and Paper. Production of chemicals, petroleum refineries, and primary metals
declined slightly between 2022 and 2023 (see Figure 3-9). From 2022 to 2023, total energy use in the
14 Some commercial customers are large enough to obtain an industrial price for natural gas and/or electricity and are
consequently grouped with the industrial end-use sector in U.S. energy statistics. These misclassifications of large
commercial customers likely cause the industrial end-use sector to appear to be more sensitive to weather conditions.
Energy 3-21
-------
industrial sector decreased by 0.9 percent, driven mainly by a 15.3 percent decrease in coal
consumption in the industrial sector. Consumption of renewables decreased 3.0 percent from 2022 to
2023. Due to the relative increases and decreases of individual indices there was an increase in natural
gas and a decrease in electricity used by this sector (see Figure 3-10). In 2023, C02, CH4, and N20
emissions from fossil fuel combustion and electricity use within the industrial end-use sector totaled
1,205.7 MMT C02 Eq., a 3.3 percent decrease from 2022 emissions.
Through EPA's Greenhouse Gas Reporting Program (GHGRP), specific industrial sector trends can be
discerned from the overall total EIA industrial fuel consumption data used for these calculations. For
example, from 2022 to 2023, the underlying EIA data showed increased consumption of natural gas,
decreased consumption of petroleum, and decreased consumption of coal in the industrial sector. The
GHGRP data highlights that several industries contributed to these trends, including chemical
manufacturing; pulp, paper and print; food processing, beverages and tobacco; minerals
manufacturing; and agriculture-forest-fisheries.15
15 Further details on industrial sector combustion emissions are provided by EPA's GHGRP. See
http://ghgdata.epa.gov/ghgp/main.do.
3-22 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Figure 3-9: Industrial Production Indices (Index 2017=100)
Energy 3-23
-------
Figure 3-10: Fuels and Electricity Used in Industrial Sector, Industrial Output, and
Total Sector C02 Emissions (Including Electricity)
Despite the growth in industrial output (65 percent) and the overall U.S. economy (125 percent) from
1990 to 2023, direct C02 emissions from fossil fuel combustion in the industrial sector decreased by 9.6
percent over the same time series. A number of factors are assumed to result in decoupling of growth in
industrial output from industrial greenhouse gas emissions, for example: (1) more rapid growth in output
from less energy-intensive industries relative to traditional manufacturing industries, and (2) energy-
intensive industries such as steel are employing new methods, such as electric arc furnaces, that are
less carbon-intensive than the older methods.
Box 3-2: Uses of Greenhouse Gas Reporting Program Data and Improvements in
Reporting Emissions from Industrial Sector Fossil Fuel Combustion
As described in the calculation methodology, total fossil fuel consumption for each year is based on
aggregated end-use sector consumption published by the EIA. The availability of facility-level
combustion emissions through EPA's GHGRP has provided an opportunity to better characterize the
industrial sector's energy consumption and emissions in the United States, through a disaggregation of
ElA's industrial sector fuel consumption data from select industries.
For GHGRP 2010 through 2023 reporting years, facility-level fossil fuel combustion emissions reported
through EPA's GHGRP were categorized and distributed to specific industry types by utilizing facility-
reported NAICS codes (as published by the U.S. Census Bureau). As noted previously in this report, the
definitions and provisions for reporting fuel types in EPA's GHGRP include some differences from the
Inventory's use of EIA national fuel statistics for national inventory reporting. The IPCC has provided
3-24 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
guidance on aligning facility-level reported fuels and fuel types published in national energy statistics,
which guided this exercise.16
As with previous Inventory reports, the current effort represents an attempt to align, reconcile, and
coordinate the facility-level reporting of fossil fuel combustion emissions under EPA's GHGRP with the
national-level approach presented in this report. Progress was made on certain fuel types for specific
industries and has been included in common data tables.17 The efforts in reconciling fuels focus on
standard, common fuel types (e.g., natural gas, distillate fuel oil) where the fuels in ElA's national
statistics aligned well with facility-level GHGRP data. For these reasons, the current information should
be viewed as an initial attempt at this exercise. Additional efforts will be made for future Inventory
reports to improve the mapping of fuel types and examine ways to reconcile and coordinate any
differences between facility-level data and national statistics. The current analysis includes the full time
series. Analyses were conducted linking GHGRP facility-level reporting with the information published
by EIA in its MECS data in order to disaggregate the full 1990 through 2023 time period. It is believed that
the current analysis has led to improvements in the presentation of data in the Inventory, but further
work will be conducted, and future improvements will be realized in subsequent Inventory reports. This
includes incorporating the latest MECS data as it becomes available.
Residential and Commercial Sectors
Total direct and indirect emissions from the residential and commercial sectors have generally
decreased since 2005. This is due in part to reduced electricity sector emissions intensity which results
in decreased indirect emissions from electricity use. For example, starting around 2014, total energy use
and emissions begin to decouple due to decarbonization of the electric power sector (see Figure 3-11).
Short-term trends in the residential and commercial sectors are often correlated with seasonal
fluctuations in energy use caused by weather conditions, rather than prevailing economic conditions.
Population growth and a trend towards larger houses has led to increasing energy use over the time
series, while population migration to warmer areas and improved energy efficiency and building
insulation have countered the increase in energy use in recent years. The shift toward energy-efficient
products and more stringent energy efficiency standards for household equipment has contributed to a
decrease in energy demand in households (EIA 2023b; Nadel, et al. 2015), resulting in a decrease in
energy-related emissions in the residential sector since 1990.
16 See Section 4 "Use of Facility-Level Data in Good Practice National Greenhouse Gas Inventories" of the IPCC meeting
report, and specifically the section on using facility-level data in conjunction with energy data, at http://www.ipcc-
nggip.iges.or.jp/public/tb/TFI Technical Bulletin 1.pdf.
17 See https://www.epa.gov/ghgemissions/inventory-us-greenhouse-gas-emissions-and-sinks.
Energy 3-25
-------
Figure 3-11: Fuels and Electricity Used in Residential and Commercial Sectors,
Heating and Cooling Degree Days, and Total Sector C02 Emissions (Including
Electricity)
25,000
20,000
15,000
8i
Z3
>¦
10,000
5,000
I Coal (TBtu)
I Renewable Energy Sources (TBtu)
I Petroleum (TBtu)
I Natural Gas (TBtu)
I Electricity Use (TBtu)
¦ Sector CCh Emissions (Index vs. 1990) [Right Axis]
¦ Heating and Cooling Degree Days (Index vs. 1990) [Right Axis]
180
160
140
120
100
80
60
40
20
CT\o\c7\ONCTvcrvONcricr>cr*oooooooooQ'>-«'«-H-^-«-^H-^H'^-<'I-«'«-H,-«,-
-------
Box 3-3: Weather and Non-Fossil Energy Effects on C02 Emissions from Fossil Fuel
Combustion Trends
The United States in 2023 experienced a warmer winter overall compared to 2022, with a 10.4 percent
decrease in heating degree days, and 2023 heating degree days were 12.4 percent below normal18 (see
Figure 3-12). Along with a warmer winter, 2023 experienced a cooler summer than 2022, with cooling
degree days 5.2 percent below 2022. However, cooling degree days were still 10.7 percent above normal
(see Figure 3-13) (EIA 2025a).19 Warmer summers can lead to increased energy use and associated
emissions to cool building spaces in the residential and commercial sectors, mostly from electricity
use. Whereas, warmer winter conditions can lead to an overall decrease in mainly direct energy use and
emissions from fossil fuel combustion in the residential and commercial sectors.
Figure 3-12: Annual Deviations from Normal Heating Degree Days for the United States
(1970-2023, Index Normal = 100)
03
E
c
o
Q
4—'
c
CL)
u
1—
QJ
Q_
30
20
10
¦10
-20
-30
Normal
(4,341 Heating Degree Days)
99% Confidence
Note: Climatological normal data are highlighted in dark red. Statistical confidence interval for "normal"
climatology period of 1991 through 2020.
T-HmLnr-vCTt!—imLnr\o>i-'—imLnr-vCTi
r\r---r\r-vr\cococococochcxicr>chcri
o^o^o^o^o^chchchchcho^o^o^G^o^
oooooi-iT-ii-ii-iT-irsirvi
oooooooooooo
fNrMfNllNfNOJfNf\lfNfN(NfM
18 The National Centers for Environmental Information of NOAA generates official U.S. climate normals every 10 years in
keeping with the needs of the user community and the requirements of the World Meteorological Organization (WMO)
and National Weather Service (NWS). The 1991-2020 U.S. Climate Normals are the latest in a series of decadal normals
first produced in the 1950s. See https://www.ncei.noaa.gov/products/land-based-station/us-climate-normals. The
variation in these normals during this time period was +16 percent and +27 percent for heating and cooling degree days,
respectively (99 percent confidence interval).
19 Degree days are relative measurements of outdoor air temperature. Heating degree days are deviations of the mean daily
temperature below 65 degrees Fahrenheit, while cooling degree days are deviations of the mean daily temperature above
65 degrees Fahrenheit. Heating degree days have a considerably greater effect on direct energy demand and related
emissions than do cooling degree days. Excludes Alaska and Hawaii.
Energy 3-27
-------
Figure 3-13: Annual Deviations from Normal Cooling Degree Days for the United States
(1970-2023, Index Normal = 100)
40
30
20
CO 4
£• 1°
ro
0
0)
CD
S1 0
Q
cn
1 -10
o
O
-20
-30
-40
Normal
(1,332 Cooling Degree Days)
99% Confidence
Note: Climatological normal data are highlighted dark blue. Statistical confidence interval for "normal"
climatology period of 1991 through 2020.
7—icni-nr-sCi!—iroLnr-s,cx»'—jroLnr-scri
r\i\r\r\r\oocooococochcricrtcrich
i—iroLnr*Ncr»7—•fOLDr-sCii—iro
OOOOOi-Ht-Ht—IfNfN
oooooooooooo
rMfNr\ir\irMrsir\ir\ir\irsjr\ir\i
U.S. Territories
Emissions from U.S. Territories are based on the fuel consumption in American Samoa, Guam, Puerto
Rico, U.S. Virgin Islands, Wake Island, and other outlying U.S. Pacific Islands. As described in the
Methodology section of C02 from Fossil Fuel Combustion, this data is collected separately from the
sectoral-level data available for the general calculations. As sectoral information is not available for U.S.
Territories, C02, CH4, and N20 emissions are not presented for U.S. Territories in the tables above by
sector, though the emissions will occur across all sectors and sources including stationary,
transportation and mobile combustion sources.
Transportation Sector and Mobile Combustion
This discussion of transportation emissions follows the alternative method of presenting combustion
emissions by allocating emissions associated with electricity generation to the transportation end-use
sector, as presented in Table 1-9. Table 1-8 presents direct C02, CH4, and N20 emissions from all
transportation sources (i.e., excluding emissions allocated to electricity consumption in the
transportation end-use sector).
The transportation end-use sector and other mobile combustion accounted for 1,800.2 MMT C02 Eq. in
2023, which represented 39 percent of C02 emissions from fossil fuel combustion, 23 percent of CH4
3-28 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
emissions from fossil fuel combustion, and 46 percent of N20 emissions from fossil fuel combustion.20
Fuel purchased in the U.S. for international aircraft and marine travel accounted for an additional 97.0
MMT C02 Eq. in 2023; these emissions are recorded as international bunkers and are not included in
U.S. totals in line with IPCC guidelines.
Transportation End-Use Sector
From 1990 to 2019, transportation emissions from fossil fuel combustion rose by 21 percent, followed
by a reduction of 13 percent from 2019 to 2020, and an increase of 13 percent from 2020 to 2023.
Overall, from 1990 to 2023, transportation emissions from fossil fuel combustion increased by 19
percent. The increase in transportation emissions from fossil fuel combustion from 1990 to 2023 was
due, in large part, to increased demand for travel (see Figure 3-14). The number of vehicle miles traveled
by light-duty motor vehicles (passenger cars and light-duty trucks) increased 49 percent from 1990 to
2023, as a result of a confluence of factors including population growth, economic growth, urban
sprawl, and relatively low fuel prices over much of this period. Between 2019 and 2020, emissions from
light-duty vehicles fell by 12 percent, primarily the result of the COVID-19 pandemic and associated
restrictions, such as people working from home and traveling less.
Commercial aircraft emissions decreased by 5 percent between 2019 and 2023 and have decreased 7
percent since 2007 (FAA 2022 and DOT 1991 through 2025).21 Decreases in jet fuel emissions (excluding
bunkers) started in 2007, due in part to improved operational efficiency that results in more direct flight
routing, improvements in aircraft and engine technologies to reduce fuel burn and emissions, and the
accelerated retirement of older, less fuel-efficient aircraft; however, the sharp decline in commercial
aircraft emissions from 2019 to 2020 and their gradual recovery since is primarily due to COVID-19
impacts on scheduled passenger air travel.
Almost all of the energy consumed for transportation was supplied by petroleum-based products, with
more than half being related to gasoline consumption in automobiles and other highway vehicles. Other
fuel uses, especially diesel fuel for freight trucks and jet fuel for aircraft, accounted for the remainder.
The primary driver of transportation-related emissions was C02 from fossil fuel combustion, which
increased by 21 percent from 1990 to 2023. Annex 3.2 presents the total emissions from all
transportation and mobile sources, including C02, N20, CH4, and HFCs.
20 Note that these totals include CO2, CH4 and N2O emissions from some sources in the U.S. Territories (ships and boats,
recreational boats, non-transportation mobile sources) and CH4 and N2O emissions from transportation rail electricity.
21 Commercial aircraft consists of passenger aircraft, cargo, and other chartered flights.
Energy 3-29
-------
Figure 3-14! Fuels Used in Transportation Sector, On-road VMT, and Total Sector C02
Emissions
Other Fuels (TBtu)
Residual Fuel (TBtu)
Natural Gas (TBtu)
I Renewable Energy (TBtu)
I Jet Fuel (Tbtu)
I Distillate Fuel (TBtu)
I Motor Gasoline (TBtu)
¦ Onroad VMT (Index vs. 1990) [Right Axis]
¦ Sector CO2 Emissions (Index vs. 1990) [Right Axis]
SF
OHtNCO^iniDNMCi
©OOiOiOiCiOiOOiOi
200
180
160
140
120
100
80
60
40
20
0
OrtNnj^ONJlOlOTHNn^ffllDNIIKJOrlNn
OOOOOOOOOOi-l-rH-rH-rt-rHi-l-rH-rH^-rH(N(N(N(N
OOOOOOOOOOOOOOOOOOOOOOOO
Csl
s
X3
Transportation Fossil Fuel Combustion C02 Emissions
Domestic transportation C02 emissions increased by 21 percent (309.5 MMT C02 Eq.) between 1990
and 2023, an annualized increase of 0.6 percent. This includes a 24 percent increase in C02 emissions
between 1990 and 2019, followed by a 13 percent decrease from 2019 to 2020. Carbon dioxide
emissions then increased by 13 percent between 2020 and 2023. Among domestic transportation
sources, light-duty vehicles (including passenger cars and light-duty trucks) represented 57 percent of
C02 emissions, medium- and heavy-duty trucks and buses 25 percent, commercial aircraft 7 percent,
and other sources 11 percent. See Table 3-13 for a detailed breakdown of transportation C02 emissions
by mode and fuel type.
Almost all of the energy consumed by the transportation sector is petroleum-based, including motor
gasoline, diesel fuel, jet fuel, and residual oil. Carbon dioxide emissions from the combustion of ethanol
and biodiesel for transportation purposes, along with the emissions associated with the agricultural and
industrial processes involved in the production of biofuel, are captured in other Inventory sectors.22
Ethanol consumption from the transportation sector has increased from 0.7 billion gallons in 1990 to
22 Biofuel estimates are presented in the Energy chapter for informational purposes only, in line with IPCC methodological
guidance. Net carbon fluxes from changes in biogenic carbon reservoirs in croplands are accounted for in the estimates
for Land Use, Land-Use Change, and Forestry (see Chapter 6). More information and additional analyses on biofuels are
available at EPA's Renewable Fuels Standards website. See https://www.epa.gov/renewable-fuel-standard-program.
3-30 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
13.2 billion gallons in 2023, while biodiesel consumption has increased from 0.01 billion gallons in 2001
to 1.6 billion gallons in 2023. For additional information, see Section 3.11 on biofuel consumption at the
end of this chapter and Table A-71 in Annex 3.2.
Carbon dioxide emissions from passenger cars and light-duty trucks totaled 1,007.9 MMT C02 in 2023,
an increase of 10 percent (95.2 MMT C02) from 1990. The increase in C02 emissions from passenger
cars and light-duty trucks from 1990 to 2023 was due, in large part, to increased demand for travel as
fleet-wide light-duty vehicle fuel economy was relatively stable (average new vehicle fuel economy
declined slowly from 1990 through 2004 and then increased more rapidly from 2005 through 2023).
Carbon dioxide emissions from passenger cars and light-duty trucks peaked at 1,146.3 MMT in 2004,
and since then have declined about 12 percent. The decline in new light-duty vehicle fuel economy
between 1990 and 2004 (see Figure 3-15) reflects the increasing market share of light-duty trucks, which
grew from about 30 percent of new vehicle sales in 1990 to 48 percent in 2004. Starting in 2005, average
new vehicle fuel economy began to increase while light-duty vehicle VMT grew only modestly for much
of the period. Light-duty vehicle VMT grew by less than one percent or declined each year between 2005
and 2013, and again between 2017 and 2019. VMT grew at faster rates of 2.3 percent from 2014 to 2015,
and 1.7 percent from 2015 to 2016. From 2019 to 2020, light-duty vehicle VMT declined by 12.0 percent
due to COVID-19 pandemic; from 2020 to 2023 light-duty vehicle VMT rebounded as a part of the
ongoing recovery from the pandemic, increasing by 12.2 percent.
Average new vehicle fuel economy has improved almost every year since 2005 while the light-duty truck
share of new vehicle sales decreased to about 33 percent of new vehicles in 2009 and has since varied
from year to year between 36 and 63 percent. Since 2014, the light-duty truck share has steadily
increased, reaching 62 percent of new vehicle sales in model year 2023. See Annex 3.2 for data by
vehicle mode and information on VMT and the share of new vehicles (in VMT).
Medium- and heavy-duty truck C02 emissions increased by 78 percent from 1990 to 2023. This increase
was largely due to a substantial growth in medium- and heavy-duty truck VMT, which increased by 84
percent between 1990 and 2023.
Carbon dioxide from the domestic operation of commercial aircraft increased by 18 percent (19.8 MMT
C02) from 1990 to 2023. Across all categories of aviation, excluding international bunkers, C02
emissions decreased by 4 percent (6.9 MMT C02) between 1990 and 20 23.23 Carbon dioxide emissions
from military aircraft decreased 68 percent between 1990 and 2023. Commercial aircraft C02 emissions
increased 27 percent between 1990 and 2007, dropped 2 percent from 2007 to 2019, dropped another
33 percent from 2019 to 2020, then increased by 30 percent from 2020 to 2023. Overall, this represents
a change of approximately 18 percent between 1990 and 2023. Transportation sources also produce
CH4 and N20; these emissions are included in Figure 3-14 and Table 3-15 and in the CH4 and N20 from
Mobile Combustion section. Annex 3.2 presents total emissions from all transportation and mobile
sources, including C02, CH4, N20, and HFCs.
23 Includes consumption of jet fuel and aviation gasoline. Does not include aircraft bunkers, which are not included in
national emission totals, in line with IPCC methodological guidance.
Energy 3-31
-------
Figure 3-15: Sales-Weighted Fuel Economy of New Passenger Cars and Light-Duty
Trucks, 1990-2023
C^CTiCT»0%C3^0,iCriC^CriC^OOOOOOOOOOi-H*-li—I *-H i—ti-Hi—I N N N N
oioio^cnoicrioioicrioioooooooooooooooooooooooo
rtHHrtHrtrtHHHNN(MN[MPJNNNfMrJNNN[MNNNIN(M(MNNIN
Source: EPA (2023).
Figure 3-16: Sales of New Passenger Cars and Light-Duty Trucks, 1990-2023
100%
90%
80%
70%
60%
50%
40%
30%
20%
—I % Light-Duty Trucks
10% ¦ % Passenger Cars
0%
o
1—1
r\l
cn
s
LO
CO
o
i—f
CM
cn
s
lo
rv
CO
CD
o
t—i
fN
cn
LO
V0
00
CD
O
i—1
CTi
cr>
CT»
CD
a*
cn
o
o
O
o
o
O
o
o
o
t—i
rH
tH
T—1
r-l
t-H
t-H
i—i
i—i
r-HI
Ov
cn
&
cn
CD
CD
(D
CD
cr\
o
o
O
o
o
o
o
o
o
o
o
o
O
o
o
o
O
o
o
O
o
O
O
o
T—(
i—I
y—1
i-H
i—(
i—t
i—I
T—1
t-H
i—i
r\j
CM
fN
rsi
fM
cm
fN
fN
rsi
fN
fN
fN
(N
(N
fN
fN
fN
rsi
fN
(N
fN
fN
fN
fN
Source: EPA (2023).
3-32 inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Table 3-13: C02 Emissions from Fossil Fuel Combustion in Transportation End-Use
Sector (MMT C02 Eq.)
Fuel/Vehicle Type
1990
2005
2019
2020
2021
2022
2023
Gasoline8
958.9
1,150.1
1,086.5
936.9
1,028.7
1,014.5
1,035.7
Passenger Cars
622.7
525.0
358.8
297.8
314.1
296.6
293.6
Light-Duty Trucks
271.7
576.1
675.2
590.5
658.9
659.2
682.4
Medium- and Heavy-Duty Trucks'3
41.8
28.8
31.5
29.5
34.7
35.8
36.3
Buses
5.1
1.8
3.6
3.1
3.5
3.9
4.1
Motorcycles
3.3
4.7
6.7
6.1
6.9
8.2
8.6
Recreational Boats0
14.3
13.7
10.7
9.9
10.6
10.8
10.8
Distillate Fuel Oil (Diesel)8
262.9
462.6
474.0
447.4
480.6
476.7
471.3
Passenger Cars
9.9
4.1
2.5
2.0
2.0
1.9
1.8
Light-Duty Trucks
8.2
30.1
32.0
26.4
27.8
27.9
27.5
Medium- and Heavy-Duty Trucks'3
192.0
356.2
373.0
361.3
389.8
384.2
381.2
Buses
7.8
14.9
20.2
16.3
17.9
19.5
19.3
Rail
35.5
46.1
36.0
31.2
32.5
32.5
30.9
Recreational Boats0
2.7
2.9
2.9
2.6
2.8
3.0
2.9
Ships and Non-Recreational Boatsd
6.8
8.4
7.5
7.6
7.8
7.8
7.7
International Bunker Fuels"
11.7
9.5
10.1
7.8
7.4
7.2
7.0
Jet Fuel
184.1
189.2
180.3
120.6
152.6
164.8
178.9
Commercial Aircraft*
109.9
132.7
136.7
91.3
119
129.7
129.7
Military Aircraft
35.7
19.8
12.2
11.7
12.5
12.4
11.5
General Aviation Aircraft
38.5
36.8
31.4
17.6
21.1
22.7
37.7
International Bunker Fuels"
38.2
60.2
78.3
39.8
50.8
66.6
66.5
International Bunker Fuels from
Commercial Aviation
30.0
55.6
75.1
36.7
47.6
63.5
63.5
Aviation Gasoline
3.1
2.4
1.6
1.4
1.5
1.5
1.5
General Aviation Aircraft
3.1
2.4
1.6
1.4
1.5
1.5
1.5
Residual Fuel Oil
22.6
19.3
14.5
7.3
24.2
22.9
16.6
Ships and Non-Recreational Boats0
22.6
19.3
14.5
7.3
24.2
22.9
16.6
International Bunker Fuels"
53.7
43.6
25.2
22.1
21.9
24.4
22.7
Natural Gass
36.0
33.1
58.9
58.8
65.2
72.3
71.7
Passenger Cars
+
+
+
+
+
+
+
Light-Duty Trucks
+
+
+
+
+
+
+
Medium- and Heavy-Duty Trucks
+
+
0.1
0.1
0.1
0.1
0.1
Buses
+
0.2
0.3
0.2
0.2
0.3
0.3
Pipeline11
36.0
32.8
58.5
58.5
64.9
72.0
71.3
LPGg
1.4
1.8
0.8
0.5
0.6
0.7
0.7
Passenger Cars
+
+
+
+
+
+
+
Light-Duty Trucks
0.1
0.1
+
+
+
0.1
0.1
Medium- and Heavy-Duty Trucks'3
1.3
0.9
0.7
0.5
0.5
0.5
0.5
Buses
+
0.7
0.1
0.1
0.1
0.1
0.1
Electricity'
3.0
4.7
4.2
3.5
4.0
4.5
5.0
Passenger Cars
+
+
0.1
0.1
0.2
0.4
0.7
Light-Duty Trucks
+
+
1.1
1.0
1.3
1.6
1.9
Energy 3-33
-------
Fuel/Vehicle Type
1990
2005
2019
2020
2021
2022
2023
Buses
+
+
+
+
+
+
+
Rail
3.0
4.7
3.1
2.4
2.5
2.5
2.4
TotaleJ
1,472.0
1,863.3
1,820.9
1,576.4
1,757.5
1,758.0
1,781.5
International Bunker Fuels
103.6
113.3
113.6
69.6
80.2
9 8.2
96.2
Biofuels-Ethanolk
4.1
21.6
78.7
68.1
75.4
75.0
76.4
Biofuels-Biodieselk
0.0
0.9
17.1
17.7
16.1
15.6
18.2
+ Does not exceed 0.05 MMT C02 Eq.
a On-road fuel consumption data from FHWATable MF-21 and MF-27 were used to determine total on-road use of motor gasoline
and diesel fuel (FHWA 1996 through 2024). Ratios developed from MOVES5 output are used to apportion FHWA fuel
consumption data to vehicle type and fuel type (see Annex 3.2 for information about the MOVES model). Onroad vehicle VMT
and fuel consumption are proxied based on the Traffic Volume Trends data for the year 2023.
b Includes medium- and heavy-duty trucks over 8,500 lbs.
c In 2014, EPA incorporated the NONROAD2008 model into the MOVES model framework. The current Inventory uses the
Nonroad component of MOVES5 for years 1999 through 2023. See Annex 3.2 for information about the MOVES model.
d Note that large year overyear fluctuations in emission estimates partially reflect nature of data collection for these sources.
e Official estimates exclude emissions from the combustion of both aviation and marine international bunker fuels; however,
estimates of international bunker fuel-related emissions are presented for informational purposes.
f Commercial aircraft, as modeled in FAA's Aviation Environmental Design Tool(AEDT), consists of passenger aircraft, cargo, and
other chartered flights.
gTransportation sector natural gas and LPG consumption are based on data from EIA(24). Prior to the 1990 to 2015 Inventory,
data from DOE TEDB were used to estimate each vehicle class's share of the total natural gas and LPG consumption. Since
TEDB does not include estimates for natural gas use by medium and heavy-duty trucks or LPG use by passenger cars, EIA
Alternative Fuel Vehicle Data (Browning 2017) is now used to determine each vehicle class's share of the total natural gas and
LPG consumption. These changes were first incorporated in the 1990 to 2016 Inventory and apply to the 1990 to 2023 time
period.
h Pipelines reflect C02 emissions from natural gas-powered pipelines transporting natural gas.
Electricity consumption by passenger cars, light-duty trucks (SUVs), and buses is based on plug-in electric vehicle sales and
engine efficiency data, as outlined in Browning (2018a). The mileage accumulation rates for electric vehicles were lowered this
year based on research by Browning (2024). In prior Inventory years, C02 emissions from electric vehicle charging were
allocated to the residential and commercial sectors. They are allocated to the transportation sector. These changes apply to
the 2010 through 2023 time period.
' Includes emissions from rail electricity.
k Ethanoland biodiesel estimates are presented for informational purposes only. See Section 3.10 of this chapter and the
estimates in Land Use, Land-Use Change, and Forestry (see Chapter 6), in line with IPCC methodological guidance, for more
information on ethanol and biodiesel.
Notes: This table does not include emissions from non-transportation mobile sources, such as agricultural equipment and
construction/mining equipment; it also does not include emissions associated with electricity consumption by pipelines or
lubricants used in transportation. In addition, this table does not include C02 emissions from U.S. Territories, since these are
covered in a separate chapter of the Inventory. Totals may not sum due to independent rounding.
Mobile Fossil Fuel Combustion CH4 and N20 Emissions
Mobile combustion includes emissions of CH4 and N20 from all transportation sources identified in the
U.S. Inventory with the exception of pipelines and electric locomotives;24 mobile sources also include
non-transportation sources such as construction/mining equipment, agricultural equipment, vehicles
used off-road, and other sources (e.g., snowmobiles, lawnmowers, etc.).25 Annex 3.2 includes a
24 Emissions of ChU from natural gas systems are reported separately. More information on the methodology used to
calculate these emissions are included in this chapter and Annex 3.4.
25 See the methodology sub-sections of the CO2 from Fossil Fuel Combustion and CH4 and N2O from Mobile Combustion
sections of this chapter. Note that N2O and CH4 emissions are reported using different categories than CO2. CO2
emissions are reported by end-use sector (transportation, industrial, commercial, residential, U.S. Territories), and
generally adhere to a top-down approach to estimating emissions. CO2 emissions from non-transportation mobile
sources (e.g., lawn and garden equipment, farm equipment, construction equipment) are allocated to their respective
3-34 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
summary of all emissions from both transportation and mobile sources. Table 3-14 and Table 3-15
provide mobile fossil fuel CH4 and N20 emission estimates in MMT C02 Eq.26
Mobile combustion was responsible for a small portion of national CH4 emissions (0.4 percent) and was
the fifth largest source of national N20 emissions (4.2 percent) in 2023. From 1990 to 2023, mobile
source CH4 emissions declined by 64.9 percent, to 2.5 MMT C02 Eq. (91 kt), due largely to emissions
control technologies employed in on-road vehicles since the mid-1990s to reduce CO, NOx, NMVOC,
and CH4 emissions. Mobile source emissions of N20 decreased by 57.0 percent, to 16.2 MMT C02 Eq.
(61 kt) in 2023. Earlier generation control technologies initially resulted in elevated N20 emissions,
causing a 32 percent increase in N20 emissions from mobile sources between 1990 and 1997.
Improvements in later-generation emission control technologies have reduced N20 output, resulting in a
67 percent decrease in mobile source N20 emissions from 1997 to 2023 (see Figure 3-17). Overall, CH4
and N20 emissions were predominantly from gasoline-fueled passenger cars, light-duty trucks, and
non-highway sources. See Annex 3.2 for data by vehicle mode and information on VMT and the share of
new vehicles.
Figure 3-17: Mobile Source CH4 and N20 Emissions
aicricricr>cricriericrier>cri00oooooooooooooooooooooo
Table 3-14: CH4 Emissions from Mobile Combustion (MMT C02 Eq.)
Fuel Type/Vehicle Type3
1990 2005
2019
2020
2021
2022
2023
Gasoline On-Roadb
5.8
3.3
1.0
0.8
0.8
0.8
0.7
Passenger Cars
3.8
1.8
0.3
0.2
0.2
0.2
0.2
Light-DutyTrucks
1.4
1.3
0.6
0.5
0.5
0.5
0.5
Medium- and Heavy-Duty Trucks and Buses
0.5
0.2
+
+
+
+
+
Motorcycles
+1
+
+
+
+
+
+
Diesel On-Roadb
+
+
0.1
0.1
0.1
0.1
0.1
end-use sector (i.e., construction equipment CO2 emissions are included in the Industrial end-use sector instead of the
transportation end-use sector). CH4 and N2O emissions are reported using the "mobile combustion" category, which
includes non-transportation mobile sources. CH4 and N2O emission estimates are bottom-up estimates, based on total
activity (fuel use, VMT) and emissions factors by source and technology type. These reporting schemes are in accordance
with IPCC guidance. For informational purposes only, CO2 emissions from non-transportation mobile sources are
presented separately from their overall end-use sector in Annex 3.2.
26 See Annex 3.2 for a complete time series of emission estimates for 1990 through 2023.
Energy 3-35
-------
Fuel Type/Vehicle Type3
1990
2005
2019
2020
2021
2022
2023
Passenger Cars
+
+
+
+
+
+
+
Light-DutyTrucks
+
+
+
+
+
+
+
Medium- and Heavy-Duty Trucks
+
+
0.1
0.1
0.1
0.1
0.1
Medium- and Heavy-Duty Buses
+
+
+
+
+
+
+
Alternative FuelOn-Road
+
+
+
+
+
+
+
Non-Road°
1.4
1.8
1.7
1.6
1.6
1.7
1.7
Ships and Boats
0.4
0.5
0.4
0.4
0.5
0.5
0.5
Raild
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Aircraft
0.1
0.1
+
+
+
+
+
Agricultural Equipment®
0.2
0.2
0.1
0.1
0.1
0.1
0.1
Construction/Mining Equipment'
0.2
0.3
0.2
0.2
0.2
0.2
0.2
Other8
0.5
0.7
0.8
0.8
0.7
0.8
0.8
Total
7-2
5.2
2.8
2.5
2.6
2.6
2.5
+ Does not exceed 0.05 MMT C02 Eq.
a See Annex 3.2 for definitions of on-road vehicle types.
b Gasoline and diesel highway vehicle mileage estimates are based on data from FHWA Highway Statistics Table VM-1. VMT for
2023 is based on FHWA's Traffic Volume Trends data series. VMT estimates from FHWA are allocated to vehicle type using ratios
of VMT per vehicle type to total VMT, derived from EPA's MOVES5 model (see Annex 3.2 for information about the MOVES model).
c Nonroad fuel consumption estimates for 2020 are adjusted to account for the COVID-19 pandemic and associated restrictions.
For agricultural equipment and airport equipment, sector specific adjustment factors were applied to the 2019 data. For all other
sectors, a 7.7 percent reduction factor is used, based on transportation diesel use (EIA 2025a).
d Rail emissions do not include emissions from electric powered locomotives. Class II and Class III diesel consumption data for
2014 to 2023 is estimated by applying the historical average fuel usage per carload factor to the annual number of carloads.
e Includes equipment, such as tractors and combines, as well as fuel consumption from trucks that are used off-road in
agriculture.
' Includes equipment, such as cranes, dumpers, and excavators, as well as fuel consumption from trucks that are used off-road
in construction.
g "Other" includes snowmobiles and other recreational equipment, logging equipment, lawn and garden equipment, railroad
equipment, airport equipment, commercial equipment, and industrial equipment, as well as fuel consumption from trucks that
are used off-road for commercial/industrial purposes.
Note: Totals may not sum due to independent rounding.
Table 3-15: N20 Emissions from Mobile Combustion (MMT C02 Eq.)
Fuel Type/Vehicle Type3
1990
2005
2019
2020
2021
2022
2023
Gasoline On-Roadb
31.4
33.4
8.2
6.3
6.1
5.4
4.8
Passenger Cars
22.3
16.4
2.6
2.0
1.8
1.5
1.4
Light-DutyTrucks
8.2
15.8
5.3
4.1
4.0
3.6
3.2
Medium- and Heavy-Duty Trucks and Buses
0.9
1.2
0.2
0.2
0.2
0.1
0.1
Motorcycles
+
+
0.1
0.1
0.1
0.1
0.1
Diesel On-Roadb
0.42
0.4
2.9
2.9
3.3
3.4
3.5
Passenger Cars
+
+
+
+
+
+
+
Light-DutyTrucks
+
+
0.2
0.2
0.2
0.2
0.2
Medium- and Heavy-Duty Trucks
0.2
0.3
2.4
2.5
2.9
2.9
3.0
Medium- and Heavy-Duty Buses
+
+
0.2
0.2
0.2
0.2
0.2
Alternative Fuel On-Road
+
+
0.1
0.1
0.1
0.1
0.1
Non-Road°
6.2
8.1
7.6
6.7
7.4
7.7
7.9
Ships and Boats
0.2
0.2
0.2
0.1
0.3
0.3
0.2
Raild
0.2
0.4
0.2
0.2
0.2
0.2
0.2
Aircraft
1.5
1.6
1.5
1.0
1.3
1.4
1.5
3-36 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Fuel Type/Vehicle Type3
1990
2005
2019
2020
2021
2022
2023
Agricultural Equipment®
1.2
1.4
1.1
1.1
1.1
1.1
1.1
Construction/Mining Equipment'
1.2
1.9
1.7
1.6
1.7
1.7
1.7
Other8
1.8
2.8
2.9
2.7
2.9
3.1
3.2
Total
37.8
42.0
18.8
16.0
16.8
16.6
16.2
+ Does not exceed 0.05 MMT C02 Eq.
a See Annex 3.2 for definitions of on-road vehicle types.
b Gasoline and diesel highway vehicle mileage estimates are based on data from FHWA Highway Statistics Table VM-1. VMT for
2023 is estimated based on trends in FHWA's Traffic Volume Trends data series. VMT estimates from FHWA are allocated to
vehicle type using ratios of VMT per vehicle type to total VMT, derived from EPA's MOVES5 model (see Annex 3.2 for information
about the MOVES model).
c Nonroad fuel consumption estimates for 2020 are adjusted to account for the COVID-19 pandemic and associated restrictions.
For agricultural equipment and airport equipment, sector specific adjustment factors were applied to the 2019 data. For all other
sectors, a 7.7 percent reduction factor is used, based on transportation diesel use (EIA 2025a).
d Rail emissions do not include emissions from electric powered locomotives. Class II and Class III diesel consumption data for
2014 through 2023 is estimated by applying the historical average fuel usage per carload factor to the annual number of
carloads.
e Includes equipment, such as tractors and combines, as well as fuel consumption from trucks that are used off-road in
agriculture.
' Includes equipment, such as cranes, dumpers, and excavators, as well as fuel consumption from trucks that are used off-road
in construction.
g "Other" includes snowmobiles and other recreational equipment, logging equipment, lawn and garden equipment, railroad
equipment, airport equipment, commercial equipment, and industrial equipment, as well as fuel consumption from trucks that
are used off-road for commercial/industrial purposes.
Note: Totals may not sum due to independent rounding.
CO2 from Fossil Fuel Combustion
Methodology and Time-Series Consistency
C02 emissions from fossil fuel combustion are estimated in line with a Tier 2 method described by the
IPCC in the 2006IPCC Guidelines for National Greenhouse Gas Inventories (IPCC 2006) Chapter 2,
Figure 2.1 decision tree and available data on energy use and country specific fuel carbon contents with
some exceptions as discussed below.27 A detailed description of the U.S. methodology is presented in
Annex 2.1, and is characterized by the following steps:
1- Determine total fuel consumption by fuel type and sector. Total fossil fuel consumption for each
year is estimated by aggregating consumption data by end-use sector (e.g., commercial,
industrial), primary fuel type (e.g., coal, petroleum, gas), and secondary fuel category (e.g.,
motor gasoline, distillate fuel oil). Fuel consumption data for the United States were obtained
directlyfrom the EIA of the U.S. Department of Energy (DOE), primarily from the Monthly Energy
Review (EIA 2025a). EIA data include fuel consumption statistics from the 50 U.S. states and the
District of Columbia, including tribal lands. The EIA does not include territories in its national
energy statistics, so fuel consumption data for territories were collected separately from ElA's
International Energy Statistics (EIA 2024a), see Annex 2.1 for more details on how Territories
data is collected.28
2. For consistency of reporting, the IPCC has recommended that countries report energy data
using the International Energy Agency (IEA) reporting convention and/or lEAdata. Data in the IEA
27 The IPCC Tier 3B methodology is used for estimating emissions from commercial aircraft.
28 Fuel consumption by U.S. Territories (i.e., American Samoa, Guam, Puerto Rico, U.S. Virgin Islands, Wake Island, and
other U.S. Pacific Islands) is included in this report and contributed total emissions of 25.3 MMT CO2 Eq. in 2023.
Energy 3-37
-------
format are presented "top down"—that is, energy consumption for fuel types and categories are
estimated from energy production data (accounting for imports, exports, stock changes, and
losses). The resulting quantities are referred to as "apparent consumption." The data collected
in the United States by EIA on an annual basis and used in this Inventory are predominantly from
mid-stream or conversion energy consumers such as refiners and electric power generators.
These annual surveys are supplemented with end-use energy consumption surveys, such as the
Manufacturing Energy Consumption Survey, that are conducted on a periodic basis (every four
years). These consumption datasets help inform the annual surveys to arrive at the national
total and sectoral breakdowns for that total.29
3. Also, note that U.S. fossil fuel energy statistics are generally presented using gross calorific
values (GCV) (i.e., higher heating values). Fuel consumption activity data presented here have
not been adjusted to correspond to international standards, which are to report energy statistics
in terms of net calorific values (NCV) (i.e., lower heating values).30
4. Subtract uses accounted for in the Industrial Processes and Product Use chapter. Portions of
the fuel consumption data for seven fuel categories—coking coal, distillate fuel, industrial other
coal, petroleum coke, natural gas, residual fuel oil, and other oil—were reallocated to the
Industrial Processes and Product Use chapter, as they were consumed during non-energy-
related industrial activity. To make these adjustments, additional data were collected from AISI
(2004 through 2021), Coffeyville (2012), U.S. Census Bureau (2001 through 2011), EIA (2024a,
2024f, 2024h), USAA (2008 through 2021), USGS (1991 through 2020), (USGS 2019), USGS (2014
through 2021 a), USGS (2014 through 2021 b), USGS (1995 through 2013), USGS (1995, 1998,
2000, 2001, 2002, 2007), USGS (2021a), USGS (1991 through 2015a), USGS (1991 through
2020), USGS (2014 through 2021 a), USGS (1991 through 2015b), USGS (2021 b), USGS (1991
through 2020).31
5. Adjust for biofuels and petroleum denaturant. Fossil fuel consumption estimates are adjusted
downward to exclude fuels with biogenic origins and avoid double counting in petroleum data
statistics. Carbon dioxide emissions from ethanol added to motor gasoline and biodiesel added
to diesel fuel are not included specifically in summing energy sector totals. Net carbon fluxes
from changes in biogenic carbon reservoirs are accounted for in the estimates for LULUCF,
therefore, fuel consumption estimates are adjusted to remove ethanol and biodiesel.32 For the
years 1993 through 2008, petroleum denaturant is currently included in EIA statistics for both
natural gasoline and finished motor gasoline. To avoid double counting, petroleum denaturant is
subtracted from finished motor gasoline for these years.33
6. Adjust for C02 sequestration. Since October 2000, the Dakota Gasification Plant has been
exporting C02 produced in the coal gasification process to Canada by pipeline. Because this
29 See IPCC Reference Approach for Estimating CO2 Emissions from Fossil Fuel Combustion in Annex 4 for a comparison of
U.S. estimates using top-down and bottom-up approaches.
30 A crude convention to convert between gross and net calorific values is to multiply the heat content of solid and liquid
fossil fuels by 0.95 and gaseous fuels by 0.9 to account for the water content of the fuels. Biomass-based fuels in U.S.
energy statistics, however, are generally presented using net calorific values.
31 See sections on Iron and Steel Production and Metallurgical Coke Production, Ammonia Production and Urea
Consumption, Petrochemical Production, Titanium Dioxide Production, Ferroalloy Production, Aluminum Production,
and Silicon Carbide Production and Consumption in the Industrial Processes and Product Use chapter.
32 Natural gas energy statistics from EIA (2024d) are already adjusted downward to account for biogas in natural gas.
33 These adjustments are explained in greater detail in Annex 2.1.
3-38 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
C02 is not emitted to the atmosphere in the United States, the associated fossil fuel (lignite
coal) that is gasified to create the exported C02 is subtracted from EIA (2024h) coal
consumption statistics that are used to calculate greenhouse gas emissions from the Energy
Sector. The associated fossil fuel is the total fossil fuel burned at the plant with the C02 capture
system multiplied by the fraction of the plant's total site-generated C02 that is recovered by the
capture system. To make these adjustments, data for C02 exports were collected from
Environment and Climate Change Canada (2025). A discussion of the methodology used to
estimate the amount of C02 captured and exported by pipeline is presented in Annex 2.1.
Additionally in 2023, the Petra Nova facility reported to the GHGRP subpart RR sequestration of
C02 that was captured from a coal fired power plant. Because the sequestered C02 is assumed
to not be admitted to the atmosphere, the C02 captured for sequestration was needed out of
C02 emissions from electric power sector coal. See Section 3.9 for more information on C02
transport, injection and geologic sequestration accounting in the Inventory.
7. Adjust sectoral allocation of distillate fuel oil and motor gasoline. EPA conducted a separate
bottom-up analysis of transportation fuel consumption based on data from the Federal Highway
Administration that indicated that the amount of distillate and motor gasoline consumption
allocated to the transportation sector in the EIA statistics should be adjusted. Therefore, for
these estimates, the transportation sector's distillate fuel and motor gasoline consumption
were adjusted to match the value obtained from the bottom-up analysis. As the total distillate
and motor gasoline consumption estimate from EIA are considered to be accurate at the
national level, the distillate and motor gasoline consumption totals for the residential,
commercial, and industrial sectors were adjusted proportionately. The data sources used in the
bottom-up analysis of transportation fuel consumption include AAR (2008 through 2022),
Benson (2002 through 2004), DOE (1993 through 2022), EIA (2007), EIA (2025a), EPA (2024e),
and FHWA (1996 through 2024).34
8. Adjust for fuels consumed for non-energy uses. U.S. aggregate energy statistics include
consumption of fossil fuels for non-energy purposes. These are fossil fuels that are
manufactured into plastics, asphalt, lubricants, or other products. Depending on the end-use,
this can result in storage of some or all of the carbon contained in the fuel for a period of time.
As the emission pathways of carbon used for non-energy purposes are vastly different than fuel
combustion (since the carbon in these fuels ends up in products instead of being combusted),
these emissions are estimated separately in Section 3.2. Therefore, the amount of fuels used for
non-energy purposes was subtracted from total fuel consumption. Data on non-fuel
consumption were provided by EIA (2025a).
9. Subtract consumption of international bunker fuels. In line with IPCC guidelines emissions from
international transport activities, or bunker fuels, should not be included in national totals. U.S.
energy consumption statistics include these bunker fuels (e.g., distillate fuel oil, residual fuel
oil, and jet fuel) as part of consumption by the transportation end-use sector, however, so
emissions from international transport activities were calculated separately following the same
procedures used to calculate emissions from consumption of all fossil fuels (i.e., estimation of
consumption, and determination of carbon content).35 The Office of the Under Secretary of
34 Bottom-up gasoline and diesel highway vehicle fuel consumption estimates are based on data from FHWA Highway
Statistics Table MF-21, MF-27, and VM-1 (FHWA 1996 through 2024).
35 See International Bunker Fuels section in this chapter for a more detailed discussion.
Energy 3-39
-------
Defense (Installations and Environment) and the Defense Logistics Agency Energy (DLA Energy)
of the U.S. Department of Defense (DoD) (DLA Energy 2025) supplied data on military jet fuel
and marine fuel use. Commercial jet fuel use was estimated based on data from FAA (2024) and
DOT (1991 through 2023); residual and distillate fuel use for civilian marine bunkers was
obtained from DOC (1991 through 2024) for 1990 through 2001 and 2007 through 2020, and
DHS (2008) for 2003 through 2006.36 Consumption of these fuels was subtracted from the
corresponding fuels totals in the transportation end-use sector. Estimates of international
bunker fuel emissions for the United States are discussed in detail in Section 3.9.
10. Determine the total carbon content of fuels consumed. Total carbon was estimated by
multiplying the amount of fuel consumed by the amount of carbon in each fuel. This total
carbon estimate defines the maximum amount of carbon that could potentially be released to
the atmosphere if all of the carbon in each fuel was converted to C02. A discussion of the
methodology and sources used to develop the carbon content coefficients are presented in
Annexes 2.1 and 2.2.
11. Estimate C02 emissions. Total C02 emissions are the product of the adjusted energy
consumption (from the previous methodology steps 1 through 7), the carbon content of the
fuels consumed, and the fraction of carbon that is oxidized. The fraction oxidized was assumed
to be 100 percent for petroleum, coal, and natural gas based on guidance in IPCC (2006) (see
Annex 2.1). Carbon emissions were multiplied by the molecular-to-atomic weight ratio of C02 to
carbon (44/12) to obtain total C02 emitted from fossil fuel combustion in million metric tons
(MMT).
12. Allocate transportation emissions by vehicle type. This report provides a more detailed
accounting of emissions from transportation because it is such a large consumer of fossil fuels
in the United States. For fuel types other than jet fuel, fuel consumption data by vehicle type and
transportation mode were used to allocate emissions by fuel type calculated for the
transportation end-use sector. Heat contents and densities were obtained from EIA (2024f) and
USAF (1998).37
• For on-road vehicles, annual estimates of combined motor gasoline and diesel fuel
consumption by vehicle category were obtained from FHWA (1996 through 2024); for each
vehicle category, the percent gasoline, diesel, and other (e.g., CNG, LPG) fuel consumption
are estimated using data from EPA's MOVES model and DOE (1993 through 2022).38 39
36 Data for 2002 were interpolated due to inconsistencies in reported fuel consumption data.
37 For a more detailed description of the data sources used for the analysis of the transportation end use sector see the
Mobile Combustion (excluding CO2) and International Bunker Fuels sections of the Energy chapter, Annex 3.2, and Annex
3.3, respectively.
38 On-road fuel consumption data from FHWA Table MF-21 and MF-27 were used to determine total on-road use of motor
gasoline and diesel fuel (FHWA 1996 through 2023). Ratios developed from MOVES5 output are used to apportion FHWA
fuel consumption data to vehicle type and fuel type (see Annex 3.2 for information about the MOVES model). Trends in
on-road vehicle VMT and fuel consumption are proxied based on the Traffic Volume Trends data for the year 2023.
39 Transportation sector naturalgas and LPG consumption are based on data from EIA (2025a). In previous Inventory years,
data from DOE (1993 through 2022) TEDB was used to estimate each vehicle class's share of the total natural gas and
LPG consumption. Since TEDB does not include estimates for natural gas use by medium- and heavy-duty trucks or LPG
use by passenger cars, EIA Alternative Fuel Vehicle Data (Browning 2017) is now used to determine each vehicle class's
share of the total natural gas and LPG consumption. These changes were first incorporated in the 1990 through 2015
Inventory and apply to the time period from 1990 to 2015.
3-40 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
• For non-road vehicles, activity data were obtained from AAR (2008 through 2023), APTA
(2007 through 2023), APTA (2006), BEA (1991 through 2015), Benson (2002 through 2004),
DLA Energy (2025), DOC (1991 through 2024), DOE (1993 through 2023), DOT (1991 through
2025), EIA (2009a), EIA (2024c), EIA(2002), EIA(1991 through 2022), EPA (2024a),40 and
Gaffney (2007).
• For jet fuel used by aircraft, C02 emissions from commercial aircraft were developed by the
U.S. Federal Aviation Administration (FAA) using a Tier 3B methodology, consistent IPCC
(2006) (see Annex 3.3). Carbon dioxide emissions from other aircraft were calculated
directly based on reported consumption of fuel as reported by EIA. Allocation to domestic
military uses was made using DoD data. General aviation jet fuel consumption is calculated
as the remainder of total jet fuel use (as determined by EIA) nets all other jet fuel use as
determined by FAA and DoD. For more information, see Annex 3.2.
Methodological recalculations were applied to the entire time series to ensure time-series consistency
from 1990 through 2023. Due to data availability and sources, some adjustments outlined in the
methodology above are not applied consistently across the full 1990 to 2023 time series. As described
in greater detail in Annex 2.1, to align with ElA's methodology for calculating motor gasoline
consumption, petroleum denaturant adjustments are applied to motor gasoline consumption only for
the period 1993 through 2008. In addition to ensuring time-series consistency, to ensure consistency in
reporting between the Inventory and the Canadian National Greenhouse Gas Inventory, the amount of
associated fossil fuel (lignite coal) that is gasified to create the exported C02 from the Dakota
Gasification Plant is adjusted to align with the Canadian National Greenhouse Gas Inventory
(Environment and Climate Change Canada 2025). This adjustment is explained in greater detail in Annex
2.1. As discussed in Annex 5, data are unavailable to include estimates of C02 emissions from any liquid
fuel used in pipeline transport or non-hazardous industrial waste incineration, but those emissions are
assumed to be insignificant.
Box 3-4: Carbon Intensity of U.S. Energy Consumption
The amount of carbon emitted from the combustion of fossil fuels is dependent upon the carbon
content of the fuel and the fraction of that carbon that is oxidized. Fossil fuels vary in their average
carbon content, ranging from about 53 MMT C02 Eq./QBtu for natural gas to upwards of 95 MMT C02
Eq./QBtu for coal and petroleum coke (see Tables A-42 and A-43 in Annex 2.1 for carbon contents of all
fuels). In general, the carbon content per unit of energy of fossil fuels is the highest for coal products,
followed by petroleum, and then natural gas. The overall carbon intensity of the U.S. economy is thus
dependent upon the quantity and combination of fuels and other energy sources employed to meet
demand.
Table 3-16 provides a time series of the carbon intensity of direct emissions for each sector of the U.S.
economy. The time series incorporates only the energy from the direct combustion of fossil fuels in each
sector. For example, the carbon intensity for the residential sector does not include the energy from or
emissions related to the use of electricity for lighting, as it is instead allocated to the electric power
sector. For the purposes of maintaining the focus of this section, renewable energy and nuclear energy
are not included in the energy totals used in Table 3-16 in order to focus attention on fossil fuel
combustion as detailed in this chapter. Looking only at this direct consumption of fossil fuels, the
40 In 2014, EPA incorporated the NONROAD2008 model into the MOVES model framework (EPA 2024b). The current
Inventory uses the Nonroad component of MOVES5 for years 1999 through 2023.
Energy 3-41
-------
residential sector exhibited the lowest carbon intensity, which is related to the large percentage of its
energy derived from natural gas for heating. The carbon intensity of the commercial sector has
predominantly declined since 1990 as commercial businesses shift from petroleum to natural gas. The
industrial sector was more dependent on petroleum and coal than either the residential or commercial
sectors, and thus had higher carbon intensities over this period. The carbon intensity of the
transportation sector was closely related to the carbon content of petroleum products (e.g., motor
gasoline and jet fuel, both around 70 MMT C02 Eq./QBtu), which were the primary sources of energy.
Lastly, the electric power sector had the highest carbon intensity due to its heavy reliance on coal for
generating electricity.
Table 3-16: Carbon Intensity from Direct Fossil Fuel Combustion by Sector (MMT C02
Eq./QBtu)
Sector
1990
2005
2019
2020
2021
2022
2023
Residential®
57.4
56.8
55.3
55.1
55.2
55.2
55.3
Commercial®
59.7
57.8
56.2
56.3
56.2
56.6
56.4
Industrial®
64.8
64.6
60.2
59.6
59.6
59.6
59.3
Transportation®
71.1
71.5
70.9
70.8
70.9
70.8
70.8
Electric Powerb
87.3
85.8
72.9
70.5
72.4
70.9
68.2
U.S. Territories0
73.1
73.4
70.8
71.5
70.1
71.6
69.9
All Sectors0
73.1
73.6
67.3
66.3
67.0
66.5
65.7
a Does not include electricity or renewable energy consumption.
b Does not include electricity produced using nuclear or renewable energy.
c Does not include nuclear or renewable energy consumption.
Notes: Excludes non-energy fuel use emissions and consumption. Totals may not sum due to independent rounding.
For the time period of 1990 through about 2008, the carbon intensity of U.S. energy consumption was
fairly constant, as the proportion of fossil fuels used by the individual sectors did not change
significantly over that time. Starting in 2008 the carbon intensity of U.S. energy consumption has
decreased, reflecting the shift from coal to natural gas in the electric power sector during that time
period. Per capita energy consumption fluctuated little from 1990 to 2007, but then started decreasing
after 2007 and, in 2023, was approximately 15.8 percent below levels in 1990 (see Table 3-17). To
differentiate these estimates from those of Table 3-16, the carbon intensity trend shown in Table 3-17
and described below includes nuclear and renewable energy EIA data to provide a comprehensive
economy-wide picture of energy consumption. Due to a general shift from a manufacturing-based
economy to a service-based economy, as well as overall increases in efficiency, energy consumption
and energy-related C02 emissions per dollar of gross domestic product (GDP) have both declined since
1990 (BEA 2024).
3-42 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Table 3-17: U.S. Energy Consumption and Energy-Related C02 Emissions Per Capita
and Per Dollar GDP
OT-H{Nn^-LouDfv.oocyio^HfNjro^:m»iDr^oocno^HrvJro^-m^£)r^oocT»Oi-HfNjro
tyityiCytOtO^CTvCTvCTiO^O^OOOOOOOOOOT—It-Ht—l-i—1"«—I'-H'i—t'i-Hi—It—lf\ir\lfMf\l
o^o^c^cncr»cr\cr\o^cno^oooooooooooooooooooooooo
HHHHHHHHHH(NJtMfNN(M(NrMfNrMN(N(NfN(NtNr\IN(N(NrMfNfNtM(N
Carbon intensity estimates were developed using nuclear and renewable energy data from EIA (2023c),
EPA (2010), and fossil fuel consumption data as discussed above and presented in Annex 2.1.
Uncertainty
For estimates of C02 from fossil fuel combustion, the amount of C02 emitted is directly related to the
amount of fuel consumed, the fraction of the fuel that is oxidized, and the carbon content of the fuel.
Therefore, a careful accounting of fossil fuel consumption by fuel type, average carbon contents of fossil
fuels consumed, and production of fossil fuel-based products with long-term carbon storage should
yield an accurate estimate of C02 emissions.
Nevertheless, there are uncertainties in the consumption data, carbon content of fuels and products,
and carbon oxidation efficiencies. For example, given the same primary fuel type (e.g., coal, petroleum,
or natural gas), the amount of carbon contained in the fuel per unit of useful energy can vary. For the
United States, however, the impact of these uncertainties on overall C02 emission estimates is believed
to be relatively small. See, for example, Marland and Pippin (1990). See also Annex 2.2 for a discussion
of uncertainties associated with fuel carbon contents. Recent updates to carbon factors for natural gas
and coal utilized the same approach as previous Inventories with updated recent data, therefore, the
uncertainty estimates around carbon contents of the different fuels as outlined in Annex 2.2 were not
impacted and the historic uncertainty ranges still apply.
Although national statistics of total fossil fuel and other energy consumption are relatively accurate, the
allocation of this consumption to individual end-use sectors (i.e., residential, commercial, industrial,
and transportation) is less certain. For example, for some fuels the sectoral allocations are based on
price rates (i.e., tariffs), but a commercial establishment may be able to negotiate an industrial rate or a
small industrial establishment may end up paying an industrial rate, leading to a misallocation of
emissions. Also, the deregulation of the natural gas industry and the more recent deregulation of the
electric power industry have likely led to some minor challenges in collecting accurate energy statistics
as firms in these industries have undergone significant restructuring.
120
110
Energy Consumption/Capita
40
Energy 3-43
-------
To calculate the total C02 emission estimate from energy-related fossil fuel combustion, the amount of
fuel used in non-energy production processes were subtracted from the total fossil fuel consumption.
The amount of C02 emissions resulting from non-energy related fossil fuel use has been calculated
separately and reported in the Carbon Emitted from Non-Energy Uses of Fossil Fuels section of this
report (Section 3.2). These factors all contribute to the uncertainty in the C02 estimates. Detailed
discussions on the uncertainties associated with carbon emitted from non-energy uses of fossil fuels
can be found within that section of this chapter.
Various sources of uncertainty surround the estimation of emissions from international bunker fuels,
which are subtracted from the U.S. totals (see the detailed discussions on these uncertainties provided
in Section 3.9). Another source of uncertainty is fuel consumption by U.S. Territories. The United States
does not collect energy statistics for its territories at the same level of detail as for the fifty states and
the District of Columbia. Therefore, estimating both emissions and bunker fuel consumption by these
territories is difficult.
Uncertainties in the emission estimates presented above also result from the data used to allocate C02
emissions from the transportation end-use sector to individual vehicle types and transport modes. In
many cases, bottom-up estimates of fuel consumption by vehicle type do not match aggregate fuel-type
estimates from EIA. Further research is planned to improve the allocation into detailed transportation
end-use sector emissions.
The uncertainty analysis was performed by primary fuel type for each end-use sector, using the IPCC-
recommended Approach 2 uncertainty estimation methodology, Monte Carlo stochastic simulation
technique, with @RISK software. For this uncertainty estimation, the inventory estimation model for C02
from fossil fuel combustion was integrated with the relevant variables from the inventory estimation
model for International Bunker Fuels, to realistically characterize the interaction (or endogenous
correlation) between the variables of these two models. About 170 input variables were modeled for
C02 from energy-related fossil fuel combustion (including about 20 for non-energy fuel consumption
and about 20 for International Bunker Fuels).
In developing the uncertainty estimation model, uniform distributions were assumed for all activity-
related input variables and emission factors, based on the SAIC/EIA (2001) report.41 Triangular
distributions were assigned for the oxidization factors (or combustion efficiencies). The uncertainty
ranges were assigned to the input variables based on the data reported in SAIC/EIA (2001) and on
conversations with various agency personnel.42
The uncertainty ranges for the activity-related input variables were typically asymmetric around their
inventory estimates; the uncertainty ranges for the emissions factors were symmetric. Bias (or
systematic uncertainties) associated with these variables accounted for much of the uncertainties
41 SAIC/EIA (2001) characterizes the underlying probability density function for the input variables as a combination of
uniform and normal distributions (the former to represent the bias component and the latter to represent the random
component). However, for purposes of the current uncertainty analysis, it was determined that uniform distribution was
more appropriate to characterize the probability density function underlying each of these variables.
42 In the SAIC/EIA (2001) report, the quantitative uncertainty estimates were developed for each of the three major fossil
fuels used within each end-use sector; the variations within the sub-fuel types within each end-use sectorwere not
modeled. However, for purposes of assigning uncertainty estimates to the sub-fuel type categories within each end-use
sector in the current uncertainty analysis, SAIC/EIA (2001 )-reported uncertainty estimates were extrapolated.
3-44 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
associated with these variables (SAIC/EIA 2001 ).43 For purposes of this uncertainty analysis, each input
variable was simulated 10,000 times through Monte Carlo sampling.
The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 3-18. Fossil fuel
combustion C02 emissions in 2023 were estimated to be between 4,465.9 and 4,753.8 MMT C02 Eq. at a
95 percent confidence level. This indicates a range of 2 percent below to 4 percent above the 2023
emission estimate of 4,559.4 MMT C02 Eq.
Table 3-18: Approach 2 Quantitative Uncertainty Estimates for C02 Emissions from
Energy-Related Fossil Fuel Combustion by Fuel Type and Sector (MMT C02 Eq. and
Percent)
Uncertainty Range Relative to Emission Estimate"
(MMT CO2 Eq.)
(%)
2023 Emission Estimate
Lower
Upper
Lower
Upper
Fuel/Sector
(MMT CO2 Eq.)
Bound
Bound
Bound
Bound
Coalb
734.7
709.8
804.2
-3% 9%
Residential
NO
NO
NO
NO
NO
Commercial
1.1
1.1
1.3
-5%
15%
Industrial
36.5
34.7
42.2
-5%
16%
Transportation
NO
NO
NO
NO
NO
Electric Power
694.6
668.0
762.0
-4%
10%
U.S. Territories
2.5
2.2
3.0
-12%
19%
Natural Gasb
1,725.8
1,704.6
1,805.0
-1%
5%
Residential
247.5
240.5
264.8
-3%
7%
Commercial
182.8
177.7
195.6
-3%
7%
Industrial
514.8
498.1
552.8
-3%
7%
Transportation
71.7
69.7
76.8
-3%
7%
Electric Power
704.5
684.2
740.6
-3%
5%
U.S. Territories
4.5
3.9
5.2
-12%
17%
Petroleumb
2,098.5
1,974.3
2,223.2
-6% 6%
Residential
59.6
56.2
62.8
-6%
5%
Commercial
60.2
56.9
63.5
-5%
5%
Industrial
241.3
186.8
296.7
-23%
23%
Transportation
1,704.7
1,597.2
1,812.3
-6%
6%
Electric Power
14.7
14.1
15.7
-4%
7%
U.S. Territories
17.9
16.6
19.7
-7%
10%
Geothermal
0.4
0.2
1.1
-46%
187%
Electric Power
0.4
0.2
1.1
-46%
187%
Total (including Geothermal)b
4,559.4
4,465.9
4,753.8
-2%
4%
NO (Not Occurring)
a Range of emission estimates predicted by Monte Carlo stochastic simulation for a 95 percent confidence interval.
bThe low and high estimates for total emissions were calculated separately through simulations and, hence, the low and high
emission estimates for the sub-source categories do not sum to total emissions.
Note: Totals may not sum due to independent rounding.
43 Although, in general, random uncertainties are the main focus of statistical uncertainty analysis, when the uncertainty
estimates are elicited from experts, their estimates include both random and systematic uncertainties. Hence, both
these types of uncertainties are represented in this uncertainty analysis.
Energy 3-45
-------
QA/QC and Verification
In order to ensure the quality of the C02 emission estimates from fossil fuel combustion, general (IPCC
Tier 1) and category-specific (Tier 2) Quality Assurance/Quality Control (QA/QC) procedures were
implemented consistent with the U.S. Inventory QA/QC plan outlined in Annex 8. The Tier 2 procedures
that were implemented involved checks specifically focusing on the activity data and methodology used
for estimating C02 emissions from fossil fuel combustion in the United States. Emission totals for the
different sectors and fuels were compared and trends were investigated to determine whether any
corrective actions were needed. Minor corrective actions were taken.
One area of QA/QC and verification is to compare the estimates and emission factors used in the
Inventory with other sources of C02 emissions reporting. Two main areas and sources of data were
considered. The first is a comparison with the EPA GHGRP combustion data (Subpart C) for stationary
combustion sources excluding the electric power sector. This mainly focused on considering carbon
factors for natural gas. The second comparison is with the EPA Air Markets Program data for electric
power production. This considered carbon factors for coal and natural gas used in electric power
production.
The EPA GHGRP collects greenhouse gas emissions data from large emitters including information on
fuel combustion. This excludes emissions from mobile sources and smaller residential and commercial
sources, those emissions are covered under supplier reporting (Subparts MM and NN) and are areas for
further research. Fuel combustion C02 data reported in 2023 was 1,969.1 MMT C02. Of that, 1,465.2
MMT C02 was from electricity production. Therefore, the non-electric power production fuel
combustion reporting was a fraction of the total covered by the Inventory under fossil fuel combustion.
Furthermore, reporters under the GHGRP can use multiple methods of calculating emissions; one
method is to use the default emission factors provided in the rule, while another is based on a Tier 3
approach using their own defined emission factors. Based on data from reporters on approach used, it
was determined that only about 10 percent of natural gas combustion emissions were based on a Tier 3
approach. Given the small sample size compared to the overall Inventory calculations for natural gas
combustion EPA determined it was not reasonable to consider the GHGRP Tier 3 natural gas factors at
this time. A more detailed analysis was done on upstream oil and gas natural gas combustion emissions
using the GHGRP data as discussed in Annex 2.2.
EPA collects detailed sulfur dioxide (S02), nitrogen oxides (NOx), and carbon dioxide (C02) emissions
data and other information from power plants across the country as part of the Acid Rain Program (ARP),
the Cross-State Air Pollution Rule (CSAPR), the CSAPR Update, and the Revised CSAPR Update (RCU).
The C02 data from these Air Market Programs (AMP) can be compared to the electric power sector
emissions calculated from the Inventory as shown in Table 3-19 for the three most recent years of data.
Table 3-19: Comparison of Electric Power Sector Emissions (MMT C02 Eq. and Percent)
CO2 Emissions (MMT CO2 Eq.)
% Change
Fuel/Sector
2021
2022
2023
2021-2022
2022-2023
Inventory Electric Power Sector
1,540.9
1,531.7
1,414.2
-0.6%
-7.7%
Coal
910.1
851.5
694.6
-6.4%
-18.4%
Natural Gas
612.8
659.3
704.5
7.6%
6.9%
Petroleum
17.7
20.5
14.7
15.9%
-28.3%
3-46 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
CO2 Emissions (MMT CO2 Eq.)
% Change
Fuel/Sector
2021
2022
2023
2021-2022
2022-2023
AMP Electric Power Sector
1,524.2
1,513.6
1,404.0
-0.7%
-7.2%
Coal
913.4
858.5
706.5
-6.0%
-17.7%
Natural Gas
609.6
652.7
695.7
7.1%
6.6%
Petroleum
1.3
2.5
1.8
83.6%
-29.1%
Note: Totals may not sum due to independent rounding.
In general, the emissions and trends from the two sources line up well. There are differences expected
based on coverage and scope of each source. The Inventory covers all emissions from the electric power
sector as defined above. The EPA AMP data covers emissions from electricity generating units of a
certain size so in some respects it could cover more sources (like electric power units at industrial
facilities that would be covered under the industrial sector in the Inventory) and not as many sources
(since smaller units are excluded). The EPA AMP data also includes heat input for different fuel types.
That data can be combined with emissions to calculate implied emission factors.44 The following Table
3-20 shows the implied emissions factors for coal and natural gas from the EPA AMP data compared to
the factors used in the Inventory for the three most recent years of data.
Table 3-20: Comparison of Emissions Factors (MMT Carbon/QBtu)
Fuel Type
2021
2022
2023
EPA AMP
Coal
25.67
25.54
25.47
Natural Gas
14.60
14.61
14.60
EPA Inventory
Electric Power Coal
26.13
26.13
26.15
Natural Gas
14.43
14.43
14.43
The factors for natural gas line up reasonably well, the EPA factors are roughly 1 percent lower than
those calculated from the EPA AMP data. For coal the EPA emissions factors are roughly 2 to 3 percent
higher than those calculated from the EPA AMP data. One possible reason for the difference is that the
EPA Inventory factors are based on all coal and natural gas used in electric power production while the
factors from the EPA AMP data are based on units where coal or natural gas is the primary source of fuel
used. There are units that use a mix of fuel sources but emissions for each fuel type could not be
calculated. This is an area of further research but given current data available the approach to develop
carbon factors as outlined in Annex 2 is still felt to be the most appropriate to represent total fuel
combustion in the United States.
A "top-down" reference approach for estimating C02 emissions from fossil fuel combustion in addition
to a "bottom-up" sectoral methodology is good practice in line with IPCC guidelines. The reference
approach (detailed in Annex 4) uses alternative methodologies and different data sources than those
contained in this section of the report. The reference approach estimates fossil fuel consumption by
adjusting national aggregate fuel production data for imports, exports, and stock changes rather than
relying on end-user consumption surveys. The reference approach assumes that once carbon-based
fuels are brought into a national economy, they are either saved in some way (e.g., stored in products,
44 These emission factors can be converted from MMT Carbon/QBtu to MMT CO2 Eq./QBtu by multiplying the emission
factor by 44/12, the molecular-to-atomic weight ratio of CO2 to C. This would assume the fraction oxidized to be 100
percent, which is the guidance in IPCC (2006) (see Annex 2.1).
Energy 3-47
-------
kept in fuel stocks, or left unoxidized in ash) or combusted, and therefore the carbon in them is oxidized
and released into the atmosphere. In the reference approach, accounting for actual consumption of
fuels at the sectoral or sub-national level is not required. One difference between the two approaches is
that emissions from carbon that was not stored during non-energy use of fuels are subtracted from the
sectoral approach and reported separately (see Section 3.2). These emissions, however, are not
subtracted in the reference approach. As a result, the reference approach emission estimates are
comparable to those of the sectoral approach, with the exception that the non-energy use (NEU) source
category emissions are included in the reference approach (see Annex 4 for more details).
Recalculations Discussion
EIA (2025a) updated natural gas consumed by all sectors in 2020 and 2022, as well as petroleum
consumed by all sectors in 2021 and 2022. Additionally, EIA (2024a) updated U.S. Territories petroleum
for the years 2020 through 2022, and U.S. Territories natural gas and coal consumption for the year
2022These updates caused total C02 emissions to increase by an annual average of 0.01 MMT C02 Eq.
(less than 0.05 percent) in the years 1990 through 2022 compared to the previous Inventory.
Planned Improvements
To reduce the uncertainty of C02 from fossil fuel combustion estimates for U.S. Territories, further expert
elicitation may be conducted to better quantify the total uncertainty associated with emissions from
U.S. Territories. Additionally, although not technically a fossil fuel, since geothermal energy-related C02
emissions are included for reporting purposes, further expert elicitation may be conducted to better
quantify the total uncertainty associated with C02 emissions from geothermal energy use.
EPA will continue to examine the availability of facility-level combustion emissions through EPA's
GHGRP to help better characterize the industrial sector's energy consumption in the United States and
further classify total industrial sector fossil fuel combustion emissions by business establishments
according to industrial economic activity type. Most methodologies used in EPA's GHGRP are consistent
with IPCC methodologies, although for EPA's GHGRP, facilities collect detailed information specific to
their operations according to detailed measurement standards, which may differ with the more
aggregated data collected for the Inventory to estimate total national U.S. emissions. In addition, and
unlike the reporting in this chapter, some facility-level fuel combustion emissions reported under the
GHGRP may also include industrial process emissions. In line with IPCC guidelines, fuel combustion
emissions are included in this chapter, while process emissions are included in the Industrial Processes
and Product Use chapter of this report. In examining data from EPA's GHGRP that would be useful to
improve the emission estimates for the C02 from fossil fuel combustion category, particular attention
will also be made to ensure time-series consistency, as the facility-level reporting data from EPA's
GHGRP are not available for all inventory years as reported in this Inventory.
Additional analyses will be conducted to align reported facility-level fuel types and IPCC fuel types per
the national energy statistics. For example, additional work will look at C02 emissions from biomass to
ensure they are separated in the facility-level reported data and maintaining consistency with national
energy statistics provided by EIA. In implementing improvements and integration of data from EPA's
GHGRP, the latest guidance from the IPCC on the use of facility-level data in national inventories will
continue to be relied upon.45
45 See http://www.ipcc-nggip.iges.or.jp/public/tb/TFI Technical Bulletin 1.pdf.
3-48 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
EPA is also evaluating the methods used to adjust for conversion of fuels and exports of C02. EPA is
including an approach used to account for C02 transport, injection, and geologic storage in this
Inventory, as part of this ongoing work there may be changes made to the accounting for C02 exports.
Finally, another ongoing planned improvement is to evaluate data availability to update the carbon and
heat content of more fuel types accounted for in this Inventory. This update will impact consumption
and emissions across all sectors and will improve consistency with EIA data as carbon and heat
contents of fuels will be accounted for as annually variable and therefore improve accuracy across the
time series. Some of the fuels considered in this effort include petroleum coke, residual fuel, and woody
biomass.
CH4and N20 from Stationary Combustion
Methodology and Time-Series Consistency
Methane and N20 emissions from stationary combustion were estimated by multiplying fossil fuel and
wood consumption data by emission factors (by sector and fuel type for industrial, residential,
commercial, and U.S. Territories; and by fuel and technology type for the electric power sector). The
electric power sector utilizes a Tier 2 methodology, whereas all other sectors utilize a Tier 1
methodology in accordance with IPCC methodological decision tree Figure 2.1 in the 2006IPCC
Guidelines for National Greenhouse Gas Inventories (IPCC 2006) and available data. The activity data
and emission factors used are described in the following subsections.
More detailed information on the methodology for calculating emissions from stationary combustion,
including emission factors and activity data, is provided in Annex 3.1.
Industrial, Residential, Commercial, and U.S. Territories
National coal, natural gas, fuel oil, and wood consumption data were grouped by sector: industrial,
commercial, residential, and U.S. Territories. For the CH4 and N20 emission estimates, consumption
data for each fuel were obtained from ElA's Monthly Energy Review (EIA 2025). Because the United
States does not include territories in its national energy statistics, fuel consumption data for territories
were provided separately by ElA's International Energy Statistics (EIA 20 24).46 Fuel consumption for the
industrial sector was adjusted to subtract out mobile source construction and agricultural use, which is
reported under mobile sources. Construction and agricultural mobile source fuel use was obtained from
EPA (2024b) and FHWA (1996 through 2024). Estimates for wood biomass consumption for fuel
combustion do not include municipal solid waste, tires, etc., that are reported as biomass by EIA. Non-
C02 emissions from combustion of the biogenic portion of municipal solid waste and tires are included
under waste incineration (Section 3.3). Estimates for natural gas combustion do not include biogas, and
therefore non-C02 emissions from biogas are not included (see the Planned Improvements section,
below). Tier 1 default emission factors for the industrial, commercial, and residential end-use sectors
were provided by the 2006 IPCC Guidelines for National Greenhouse Gas Inventories (IPCC 2006). U.S.
Territories' emission factors were estimated using the U.S. emission factors for the primary sector in
which each fuel was combusted.
46 U.S. Territories data also include combustion from mobile activities because data to allocate territories' energy use were
unavailable. For this reason, ChU and N2O emissions from combustion by U.S. Territories are only included in the
stationary combustion totals.
Energy 3-49
-------
Electric Power Sector
The electric power sector uses a Tier 2 emission estimation methodology as fuel consumption for the
electric power sector by control-technology type is based on EPA's Acid Rain Program Dataset (EPA
2024). Total fuel consumption in the electric power sector from EIA (2025) was apportioned to each
combustion technology type and fuel combination using a ratio of fuel consumption by technology type
derived from EPA (2024a) data. The combustion technology and fuel use data by facility obtained from
EPA (2024a) were only available from 1996 to 2023 so the consumption estimates from 1990 to 1995
were estimated by applying the 1996 consumption ratio by combustion technology type from EPA
(2024a) to the total EIA (2024a) consumption for each year from 1990 to 1995.
Emissions were estimated by multiplying fossil fuel and wood consumption by technology-, fuel-, and
country-specific Tier 2 emission factors. The Tier 2 emission factors used are based in part on emission
factors published by EPA, and EPA's Compilation of Air Pollutant Emission Factors, AP-42 (EPA 1997) for
coal wall-fired boilers, residual fuel oil, diesel oil and wood boilers, natural gas-fired turbines, and
combined cycle natural gas units.47
As discussed in Annex 5, data are unavailable to include estimates of CH4 and N20 emissions from
biomass use in Territories, but those emissions are assumed to be insignificant.
Uncertainty
Methane emission estimates from stationary sources exhibit high uncertainty, primarily due to
difficulties in calculating emissions from wood combustion (i.e., fireplaces and wood stoves). The
estimates of CH4 and N20 emissions presented are based on broad indicators of emissions (i.e., fuel use
multiplied by an aggregate emission factor for different sectors), rather than specific emission
processes (i.e., by combustion technology and type of emission control).
An uncertainty analysis was performed by primary fuel type for each end-use sector, using the IPCC-
recommended Approach 2 uncertainty estimation methodology, Monte Carlo stochastic simulation
technique, with @RISK software.
The uncertainty estimation model for this source category was developed by integrating the CH4 and
N20 stationary source inventory estimation models with the model for C02 from fossil fuel combustion
to realistically characterize the interaction (or endogenous correlation) between the variables of these
three models. About 55 input variables were simulated for the uncertainty analysis of this source
category (about 20 from the C02 emissions from fossil fuel combustion inventory estimation model and
about 35 from the stationary source inventory models).
In developing the uncertainty estimation model, uniform distribution was assumed for all activity-
related input variables and N20 emission factors, based on the SAIC/EIA (2001) report. 48 For these
variables, the uncertainty ranges were assigned to the input variables based on the data reported in
47 Several of the U.S. Tier 2 emission factors were used in IPCC (2006) as Tier 1 emission factors. See Table A-68 in Annex
3.1 for emission factors by technology type and fuel type for the electric power sector.
48 SAIC/EIA (2001) characterizes the underlying probability density function for the input variables as a combination of
uniform and normal distributions (the former distribution to represent the bias component and the latter to represent the
random component). However, for purposes of the current uncertainty analysis, it was determined that uniform
distribution was more appropriate to characterize the probability density function underlying each of these variables.
3-50 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
SAIC/EIA (2001 )..49 However, the CH4 emission factors differ from those used by EIA. These factors and
uncertainty ranges are based on IPCC default uncertainty estimates (IPCC 2006).
The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 3-101.
Stationary combustion CH4 emissions in 2023 (including biomass) were estimated to be between 5.8
and 19.7 MMT C02 Eq. at a 95 percent confidence level. This indicates a range of 34 percent below to
125 percent above the 2023 emission estimate of 8.8 MMT C02 Eq.50 Stationary combustion N20
emissions in 2023 (including biomass) were estimated to be between 15.0 and 29.5 MMT C02 Eq. at a 95
percent confidence level. This indicates a range of 23 percent below to 51 percent above the 2023
emission estimate of 19.6 MMT C02 Eq.
Table 3-21: Approach 2 Quantitative Uncertainty Estimates for CH4 and N20 Emissions
from Energy-Related Stationary Combustion, Including Biomass (MMT C02 Eq. and
Percent)
Uncertainty Range Relative to Emission Estimate"
2023 Emission
(MMT CO2 Eq.)
(%)
Source
Gas
Estimate
(MMT CO2 Eq.)
Lower
Bound
Upper
Bound
Lower Upper
Bound Bound
Stationary Combustion
CH4
8.8
5.8
19.7
-34% +125%
Stationary Combustion
n2o
19.6
15.0
29.5
-23% +51%
a Range of emission estimates predicted by Monte Carlo stochastic simulation for a 95 percent confidence interval.
The uncertainties associated with the emission estimates of CH4 and N20 are greater than those
associated with estimates of C02 from fossil fuel combustion, which mainly rely on the carbon content
of the fuel combusted. Uncertainties in both CH4 and N20 estimates are due to the fact that emissions
are estimated based on emission factors representing only a limited subset of combustion conditions.
For the indirect greenhouse gases, uncertainties are partly due to assumptions concerning combustion
technology types, age of equipment, emission factors used, and activity data projections.
QA/QC and Verification
In order to ensure the quality of the non-C02 emission estimates from stationary combustion, general
(IPCC Tier 1) and category-specific (Tier 2) Quality Assurance/Quality Control (QA/QC) procedures were
implemented consistent with the U.S. Inventory QA/QC plan outlined in Annex 8. The Tier 2 procedures
that were implemented involved checks specifically focusing on the activity data and emission factor
sources and methodology used for estimating CH4, N20, and the greenhouse gas precursors from
stationary combustion in the United States. Emission totals for the different sectors and fuels were
compared and trends were investigated.
49 In the SAIC/EIA (2001) report, the quantitative uncertainty estimates were developed for each of the three major fossil
fuels used within each end-use sector; the variations within the sub-fuel types within each end-use sectorwere not
modeled. However, for purposes of assigning uncertainty estimates to the sub-fuel type categories within each end-use
sector in the current uncertainty analysis, SAIC/EIA (2001 )-reported uncertainty estimates were extrapolated.
50 The low emission estimates reported in this section have been rounded down to the nearest integer values and the high
emission estimates have been rounded up to the nearest integer values.
Energy 3-51
-------
Recalculations Discussion
EPA adjusted the share of total consumption apportioned to each combustion technology type for 2022
to correct a previous error. EIA (2025) updated the heat constant of bituminous coal for the time series.
EIA (2025) updated natural gas consumed by all sectors in 2020 and 2022, as well as petroleum
consumed by all sectors in 2021 and 2022. EIA (2025) also updated electricity statistics which affected
commercial sector wood consumption for the years 2014 through 2022. Additionally, EIA (2024) updated
U.S. Territories petroleum for the years 2020 through 2022, and U.S. Territories natural gas and coal
consumption for the year 2022. These updates resulted in an average annual decrease of less than 0.5
MMT C02 Eq. (0.1 percent) in CH4 emissions and an average annual decrease of 0.1 MMT C02 Eq. (0.3
percent) in N20 emissions across the time series compared to the previous Inventory.
Planned Improvements
Several items are being evaluated to improve the CH4 and N20 emission estimates from stationary
combustion and to reduce uncertainty for U.S. Territories. Efforts will be taken to work with EIA and other
agencies to improve the quality of the U.S. Territories data. Because these data are not broken out by
stationary and mobile uses, further research will be aimed at trying to allocate consumption
appropriately. In addition, the uncertainty of biomass emissions will be further investigated because it
was expected that the exclusion of biomass from the estimates would reduce the uncertainty; and in
actuality the exclusion of biomass increases the uncertainty. The CH4 emission factor for residential
wood combustion developed by NESCAUM (2024) will also be reviewed and potentially incorporated
based on this review. Factors for methane slip will also be reviewed. These improvements are not all-
inclusive but are part of an ongoing analysis and efforts to continually improve these stationary
combustion estimates from U.S. Territories.
Other forms of biomass-based gas consumption include biogas. As an additional planned
improvement, EPA will examine EIA and GHGRP data on biogas collected and burned for energy use and
determine if CH4 and N20 emissions from biogas can be included in future Inventories. EIA (2024a)
natural gas data already deducts biogas used in the natural gas supply, so no adjustments are needed to
the natural gas fuel consumption data to account for biogas.
CH4 and N20 from Mobile Combustion
Methodology and Time-Series Consistency
Estimates of CH4 and N20 emissions from mobile combustion were calculated by multiplying emission
factors by measures of activity for each fuel and vehicle type (e.g., light-duty gasoline trucks). Activity
data included vehicle miles traveled (VMT) for on-road vehicles and fuel consumption for non-road
mobile sources. The activity data and emission factors used in the calculations are described in the
subsections that follow. A complete discussion of the methodology used to estimate CH4 and N20
emissions from mobile combustion and the emission factors used in the calculations is provided in
Annex 3.2.
On-Road Vehicles
Estimates of CH4 and N20 emissions from gasoline and diesel on-road vehicles are based on VMT and
emission factors (in grams of CH4 and N20 per mile) by vehicle type, fuel type, model year, and emission
3-52 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
control technology. Emission estimates for alternative fuel vehicles (AFVs) are based on VMT and
emission factors (in grams of CH4 and N20 per mile) by vehicle and fuel type.51
CH4 and N20 emissions factors by vehicle type and emission tier for newer (starting with modelyear
2004) on-road gasoline vehicles were calculated by Browning (2019) from annual vehicle certification
data compiled by EPA. CH4 and N20 emissions factors for older (modelyear 2003 and earlier) on-road
gasoline vehicles were developed by ICF (2004). These earlier emission factors were derived from EPA,
California Air Resources Board (CARB) and Environment and Climate Change Canada (ECCC) laboratory
test results of different vehicle and control technology types. The EPA, CARB and ECCC tests were
designed following the Federal Test Procedure (FTP). The procedure covers three separate driving
segments, since vehicles emit varying amounts of greenhouse gases depending on the driving segment.
These driving segments are: (1) a transient driving cycle that includes cold start and running emissions,
(2) a cycle that represents running emissions only, and (3) a transient driving cycle that includes hot start
and running emissions. For each test run, a bag was affixed to the tailpipe of the vehicle and the exhaust
was collected; the content of this bag was then analyzed to determine quantities of gases present. The
emissions characteristics of driving segment 2 tests were used to define running emissions. Running
emissions were subtracted from the total FTP emissions to determine start emissions. These were then
recombined to approximate average driving characteristics, based upon the ratio of start to running
emissions for each vehicle class from MOBILE6.2, an EPA emission factor model that predicts grams per
mile emissions of C02, CO, HC, NOx, and PM from vehicles under various conditions.52
Diesel on-road vehicle emission factors were developed by ICF (2006). CH4 and N20 emissions factors
for newer (starting with modelyear 2007) on-road diesel vehicles (those using engine aftertreatment
systems) were calculated from annual vehicle certification data compiled by EPA.
CH4 and N20 emission factors for AFVs were developed based on the 2023 Greenhouse gases,
Regulated Emissions, and Energy use in Transportation (GREET) model (ANL 2023). For light-duty trucks,
EPA used travel fractions for LDT1 and LDT2 (MOVES Source Type 31 for LDT1 and MOVES Source Type
32 for LDT2; see Annex 3.2 for information about the MOVES model) to determine emission factors. For
medium-duty vehicles, EPA used emission factors for light heavy-duty vocational trucks. For heavy-duty
vehicles, EPA used emission factors for long-haul combination trucks. For buses, EPA used emission
factors for transit buses. These values represent vehicle operations only (tank-to-wheels); upstream
well-to-tank emissions are calculated elsewhere in the Inventory. Biodiesel CH4 emission factors were
corrected from GREET values to be the same as CH4 emission factors for diesel vehicles. GREET
overestimated biodiesel CH4 emission factors based upon an incorrect CH4-to-THC ratio for diesel
vehicles with aftertreatment technology.
Annual VMT data for 1990 through 2022 were obtained from the Federal Highway Administration's
(FHWA) Highway Performance Monitoring System database as reported in Highway Statistics (FHWA
1996 through 2024). VMT data for 2023 was proxied based on FHWA's Traffic Volume Trends Data for
2023. VMT estimates were then allocated to vehicle type using ratios of VMT per vehicle type to total
VMT, derived from EPA's MOVES5 model (see Annex 3.2 for information about the MOVES model). This
corrects time series inconsistencies in FHWA definitions of vehicle types (Browning 2022a). VMT for
alternative fuel vehicles (AFVs) were estimated based on Browning (2024). The age distributions of the
51 Alternative fuel and advanced technology vehicles are those that can operate using a motor fuel other than gasoline or
diesel. This includes electric or other bi-fuel or dual-fuel vehicles that may be partially powered by gasoline or diesel.
52 Additional information regarding the MOBILE model can be found at https://www.epa.gov/moves/description-and-
history-mobile-highway-vehicle-emission-factor-model.
Energy 3-53
-------
U.S. vehicle fleet were obtained from EPA (2004, 2024a), and the average annual age-specific vehicle
mileage accumulation of U.S. vehicles were obtained from EPA (2024a).
Control technology and standards data for on-road vehicles were obtained from EPA's Office of
Transportation and Air Quality (EPA 1998, 2022b, 2023, and 2024) and Browning (2005). These
technologies and standards are defined in Annex 3.2, and were compiled from EPA (1994a, 1994b, 1998,
1999) and IPCC (2006) sources.
Non-Road Mobile Sources
The nonroad mobile category for CH4 and N20 includes ships and boats, aircraft, locomotives, and other
mobile non-road sources (e.g., construction or agricultural equipment). For locomotives, aircraft, ships,
and non-recreational boats, fuel-based emission factors are applied to data on fuel consumption,
following the IPCC Tier 1 approach, The Tier 2 approach for these sources would require separate fuel-
based emissions factors by technology, for which data are not currently available. For other non-road
sources, EPA uses the Nonroad component of the MOVES model to estimate fuel use. Emission factors
by horsepower bin are estimated from EPA engine certification data. Because separate emission factors
are applied to specific engine technologies; these non-road sources utilize a Tier 2 approach.
To estimate CH4 and N20 emissions from non-road mobile sources, fuel consumption data were
employed as a measure of activity and multiplied by fuel-specific emission factors (in grams of N20 and
CH4 per kilogram of fuel consumed).53 Activity data were obtained from AAR (2008 through 2024), APTA
(2007 through 2024), Rail Inc (2014 through 2024), APTA (2006), BEA(1991 through 2015), Benson (2002
through 2004), Bureau of Transportation Statistics (BTS; 2023), DLA Energy (2025), DOC (1991 through
2024), DOE (1993 through 2022), DOT (1991 through 2024), EIA (2002, 2007, 2024, 2023), EIA (1991
through 2023), EPA (2024a), Esser (2003 through 2004), FAA (2022), FHWA (1996 through 2024),54
Gaffney (2007), FTA (2023), and Whorton (2006 through 2014). Fuel consumption data regarding jet fuel,
on-road vehicles, and diesel consumption in US territories and vessel bunking were proxied from 2022,
awaiting publication of updated data. Fuel consumption data for boats and vessels in U.S. Territories
data and vessel domestic vessel bunkering is proxied from 2022 proxy data. Emission factors for non-
road modes were taken from IPCC (2006) and Browning (2020 and 2018).
Uncertainty
A quantitative uncertainty analysis was conducted for the mobile source sector using the IPCC-
recommended Approach 2 uncertainty estimation methodology, Monte Carlo stochastic simulation
technique, using @RISK software. The uncertainty analysis was performed on 2023 estimates of CH4
and N20 emissions, incorporating probability distribution functions associated with the major input
53 The consumption of international bunker fuels is not included in these activity data, but emissions related to the
consumption of international bunker fuels are estimated separately under the International Bunker Fuels source
category.
54 This Inventory uses FHWA's Agriculture, Construction, and Commercial/Industrial MF-24 fuel volumes along with the
MOVES model gasoline volumes to estimate non-road mobile source CFU and N2O emissions for these categories. For
agriculture, the MF-24 gasoline volume is used directly because it includes both non-road trucks and equipment. For
construction and commercial/industrial category gasoline estimates, the 2014 and older MF-24 volumes represented
non-road trucks only; therefore, the MOVES gasoline volumes for construction and commercial/industrial categories are
added to the respective categories in the Inventory. Beginning in 2015, this addition is no longer necessary since the
FHWA updated its methods for estimating on-road and non-road gasoline consumption. Among the method updates,
FHWA now incorporates MOVES equipment gasoline volumes in the construction and commercial/industrial categories.
3-54 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
variables. For the purposes of this analysis, the uncertainty was modeled for the following four major
sets of input variables: (1) VMT data, by on-road vehicle and fuel type, (2) emission factor data, by on-
road vehicle, fuel, and control technology type, (3) fuel consumption, data, by non-road vehicle and
equipment type, and (4) emission factor data, by non-road vehicle and equipment type.
Uncertainty analyses were not conducted for NOx, CO, or NMVOC emissions. Emission factors for these
gases have been extensively researched because emissions of these gases from motor vehicles are
regulated in the United States, and the uncertainty in these emission estimates is believed to be
relatively low. For more information, see Section 3.11. However, a much higher level of uncertainty is
associated with CH4 and N20 emission factors due to limited emission test data, and because, unlike
C02 emissions, the emission pathways of CH4 and N20 are highly complex.
Based on the uncertainty analysis, mobile combustion CH4 emissions from all mobile sources in 2023
were estimated to be between 2.5 and 3.3 MMT C02 Eq. at a 95 percent confidence level. This indicates
a range of 4 percent below to 30 percent above the corresponding 2023 emission estimate of 2.5 MMT
C02 Eq. Mobile combustion N20 emissions from mobile sources in 2023 were estimated to be between
15.1 and 19.8 MMT C02 Eq. at a 95 percent confidence level. This indicates a range of 7 percent below to
22 percent above the corresponding 2023 emission estimate of 16.2 MMT C02 Eq.
Table 3-22: Approach 2 Quantitative Uncertainty Estimates for CH4 and N20 Emissions
from Mobile Sources (MMT C02 Eq. and Percent)
Uncertainty Range Relative to Emission Estimate"
2023 Emission
(MMT CO2 Eq.)
(%)
Estimate
Lower
Upper
Lower
Upper
Source
Gas
(MMT CO2 Eq.)
Bound
Bound
Bound
Bound
Mobile Sources
CH4
2.5
2.5
3.3
-4%
+30%
Mobile Sources
n2o
16.2
15.1
19.8
-7%
+22%
a Range of emission estimates predicted by Monte Carlo stochastic simulation for a 95 percent confidence interval.
This uncertainty analysis is a continuation of a multi-year process for developing quantitative
uncertainty estimates for this source category using the IPCC Approach 2 uncertainty estimation
methodology. As a result, as new information becomes available, uncertainty characterization of input
variables may be improved and revised. For additional information regarding uncertainty in emission
estimates for CH4 and N20 please refer to the Uncertainty Annex. As discussed in Annex 5, data are
unavailable to include estimates of CH4 and N20 emissions from any liquid fuel used in pipeline
transport or some biomass used in transportation sources, but those emissions are assumed to be
insignificant.
QA/QC and Verification
In order to ensure the quality of the emission estimates from mobile combustion, general (IPCC Tier 1)
and category-specific (Tier 2) Quality Assurance/Quality Control (QA/QC) procedures were
implemented consistent with the U.S. Inventory QA/QC plan outlined in Annex 8. The specific plan used
for mobile combustion was updated prior to collection and analysis of this current year of data. The Tier
2 procedures focused on the emission factor and activity data sources, as well as the methodology used
for estimating emissions. These procedures included a qualitative assessment of the emission
estimates to determine whether they appear consistent with the most recent activity data and emission
factors available. A comparison of historical emissions between the current Inventory and the previous
Energy 3-55
-------
Inventory was also conducted to ensure that the changes in estimates were consistent with the changes
in activity data and emission factors.
Recalculations Discussion
New data on the activity of battery and plug-in hybrid electric vehicles was used to estimate electric
vehicle mileage (Browning, 2024). Past inventories estimated that electric vehicles had similar mileage
accumulation to conventionally fueled vehicles. The current inventory uses more recent data that shows
electric vehicles drive fewer miles annually than the average conventionally fueled vehicle. This annual
mileage update resulted in total C02 emissions from electricity consumption by electric vehicles to
decrease by 1.6 MMT (26 percent) in 2022 and 1.1 MMT in 2021 (22 percent), compared to the previous
Inventory. CH4 and N20 emissions are not impacted by this update.
Updated alternative fuel emissions factors based on the latest GREET model (GREET 2023) were also
included. As a result of these updates, CH4 emissions from alternative fuel vehicles increased by an
annual average of .01 MMT C02 Eq. (27 to 30 percent) in the years 2020 through 2022 compared to the
previous Inventory. Alternative fuel vehicle N20 emissions decreased by .01 MMT C02 Eq. (8 percent) in
2022 and changes in N20 emissions in 2020 and 2021 were less than .01 MMT C02 Eq., relative to the
previous Inventory.
Output from the recently released MOVES5 model was used to update the vehicle fleet composition by
age and type and the estimated mileage and fuel use by vehicle type and model year. This change
affects the historical time series of emissions by vehicle type. A significant component of this change is
replacing projections based on MOVES3, with actual vehicle fleet data from MOVES5. MOVES5 also
includes more accurate projections based on the most recent data. Due to this update, N20 emissions
from gasoline powered highway vehicles increased by 0.03 MMT C02 Eq. (0.4 percent) in 2021 and 0.08
MMT C02 Eq. (1.5 percent) in 2022 compared to the previous Inventory. Total N20 emissions for diesel
highway vehicles increased by less than 0.01 MMT C02 Eq. (0.1 percent) in 2021 and decreased by 0.1
MMT C02 Eq (2.9 percent) in 2022 compared to the previous Inventory. Changes in CH4 emissions for
gasoline highway vehicles were small for 2021 and 2022, between -0.01 MMT C02 Eq. (-1.1 percent) in
2021 and less than 0.01 MMT C02 Eq. (0.4 percent) in 20222. Changes in CH4 emissions from diesel
highway vehicles were small, decreasing 0.01 MMT C02 Eq. (9.4 percent) in 2021 and 0.02 (14.1 percent)
in 2022.
Together, these updates resulted in an average annual increase of 0.4 MMT C02 Eq. (7.9 percent) in CH4
emissions and an average annual increase 1.4 MMT C02 Eq. (4.8 percent) in N20 emissions across the
time series compared to the previous Inventory.
Planned Improvements
While the data used for this report represent the most accurate information available, several areas for
improvement have been identified.
• Improve estimates of electric vehicle activity and energy use. EIA publishes output from a model
that estimates electric vehicle energy consumption starting in 2018 for passenger vehicles.
Model results from EIA could be used to improve the estimates of electric vehicle activity and
energy use in the Inventory.
3-56 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
• Update emission factors for ships and non-recreational boats using residual fuel and distillate
fuel. Develop emission factors for locomotives using ultra-low sulfur diesel and emission
factors for aircraft using jet fuel. The Inventory currently uses IPCC default values for these
emission factors.
• Continue to explore potential improvements to estimates of domestic waterborne fuel
consumption for future Inventories. The Inventory estimates for residual and distillate fuel used
by ships and boats is based in part on data on bunker fuel use from the U.S. Department of
Commerce. Domestic fuel consumption is estimated by subtracting fuel sold for international
use from the total sold in the United States. Since 2015, all ships travelling within 200 nautical
miles of the U.S. coastlines must use distillate fuels, thereby overestimating the residual fuel
used by U.S. vessels and underestimating distillate fuel use in these ships. Additionally, the EIA
has stopped publishing the Fuel Oil and Kerosene Sales report, which reported data on distillate
marine fuel use in the U.S. and the territories. This affects the volume of fuel and emissions that
are allocated to the domestic ships and boats source, although top-down data is still available
from the Monthly Energy Review that will be used to estimate total domestic emission from
diesel fuel use. New data and methods are being explored to improve the diesel ships and boats
emissions estimates going forward.
3.2 Carbon Emitted from Non-Energy Uses
of Fossil Fuels (Source Category 1 A)
In addition to being combusted for energy, fossil fuels are also consumed for non-energy uses (NEU) in
the United States. The fuels used for these purposes are diverse, including natural gas, hydrocarbon gas
liquids (HGL),55asphalt (a viscous liquid mixture of heavy crude oil distillates), petroleum coke
(manufactured from heavy oil), and coal (metallurgical) coke (manufactured from coking coal). The non-
energy applications of these fuels are equally diverse, including feedstocks for the manufacture of
plastics, rubber, synthetic fibers and other materials; reducing agents for the production of various
metals and inorganic products; and products such as lubricants, waxes, and asphalt (IPCC 2006).
Emissions from non-energy use of lubricants, paraffin waxes, bitumen/asphalt, and solvents are
reported in the Energy sector, as opposed to the Industrial Processes and Product Use (IPPU) sector, to
reflect national circumstances in its choice of methodology and to increase transparency of this source
category's unique country-specific data sources and methodology (see Box 3-5). In addition, estimates
of non-energy use emissions included here do not include emissions already reflected in the IPPU
sector, e.g., fuels used as reducing agents. To avoid double counting, the "raw" non-energy fuel
consumption data reported by EIA are reduced to account for these emissions already included under
IPPU.
Carbon dioxide emissions arise from non-energy uses via several pathways. Emissions may occur
during the manufacture of a product, as is the case in producing plastics or rubber from fuel-derived
55 HGL (formerly referred to as liquefied petroleum gas, or LPG) are hydrocarbons that occur as gases at atmospheric
pressure and as liquids under higher pressures. HGLs include paraffins, such as ethane, propane, butanes, isobutane,
and natural gasoline (formerly referred to as pentanes plus), and HGLs include olefins, such as ethylene, propylene,
butylene and isobutylene.
Energy 3-57
-------
feedstocks. Additionally, emissions may occur during the product's lifetime, such as during solvent use.
Overall, throughout the time series and across all uses, about 64 percent of the total carbon consumed
for non-energy purposes was stored in products (e.g., plastics), and not released to the atmosphere; the
remaining 36 percent was emitted.
There are several areas in which non-energy uses of fossil fuels are closely related to other parts of this
Inventory. For example, some of the non-energy use products release C02 at the end of their
commercial life when they are combusted after disposal; these emissions are reported separately
within the Energy chapter in the Incineration of Waste source category. There are also net exports of
petrochemical intermediate products that are not completely accounted for in the EIA data, and the
Inventory calculations adjust for the effect of net exports on the mass of carbon in non-energy
applications.
As shown in Table 3-23, fossil fuel emissions in 2023 from the non-energy uses of fossil fuels were 107.1
MMT C02 Eq., which constituted approximately 2.2 percent of overall fossil fuel emissions. In 2023, the
consumption of fuels for non-energy uses (after the adjustments described above) was 5,570.9 TBtu
(see Table 3-24). A portion of the carbon in the 5,570.9 TBtu of fuels was stored (240.9 MMT C02 Eq.),
while the remaining portion was emitted (107.1 MMT C02 Eq.). Non-energy use emissions increased by
5.3 percent from 2022 to 2023, primarily due to increases in HGL production, industry lubricants, and
transportation lubricants. See Annex 2.3 for more details.
Table 3-23: C02 Emissions from Non-Energy Use Fossil Fuel Consumption (MMT C02
Eq. and Percent C)
Year
1990
2005
2019
2020
2021
2022
2023
Potential Emissions
292.5
357.7
334.7
328.7
345.1
338.5
348.0
C Stored
193.4
232.7
228.2
230.8
233.4
236.8
240.9
Emissions as a % of Potential
34%
35%
32%
30%
32%
30%
31%
C Emitted
99.1
125.0
106.5
97.9
111.7
101.7
107.1
Notes: NEU emissions presented in this table differ from the NEU emissions presented in the common data tables since those
report NEU emissions from U.S. Territories under the U.S. Territories category and not under the NEU category. Totals may not
sum due to independent rounding.
Methodology and Time-Series Consistency
As per discussion of methodology for estimating C02 emissions from fossil fuel combustion, NEU
emissions are estimated in line with a Tier 2 method described by the IPCC in the 2006IPCC Guidelines
for National Greenhouse Gas Inventories (IPCC 2006) Chapter 2, Figure 2.1 decision tree and available
data on energy use and country specific fuel carbon contents. The first step in estimating carbon stored
in products was to determine the aggregate quantity of fossil fuels consumed for non-energy uses. The
carbon content of these feedstock fuels is equivalent to potential emissions, or the product of
consumption and the fuel-specific carbon content values. Both the non-energy fuel consumption and
carbon content data were supplied by the EIA (2024) (see Annex 2.1). Consumption values for industrial
coking coal, petroleum coke, other oils, and natural gas in Table 3-24 and Table 3-25 have been adjusted
to subtract non-energy uses that are included in the source categories of the Industrial Processes and
3-58 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Product Use chapter.56 Consumption of natural gas, HGL, naphthas, other oils, and special naphtha
were adjusted to subtract out net exports of these products that are not reflected in the raw data from
EIA. Consumption values were also adjusted to subtract net exports of HGL components (e.g.,
propylene, ethane).
For the remaining non-energy uses, the quantity of carbon stored was estimated by multiplying the
potential emissions by a storage factor.
• For several fuel types—petrochemical feedstocks (including natural gas for non-fertilizer uses,
HGL, naphthas, other oils, still gas, special naphtha, and industrial other coal), asphalt and road
oil, lubricants, and waxes—U.S. data on carbon stocks and flows were used to develop carbon
storage factors, calculated as the ratio of (a) the carbon stored by the fuel's non-energy
products to (b) the total carbon content of the fuel consumed. A lifecycle approach was used in
the development of these factors in order to account for losses in the production process and
during use. Because losses associated with municipal solid waste management are handled
separately in the Energy sector under the Incineration of Waste source category, the storage
factors do not account for losses at the disposal end of the life cycle.
• For industrial coking coal and distillate fuel oil, storage factors were taken from Marland and
Rotty (1984).
• For the remaining fuel types (petroleum coke, miscellaneous products and other petroleum),
IPCC (2006) does not provide guidance on storage factors, and assumptions were made based
on the potential fate of carbon in the respective non-energy use products. Carbon dioxide
emissions from carbide production are implicitly accounted for in the storage factor calculation
for the non-energy use of petroleum coke.
Table 3-24: Adjusted Consumption of Fossil Fuels for Non-Energy Uses (TBtu)
Year
1990
2005
2019
2020
2021
2022
2023
Industry
4,110.2
4,961.2
5,144.0
5,096.3
5,342.2
5,281.0
5,472.4
Industrial Coking Coal
NO
80.4
113.0
79.4
77.5
46.4
65.4
Industrial Other Coal
7.6
11.0
9.5
9.5
9.5
9.5
9.5
Natural Gas to Chemical Plants
280.6
260.7
663.4
660.5
663.6
654.9
663.7
Asphalt & Road Oil
1,170.2
1,323.2
843.9
832.3
898.1
916.1
891.8
HGLa
1,135.0
1,554.3
2,372.8
2,469.5
2,638.6
2,742.4
2,968.1
Lubricants
186.3
160.2
118.3
111.1
113.5
115.0
86.3
Natural Gasolineb
NO
NO
NO
NO
NO
NO
NO
Naphtha (<401 °F)
325.4
679.2
367.7
327.8
329.2
244.1
252.8
Other Oil (>401 °F)
660.4
499.2
211.1
194.7
195.3
111.0
104.6
Still Gas
36.7
67.7
158.7
145.4
152.8
157.1
155.8
Petroleum Coke
29.1
104.2
NO
NO
NO
NO
NO
Special Naphtha
100.6
60.9
89.1
80.4
75.7
82.4
83.4
Distillate Fuel Oil
7.0
16.0
5.8
5.8
5.8
5.8
5.8
Waxes
33.3
31.4
10.4
9.2
11.8
13.0
9.0
56 These source categories include iron and steel production, lead production, zinc production, ammonia manufacture,
carbon black manufacture (included in petrochemical production), titanium dioxide production, ferroalloy production,
silicon carbide production, and aluminum production.
Energy 3-59
-------
Year
1990
2005
2019
2020
2021
2022
2023
Miscellaneous Products
137.8
112.8
180.2
170.7
170.8
183.4
176.2
Transportation
176.0
151.3
131.3
115.6
119.0
129.9
97.5
Lubricants
176.0
151.3
131.3
115.6
119.0
129.9
97.5
U.S. Territories
50.8
114.9
3.6
3.5
3.5
1.0
1.0
Lubricants
0.7
4.6
1.0
1.0
1.0
1.0
1.0
Other Petroleum (Misc. Prod.)
50.1
110.31
2.6
2.5
2.5
NO
NO
Total
4,337.1
5,227.51
5,278.9
5,215.4
5,464.7
5,412.0
5,570.9
NO (Not Occurring)
a Excludes natural gasoline.
b Formerly referred to as"pentanes plus." This source has been adjusted and is reported separately from HGL to align with
historic data and revised EIA terminology.
Note: Totals may not sum due to independent rounding.
Table 3-25: 2023 Adjusted Non-Energy Use Fossil Fuel Consumption, Storage, and
Emissions
Adjusted
Carbon
Non-
Content
Carbon
Energy
Coefficient
Potential
Storage
Carbon
Carbon
Emissions
Use3
(MMT
Carbon
Factor
Stored
Emissions
(MMT CO2
Sector/Fuel Type
(TBtu)
C/QBtu)
(MMT C)
(MMT C)
(MMT C)
(MMT C)
Eq.)
Industry
5,472.4
NA
92.9
NA
65.5
27.4
100.4
Industrial Coking Coal
65.4
25.6
1.7
0.1
0.2
1.5
5.5
Industrial Other Coal
9.5
26.1
0.2
0.7
0.2
0.1
0.3
Natural Gas to Chemical
Plants
663.7
14.5
9.6
0.7
6.3
3.3
11.9
Asphalt & Road Oil
891.8
20.6
18.3
1.0
18.2
0.1
0.3
HGLb
2,968.1
16.8
49.9
0.7
32.9
17.0
62.2
Lubricants
86.3
20.2
1.7
0.1
0.2
1.6
5.8
Natural Gasoline0
NO
18.2
NO
0.7
NO
NO
NO
Naphtha (<401° F)
252.8
18.6
4.7
0.7
3.1
1.6
5.8
Other Oil (>401° F)
104.6
20.2
2.1
0.7
1.4
0.7
2.6
Still Gas
155.8
17.5
2.7
0.7
1.8
0.9
3.4
Petroleum Coke
NO
27.8
NO
0.3
NO
NO
NO
Special Naphtha
83.4
19.7
1.6
0.7
1.1
0.6
2.1
Distillate Fuel Oil
5.8
20.2
0.1
0.5
0.1
0.1
0.2
Waxes
9.0
19.8
0.2
0.6
0.1
0.1
0.3
Miscellaneous Products
176.2
NO
NO
NO
NO
NO
NO
Transportation
97.5
NA
2.0
NA
0.2
1.8
6.6
Lubricants
97.5
20.2
2.0
0.1
0.2
1.8
6.6
U.S. Territories
1.0
NA
+
NA
+
+
0.1
Lubricants
1.0
20.2
+
0.1
+
+
0.1
Other Petroleum (Misc. Prod.)
+
20.0
+
0.1
+
+
+
Total
5,570.9
94.9
65.7
29.2
107.1
+ Does not exceed 0.05 TBtu, MMT C, or MMT C02 Eq.
NA (Not Applicable)
NO (Not Occurring)
aTo avoid double counting, net exports have been deducted.
3-60 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
b Excludes natural gasoline.
c Formerly referred to as"pentanes plus." this source has been adjusted and is reported separately from HGL to align with historic
data and revised EIA terminology.
Note: Totals may not sum due to independent rounding.
Lastly, emissions were estimated by subtracting the carbon stored from the potential emissions (see
Table 3-23). More detail on the methodology for calculating storage and emissions from each of these
sources is provided in Annex 2.3.
Where storage factors were calculated specifically for the United States, data were obtained on (1)
products such as asphalt, plastics, synthetic rubber, synthetic fibers, cleansers (soaps and detergents),
pesticides, food additives, antifreeze and deicers (glycols), and silicones; and (2) industrial releases
including energy recovery (waste gas from chemicals), Toxics Release Inventory (TRI) releases,
hazardous waste incineration, and volatile organic compound, solvent, and non-combustion CO
emissions. Data were taken from a variety of industry sources, government reports, and expert
communications. Sources include EPA reports and databases such as compilations of air emission
factors (EPA 2001), EPA's Emissions Inventory System (EIS) to National Inventory Report (NIR) Mapping
file (EPA 2025), Toxics Release Inventory, 1998 (EPA 2000b), Biennial Reporting System (EPA 2000a,
2009), Resource Conservation and Recovery Act Information System (EPA 2013b, 2015, 2016b, 2018b,
2021, 2024b), pesticide sales and use estimates (EPA 1998, 1999, 2002, 2004, 2011, 2017), and the
Chemical Data Access Tool (EPA 2014b); the EIA Manufacturer's Energy Consumption Survey (MECS)
(EIA 1994,1997, 2001, 2005, 2010, 2013, 2017, 2021); the National Petrochemical & Refiners
Association (NPRA2002); the U.S. Census Bureau (1999, 2004, 2009, 2014, 2021); Bank of Canada
(2012, 2013, 2014, 2016, 2017, 2018, 2019, 2020, 2021, 2022, 2023, 2024); Financial Planning
Association (2006); INEGI (2006); the United States International Trade Commission (2024); Gosselin,
Smith, and Hodge (1984); EPA's Municipal Solid Waste (MSW) Facts and Figures (EPA 2013, 2014a,
2016a, 2018a, 2019); the U.S. Tire Manufacturers Association (USTMA 2012, 2013, 2014, 2016, 2018,
2020, 2022, 2024); the International Institute of Synthetic Rubber Products (IISRP 2000, 2003); the Fiber
Economics Bureau (FEB 2001, 2003, 2005, 2007, 2009, 2010, 2011, 2012, 2013); the Independent
Chemical Information Service (ICIS 2008, 2016); the EPA Chemical Data Access Tool (CDAT) (EPA
2014b); the American Chemistry Council (ACC 2003 through 2011, 2013, 2014, 2015, 2016, 2017, 2018,
2019, 2020, 2021, 2022, 2023, 2024a); the Guide to the Business of Chemistry (ACC 2024b); and the
Chemistry Industry Association of Canada (CIAC 2024). Specific data sources are listed in full detail in
Annex 2.3.
Methodological recalculations were applied to the entire time series to ensure time-series consistency
from 1990 through 2023 as discussed below.
Box 3-5: Reporting of Lubricants, Waxes, and Asphalt and Road Oil Product Use in
Energy Sector
IPCC (2006) provides methodological guidance to estimate emissions from the first use of fossil fuels as
a product for primary purposes other than combustion for energy purposes (including lubricants,
paraffin waxes, bitumen / asphalt, and solvents) under the IPPU sector. In this Inventory, carbon storage
and carbon emissions from product use of lubricants, waxes, and asphalt and road oil are reported
under the Energy sector in the Carbon Emitted from Non-Energy Uses of Fossil Fuels source category
(Source Category 1A5).
The emissions are reported in the Energy sector, as opposed to the IPPU sector, to reflect national
circumstances in its choice of methodology and to increase transparency of this source category's
Energy 3-61
-------
unique country-specific data sources and methodology. Although emissions from these non-energy
uses are reported in the Energy chapter the methodologies used to determine emissions are compatible
with the 2006 IPCC Guidelines. The country-specific methodology used for the Carbon Emitted from
Non-Energy Uses of Fossil Fuels source category is based on a carbon balance (i.e., carbon inputs-
outputs) calculation of the aggregate amount of fossil fuels used for non-energy uses, including inputs
of lubricants, waxes, asphalt and road oil (see Table 3-25).
For those inputs, U.S. country-specific data on carbon stocks and flows are used to develop carbon
storage factors, which are calculated as the ratio of the carbon stored by the fossil fuel non-energy
products to the total carbon content of the fuel consumed, taking into account losses in the production
process and during product use. The country-specific methodology to reflect national circumstances
starts with the aggregate amount of fossil fuels used for non-energy uses and applies a carbon balance
calculation, breaking out the carbon emissions from non-energy use of lubricants, waxes, and asphalt
and road oil. The emissions are reported under the Energy chapter to improve transparency, report a
more complete carbon balance and to avoid double counting. Due to U.S. national circumstances,
reporting these carbon emissions separately under IPPU would involve making artificial adjustments to
allocate both the carbon inputs and carbon outputs of the non-energy use carbon balance. For example,
only the emissions from the first use of lubricants and waxes are to be reported under the IPPU sector,
emissions from use of lubricants in 2-stroke engines and emissions from secondary use of lubricants
and waxes in waste incineration with energy recovery are to be reported under the Energy sector.
Reporting these non-energy use emissions from only first use of lubricants and waxes under IPPU would
involve making artificial adjustments to the non-energy use carbon balance and could potentially result
in double counting of emissions. These artificial adjustments would also be required for asphalt and
road oil and solvents (which are captured as part of petrochemical feedstock emissions) and could also
potentially result in double counting of emissions. To avoid presenting an incomplete carbon balance
and a less transparent approach for the Carbon Emitted from Non-Energy Uses of Fossil Fuels source
category calculation, the entire calculation of carbon storage and carbon emissions is therefore
conducted in the Non-Energy Uses of Fossil Fuels category calculation methodology, and both the
carbon storage and carbon emissions for lubricants, waxes, and asphalt and road oil are reported under
the Energy sector.
However, emissions from non-energy uses of fossil fuels as feedstocks or reducing agents (e.g.,
petrochemical production, aluminum production, titanium dioxide, and zinc production) are reported in
the IPPU chapter, unless otherwise noted due to specific national circumstances.
Uncertainty
An uncertainty analysis was conducted to quantify the uncertainty surrounding the estimates of
emissions and storage factors from non-energy uses. This analysis, performed using @RISK software
and the IPCC-recommended Approach 2 methodology (Monte Carlo stochastic simulation technique),
provides for the specification of probability density functions for key variables within a computational
structure that mirrors the calculation of the inventory estimate. The results presented below provide the
95 percent confidence interval, the range of values within which emissions are likely to fall, for this
source category.
As noted above, the non-energy use analysis is based on U.S.-specific storage factors for (1) feedstock
materials (natural gas, HGL, natural gasoline, naphthas, other oils, still gas, special naphthas, and other
3-62 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
industrial coal), (2) asphalt, (3) lubricants, and (4) waxes. For the remaining fuel types (the "other"
category in Table 3-24 and Table 3-25) the storage factors were taken directly from IPCC (2006), where
available, and otherwise assumptions were made based on the potential fate of carbon in the respective
NEU products. To characterize uncertainty, five separate analyses were conducted, corresponding to
each of the five categories. In all cases, statistical analyses or expert judgments of uncertainty were not
available directly from the information sources for all the activity variables; thus, uncertainty estimates
were determined using assumptions based on source category knowledge.
The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 3-26
(emissions) and Table 3-27 (storage factors). Carbon emitted from non-energy uses of fossil fuels in
2023 was estimated to be between 68.2 and 173.6 MMT C02 Eq. at a 95 percent confidence level. This
indicates a range of 36 percent below to 62 percent above the 2023 emission estimate of 107.1 MMT
C02 Eq. The uncertainty in the emission estimates is a function of uncertainty in both the quantity of fuel
used for non-energy purposes and the storage factor.
Table 3-26: Approach 2 Quantitative Uncertainty Estimates for C02 Emissions from
Non-Energy Uses of Fossil Fuels (MMT C02 Eq. and Percent)
Uncertainty Range Relative to Emission Estimate"
2023 Emission
(MMT CO2 Eq.)
(%)
Estimate
Lower
Upper
Lower
Upper
Source
Gas
(MMT CO2 Eq.)
Bound
Bound
Bound
Bound
Feedstocks
C02
88.3
51.7
156.0
-42%
+77%
Asphalt
CO2
0.3
0.1
0.6
-58%
+ 116%
Lubricants
CO2
12.4
10.3
14.4
-17%
+ 16%
Waxes
CO2
0.3
0.2
0.6
-26%
+ 103%
Other
CO2
5.7
1.1
6.7
-81%
+ 17%
Total
CO2
107.1
68.2
173.6
-36%
+62%
a Range of emission estimates predicted by Monte Carlo stochastic simulation for a 95 percent confidence interval.
Note: Totals may not sum due to independent rounding.
Table 3-27: Approach 2 Quantitative Uncertainty Estimates for Storage Factors of Non-
Energy Uses of Fossil Fuels (Percent)
Uncertainty Range Relative to Emission Estimate"
2023 Storage
(%)
(%, Relative)
Factor
Lower
Upper
Lower
Upper
Source
Gas
(%)
Bound
Bound
Bound
Bound
Feedstocks
CO2
66.0%
52%
74%
-22%
+ 13%
Asphalt
CO2
99.6%
99%
100%
-0.5%
+0.3%
Lubricants
CO2
9.2%
4%
17%
-59%
+91%
Waxes
CO2
57.8%
47%
68%
-18%
+ 17%
Other
CO2
12.6%
7%
83%
-42%
+555%
a Range of emission estimates predicted by Monte Carlo stochastic simulation for a 95 percent confidence interval, as a
percentage of the inventory value (also expressed in percent terms).
As shown in Table 3-27, waxes and asphalt contribute least to overall storage factor uncertainty on a
percentage basis. Although the feedstocks category—the largest use category in terms of total carbon
flows—also appears to have relatively tight confidence limits, this is to some extent an artifact of the
Energy 3-63
-------
way the uncertainty analysis was structured. As discussed in Annex 2.3, the storage factor for
feedstocks is based on an analysis of six fates that result in long-term storage (e.g., plastics production),
and 11 that result in emissions (e.g., volatile organic compound emissions). Rather than modeling the
total uncertainty around all of these fate processes, the current analysis addresses only the storage
fates, and assumes that all carbon that is not stored is emitted. As the production statistics that drive
the storage values are relatively well-characterized, this approach yields a result that is probably biased
toward understating uncertainty.
As is the case with the other uncertainty analyses discussed throughout this document, the uncertainty
results above address only those factors that can be readily quantified. More details on the uncertainty
analysis are provided in Annex 2.3.
QA/QC and Verification
In order to ensure the quality of the emission estimates from non-energy uses of fossil fuels, general
(IPCC Tier 1) and category-specific (Tier 2) Quality Assurance/Quality Control (QA/QC) procedures were
implemented consistent with the U.S. Inventory QA/QC plan outlined in Annex 8. This effort included a
general analysis, as well as portions of a category specific analysis for non-energy uses involving
petrochemical feedstocks and for imports and exports. The Tier 2 procedures that were implemented
involved checks specifically focusing on the activity data and methodology for estimating the fate of
carbon (in terms of storage and emissions) across the various end-uses of fossil C. Emission and
storage totals for the different subcategories were compared, and trends across the time series were
analyzed to determine whether any corrective actions were needed. Corrective actions were taken to
rectify minor errors and to improve the transparency of the calculations, facilitating future QA/QC.
For petrochemical import and export data, special attention was paid to NAICS numbers and titles to
verify that none had changed or been removed. Import and export totals were compared with 2022
totals as well as their trends across the time series.
It is important to ensure no double counting of emissions between fuel combustion, non-energy use of
fuels and industrial process emissions. For petrochemical feedstock production, our review of the
categories suggests this is not a significant issue since the non-energy use industrial release data
includes different categories of sources and sectors than those included in the Industrial Processes and
Product Use (IPPU) emissions category for petrochemicals. Further data integration is not available at
this time because feedstock data from the EIA used to estimate non-energy uses of fuels are aggregated
by fuel type, rather than disaggregated by both fuel type and particular industries. Also, GHGRP-reported
data on quantities of fuel consumed as feedstocks by petrochemical producers are unable to be used
due to the data failing GHGRP CBI aggregation criteria. This country-specific approach taken is better
able to reflect the national situation because it accounts for secondary product imports and exports
that are not included directly in the national energy statistics. Furthermore, it is compatible with the
2006 IPCC Guidelines as discussed in Box 3-4 above, but also as the NEU emissions are here represent
different emissions from those covered in the IPPU petrochemical production category.
Recalculations Discussion
Several updates to activity data factors lead to recalculations of previous year results. The major
updates are as follows:
3-64 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
• ACC (2023b) updated adipic acid, and acetic acid production in 2022, which resulted in a slight
decrease in emissions relative to the previous Inventory.
• U.S. International Trade Commission (2024) updated historical import and export data from
2020 to 2022 for cleansers and antifreeze, resulting in slight changes from the previous
Inventory.
• EIA (2025) updated historical fuel consumption data for HGL, industrial coking coal, and
lubricants, resulting in a decrease in emissions for the period 2019 through 2022.
• EPA (2024) published new data on the quantity of hazardous waste incinerated in 2021 and
2023, resulting in a slight decrease in emissions for the period 2020 through 2022.
• EIA (2024) published new data on the consumption of other petroleum liquids by U.S. Pacific
Islands and Wake Island for years 2020 through 2022, resulting in a slight increase in emissions
relative to the previous Inventory.
Overall, these changes resulted in an average annual decrease of less than 0.05 MMT C02 Eq. (less than
0.05 percent) in carbon emissions from non-energy uses of fossil fuels for the period 1990 through 2022,
relative to the previous Inventory. This change was driven by slight increases in emissions for the period
2019 through 2021, and a decrease in emissions in 2022.
Planned Improvements
There are several future improvements planned:
• More accurate accounting of carbon in petrochemical feedstocks. EPA has worked with EIA to
determine the cause of input/output discrepancies in the carbon mass balance contained
within the NEU model. In the future, two strategies to reduce or eliminate this discrepancy will
continue to be pursued as part of quality control procedures. First, accounting of carbon in
imports and exports will be improved. The import/export adjustment methodology will be
examined to ensure that net exports of intermediaries such as ethylene and propylene are fully
accounted for. Second, the use of top-down carbon input calculation in estimating emissions
will be reconsidered. Alternative approaches that rely more substantially on the bottom-up
carbon output calculation will be considered instead.
• Improving the uncertainty analysis. Most of the input parameter distributions are based on
professional judgment rather than rigorous statistical characterizations of uncertainty.
• Better characterizing flows of fossil carbon. Additional fates may be researched, including the
fossil carbon load in organic chemical wastewaters, plasticizers, adhesives, films, paints, and
coatings. There is also a need to further clarify the treatment of fuel additives and backflows
(especially methyl tert-butyl ether, MTBE).
• Reviewing the trends in fossil fuel consumption for non-energy uses. Annual consumption for
several fuel types is highly variable across the time series, including industrial coking coal and
other petroleum. A better understanding of these trends will be pursued to identify any
mischaracterized or misreported fuel consumption for non-energy uses.
• Updating the average carbon content of solvents was researched, since the entire time series
depends on one year's worth of solvent composition data. The data on carbon emissions from
Energy 3-65
-------
solvents that were readily available do not provide composition data for all categories of solvent
emissions and also have conflicting definitions for volatile organic compounds, the source of
emissive carbon in solvents. Additional sources of solvents data will be investigated in order to
update the carbon content assumptions.
• Updating the average carbon content of cleansers (soaps and detergents) was researched;
although production and consumption data for cleansers are published every 5 years by the
Census Bureau, the composition (C content) of cleansers has not been recently updated.
Recently available composition data sources may facilitate updating the average carbon
content for this category.
• Revising the methodology for consumption, production, and carbon content of plastics was
researched; because of recent changes to the type of data publicly available for plastics, the
NEU model for plastics applies data obtained from personal communications. Potential
revisions to the plastics methodology to account for the recent changes in published data will
be investigated.
• Although U.S.-specific storage factors have been developed for feedstocks, asphalt, lubricants,
and waxes, default values from IPCC are still used for two of the non-energy fuel types
(industrial coking coal, distillate oil), and broad assumptions are being used for miscellaneous
products and other petroleum. Over the long term, there are plans to improve these storage
factors by analyzing carbon fate similar to those described in Annex 2.3 or deferring to more
updated default storage factors from IPCC where available.
• Reviewing the storage of carbon black across various sectors in the Inventory, in particular, the
carbon black abraded and stored in tires.
• Assess the current method and/or identify new data sources (e.g., EIA) for estimating emissions
from ammonia/fertilizer use of natural gas.
Investigate EIA NEU and MECS data to update, as needed, adjustments made for ammonia production
and "natural gas to chemical plants, other uses" and "natural gas to other" non-energy uses, including
iron and steel production, in energy uses and IPPU.
3-66 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
3.3 Incineration of Waste (Source
Category 1A)
Combustion is used to manage about 7 to 19 percent of the solid wastes generated in the United States,
depending on the source of the estimate and the scope of materials included in the definition of solid
waste (EPA 2000; EPA 2020; Goldstein and Madtes 2001; Kaufman et al. 2004; Simmons et al. 2006; van
Haaren et al. 2010). In the context of this section, waste includes all municipal solid waste (MSW) as
well as scrap tires. In the United States, combustion of MSW tends to occur at waste-to-energy facilities
or industrial facilities where useful energy is recovered, and thus emissions from waste combustion are
accounted for in the Energy chapter. Similarly, scrap tires are combusted for energy recovery in
industrial and utility boilers, pulp and paper mills, and cement kilns. Combustion of waste results in
conversion of the organic inputs to C02. According to the 2006IPCC Guidelines, when the C02 emitted
is of fossil origin, it is counted as a net anthropogenic emission of C02 to the atmosphere. Thus, the
emissions from waste combustion are calculated by estimating the quantity of waste combusted and
the fraction of the waste that is carbon derived from fossil sources.
Most of the organic materials in MSW are of biogenic origin (e.g., paper, yard trimmings), and have their
net carbon flows accounted for under the Land Use, Land-Use Change, and Forestry chapter. However,
some components of MSW and scrap tires—plastics, synthetic rubber, synthetic fibers, and carbon
black—are of fossil origin. Plastics in the U.S. waste stream are primarily in the form of containers,
packaging, and durable goods. Rubber is found in durable goods, such as carpets, and in non-durable
goods, such as clothing and footwear. Fibers in MSW are predominantly from clothing and home
furnishings. As noted above, scrap tires (which contain synthetic rubber and carbon black) are also
considered a "non-hazardous" waste and are included in the waste combustion estimate, though waste
disposal practices for tires differ from MSW. Estimates on emissions from hazardous waste combustion
can be found in Annex 2.3 and are accounted for as part of the carbon mass balance for non-energy
uses of fossil fuels.
Approximately 25.7 million metric tons of MSW were combusted in 2023 (EPA 2024). Carbon dioxide
emissions from combustion of waste decreased 3.7 percent since 1990, to an estimated 12.4 MMT C02
(12,425 kt) in 2023. Emissions across the time series are shown in Table 3-28 and Table 3-29.
Waste combustion is also a source of CH4 and N20 emissions (De Soete 1993; IPCC 2006). Methane
emissions from the combustion of waste were estimated to be less than 0.5 MMT C02 Eq. (less than
0.05 kt CH4) in 2023 and have remained steady since 1990. Nitrous oxide emissions from the
combustion of waste were estimated to be 0.3 MMT C02 Eq. (1.2 kt N20) in 2023 and have decreased by
19 percent since 1990. This decrease is driven by the decrease in total MSW combusted.
Energy 3-67
-------
Table 3-28: C02, CH4, and N20 Emissions from the Combustion of Waste (MMT C02 Eq.)
Gas
1990
2005
2019
2020
2021
2022
2023
C02
12.9
13.3
12.9
12.9
12.5
12.5
12.4
cm
+
+
+
+
+
+
+
n2o
0.41
0.31
0.4
0.3
0.4
0.3
0.3
Total
13.3
13.6
13.3
13.3
12.8
12.8
12.8
+ Does not exceed 0.05 MMT C02 Eq.
Note: Totals may not sum due to independent rounding.
Table 3-29: C02, CH4,
and N20 Emissions from the Combustion of Waste (kt)
Gas
1990 20051 2019
2020
2021
2022
2023
CO2
12,9001
13,2541
12,948
12,921
12,476
12,484
12,425
CH4
-
-
+
+
+
+
+
N2O
2I
I 'I
I 1
1
1
1
1
+ Does not exceed 0.05 kt.
Methodology and Time-Series Consistency
Municipal Solid Waste Combustion
ATier 2 approach is used to determine both C02 and non-C02 emissions from the combustion of waste,
the method uses tonnage of waste combusted and an estimated country specific emissions factor.
Emission estimates from the combustion of tires are discussed separately. Data for total waste
combusted was derived from BioCycle (van Haaren et al. 2010), EPA Facts and Figures Report, Energy
Recovery Council (ERC), EPA's Greenhouse Gas Reporting Program (GHGRP), and the U.S. Energy
Information Administration (EIA). Multiple sources were used to ensure a complete, quality dataset, as
each source encompasses a different timeframe.
EPA determined the MSW tonnages based on data availability and accuracy throughout the time series.
• 1990-2006: MSW combustion tonnages are from Biocycle combustion data. Tire combustion
data from the U.S. Tire Manufacturers Association (USTMA) are removed to arrive at MSW
combusted without tires.
• 2006-2010: MSW combustion tonnages are an average of Biocycle (with USTMA tire data
tonnage removed), U.S. EPA Facts and Figures, EIA, and Energy Recovery Council data (with
USTMA tire data tonnage removed).
• 2011-2023: MSW combustion tonnages are from EPA's GHGRP data.
Table 3-30 provides the estimated tons of MSW combusted including and excluding tires.
Table 3-30: Municipal Solid Waste Combusted (Short Tons)
19901
20051 2019 2020 2021 2022 2023
Waste Combusted
(excluding tires) 33,344,8391
26,486,414
28,174,311 27,586,271 27,867,446 26,338,130 25,676,432
Waste Combusted
(including tires) 33,766,239
28,631,054
29,821,141 29,106,686 29,261,446 27,808,130 27,222,432
Sources: BioCycle, EPA Facts and Figures, ERC, GHGRP, EIA, USTMA.
3-68 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
C02 Emissions from MSW Excluding Scrap Tires
Fossil C02 emission factors were calculated from EPA's GHGRP data for non-biogenic sources. Using
GHGRP-reported emissions for CH4and N20 and assumed emission factors, the tonnage of waste
combusted, excluding tires, was derived. Methane and N20 emissions and assumed emission factors
were used to estimate the amount of MSW combusted in terms of energy content. The energy content of
MSW combusted was then converted into tonnage based on assumed MSW heating value. Two
estimates were generated (one for CH4 and one for N20) and the two were averaged together. Dividing
fossil C02 emissions from GHGRP FLIGHT data for MSW combustors by this estimated tonnage yielded
an annual C02 emission factor. As this data was only available following 2011, all years prior use an
average of the emission factors from 2011 through 2015. See Annex 3.7 for more detail on how MSW
carbon factors were calculated.
Finally, C02 emissions were calculated by multiplying the annual tonnage estimates, excluding tires, by
the calculated emissions factor. Calculated fossil C02 emission factors are shown in Table 3-31.
Table 3-31: Calculated Fossil C02 Content per Ton Waste Combusted (kg C02/Short
Ton Combusted)
Year
1990 |
2005
2019
2020
2021
2022
2023
CO2 Emission Factors
3661
3661
363
377
365
382
384
CO2 Emissions from Scrap Tires
Scrap tires contain several types of synthetic rubber, carbon black, and synthetic fibers. Each type of
synthetic rubber has a discrete carbon content, and carbon black is 100 percent C. For synthetic rubber
and carbon black in scrap tires, information on average weight, disposal percentage, and total tires
incinerated for energy was obtained biannually from U.S. Scrap Tire Management Summary for 2005
through 2023 data (USTMA 2024). Information about scrap tire composition was taken from the Rubber
Manufacturers' Association internet site (USTMA 2012a). Emissions of C02 were calculated based on
the amount of scrap tires used for fuel and the synthetic rubber and carbon black content of scrap tires.
The mass of combusted material is multiplied by its carbon content to calculate the total amount of
carbon stored. More detail on the methodology for calculating emissions from each of these waste
combustion sources is provided in Annex 3.7. Table 3-32 provides C02 emissions from combustion of
waste tires.
Table 3-32: C02 Emissions from Combustion of Tires (MMT C02 Eq.)
Year
1990
2005
2019
2020
2021
2022
2023
Synthetic Rubber
0.3
1.6
1.2
1.1
1.0
1.2
1.2
C Black
0.4
2.0
1.5
1.4
1.3
1.4
1.4
Total
0.7
3.6
2.7
2.5
2.3
2.6
2.6
Note: Totals may not sum due to independent rounding.
Non-C02 Emissions
Combustion of waste also results in emissions of CH4and N20. These emissions were calculated by
multiplying the total estimated mass of waste combusted, including tires, by the respective emission
factors. The emission factors for CH4 and N20 emissions per quantity of MSW combusted are default
Energy 3-69
-------
emission factors for the default continuously-fed stoker unit MSW combustion technology type and
were taken from IPCC (2006).
Uncertainty
An Approach 2 Monte Carlo analysis was performed to determine the level of uncertainty surrounding
the estimates of C02 emissions and N20 emissions from the incineration of waste (given the very low
emissions for CH4, no uncertainty estimate was derived). IPCC Approach 2 analysis allows the
specification of probability density functions for key variables within a computational structure that
mirrors the calculation of the Inventory estimate. Statistical analyses or expert judgments of uncertainty
were not available directly from the information sources for most variables; thus, uncertainty estimates
for these variables were determined using assumptions based on source category knowledge and the
known uncertainty estimates for the waste generation variables.
The uncertainties in the waste incineration emission estimates arise from both the assumptions applied
to the data and from the quality of the data. Key factors include reported C02 emissions; N20 and CH4
emissions factors, and tire synthetic rubber and black carbon contents. The highest levels of uncertainty
surround the reported emissions from GHGRP; the lowest levels of uncertainty surround variables that
were determined by quantitative measurements (e.g., combustion efficiency, carbon content of carbon
black).
The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 3-23. Waste
incineration C02 emissions in 2023 were estimated to be between 10.3 and 15.0 MMT C02 Eq. at a 95
percent confidence level. This indicates a range of 17 percent below to 20 percent above the 2023
emission estimate of 12.4 MMT C02 Eq. Waste incineration CH4 emissions in 2023 were estimated to be
between less than 0.00005 and less than 0.0005 MMT C02 Eq. at a 95 percent confident level. This
indicates a range of 102 percent below to 103 percent above the 2023 emission estimate of less than
0.0005 MMT C02 Eq. Also at a 95 percent confidence level, waste incineration N20 emissions in 2023
were estimated to be between 0.2 and 0.9 MMT C02 Eq. This indicates a range of 53 percent below to
161 percent above the 2023 emission estimate of 0.3 MMT C02 Eq.
Table 3-33: Approach 2 Quantitative Uncertainty Estimates for C02 and N20 from the
Incineration of Waste (MMT C02 Eq. and Percent)
Source
Gas
2023 Emission
Estimate
(MMT CO2 Eq.)
Uncertainty Range Relative to Emission Estimate"
(MMT CO2 Eq.)
(%)
Lower
Bound
Upper
Bound
Lower
Bound
Upper
Bound
Incineration ofWaste
C02
12.4
10.3
15.0
-17%
+20%
Incineration ofWaste
cm
+
+
+
-102%
+ 103%
Incineration ofWaste
n2o
0.3
0.2
0.9
-53%
+ 161%
+ Does not exceed 0.05 MMT C02 Eq.
a Range of emission estimates predicted by Monte Carlo Simulation for a 95 percent confidence interval.
QA/QC and Verification
In order to ensure the quality of the emission estimates from waste combustion, general (IPCC Tier 1)
and category-specific (Tier 2) Quality Assurance/Quality Control (QA/QC) procedures were
3-70 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
implemented consistent with the U.S. Inventory QA/QC plan outlined in Annex 8. The Tier 2 procedures
that were implemented involved checks specifically focusing on the activity data and specifically
focused on the emission factor and activity data sources and methodology used for estimating
emissions from combustion of waste. Trends across the time series were analyzed to determine
whether any corrective actions were needed. Corrective actions were taken to rectify minor errors in the
use of activity data.
Recalculations Discussion
No recalculations were performed for the current Inventory.
Planned Improvements
No planned improvements for waste combustion were identified.
3.4 Coal Mining (Source Category 1B1 a)
Three types of coal mining-related activities release CH4 and C02 to the atmosphere: underground
mining, surface mining, and post-mining (i.e., coal-handling) activities. While surface coal mines
account for the majority of U.S. coal production, underground coal mines contribute the largest share of
fugitive CH4 emissions (see Table 3-35 and Table 3-36) due to the higher CH4 content of coal in the
deeper underground coal seams. In 2023,190 underground coal mines and 362 surface mines were
operating in the United States (EIA 2024). In recent years, the total number of active coal mines in the
United States has declined.
Table 3-34: Coal Production (kt)
Year
19901
2005
2019
2020
2021
2022
2023
Underground
Number of Mines
1,683
586
226
196
174
185
190
Production
384,244
334,399
242,557
177,380
200,122
201,525
197,701
Surface
Number of Mines
1,656
789
432
350
332
354
362
Production
546,808 |
691,447
397,750
307,944
323,142
336,990
326,340
Total
Number of Mines
3,3391
1,398
658
546
506
539
552
Production
931,052 |
1,025,8461
640,307
485,324
523,264
538,515
524,041
Note: Totals may not sum due to independent rounding.
Fugitive CH4 Emissions
Underground coal mines liberate CH4 from ventilation systems and from degasification systems.
Ventilation systems pump air through the mine workings to dilute noxious gases and ensure worker
safety; these systems can exhaust significant amounts of CH4 to the atmosphere in low concentrations.
Degasification systems are wells drilled from the surface or boreholes drilled inside the mine that
Energy 3-71
-------
remove large, often highly concentrated volumes of CH4 before, during, or after mining. Some mines
recover and use CH4 generated from ventilation and degasification systems, thereby reducing emissions
to the atmosphere.
Surface coal mines liberate CH4 as the overburden is removed and the coal is exposed to the
atmosphere. Methane emissions are normally a function of coal rank (a classification related to the
percentage of carbon in the coal) and depth. Surface coal mines typically produce lower-rank coals and
remove less than 250 feet of overburden, so their level of emissions is much lower than from
underground mines.
In addition, CH4 is released during post-mining activities, as the coal is processed, transported, and
stored for use.
Total CH4 emissions in 2023 were estimated to be 1,623 kt (45.4 MMT C02 Eq.), a decline of
approximately 58 percent since 1990 (see Table 3-35 and Table 3-36). In 2023, underground mines
accounted for approximately 74 percent of total emissions, surface mines accounted for 13 percent,
and post-mining activities accounted for 13 percent. In 2023, total CH4 emissions from coal mining
increased by approximately 4 percent relative to the previous year. Total coal production in 2023
decreased by 3 percent compared to 2022. This resulted in a decrease of 1 percent in CH4 emissions
from surface mining and post-mining activities in 2023. However, surface mining and post-mining
activities have a lower impact on total CH4 compared to underground mining (74 percent of total
emissions in 2023). The number of operating underground mines increased in 2023 and the amount of
CH4 recovered and used in 2023 increased by 6 percent compared to 2022. In 2023, the amount of CH4
from underground mining activities increased by 6 percent compared to 2022.
Table 3-35: CH4 Emissions from Coal Mining (MMT C02 Eq.)
Activity
1990
2005
2019
2020
2021
2022
2023
Underground (UG) Mining
83.1
46.7
38.5
35.2
32.9
31.5
33.4
Liberated
90.6
66.9
56.6
53.7
52.3
56.1
59.4
Recovered & Used
(7.5) I
(20.1)1
(18.1)
(18.5)
(19.4)
(24.6)
(26.0)
Surface Mining
12.0
13.3
7.2
5.4
5.7
6.0
5.9
Post-Mining (UG)
10.3
8.6
5.8
4.3
4.8
4.8
4.8
Post-Mining (Surface)
2.6
2.9
1.5
1.2
1.2
1.3
1.3
Total
108.1
71.5
53.0
46.2
44.7
43.6
45.4
Notes: Parentheses in above emissions tables indicate negative values. Totals may not sum due to independent rounding.
Table 3-36: CH4 Emissions from Coal Mining (kt)
Activity
19901
2005
2019
2020
2021
2022
2023
Underground (UG) Mining
2,968
1,669
1,375
1,257
1,176
1,124
1,193
Liberated
3,237
2,388
2,022
1,917
1,868
2,003
2,122
Recovered & Used
(269)
(719)
(646)
(660)
(692)
(880)
(928)
Surface Mining
430
475
255
194
205
215
211
Post-Mining (UG)
368
306
206
155
170
173
172
Post-Mining (Surface)
931
103
55
42
44
47
46
Total
3,8601
2,552
1,892
1,648
1,595
1,558
1,623
Notes: Parentheses in above emissions tables indicate negative values. Totals may not sum due to independent rounding.
3-72 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Methodology and Time-Series Consistency
EPA uses an IPCC Tier 3 method for estimating CH4 emissions from underground coal mining and an
IPCC Tier 2 method for estimating CH4 emissions from surface mining and post-mining activities (for
coal production from both underground mines and surface mines) in accordance with methodological
decisions trees in IPCC guidelines (Volume 2, Chapter 4, Figure 4.1.1 and 4.1.2) and available data
(IPCC 2006). The methodology for estimating CH4 emissions from coal mining consists of two steps:
• Estimate CH4 emissions from underground mines. These emissions have two sources:
ventilation systems and degasification systems. They are estimated using mine-specific data,
then summed to determine total CH4 liberated. The CH4 recovered and used is then subtracted
from this total, resulting in an estimate of net emissions to the atmosphere.
• Estimate CH4 emissions from surface mines and post-mining activities. Unlike the methodology
for underground mines, which uses mine-specific data, the methodology for estimating
emissions from surface mines and post-mining activities consists of multiplying basin-specific
coal production by basin-specific gas content and an emission factor.
Step 1: Estimate CH4 Liberated and CH4 Emitted from Underground Mines
Underground mines generate CH4 from ventilation systems and degasification systems. Some mines
recover and use the liberated CH4, thereby reducing emissions to the atmosphere. Total CH4 emitted
from underground mines equals the CH4 liberated from ventilation systems, plus the CH4 liberated from
degasification systems, minus the CH4 recovered and used.
Step 1.1: Estimate CH4 Liberated from Ventilation Systems
To estimate CH4 liberated from ventilation systems, EPA uses data collected through its Greenhouse
Gas Reporting Program (GHGRP)57 (Subpart FF, "Underground Coal Mines"), data provided by the U.S.
Mine Safety and Health Administration (MSHA) (MSHA 2024), and occasionally data collected from
other sources on a site-specific level (e.g., state gas production databases). Since 2011, the nation's
"gassiest" underground coal mines—those that liberate more than 36,500,000 actual cubic feet of CH4
per year (about 17,525 MT C02 Eq.)—have been required to report to EPA's GHGRP (EPA 20 24).58 Mines
that report to EPA's GHGRP must report quarterly measurements of CH4 emissions from ventilation
systems; they have the option of recording and reporting their own measurements, or using the
measurements taken by MSHA as part of that agency's quarterly safety inspections of all mines in the
United States with detectable CH4 concentrations.59
Since 2013, ventilation CH4 emission estimates have been calculated based on both quarterly GHGRP
data submitted by underground mines and on quarterly measurement data obtained directly from
MSHA. Because not all mines report under EPA's GHGRP, the emissions of the mines that do not report
must be calculated using MSHA data. The MSHA data also serves as a quality assurance tool for
validating GHGRP data. For GHGRP data, reported quarterly ventilation methane emissions (metric
57 In implementing improvements and integrating data from EPA's GHGRP, EPA followed the latest guidance from the IPCC
on the use of facility-level data in national inventories (IPCC 2011).
58 Underground coal mines report to EPA under Subpart FF of the GHGRP (40 CFR Part 98). In 2023, 58 underground coal
mines reported to the program.
59 MSHA records coal mine CH4 readings with concentrations of greater than 50 ppm (parts per million) CH4. Readings
below this threshold are considered non-detectable.
Energy 3-73
-------
tons) are summed for each mine to develop mine-specific annual ventilation emissions. For MSHA data,
the average daily CH4 emission rate for each mine is determined using the CH4 total for all data
measurement events conducted during the calendar year and total duration of all data measurement
events (in days). The calculated average daily CH4 emission rate is then multiplied by 365 days to
estimate annual ventilation CH4 emissions for the MSHA dataset.
Step 1.2: Estimate CH4 Liberated from Degasification Systems
Particularly gassy underground mines also use degasification systems (e.g., wells or boreholes) to
remove CH4 before, during, or after mining. This CH4 can then be collected for use or vented to the
atmosphere. Nineteen mines used degasification systems in 2023 and all of these mines reported the
CH4 removed through these systems to EPA's GHGRP under Subpart FF (EPA 2024). Based on the weekly
measurements reported to EPA's GHGRP, degasification data summaries for each mine are added to
estimate the CH4 liberated from degasification systems. Eleven of the 19 mines with degasification
systems had operational CH4 recovery and use projects, including one mine with two recovery and use
projects (see step 1.3 below).60
Degasification data reported to EPA's GHGRP by underground coal mines is the primary source of data
used to develop estimates of CH4 liberated from degasification systems. Data reported to EPA's GHGRP
were used exclusively to estimate CH4 liberated from degasification systems at 14 of the 19 mines that
used degasification systems in 2023. Data from state gas well production databases were used to
supplement GHGRP degasification data for the remaining 5 mines (DMME 2024; ERG 2024; GSA 2024;
VWGES 2024).
For pre-mining wells, cumulative degasification volumes that occur prior to the well being mined
through are attributed to the mine in the inventory year in which the well is mined through.61 EPA's
GHGRP does not require gas production from virgin coal seams (coalbed methane) to be reported by
coal mines under Subpart FF.62 Most pre-mining wells drilled from the surface are considered coalbed
methane wells prior to mine-through and associated CH4 emissions are reported under another subpart
of the GHGRP (Subpart W, "Petroleum and Natural Gas Systems"). As a result, GHGRP data must be
supplemented to estimate cumulative degasification volumes that occurred prior to well mine-through.
There were five mines with degasification systems that include pre-mining wells that were mined
through in 2023. For all of these mines, GHGRP data were supplemented with historical data from state
gas well production databases (ERG 2024; GSA 2024), as well as with mine-specific information
regarding the locations and dates on which the pre-mining wells were mined through (JWR 2010; El Paso
2009; ERG 2024).
Step 1.3: Estimate CH4 Recovered from Ventilation and Degasification Systems, and
Utilized or Destroyed (Emissions Avoided)
Eleven mines had a total of 12 CH4 recovery and use projects in place in 2023, including one mine that
had two recovery and use projects. All of these projects involved degasification systems. Ten of these
60 Several of the mines venting ChUfrom degasification systems use a small portion of the gas to fuel gob well blowers in
remote locations where electricity is not available. However, this ChU use is not considered to be a formal recovery and
use project.
61 A well is "mined through" when coal mining development or the working face intersects the borehole or well.
62 This applies for pre-drainage in years prior to the well being mined through. Beginning with the year the well is mined
through, the annual volume of ChU liberated from a pre-drainage well is reported under Subpart FF of EPA's GHGRP.
3-74 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
mines sold the recovered CH4 to a pipeline, including one that also used CH4 to fuel a thermal coal
dryer. One mine destroyed the recovered CH4 using enclosed flares.
The CH4 recovered and used (or destroyed) at the 11 mines described above are estimated using the
following methods:
• EPA's GHGRP data was exclusively used to estimate the CH4 recovered and used from six of the
11 mines that deployed degasification systems in 2023. Based on quarterly measurements, the
GHGRP degasification destruction data summaries for each mine are added together to
estimate the CH4 recovered and used from degasification systems.
• State sales data were used to supplement GHGRP data to estimate CH4 recovered and used
from five mines that deployed degasification systems in 2023 (DMME 2024, ERG 2024, GSA
2024, and WVGES 2024). Supplemental information is used for these mines because estimating
CH4 recovery and use from pre-mining wells requires additional data not reported under Subpart
FF of EPA's GHGRP (see discussion in step 1.2 above) to account for the emissions avoided prior
to the well being mined through. The supplemental data is obtained from state gas production
databases as well as mine-specific information on the location and timing of mined-through
pre-mining wells.
Step 2: Estimate CH4 Emitted from Surface Mines and Post-Mining
Activities
Mine-specific data are not available for estimating CH4 emissions from surface coal mines or for post-
mining activities. For surface mines, basin-specific coal production obtained from the Energy
Information Administration's Annual Coal Report (EIA 2024) is multiplied by basin-specific CH4 contents
(EPA 1996, 2005) and a 150 percent emission factor (to account for CH4from over- and under-burden) to
estimate CH4 emissions (King 1994; Saghafi 2013). For post-mining activities, basin-specific coal
production is multiplied by basin-specific CH4 contents and a mid-range 32.5 percent emission factor
for CH4 desorption during coal transportation and storage (Creedy 1993). Basin-specific in situ gas
content data were compiled from AAPG (1984) and USBM (1986).
Fugitive CO2 Emissions
Methane and C02 are naturally occurring in coal seams and are collectively referred to as coal seam
gas. These gases remain trapped in the coal seam until coal is mined (i.e., coal seam is exposed and
fractured during mining operations). Fugitive C02 emissions occur during underground coal mining,
surface coal mining, and post-mining activities. Methods and data to estimate fugitive C02 emissions
from underground and surface coal mining are presented in the sections below. Fugitive C02 emissions
from post-mining activities were not estimated due to the lack of an IPCC method and unavailability of
data.
Total fugitive C02 emissions in 2023 were estimated to be 2,404 kt (2.4 MMT C02 Eq.), a decline of
approximately 48 percent since 1990. In 2023, underground mines accounted for approximately 89
percent of total fugitive C02 emissions. In 2023, total fugitive C02 emissions from coal mining
decreased by approximately 3 percent relative to the previous year. This decrease was due to a decrease
in annual coal production.
Energy 3-75
-------
Table 3-37: C02 Emissions from Coal Mining (MMT C02 Eq.)
Activity
1990
2005
2019
2020
2021
2022
2023
Underground (UG) Mining
4.2
3.6
2.7
1.9
2.2
2.2
2.1
Liberated
4.2
3.6
2.6
1.9
2.2
2.2
2.1
Recovered & Used
{+)
{+)
{+)
{+)
{+)
{+)
{+)
Flaring
NO
NO
0.1
+
+
+
+
Surface Mining
0.4
0.6
0.3
0.2
0.3
0.3
0.3
Total
4.6
4.2
3.0
2.2
2.5
2.5
2.4
+ Does not exceed 0.05 MMT C02 Eq.
NO (Not Occurring)
Notes: Parentheses indicate negative values. Totals may not sum due to independent rounding.
Table 3-38: C02 Emissions from Coal Mining (kt)
Activity
1990
2005
2019
2020
2021
2022
2023
Underground (UG) Mining
4,164
I I
3,610
2,670
1,948
2,193
2,201
2,140
Liberated
4,171
3,630
2,633
1,926
2,173
2,188
2,146
Recovered & Used
(8)
(21)
(18)
(19)
(19)
(25)
(26)
Flaring
NO
NO
55
41
40
38
20
Surface Mining
443
560
322
249
262
273
264
Total
4,606
4,169
2,992
2,197
2,455
2,474
2,404
NO (Not Occurring)
Notes: Parentheses indicate negative values. Totals may not sum due to independent rounding.
Methodology and Time-Series Consistency
EPA uses an IPCC Tier 1 method for estimating fugitive C02 emissions from underground coal mining
and surface mining in accordance with methodological decisions trees in IPCC guidelines (Volume 2,
Chapter 4, Figure 4.1.1 a) and available data (IPCC 2019). IPCC methods and data to estimate fugitive
C02 emissions from post-mining activities (for both underground and surface coal mining) are currently
not available.
Step 1: Underground Mining
EPA used the following overarching IPCC equation to estimate fugitive C02 emissions from underground
coal mines (IPCC 2019):
Equation 3-1: Estimating Fugitive C02 Emissions from Underground Mines
Total C02 from Underground Mines
= C02 from underground mining — Amount of C02 in gas recovered
+ C02 from methane flaring
Step 1.1: Estimate Fugitive C02 Emissions from Underground Mining
EPA estimated fugitive C02 emissions from underground mining using the IPCC Tier 1 emission factor
(5.9 m3/metric ton) and annual coal production from underground mines (EIA 2024). The underground
mining default emission factor accounts for all the fugitive C02 likely to be emitted from underground
3-76 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
coal mining. Therefore, the amount of C02 from coal seam gas recovered and utilized for energy is
subtracted from underground mining estimates in Step 2, below. Under IPCC methods, the C02
emissions from gas recovered and utilized for energy use (e.g., injected into a natural gas pipeline) are
reported under other sectors of the Inventory (e.g., stationary combustion of fossil fuel or oil and natural
gas systems) and not under the coal mining sector.
Step 1.2: Estimate Amount ofC02 In Coal Seam Gas Recovered for Energy Purposes
EPA estimated fugitive C02 emissions from coal seam gas recovered and utilized for energy purposes by
using the IPCC Tier 1 default emission factor (19.57 metric tons C02/million cubic meters of coal bed
methane (CBM) produced) and quantity of coal seam gas recovered and utilized. Data on annual
quantity of coal seam gas recovered and utilized are available from GHGRP and state sales data (EPA
2024; DMME 2024; ERG 2024; GSA2024; WVGES 2024). The quantity of coal seam gas recovered and
destroyed without energy recovery (e.g., flaring) is deducted from the total coal seam gas recovered
quantity (EPA 2024).
Step 1.3: Estimate Fugitive C02 Emissions from Flaring
The IPCC method includes combustion C02 emissions from gas recovered for non-energy uses (i.e.,
flaring, or catalytic oxidation) under fugitive C02 emission estimates for underground coal mining. In
effect, these emissions, though occurring through stationary combustion, are categorized as fugitive
emissions in the Inventory. EPA estimated C02 emissions from methane flaring using the following
equation:
Equation 3-2: Estimating C02 Emissions from Drained Methane Flared or Catalytically
Oxidized
C02 from flaring
= 0.98 x Volume of methane flared x Conversion Factor
x Stoichiometric Mass Factor
In 2023, there was a single mine that reported destruction of recovered methane through flaring without
energy use. Annual data for 2023 for this mine were obtained from the GHGRP (EPA 2024).
Step 2: Surface Mining
EPA estimated fugitive C02 emissions from surface mining using the IPCC Tier 1 emission factor (0.44
m3/metric ton) and annual coal production from surface mines (EIA 2024).
Uncertainty
A quantitative uncertainty analysis was conducted for the coal mining source category using the IPCC-
recommended Approach 2 uncertainty estimation methodology. Because emission estimates of CH4
from underground ventilation systems were based on actual measurement data from EPA's GHGRP or
from MSHA, uncertainty is relatively low. A degree of imprecision was introduced because the
ventilation air measurements used were not continuous but rather quarterly instantaneous readings
that were used to determine the average annual emission rates. Additionally, the measurement
equipment used can be expected to have resulted in an average of 10 percent overestimation of annual
Energy 3-77
-------
CH4 emissions (Mutmansky & Wang 2000). Equipment measurement uncertainty is applied to GHGRP
data.
Estimates of CH4 liberated and recovered by degasification systems are relatively certain for utilized CH4
because of the availability of EPA's GHGRP data and state gas sales information. Many of the liberation
and recovery estimates use data on wells within 100 feet of a mined area. However, uncertainty exists
concerning the radius of influence of each well. The number of wells counted, and thus the liberated
CH4 and avoided emissions, may vary if the drainage area is found to be larger or smaller than
estimated.
EPA's GHGRP requires weekly CH4 monitoring of mines that report degasification systems, and
continuous CH4 monitoring is required for CH4 utilized on- or off-site. Since 2012, GHGRP data have
been used to estimate CH4 emissions from vented degasification wells, reducing the uncertainty
associated with prior MSHA estimates used for this sub-source. Beginning in 2013, GHGRP data were
also used for determining CH4 recovery and use at mines without publicly available gas usage or sales
records, which has reduced the uncertainty from previous estimation methods that were based on
information from coal industry contacts.
Surface mining and post-mining emissions are associated with considerably more uncertainty than
underground mines, because of the difficulty in developing accurate emission factors from field
measurements. However, since underground coal mining, as a general matter, results in significantly
larger CH4 emissions due to production of higher-rank coal and greater depth, and estimated emissions
from underground mining constitute the majority of estimated total coal mining CH4 emissions, the
uncertainty associated with underground emissions is the primary factor that determines overall
uncertainty.
The major sources of uncertainty for estimates of fugitive C02 emissions are the Tier 1 IPCC default
emission factors used for underground mining (-50 percent to +100 percent) and surface mining (-67
percent to +200 percent) (IPCC 2019). Additional sources of uncertainty for fugitive C02 emission
estimates include ElA's annual coal production data and data used for gas recovery projects, such as
GHGRP data, state gas sales data, and VAM estimates for the single mine that operates an active VAM
project. Uncertainty ranges for these additional data sources are already available, as these are the
same data sources used for CH4 emission estimates.
The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 3-39. Coal
mining CH4 emissions in 2023 were estimated to be between 40.7 and 55.2 MMT C02 Eq. at a 95 percent
confidence level. This indicates a range of 10 percent below to 21 percent above the 2023 emission
estimate of 45.4 MMT C02 Eq. Coal mining fugitive C02 emissions in 2023 were estimated to be between
0.8 and 4.2 MMT C02 Eq. at a 95 percent confidence level. This indicates a range of 68 percent below to
76 percent above the 2023 emission estimate of 2.4 MMT C02 Eq.
Table 3-39: Approach 2 Quantitative Uncertainty Estimates for CH4 and C02 Emissions
from Coal Mining (MMT C02 Eq. and Percent)
2023 Emission
Source
Gas
Estimate
Uncertainty Range Relative to Emission Estimate"
(MMT CO2 Eq.)
(MMT CO2 Eq.)
(%)
Lower
Upper
Lower
Upper
Bound
Bound
Bound
Bound
Coal Mining
CH4
45.4
40.7
55.2
-10%
+21%
Coal Mining
CO2
2.4
0.8
4.2
-68%
+76%
a Range of emission estimates predicted by Monte Carlo stochastic simulation for a 95 percent confidence interval.
3-78 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
QA/QC and Verification
To ensure the quality of the emission estimates for coal mining, general (IPCC Tier 1) and category-
specific (Tier 2) Quality Assurance/Quality Control (QA/QC) procedures were implemented consistent
with the U.S. Inventory QA/QC plan outlined in Annex 8. Trends across the time series were analyzed to
determine whether any corrective actions were needed.
Emission estimates for coal mining rely in large part on data reported by coal mines to EPA's GHGRP.
EPA verifies annual facility-level reports through a multi-step process to identify potential errors and
ensure that data submitted to EPA are accurate, complete, and consistent. All reports submitted to EPA
are evaluated by electronic validation and verification checks. If potential errors are identified, EPA will
notify the reporter, who can resolve the issue either by providing an acceptable response describing why
the flagged issue is not an error or by correcting the flagged issue and resubmitting their annual report.
Additional QA/QC and verification procedures occur for each GHGRP subpart. No QA/QC issues or
errors were identified in the 2023 Subpart FF data.
Recalculations Discussion
Time series recalculations were performed due to revised historical data from state natural gas sales
databases for three mines, which are used to estimate avoided CH4 emissions from CH4 recovered and
used. As a result of recalculations, CH4 emissions decreased by an average of less than 0.001 percent
across the time series, compared to the previous Inventory. The biggest increase in CH4 emissions was
in 1991 where emissions increased by 0.004 percent, compared to the previous Inventory. The biggest
decrease in CH4 emissions was in 2011 (less than 0.001 percent). As a result of recalculations, there
was a very minor increase in CH4 emissions in 2022 (less than 0.001 percent), compared to the previous
Inventory.
Planned Improvements
EPA is assessing planned improvements for future reports, but currently has no specific planned
improvements for estimating CH4 and C02 emissions from underground and surface mining and CH4
emissions from post-mining.
3.5 Abandoned Underground Coal Mines
(Source Category 1B1 a)
Underground coal mines contribute the largest share of coal mine methane (CMM) emissions, with
active underground mines the leading source of underground emissions. However, mines also continue
to release CH4 after closure. As mines mature and coal seams are mined through, mines are closed and
abandoned. Many are sealed and some flood through intrusion of groundwater or surface water into the
void. Shafts or portals are generally filled with gravel and capped with a concrete seal, while vent pipes
and boreholes are plugged in a manner similar to oil and gas wells. Some abandoned mines are vented
to the atmosphere to prevent the buildup of CH4 that may find its way to surface structures through
overburden fractures. As work stops within the mines, CH4 liberation decreases but it does not stop
Energy 3-79
-------
completely. Following an initial decline, abandoned mines can liberate CH4 at a near-steady rate over an
extended period of time, or if flooded, produce gas for only a few years. The gas can migrate to the
surface through the conduits described above, particularly if they have not been sealed adequately. In
addition, diffuse emissions can occur when CH4 migrates to the surface through cracks and fissures in
the strata overlying the coal mine. The following factors influence abandoned mine emissions:
• Time since abandonment;
• Gas content and adsorption characteristics of coal;
• CH4 flow capacity of the mine;
• Mine flooding;
• Presence of vent holes; and
• Mine seals.
Annual gross abandoned mine CH4 emissions ranged from 8.1 to 12.1 MMT C02 Eq. from 1990 to 2023,
varying, in general, by less than 1 percent to approximately 19 percent from year to year. Fluctuations
were due mainly to the number of mines closed during a given year as well as the magnitude of the
emissions from those mines when active. Gross abandoned mine emissions peaked in 1996 (12.1 MMT
C02 Eq.) due to the large number of gassy mine63 closures from 1994 to 1996 (70 gassy mines closed
during the three-year period). In spite of this rapid rise, abandoned mine emissions have been generally
on the decline since 1996. Since 2002, there have been fewer than twelve gassy mine closures each
year. In 2023 there was one gassy mine closure. Gross abandoned mine emissions decreased slightly
from 9.1 MMT C02 Eq. (324 kt CH4) in 2022 to 9.0 (323 kt CH4) MMT C02 Eq. in 2023 (see Table 3-40 and
Table 3-41). Gross emissions are reduced by CH4 recovered and used at 62 mines, resulting in net
emissions in 2023 of 6.1 MMT C02 Eq. (219 kt CH4).
Table 3-40: CH4 Emissions from Abandoned Coal Mines (MMT C02 Eq.)
Activity 19901
2005
2019
2020
2021
2022
2023
I
Abandoned Underground Mines 8.1 1
1
9.3
9.6
9.4
9.2
9.1
9.0
Recovered & Used NO
(2.0)
(2.9)
(2.9)
(3.0)
(3.0)
(2.9)
Total 8.1|
7.4
6.6
6.5
6.2
6.1
6.1
NO (Not Occurring)
Notes: Parentheses indicate negative values. Totals may not sum due to independent rounding.
Table 3-41: CH4 Emissions from Abandoned Coal Mines (kt)
Activity 19901
2005
2019
2020
2021
2022
2023
Abandoned Underground Mines 288
334
341
335
330
324
323
Recovered & Used NO
(70)
(104)
(103)
(109)
(106)
(104)
Total 288|
264
237
232
221
218
219
NO (Not Occurring)
Note: Parentheses indicate negative values. Totals may not sum due to independent rounding.
63 A mine is considered a "gassy" mine if it emits more than 100 thousand cubic feet of ChU per day (100 Mcfd).
3-80 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Methodology and Time-Series Consistency
Estimating CH4 emissions from an abandoned coal mine requires predicting the emissions of a mine
from the time of abandonment through the inventory year of interest. The flow of CH4from the coal to
the mine void is primarily dependent on the mine's emissions when active and the extent to which the
mine is flooded or sealed. The CH4 emission rate before abandonment reflects the gas content of the
coal, the rate of coal mining, and the flow capacity of the mine in much the same way as the initial rate
of a water-free conventional gas well reflects the gas content of the producing formation and the flow
capacity of the well. A well or a mine that produces gas from a coal seam and the surrounding strata will
produce less gas through time as the reservoir of gas is depleted. Depletion of a reservoir will follow a
predictable pattern depending on the interplay of a variety of natural physical conditions imposed on the
reservoir. The depletion of a reservoir is commonly modeled by mathematical equations and mapped as
a type curve. Type curves, which are referred to as decline curves, have been developed for abandoned
coal mines. Existing data on abandoned mine emissions through time, although sparse, appear to fit the
hyperbolic type of decline curve used in forecasting production from natural gas wells.
There are sufficient mine level data available to establish decline curves for individual gassy mines
abandoned since 1972. For mines abandoned prior to 1972, county level data are available. Mine status
information (i.e., whether a mine is sealed, venting, or flooded) is not available for all the abandoned
gassy mines. Therefore, a hybrid Tier 2/Tier 3 method was developed to model abandoned gassy mine
emissions using Monte Carlo simulations. Tier 3 calculations are used for mines with known status
information where decline curves can be used to directly estimate abandoned mine emissions. For
mines with unknown status, a Tier 2 approach that estimates basin level emissions is used. This Tier 2
approach relies on data from other mines with known status and located within the same basin as the
unknown status mines. This approach is consistent with the IPCC 2006 Guidelines as underground
mines can be considered point sources and measurement methods are available.
To estimate CH4 emissions over time for a given abandoned mine, it is necessary to apply a decline
function, initiated upon abandonment, to that mine. In the analysis, mines were grouped by coal basin
with the assumption that they will generally have the same initial pressures, permeability, and isotherm.
As CH4 leaves the system, the reservoir pressure (Pr) declines as described by the isotherm's
characteristics. The emission rate declines because the mine pressure (Pw) is essentially constant at
atmospheric pressure for a vented mine, and the productivity index (PI), which is expressed as the flow
rate per unit of pressure change, is essentially constant at the pressures of interest (atmospheric to 30
psia). The CH4 flow rate is determined by the laws of gas flow through porous media, such as Darcy's
Law. A rate-time equation can be generated that can be used to predict future emissions. This decline
through time is hyperbolic in nature and can be empirically expressed as:
Equation 3-3: Decline Function to Estimate Venting Abandoned Mine Methane Emissions
where,
q
q.
b
Di
t
Gas flow rate at time t in million cubic feet per day (mmcfd)
Initial gas flow rate at time zero (t0), mmcfd
The hyperbolic exponent, dimensionless
Initial decline rate, 1/year
Elapsed time from t0 (years)
Energy 3-81
-------
This equation is applied to mines of various initial emission rates that have similar initial pressures,
permeability, and adsorption isotherms (EPA 2004).
The decline curves created to model the gas emission rate of coal mines must account for factors that
decrease the rate of emissions after mining activities cease, such as sealing and flooding. Based on field
measurement data, it was assumed that most U.S. mines prone to flooding will become completely
flooded within eight years and therefore will no longer have any measurable CH4 emissions. Based on
this assumption, an average decline rate for flooded mines was established by fitting a decline curve to
emissions from field measurements. An exponential equation was developed from emissions data
measured at eight abandoned mines known to be filling with water located in two of the five basins.
Using a least squares, curve-fitting algorithm, emissions data were matched to the exponential equation
shown below. For this analysis of flooded abandoned mines, there was not enough data to establish
basin-specific equations, as was done with the vented, non-flooding mines (EPA 2004). This decline
through time can be empirically expressed as:
Equation 3-4: Decline Function to Estimate Flooded Abandoned Mine Methane Emissions
Seals have an inhibiting effect on the rate of flow of CH4 into the atmosphere compared to the flow rate
that would exist if the mine had an open vent. The total volume emitted will be the same, but emissions
will occur over a longer period of time. The methodology, therefore, treats the emissions prediction from
a sealed mine similarly to the emissions prediction from a vented mine, but uses a lower initial rate
depending on the degree of sealing. A computational fluid dynamics simulator was used with the
conceptual abandoned mine model to predict the decline curve for inhibited flow. The percent sealed is
defined as 100 * (1 - [initial emissions from sealed mine / emission rate at abandonment prior to
sealing]). Significant differences are seen between 50 percent, 80 percent, and 95 percent closure.
These decline curves were therefore used as the high, middle, and low values for emissions from sealed
mines (EPA2004).
For active coal mines, those mines producing over 100 thousand cubic feet per day (Mcfd) of CH4
account for about 98 percent of all CH4 emissions. This same relationship is assumed for abandoned
mines. It was determined that the 532 abandoned mines closed since 1972 produced CH4 emissions
greater than 100 Mcfd when active. Further, the status of 308 of the 532 mines (or 58 percent) is known
to be either: 1) vented to the atmosphere; 2) sealed to some degree (either earthen or concrete seals); or
3) flooded (enough to inhibit CH4 flow to the atmosphere). The remaining 42 percent of the mines whose
status is unknown were placed in one of these three categories by applying a probability distribution
analysis based on the known status of other mines located in the same coal basin (EPA 2004). Table
3-42 presents the count of mines by post-abandonment state, based on EPA's probability distribution
analysis.
q = qte( Dt)
where,
q
q.
D
t
Gas flow rate at time t in mmcfd
Initial gas flow rate at time zero (t0), mmcfd
Decline rate, 1/year
Elapsed time from t0 (years)
3-82 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Table 3-42: Number of Gassy Abandoned Mines Present in U.S. Basins in 2023,
Grouped by Class According to Post-Abandonment State
Basin
Sealed
Vented
Flooded
Total Known
Unknown
Total Mines
Central Appl.
43
25
50
118
145
263
Illinois
35
3
14
52
31
83
Northern Appl.
50
23
15
88
38
126
Warrior Basin
0
0
16
16
0
16
Western Basins
28
4
2
34
10
44
Total
156
55
97
308
224
532
Inputs to the decline equation require the average CH4 emission rate prior to abandonment and the date
of abandonment. Generally, these data are available for mines abandoned after 1971; however, such
data are largely unknown for mines closed before 1972. Information that is readily available, such as
coal production by state and county, is helpful but does not provide enough data to directly employ the
methodology used to calculate emissions from mines abandoned before 1972. It is assumed that pre-
1972 mines are governed by the same physical, geologic, and hydrologic constraints that apply to post-
1971 mines; thus, their emissions may be characterized by the same decline curves.
During the 1970s, 78 percent of CH4 emissions from coal mining came from seventeen counties in seven
states. Mine closure dates were obtained for two states, Colorado and Illinois, for the hundred-year
period extending from 1900 through 1999. The data were used to establish a frequency of mine closure
histogram (by decade) and applied to the other five states with gassy mine closures. As a result, basin-
specific decline curve equations were applied to the 145 gassy coal mines estimated to have closed
between 1920 and 1971 in the United States, representing 78 percent of the emissions. State-specific,
initial emission rates were used based on average coal mine CH4 emission rates during the 1970s (EPA
2004).
Abandoned mine emission estimates are based on all closed mines known to have active mine CH4
ventilation emission rates greater than 100 Mcfd at the time of abandonment. For example, for 1990 the
analysis included 145 mines closed before 1972 and 258 mines closed between 1972 and 1990. Initial
emission rates based on MSHA reports, time of abandonment, and basin-specific decline curves
influenced by a number of factors were used to calculate annual emissions for each mine in the
database (MSHA 2023). Coal mine degasification data are not available for years prior to 1990, thus the
initial emission rates used reflect only ventilation emissions for pre-1990 closures. Methane
degasification amounts were added to the quantity of CH4 vented to determine the total CH4 liberation
rate for all mines that closed between 1992 and 2023. Since the sample of gassy mines described above
is assumed to account for 78 percent of the pre-1972 and 98 percent of the post-1971 abandoned mine
emissions, the modeled results were multiplied by 1.22 and 1.02, respectively, to account for all U.S.
abandoned mine emissions.
From 1993 through 2023, emission totals were downwardly adjusted to reflect CH4 emissions avoided
from abandoned mines with CH4 recovery and use or destruction systems. Currently, there are 62
abandoned mines with recovery projects, including 11 projects at mines abandoned before 1972 (pre-
1972 mines) (EPA 2004, CMOP 2024). Because CH4 recovered by these projects is expected to decline
with the age of the mine, CH4 recovery is assumed to be the total estimated CH4 liberated based on the
Energy 3-83
-------
mine's decline function except for three recovery projects where additional data are available (COGIS
2018, MSHA 2024).64
The Inventory totals were not adjusted for abandoned mine CH4 emissions avoided from 1990 through
1992 due to unavailability of data. Avoided CH4 emissions from pre-1972 abandoned mines are
estimated by multiplying the total estimated emissions from these mines in each decade by the fraction
of mines with recovery projects in that decade. For recovery projects at pre-1972 abandoned mines, four
projects are at mines abandoned in the 1920s, three in the 1930s, two in the 1950s, and two in the 1960s
(EPA 2004).
Reviewing Coalbed Methane Outreach Program data (CMOP 2024) revealed five additional recovery
projects starting in 2021 that were added to the recovery project list. In addition to reviewing CMOP
data, the recovery project list was checked against the Global Methane Initiative International Coal Mine
Methane Project List Database (GMI) and the American Carbon Registry (ACR) (GMI 2024, ACR 2024). Of
the 44 operational recovery projects for U.S. abandoned coal mines currently available in the GMI
dataset, 35 are already included in the AMM model. Three new projects from this dataset were added to
the recovery list (one project contains three mines). The remaining projects in the GMI dataset are for
mines that are not yet abandoned according to MSHA records or were in abandoned in 2024 and will be
included in next year's Inventory (MSHA 2024). The ACR Registry had one additional recovery project not
listed in the other datasets that was added to the AMM model (ACR 2024).
Uncertainty
A quantitative uncertainty analysis was conducted for the abandoned coal mine source category using
the IPCC-recommended Approach 2 uncertainty estimation methodology. The uncertainty analysis
provides for the specification of probability density functions for key variables within a computational
structure that mirrors the calculation of the Inventory estimate. The results provide the range within
which, with 95 percent certainty, emissions from this source category are likely to fall.
As discussed above, the parameters for which values must be estimated for each mine to predict its
decline curve are: 1) the coal's adsorption isotherm; 2) CH4 flow capacity as expressed by permeability;
and 3) pressure at abandonment. Because these parameters are not available for each mine, a
methodological approach to estimating emissions was used that generates a probability distribution of
potential outcomes based on the most likely value and the probable range of values for each parameter.
The range of values is not meant to capture the extreme values, but rather values that represent the
highest and lowest quartile of the cumulative probability density function of each parameter. Once the
low, mid, and high values are selected, they are applied to a probability density function.
The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 3-43. Annual
abandoned coal mine CH4 emissions in 2023 were estimated to be between 4.9 and 7.6 MMT C02 Eq. at
a 95 percent confidence level. This indicates a range of 20 percent below to 24 percent above the 2023
emission estimate of 6.1 MMT C02 Eq. One of the reasons for the relatively narrow range is that mine-
specific data is available for use in the methodology for mines closed in 1972 and later years. Emissions
from mines closed prior to 1972 have the largest degree of uncertainty because no mine-specific CH4
liberation rates at the time of abandonment exist.
64 Data from a state oil and gas database (COGIS) is used for one project and the mine status information from MSHA for
two mines (sealed and flooded) indicate zero recovery emissions for these projects.
3-84 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Table 3-43: Approach 2 Quantitative Uncertainty Estimates for CH4 Emissions from
Abandoned Underground Coal Mines (MMT C02 Eq. and Percent)
Source
Gas
2023
Emission
Estimate
(MMT CO2 Eq.)
Uncertainty Range Relative to Emission Estimate"
(MMT CO2 Eq.)
(%)
Lower Upper
Bound Bound
Lower Upper
Bound Bound
Abandoned Underground Coal Mines
CH4
6.1
4.9 7.6
-20% +24%
a Range of emission estimates predicted by Monte Carlo Simulation for a 95 percent confidence interval.
QA/QC and Verification
To ensure the quality of the emission estimates for abandoned coal mines, general (IPCC Tier 1) and
category-specific (Tier 2) Quality Assurance/Quality Control (QA/QC) procedures were implemented
consistent with the U.S. Inventory QA/QC plan outlined in Annex 8. Trends across the time series were
analyzed to determine whether any corrective actions were needed.
Recalculations Discussion
Eleven new abandoned mine methane recovery projects were added to the AMM model during the
current Inventory (CMOP 2024, GMI 2024, ACR 2024). CMOP, GMI, and ACR data indicate 10 of these
recovery projects were started in 2021 and one in 2022. Time series recalculations were performed for
2021 and 2022 by adding in the recovery project(s) and rerunning the 2021 and 2022 AMM models. As a
result of recalculations, CH4 emissions decreased by one percent in 2021 and three percent in 2022,
compared to the previous Inventory.
3.6 Petroleum Systems (Source Category
1B2a)
This category (1 B2a) is defined in the IPCC methodological guidance as fugitive emissions from
petroleum systems, which per IPCC guidelines include emissions from leaks, venting, and flaring.
Methane emissions from petroleum systems are primarily associated with onshore and offshore crude
oil exploration, production, transportation, and refining operations. During these activities, CH4 is
released to the atmosphere as emissions from leaks, venting (including emissions from operational
upsets), and flaring. Carbon dioxide emissions from petroleum systems are primarily associated with
onshore and offshore crude oil production and refining operations. Note, C02 emissions in petroleum
systems exclude all combustion emissions (e.g., engine combustion) except for flaring C02 emissions.
All combustion C02 emissions (except for flaring) are accounted for in the fossil fuel combustion
chapter (see Section 3.1). Emissions of N20 from petroleum systems are primarily associated with
flaring.
Total greenhouse gas emissions (CH4, C02, and N20) from petroleum systems in 2023 were 61.3 MMT
C02 Eq., an increase of 3 percent from 1990, primarily due to increases in C02 emissions. Total
emissions decreased by 7 percent from 2010 levels and have increased by 5 percent since 2022. Total
C02 emissions from petroleum systems in 2023 were 23.3 MMT C02 (23,272 kt C02), 2.4 times higher
Energy 3-85
-------
than in 1990. Total C02 emissions in 2023 were 1.7 times higher than in 2010 and 5 percent higher than
in 2022. Total CH4 emissions from petroleum systems in 2023 were 38.0 MMT C02 Eq. (1,358 kt CH4), a
decrease of 24 percent from 1990. Since 2010, total CH4 emissions decreased by 27 percent; and since
2022, CH4 emissions increased by 5 percent. Total N20 emissions from petroleum systems in 2023 were
0.022 MMT C02 Eq. (0.083 kt N20), 1.6 times higher than in 1990,1.2 times higher than in 2010, and 54
percent lower than in 2022. Since 1990, U.S. oil production has increased by 69 percent. In 2023, U.S. oil
production was 186 percent higher than in 2010 and 8 percent higher than in 2022.
Each year, some estimates in the Inventory are recalculated with improved methods and/or data. These
improvements are implemented consistently across the entire Inventory's time series (i.e., 1990 to 2023)
to ensure that the trend is representative of changes in emissions levels. Recalculations in petroleum
systems in this year's Inventory include:
• Updates to oil and gas well counts, oil and gas production volumes, and produced water
production volumes using the most recent data from Enverus.
• Updates to oil and gas production volumes using the most recent data from the United States
Energy Information Administration (EIA).
• Recalculations due to Greenhouse Gas Reporting Program (GHGRP) submission revisions.
• Methodological updates for offshore production in the Gulf of America.
The Recalculations Discussion section below provides more details on the updated methods.
Exploration. Exploration includes well drilling, testing, and completions. Exploration accounts for less
than 0.5 percent of total CH4 emissions (including leaks, vents, and flaring) from petroleum systems in
2023. The predominant sources of CH4 emissions from exploration are hydraulically fractured oil well
completions. Other sources include well testing, well drilling, and well completions without hydraulic
fracturing. Since 1990, exploration CH4 emissions have decreased 95 percent, and while the number of
hydraulically fractured wells completed increased 64 percent, there were decreases in the fraction of
such completions without reduced emissions completions (RECs) or flaring. Emissions of CH4 from
exploration were highest in 2008, over 62 times higher than in 2023; and lowest in 2022. Emissions of
CH4 from exploration increased 10 percent from 2022 to 2023, due to an increase in emissions from
hydraulically fractured oil well completions without RECs. Exploration accounts for 2 percent of total
C02 emissions (including leaks, vents, and flaring) from petroleum systems in 2023. Emissions of C02
from exploration in 2023 were 21 percent higher than in 1990, and increased by 50 percent from 2022,
largely due to an increase in emissions from hydraulically fractured oil well completions with REC and
flaring (by 78 percent from 2022). Emissions of C02 from exploration were highest in 2014, over 8 times
higher than in 2023. Exploration accounts for 1 percent of total N20 emissions from petroleum systems
in 2023. Emissions of N20 from exploration in 2023 are 22 percent higher than in 1990, and 59 percent
higher than in 2022, due to hydraulically fractured oil well completions with flaring.
Production. Production accounts for 97 percent of total CH4 emissions (including leaks, vents, and
flaring) from petroleum systems in 2023. The predominant sources of emissions from production field
operations are pneumatic controllers, equipment leaks, offshore oil platforms, produced water, gas
engines, chemical injection pumps, and associated gas flaring. In 2023, these seven sources together
accounted for 91 percent of the CH4 emissions from production. Since 1990, CH4emissions from
production have decreased by 20 percent primarily due to decreases in emissions from offshore
production. Overall, production segment CH4 emissions increased by 5 percent from 2022 levels due
primarily to equipment leaks. Production emissions account for 86 percent of the total C02 emissions
3-86 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
(including leaks, vents, and flaring) from petroleum systems in 2023. The principal sources of C02
emissions are associated gas flaring, miscellaneous production flaring, and oil tanks with flares. In
2023, these three sources together accounted for 97 percent of the C02 emissions from production. In
2023, C02 emissions from production were 3.3 times higher than in 1990, due to increases in flaring
emissions from associated gas flaring, miscellaneous production flaring, and tanks. Overall, in 2023,
production segment C02 emissions increased by 5 percent from 2022 levels primarily due to increases
in miscellaneous production flaring in the Permian Basin. Production emissions accounted for 65
percent of the total N20 emissions from petroleum systems in 2023. The principal sources of N20
emissions are oil tanks with flares, associated gas flaring, and miscellaneous production flaring,
accountingfor 83 percent of N20 emissions from the production segment in 2023. In 2023, N20
emissions from production were 2.2 times higher than in 1990 and were 65 percent lower than in 2022.
Crude Oil Transportation. Emissions from crude oil transportation account for a very small percentage of
the total emissions (including leaks, vents, and flaring) from petroleum systems. Crude oil
transportation activities account for 0.7 percent of total CH4 emissions from petroleum systems.
Emissions from tanks, marine loading, and truck loading operations accounted for 81 percent of CH4
emissions from crude oil transportation in 2023. Since 1990, CH4 emissions from transportation have
increased by 37 percent. In 2023, CH4 emissions from transportation increased by 6 percent from 2022
levels. Crude oil transportation activities account for less than 0.01 percent of total C02 emissions from
petroleum systems. Emissions from tanks, marine loading, and truck loading operations account for 81
percent of C02 emissions from crude oil transportation.
Crude Oil Refining. Crude oil refining processes and systems account for 2 percent of total CH4
emissions from petroleum systems in 2023. This low share is because most of the CH4 in crude oil is
removed or escapes before the crude oil is delivered to the refineries. There is a negligible amount of
CH4 in all refined products. Within refineries, flaring accounts for 45 percent of the CH4 emissions, while
delayed cokers, uncontrolled blowdowns, and equipment leaks account for 17,13 and 11 percent,
respectively. CH4 emissions from refining of crude oil have decreased by 12 percent since 1990, and
decreased by 4 percent from 2022; however, like the transportation subcategory, this increase has had
little effect on the overall emissions of CH4 from petroleum systems. Crude oil refining processes and
systems account for 12 percent of total C02 emissions from petroleum systems. Of the total C02
emissions from refining, almost all (about 99 percent) of it comes from flaring.65 Since 1990, refinery
C02 emissions decreased by 10 percent and have increased by 1 percent from 2022 levels, due to
changes in flaring emissions. Flaring occurring at crude oil refining processes and systems accounts for
34 percent of total N20 emissions from petroleum systems. In 2023, refinery N20 emissions increased
by 3 percent since 1990 and increased by 1 percent from 2022 levels.
Table 3-44: Total Greenhouse Gas Emissions (C02, CH4, and N20) from Petroleum
Systems (MMT C02 Eq.)
Activity
1990
2005 | 2019
2020
2021
2022
2023
Exploration
3.4
5.8 I
2.5
1.1
0.8
0.4
0.6
Production
52.2
48.3
89.0
74.5
64.5
54.2
56.9
Transportation
0.2
0.1 |
00
0
0.2
0.2
0.2
0.3
65 Petroleum systems includes emissions from leaks, venting, and flaring. In many industries, including petroleum
refineries, the largest source of onsite CO2 emissions is often fossil fuel combustion, which is covered in Section 3.1 of
this chapter.
Energy 3-87
-------
Activity
1990
2005
2019
2020
2021
2022
2023
Crude Refining
3.9
4.51
4.4
3.6
3.7
3.5
3.5
Total
59.6
58.71
96.2
79.5
69.2
58.4
61.3
Note: Totals may not sum due to independent rounding.
Table 3-45: CH4 Emissions from Petroleum Systems (MMT C02
Eq.)
Activity
1990 2005 2019
2020
2021
2022
2023
Exploration
3.0
5.3
0.5
0.3
0.2
0.1
0.1
Production
46.1
42.2
49.2
49.3
44.0
35.2
37.0
Pneumatic Controllers
21.3
22.8
24.3
30.8
27.1
18.5
16.3
Offshore Production
10.5
7.3
4.3
3.0
2.9
2.9
2.9
Equipment Leaks
2.3
2.8
3.9
3.2
3.2
3.1
6.7
Gas Engines
2.3
2.0
2.6
2.5
2.5
2.5
2.5
Produced Water
2.6
1.8
2.8
2.5
2.3
2.4
2.5
Chemical Injection Pumps
1.3
2.2
3.3
2.6
2.3
2.1
1.8
Assoc Gas Flaring
0.6
0.4
2.5
1.3
1.0
0.8
0.9
Other Sources
5.3
2.8
5.4
3.4
2.8
2.8
3.3
Crude Oil Transportation
0.2
0.1
0.3
0.2
0.2
0.2
0.3
Refining
0.7
0.9
0.9
0.7
0.7
0.7
0.7
Total
50.0
48.4
50.8
50.6
45.1
36.3
38.0
Note: Totals may not sum due to independent rounding.
Table 3-46: CH4 Emissions from Petroleum Systems (kt CH4)
Activity
1990
2005
2019
2020
2021
2022
2023
Exploration
106
189
16
12
7
4
5
Production
1,648
1,505
1,756
1,762
1,571
1,258
1,321
Pneumatic Controllers
761
814
868
1,099
967
662
581
Offshore Production
374
261
155
108
104
104
104
Equipment Leaks
82
101
139
113
113
112
239
Gas Engines
81
70
93
89
88
89
90
Produced Water
92
64
100
91
81
85
90
Chemical Injection Pumps
47
80
118
93
83
76
64
Assoc Gas Flaring
20
15
91
47
35
30
34
Other Sources
189
100
193
122
101
100
119
Crude Oil Transportation
7
5
9
8
8
8
9
Refining
27
30
31
26
25
25
24
Total
1,787
1,730
1,813
1,807
1,611
1,295
1,358
Note: Totals may not sum due to independent rounding.
Table 3-47: C02 Emissions from Petroleum Systems (MMT C02)
Activity
1990
2005
2019
2020
2021
2022
2023
Exploration
0.4
0.5
2.1
0.8
0.6
0.3
0.5
Production
6.0
6.2
39.8
25.2
20.5
18.9
19.9
Transportation
+1
I
+1
+
+
+
+
+
Crude Refining
3.2
3.6
3.6
2.9
3.0
2.8
2.9
3-88 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Activity
1990
2005
2019
2020
2021
2022
2023
Total
9.6
10.2
45.4
28.9
24.1
22.1
23.3
+ Does not exceed 0.05 MMT C02
Note: Totals may not sum due to independent rounding.
Table 3-48: C02 Emissions from Petroleum Systems (kt C02)
Activity
1990
2005
2019
2020
2021
2022
2023
Exploration
398
465
2,053
798
602
321
481
Production
6,024
6,153
39,830
25,203
20,487
18,941
19,928
Transportation
0.9
0.7
1.3
1.2
1.1
1.2
1.3
Crude Refining
3,174
3,602
3,560
2,874
3,001
2,820
2,862
Total
9,597 10,222 45,445
28,876
24,091
22,084
23,272
Note: Totals may not sum due to independent rounding.
Table 3-49: N20 Emissions from Petroleum Systems (Metric Tons C02 Eq.)
Activity
1990
2005
2019
2020
2021
2022
2023
Exploration
181
209
822
353
290
138
220
Production
6,635
6,168
22,120
13,875
11,977
40,649
14,330
Transportation
NE
NE
NE
NE
NE
NE
NE
Crude Refining
7,262
8,243
9,283
7,523
7,867
7,387
7,496
Total
14,078
14,621
32,225
21,751
20,134
48,174
22,046
NE (Not Estimated)
Note: Totals may not sum due to independent rounding.
Table 3-50: N20 Emissions from Petroleum Systems (Metric Tons N20)
Activity
1990
2005
2019
2020
2021
2022
2023
Exploration
0.7
0.8
3.1
1.3
1.1
0.5
0.8
Production
25.0
23.3
83.5
52.4
45.2
153.4
54.1
Transportation
NE
NE
NE
NE
NE
NE
NE
Crude Refining
27.4
31.1
35.0
28.4
29.7
27.9
28.3
Total
53.1 |
55.2
121.6
82.1
76.0
181.8
83.2
NE (Not Estimated)
Note: Totals may not sum due to independent rounding.
Methodology and Time-Series Consistency
See Annex 3.5 for the full time series of emissions data, activity data, emission factors, and additional
information on methods and data sources.
Petroleum systems includes emission estimates for activities occurring in petroleum systems from the
oil wellhead through crude oil refining, including activities for crude oil exploration, production field
operations, crude oil transportation activities, and refining operations. Generally, emissions are
estimated for each activity by multiplying emission factors (e.g., emission rate per equipment or per
activity) by corresponding activity data (e.g., equipment count or frequency of activity). Certain sources
within petroleum refineries are developed using an IPCC Tier 3 approach (i.e., all refineries in the nation
report facility-level emissions data to the GHGRP, which are included directly in the national emissions
estimates here). Other estimates are developed with a Tier 2 approach. Tier 1 approaches are not used.
Energy 3-89
-------
Emission Factors. Key references for emission factors include Methane Emissions from the Natural Gas
Industry by the Gas Research Institute and EPA (GRI/EPA 1996), Estimates of Methane Emissions from
the U.S. Oil Industry (EPA 1999), Compilation of Air Pollutant Emission Factors, AP-42 (EPA 1997), Global
Emissions of Methane from Petroleum Sources (AP11992), consensus of industry peer review panels,
Bureau of Ocean Energy Management (BOEM) reports, Nonpoint Oil and Gas Emission Estimation Tool
(EPA 2017), and analysis of GHGRP data (EPA 2024).
Emission factors for hydraulically fractured (HF) oil well completions and workovers (in four control
categories) were developed at the basin level using EPA's GHGRP data; year-specific data were used to
calculate basin-specific emission factors from 2016-forward and the year 2016 emission factors were
applied to all prior years in the time series. For basins not reporting to the GHGRP, Subpart W average
emission factors were used. For more information, please see the 2023 memoranda available online.66
The emission factors for well testing and associated gas venting and flaring were developed using year-
specific GHGRP data for years 2015 forward; earlier years in the time series use 2015 emission factors.
For miscellaneous production flaring, year-specific emission factors were developed for years 2015
forward from GHGRP data, an emission factor of 0 (assumption of no flaring) was assumed for 1990
through 1992, and linear interpolation was applied to develop emission factors for 1993 through 2014.
For more information, please see memoranda available online.67 For offshore oil production, emission
factors were calculated using BOEM data for offshore facilities in federal waters of the Gulf of America
(and these data were also applied to facilities located in state waters of the Gulf of America) and GHGRP
data for offshore facilities off the coasts of California and Alaska. For many other sources, emission
factors were held constant for the period 1990 through 2023, and trends in emissions reflect changes in
activity levels. Emission factors from EPA 1999 are used for all other production and transportation
activities.
For associated gas venting and flaring and miscellaneous production flaring, emission factors were
developed on a production basis (i.e., emissions per unit oil produced). Additionally, for these two
sources, basin-specific activity and emission factors were developed for each basin that in any year
from 2011 forward contributed at least 10 percent of total source emissions (on a C02 Eq. basis) in the
GHGRP. For associated gas venting and flaring, basin-specific factors were developed for four basins:
Williston, Permian, Gulf Coast, and Anadarko. For miscellaneous production flaring, basin-specific
factors were developed for three basins: Williston, Permian, and Gulf Coast. For each source, data from
all other basins were combined, and activity and emission factors were developed for the other basins
as a single group.
For pneumatic controllers and tanks, basin-specific emission factors were calculated for all the basins
reporting to the GHGRP. These emission factors were calculated for all the years with applicable GHGRP
data (i.e., 2011 - 2023 or 2015 - 2023). For the remaining basins (i.e., basins not reporting to the GHGRP),
Subpart W average emission factors were used. For more information, please see memoranda available
online.
For the exploration and production segments, in general, C02 emissions for each source were estimated
with GHGRP data or by multiplying C02 content factors by the corresponding CH4 data, as the C02
content of gas relates to its CH4 content. Sources with C02 emission estimates calculated using GHGRP
data include HF completions and workovers, associated gas venting and flaring, tanks, well testing,
66 See https://www.epa.gov/ghgemissions/natijral-gas-and-petroleum-systems.
67 See https://www.epa.gov/ghgemissions/natijral-gas-and-petroleum-systems.
3-90 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
pneumatic controllers, chemical injection pumps, miscellaneous production flaring, and certain
offshore production facilities (those located off the coasts of California and Alaska). For these sources,
C02 was calculated using the same methods as used for CH4. Carbon dioxide emission factors for
offshore oil production in the Gulf of America were derived using data from BOEM, following the same
methods as used for CH4 estimates. For other sources, the production field operations emission factors
for C02 are generally estimated by multiplying the CH4 emission factors by a conversion factor, which is
the ratio of C02 content and CH4 content in produced associated gas.
For the exploration and production segments, N20 emissions were estimated for flaring sources using
GHGRP or BOEM OGOR-B data and the same method used for C02. Sources with N20 emissions in the
exploration segment include well testing and HF completions with flaring. Sources with N20 emissions
in the production segment include associated gas flaring, tank flaring, miscellaneous production flaring,
HF workovers with flaring, and flaring from offshore production sources.
For crude oil transportation, emission factors for CH4 were largely developed using data from EPA
(1997), API (1992), and EPA (1999). Emission factors for C02 were estimated by multiplying the CH4
emission factors by a conversion factor, which is the ratio of C02 content and CH4 content in whole
crude post-separator.
For petroleum refining activities, year-specific emissions from 2010 forward were directly obtained from
EPA's GHGRP. All U.S. refineries have been required to report CH4, C02, and N20 emissions for all major
activities starting with emissions that occurred in 2010. The reported total CH4, C02, and N20 emissions
for each activity was used for the emissions in each year from 2010 forward. To estimate emissions for
1990 to 2009, the 2010 to 2013 emissions data from GHGRP along with the refinery feed data for 2010 to
2013 were used to derive CH4 and C02 emission factors (i.e., sum of activity emissions/sum of refinery
feed) and 2010 to 2017 data were used to derive N20 emission factors; these emission factors were then
applied to the annual refinery feed in years 1990 to 2009. GHGRP delayed coker CH4 emissions for 2010
through 2017 were increased using the ratio of certain reported emissions for 2018 to 2017, to account
for a more accurate GHGRP calculation methodology that was implemented starting in reporting year
2018.
A complete list of references for emission factors and activity data by emission source is provided in
Annex 3.5.
Activity Data. References for activity data include Enverus data (Enverus 2025), Energy Information
Administration (EIA) reports, Methane Emissions from the Natural Gas Industry by the Gas Research
Institute and EPA (EPA/GR11996), Estimates of Methane Emissions from the U.S. Oil Industry (EPA 1999),
consensus of industry peer review panels, BOEM reports, the Oil & Gas Journal, the Interstate Oil and
Gas Compact Commission, the United States Army Corps of Engineers, and analysis of GHGRP data
(EPA 2024).
For HF oil well completions and workovers, pneumatic controllers, equipment leaks, chemical injection
pumps, and tanks, basin-specific activity factors were calculated for all the basins reporting to the
GHGRP. These factors were calculated for all the years with applicable GHGRP data (i.e., 2011 through
2023, 2016 through 2023, or 2015 through 2023). For the remaining basins (i.e., basins not reporting to
the GHGRP), GHGRP average activity factors were used. For more information, please see memoranda
available online.68
68 See https://www.epa.gov/ghgemissions/natijral-gas-and-petroleum-systems.
Energy 3-91
-------
For many sources, complete activity data were not available for all years of the time series. In such
cases, one of three approaches was employed to estimate values, consistent with IPCC good practice.
Where appropriate, the activity data were calculated from related statistics using ratios developed
based on EPA/GRI (1996) and/or GHGRP data. In some cases, activity data are developed by
interpolating between recent data points (such as from GHGRP) and earlier data points, such as from
EPA/GRI (1996). Lastly, in limited instances the previous year's data were used if current year data were
not yet available.
A complete list of references for emission factors and activity data by emission source is provided in
Annex 3.5. The notation key "IE" is used for C02 and CH4 emissions from venting and flaring in common
data tables category 1 .B.2. Disaggregating flaring and venting estimates across the Inventory would
involve the application of assumptions and could result in inconsistent reporting and, potentially,
decreased transparency. Data availability varies across segments within oil and gas activities systems,
and emission factor data available for activities that include flaring can include emissions from multiple
sources (flaring, venting and leaks).
As noted above, EPA's GHGRP data, available starting in 2010 for refineries and in 2011 for other
sources, have improved estimates of emissions from petroleum systems. Many of the previously
available datasets were collected in the 1990s. To develop a consistent time series for sources with new
data, EPA reviewed available information on factors that may have resulted in changes over the time
series (e.g., regulations, voluntary actions) and requested stakeholder feedback on trends as well. For
most sources, EPA developed annual data for 1993 through 2009 or 2014 by interpolating activity data or
emission factors or both between 1992 (when GRI/EPA data are available) and 2010 or 2015 data points.
Information on time-series consistency for sources updated in this year's Inventory can be found in the
Recalculations Discussion below, with additional detail provided in supporting memos (relevant memos
are cited in the Recalculations Discussion). For information on other sources, please see the
Methodology and Time-Series Consistency discussion above and Annex 3.5.
Methodological recalculations were applied to the entire time series to ensure time-series consistency
from 1990 through 2023.
Uncertainty
EPA conducted a quantitative uncertainty analysis using the IPCC Approach 2 methodology (Monte
Carlo Simulation technique) to characterize uncertainty for petroleum systems. For more information on
the approach, please see the memoranda Inventory of U.S. Greenhouse Gas Emissions and Sinks 1990-
2016: Natural Gas and Petroleum Systems Uncertainty Estimates (2018 Uncertainty memo) and
Inventory of U.S. Greenhouse Gas Emissions and Sinks 1990-2019: Update for Natural Gas and
Petroleum Systems C02 Uncertainty Estimates (2021 Uncertainty memo).69
EPA used Palisade's @RISK add-in tool for Microsoft Excel to estimate the 95 percent confidence bound
around CH4 and C02 emissions from petroleum systems for the current Inventory. For the CH4
uncertainty analysis, EPA focused on the three highest methane-emitting sources for the year 2023,
which together emitted 51 percent of methane from petroleum systems in 2023, and extrapolated the
estimated uncertainty for the remaining sources. For the C02 uncertainty analysis, EPA focused on the
four highest-emitting sources for the year 2023 which together emitted 52 percent of C02 from
69 See https://www.epa.gov/ghgemissions/natijral-gas-and-petroleum-systems.
3-92 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
petroleum systems in 2023, and extrapolated the estimated uncertainty for the remaining sources. The
@RISK add-in provides for the specification of probability density functions (PDFs) for key variables
within a computational structure that mirrors the calculation of the inventory estimate. For emission
factors that are derived from methane emissions measurement studies, the PDFs are commonly
determined to be lognormally distributed (GRI/EPA 1996; EPA 1999). For activity data that are derived
from national datasets, the PDFs are set to a uniform distribution (see 2018 and 2021 Uncertainty
memos). Many emission factors and activity factors are calculated using subpart W data, and for these,
the @RISK add-in determines the best fitting PDF (e.g., lognormal, gaussian), based on bootstrapping of
the underlying data (see 2018 and 2021 Uncertainty memos). The IPCC guidance notes that in using this
Approach 2 method, "some uncertainties that are not addressed by statistical means may exist,
including those arising from omissions or double counting, or other conceptual errors, or from
incomplete understanding of the processes that may lead to inaccuracies in estimates developed from
models." As a result, the understanding of the uncertainty of emission estimates for this category
evolves and improves as the underlying methodologies and datasets improve. The uncertainty bounds
reported below only reflect those uncertainties that EPA has been able to quantify and do not
incorporate considerations such as modeling uncertainty, data representativeness, measurement
errors, misreportingor misclassification. To estimate uncertainty for N20, EPA applied the uncertainty
bounds calculated for C02. EPA will seek to refine this estimate in future Inventories.
The results presented below provide the 95 percent confidence bound within which actual emissions
from this source category are likely to fall for the year 2023, using the recommended IPCC methodology.
The results of the Approach 2 uncertainty analysis are summarized in Table 3-51. Petroleum systems
CH4 emissions in 2023 were estimated to be between 33.1 and 47.2 MMT C02 Eq., while C02 emissions
were estimated to be between 19.0 and 28.5 MMT C02 Eq. at a 95 percent confidence level. Petroleum
systems N20 emissions in 2023 were estimated to be between 0.018 and 0.027 MMT C02 Eq. at a 95
percent confidence level.
Uncertainty bounds for other years of the time series have not been calculated, but uncertainty is
expected to vary over the time series. For example, years where many emission sources are calculated
with interpolated data would likely have higher uncertainty than years with predominantly year-specific
data. In addition, the emission sources that contribute the most to CH4 and C02 emissions are different
over the time series, particularly when comparing recent years to early years in the time series. For
example, associated gas venting emissions were higher and flaring emissions were lower in early years
of the time series, compared to recent years. Technologies also changed over the time series (e.g.,
reduced emissions completions were not used early in the time series).
Table 3-51: Approach 2 Quantitative Uncertainty Estimates for CH4 and C02 Emissions
from Petroleum Systems (MMT C02 Eq. and Percent)
2023 Emission
Uncertainty Range Relative to Emission Estimate"
Source
Gas
Estimate
(MMT CO2 Eq.)b
(MMT CO2 Eq.)
(%)
Lower
Upper
Lower
Upper
Bound
Bound
Bound
Bound
Petroleum Systems
CH4
38.0
33.1
47.2
-13%
+24%
Petroleum Systems
CO2
23.3
19.0
28.5
-18%
+22%
Petroleum Systems
N2O
0.022
0.018
0.027
-18%
+22%
a Range of emission estimates estimated by applying the 95 percent confidence intervals obtained from the Monte Carlo
Energy 3-93
-------
Simulation analysis conducted for the year 2023 ChU and C02 emissions.
b All reported values are rounded after calculation. As a result, lower and upper bounds may not be duplicable from other
rounded values as shown in table.
QA/QC and Verification Discussion
In order to ensure the quality of the emission estimates for petroleum systems, general (IPCC Tier 1)
Quality Assurance/Quality Control (QA/QC) procedures were implemented consistent with the U.S.
Inventory QA/QC plan outlined in Annex 8.
The petroleum systems emission estimates in the Inventory are continually being reviewed and
assessed to determine whether emission factors and activity factors accurately reflect current industry
practices. A QA/QC analysis was performed for data gathering and input, documentation, and
calculation. QA/QC checks are consistently conducted to minimize human error in the emission
calculations. EPA performs a thorough review of information associated with new studies, GHGRP data,
regulations, public webcasts, and the Natural Gas STAR Program to assess whether the assumptions in
the Inventory are consistent with current industry practices. EPA has a multi-step data verification
process for GHGRP data, including automatic checks during data-entry, statistical analyses on
completed reports, and staff review of the reported data. Based on the results of the verification
process, EPA follows up with facilities to resolve mistakes that may have occurred.70
As in previous years, EPA conducted early engagement and communication with stakeholders on
updates prior to public review of the current Inventory. EPA held a stakeholder webinar on BOEM
offshore data updates and greenhouse gas data for oil and gas in November of 2024. EPA released
memos detailing updates under consideration and requesting stakeholder feedback. EPA then released
a final memorandum documenting the methodology implemented in the current Inventory.ln recent
years, several studies have measured emissions at the source level and at the national or regional level
and calculated emission estimates that may differ from the Inventory. There are a variety of potential
uses of data from new studies, including replacing a previous estimate or factor, verifying or QA of an
existing estimate or factor, and identifying areas for updates. In general, there are two major types of
studies related to oil and gas greenhouse gas data: studies that focus on measurement or quantification
of emissions from specific activities, processes, and equipment, and studies that use tools such as
inverse modeling to estimate the level of overall emissions needed to account for measured
atmospheric concentrations of greenhouse gases at various scales. The first type of study can lead to
direct improvements to or verification of Inventory estimates. In the past few years, EPA has reviewed,
and in many cases, incorporated data from these data sources. The second type of study can provide
general indications on potential over- and under-estimates.
A key challenge in using these types of studies to assess Inventory results is having a relevant basis for
comparison (e.g., the two data sets should have comparable time frames and geographic coverage, and
the independent study should assess data from the Inventory and not another data set, such as the
Emissions Database for Global Atmospheric Research, or "EDGAR"). In an effort to improve the ability to
compare the national-level Inventory with measurement results that may be at other spatial and
temporal scales, EPA has developed a gridded inventory of U.S. anthropogenic methane emissions with
0.1 degree x 0.1 degree spatial resolution, monthly temporal resolution, and detailed scale-dependent
70 See https://www.epa.gov/sites/prodijction/files/9015-07/documents/ghgrp verification factsheet.pdf.
3-94 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
error characterization.71 The most recent version of the gridded methane inventory is designed to be
consistent with the U.S. EPA's Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2018
estimates for the years 2012 through 2018. The gridded inventory improves efforts to compare results of
this Inventory with atmospheric studies.
As discussed above, refinery emissions are quantified by using the total emissions reported to GHGRP
for the refinery emission categories included in petroleum systems. Subpart Y has provisions that
refineries are not required to report under Subpart Y if their emissions fall below certain thresholds.
Each year, a review is conducted to determine whether an adjustment is needed to the Inventory
emissions to include emissions from refineries that stopped reporting to the GHGRP. Based on the
review of the most recent GHGRP data, EPA did not identify any additional refineries that would require
gap filling. There are a total of 6 refineries that EPA previously identified (i.e., during the 1990 through
2022 Inventory and prior versions) as not reporting to the GHGRP and continued to gap fill annual
emissions for these refineries. EPA used the last reported emissions (by source) for these refineries as
proxy to gap fill annual emissions.
Recalculations Discussion
EPA received information and data related to the emission estimates through GHGRP reporting and
presented information to stakeholders regarding the updates under consideration. In December 2024,
EPA released a draft memorandum that discussed changes under consideration and requested
stakeholder feedback on those changes. EPA then released a final memorandum documenting the
methodology implemented in the current Inventory.72 The memorandum cited in the Recalculations
Discussion below is: Inventory of U.S. Greenhouse Gas Emissions and Sinks 1990-2023: Updates to Use
New Offshore Data (Offshore Production memo), presented information to stakeholders regarding the
updates under consideration. In December 2024, EPA released a draft memorandum that discussed
changes under consideration and requested stakeholder feedback on those changes. EPA then released
a final memorandum documenting the methodology implemented in the current Inventory.73 The
memorandum cited in the Recalculations Discussion below is: Inventory of U.S. Greenhouse Gas
Emissions and Sinks 1990-2023: Updates to Use New Offshore Data (Offshore Production memo).
EPA evaluated relevant information available and made updates to the Inventory for offshore production
sources in Gulf of America (GOA) federal and state waters. General information for these source specific
recalculations is presented below and details are available in the Offshore Production memo.
In addition to the updates to the offshore production sources mentioned above, for certain sources, CH4
and/or C02 emissions changed by greater than 0.05 MMT C02 Eq., comparing the previous estimate for
2022 to the current (recalculated) estimate for 2022. The emissions changes were mostly due to GHGRP
data submission revisions and updated Enverus data. These sources are discussed below and include
pneumatic controllers, chemical injection pumps, produced water, production storage tanks,
miscellaneous production flaring, and refinery flaring.
The combined impact of revisions to 2022 petroleum systems CH4 emission estimates on a C02-
equivalent basis, compared to the previous Inventory, is a decrease from 39.6 to 36.3 MMT C02 Eq. (3.4
71 See https://www.epa.gov/ghgemissionsAjs-gridded-methane-emissions.
73 Stakeholder materials including draft and final memoranda for the current (i.e., 1990 to 2023) Inventory are available at
https://www.epa.gov/ghgemissions/natural-gas-and-petroleum-svstems.
Energy 3-95
-------
MMT C02 Eq., or 8 percent). The recalculations resulted in lower CH4 emission estimates on average
across the 1990 through 2022 time series, compared to the previous Inventory, by 0.08 MMT C02 Eq., or
0.2 percent.
The combined impact of revisions to 2022 petroleum systems C02 emission estimates, compared to the
previous Inventory, is a slight increase from 21.97 to 22.08 MMT C02 (0.12 MMTC02, or 0.5 percent). The
recalculations resulted in higher emission estimates on average across the 1990 through 2022 time
series, compared to the previous Inventory, by less than 0.005 MMT C02 Eq., or less than 0.1 percent.
The combined impact of revisions to 2022 petroleum systems N20 emission estimates on a C02-
equivalent basis, compared to the previous Inventory, is an increase of 0.001 MMT C02, Eq. or 1 percent.
The recalculations resulted in an average increase in emission estimates across the 1990 through 2022
time series, compared to the previous Inventory, of 0.001 MMT C02 Eq., or 9 percent.
Table 3-52and Table 3-53 below are categories in petroleum systems with updated methodologies or
with recalculations resulting in a change of greater than 0.05 MMT C02 Eq., comparing the previous
estimate for 2022 to the current (recalculated) estimate for 2022. For more information, please see the
discussion below.
Table 3-52: Recalculations of C02 in Petroleum Systems (MMT C02)
Segment/Source
Previous Estimate
Year 2022,
2024 Inventory
Current Estimate
Year 2022,
2025 Inventory
Current Estimate
Year 2023,
2025 Inventory
Exploration
0.3
0.3
0.5
Production
18.8
18.9
19.9
Tanks
4.5
4.6
4.6
Miscellaneous Production Flaring
5.0
5.1
6.1
Offshore Production - GOA Federal Waters
+
+
+
Offshore Production - GOA State Waters
+
+
+
Transportation
+
+
+
Refining
2.9
2.8
2.9
Flares
2.8
2.8
2.8
Petroleum Systems Total
22.0
22.1
23.3
+ Does not exceed 0.05 MMT C02 Eq.
Note: Totals may not sum due to independent rounding.
Table 3-53: Recalculations of CH4
in Petroleum Systems (MMT C02 Eq.)
Segment/Source
Previous Estimate
Year2022,
2024 Inventory
Current Estimate
Year 2022,
2025 Inventory
Current Estimate
Year 2023,
2025 Inventory
Exploration
0.1
0.1
0.1
Production
38.6
35.2
37.0
Pneumatic Controllers
19.4
18.5
16.3
Chemical Injection Pumps
2.2
2.1
1.8
Produced Water
2.7
2.4
2.5
Offshore Production - GOA Federal Waters
4.6
2.4
2.4
Offshore Production - GOA State Waters
+
+
+
Transportation
0.2
0.2
0.3
3-96 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Previous Estimate
Current Estimate
Current Estimate
Year2022,
Year 2022,
Year 2023,
Segment/Source
2024 Inventory
2025 Inventory
2025 Inventory
Refining
0.7
0.7
0.7
Petroleum Systems Total
39.6
36.3
38.0
+ Does not exceed 0.05 MMT C02 Eq.
Note: Totals may not sum due to independent rounding.
Exploration
There were no methodological updates for exploration and recalculations for the exploration segment
have resulted in minor changes in calculated CH4 and C02 emissions over the time series. Methane
emissions have decreased by an average of 0.1 percent and C02 emissions have increased by an
average of 0.1 percent across the time series, compared to the previous Inventory.
Production
Offshore Production in Gulf of America (Methodological Update)
EPA updated the calculation methodology for offshore production in the Gulf of America (GOA) to use
new emission factors calculated from year 2021 data from the Bureau of Ocean Energy Management
(BOEM). Offshore production in the GOA occurs in two areas, federal waters and state waters. BOEM
provides periodic emission inventories which account for emissions specific to GOA federal waters
production. Previously, year 2017 BOEM data were the most recent that were incorporated into the
Inventory's calculation methodology. EPA previously applied emission source-specific emission factors
calculated from BOEM's 2017 dataset to calculate emissions for all years from 2016 - 2022 for GOA
federal waters. EPA then calculated GOA state waters emissions using the federal waters emissions,
assuming the emissions were equivalent on a production basis. With the release of the BOEM 2021
dataset, EPA calculated new emission source-specific emission factors. EPA applied the same
approach to calculate emission factors from the 2021 BOEM dataset as it did for the prior BOEM
datasets. EPA applied the emission factors calculated from the BOEM 2021 dataset for years 2020 -
2023, maintained the emission factors from the BOEM 2017 dataset for 2016 - 2018, and calculated
emission factors that average both BOEM datasets together for year 2019. This update impacts sources
of vent and leak emissions only, flaring emissions are not affected. Details and additional
considerations for this update are available in the Offshore Production memo.
As a result of this methodological update, CH4 emissions estimates for offshore production in the GOA
are on average 40 percent lower for 2019 to 2022 compared to the previous Inventory. The 2022 CH4
emissions estimate is 48 percent lower than in the previous Inventory. The update resulted in C02
emissions estimates for offshore production in the GOA that are on average 30 percent lower for 2019 to
2022 compared to the previous Inventory. The 2022 C02 emissions estimate is 36 percent lower than in
the previous Inventory. This methodological update impacted CH4 and C02 estimates for 2019 to 2022,
compared to the previous Inventory. The methodological update did not impact emissions for years
prior to 2019; differences in emissions compared to the previous Inventory for years prior to 2019 are
due to changes in underlying activity data (e.g., number of offshore complexes, oil and gas production).
Energy 3-97
-------
Table 3-54: GOA Offshore Production Vent and Leak National CH4 Emissions (Metric
Tons CH4)
Source
1990
2005
2019
2020
2021
2022
2023
GOA Federal Waters - Major Complexes
271,602
187,140
119,930
76,088
73,303
71,609
72,838
GOA Federal Waters - Minor Complexes
31,672
32,356
19,204
15,572
14,580
14,210
14,533
GOA State Waters
26,1981
2,861 |
842
484
385
454
424
Total Emissions
329,472
| 222,357
| 139,976
92,144
88,267
86,272
87,795
Previous Estimate
308,5431
| 219,8931
| 178,558
167,001
165,720
164,395
NA
NA (Not Applicable)
Note: Totals may not sum due to independent rounding.
Table 3-55: GOA Offshore Production Vent and Leak National C02 Emissions (Metric
Tons C02)
Source
1990
2005
2019
2020
2021
2022
2023
GOA Federal Waters - Major Complexes
3,010
2,299
1,624
902
869
848
863
GOA Federal Waters - Minor Complexes
402
411
489
682
639
622
637
GOA State Waters
2951
35
13
8
7
8
7
Total Emissions
3,707
2,745
2,125
1,592
1,514
1,479
1,507
Previous Estimate
3,7021
2,741
2,525
2,359
2,341
2,324
NA
NA (Not Applicable)
Note: Totals may not sum due to independent rounding.
Pneumatic Controllers (Recalculation with Updated Data)
Methane emissions from onshore production pneumatic controllers are on average 0.1 percent lower
across the time series and 5 percent lower in 2022, compared to the previous Inventory. The emission
changes were due to GHGRP data submission revisions and updated oil well counts.
Table 3-56: Pneumatic Controllers National CH4 Emissions (Metric Tons CH4)
Source
1990
2005
2019
2020
2021
2022
2023
High Bleed Controllers
709,646
483,896
73,071
87,173
45,577
24,282
14,184
Low Bleed Controllers
51,050
62,291
49,997
36,751
45,991
35,659
31,292
Intermittent Bleed Controllers
NO |
267,9081
744,789
975,122
875,857
602,017
535,139
Total Emissions
760,696
814,0951
867,857
1,099,046
967,425
661,958
580,615
Previous Estimate
760,9251
811,142 j
881,203
1,119,352
1,003,063
693,551
NA
NA (Not Applicable)
Note: Totals may not sum due to independent rounding.
Chemical Injection Pumps (Recalculation with Updated Data)
Methane emissions from chemical injection pumps are on average 0.2 percent lower across the time
series and 5 percent lower in 2022, compared to the previous Inventory. The emission changes were due
to GHGRP data submission revisions and updated oil well counts.
3-98 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Table 3-57: Chemical Injection Pump National CH4 Emissions (Metric Tons CH4)
Source
1990
20051
2019
2020
2021
2022
2023
Chemical Injection Pumps
47,480
80,259
117,570
93,396
82,592
75,695
64,163
Previous Estimate
47,425
79,968\
122,967
96,186
85,494
79,712
NA
NA (Not Applicable)
Produced Water (Recalculation with Updated Data)
Methane emissions from produced water are on average 0.6 percent lower across the time series and 10
percent lower in 2022, compared to the previous Inventory. The emission changes were due to updated
produced water volumes.
Table 3-58: Produced Water National CH4 Emissions (Metric Tons CH4)
Source
1990
2005
2019
2020
2021
2022
2023
Produced Water - Regular
Pressure Wells
71,923
49,898
77,459
70,460
62,718
66,488
70,204
Produced Water - Low Pressure
Wells
20,502
14,224
22,080
20,085
17,878
18,953
20,012
Total Emissions
92,425
64,122
99,539
90,545
80,596
85,440
90,216
Previous Estimate
92,336
64,047
99,425
90,435
92,201
94,663
NA
NA (Not Applicable)
Note: Totals may not sum due to independent rounding.
Storage Tanks (Recalculation with Updated Data)
Carbon dioxide emissions from production storage tanks are on average 1 percent higher across the
time series, compared to the previous Inventory. Carbon dioxide emission estimates for 2022 are 1
percent higher than in the previous Inventory, which is primarily due to large tanks with flares. The
emission changes were due to updated oil production volumes.
Table 3-59: Storage Tanks National C02 Emissions (kt C02)
Source
1990
2005
2019
2020
2021
2022
2023
La rge Ta n ks w/Fla res
0
718
6,213
5,805
5,581
4,567
4,594
La rge Ta n ks w/VRU
0
3
9
2
1
1
0
Large Tanks w/o Control
24
8
9
5
4
2
3
Small Tanks w/Fla res
0
3
9
10
10
11
10
Small Tanks w/o Flares
12
5
4
4
5
5
4
Malfunctioning Separator
Dump Valves
12
13
26
21
34
8
11
Total Emissions
48
750
6,270
5,848
5,636
4,593
4,622
Previous Estimate
47
748
6,024
5,255
5,439
4,539
NA
NA (Not Applicable)
Note: Totals may not sum due to independent rounding.
Miscellaneous Production Flaring (Recalculation with Updated Data)
Carbon dioxide emissions from miscellaneous production flaring are on average 0.1 percent higher
across the time series and 1 percent higher in 2022, compared to the previous Inventory. The emission
changes were due to updated oil production volumes.
Energy 3-99
-------
Table 3-60: Miscellaneous Production Flaring National C02 Emissions (kt C02)
Source
1990
2005
2019
2020
2021
2022
2023
220 - Gulf Coast Basin (LA, TX)
0
103
608
652
802
656
1,069
395-Williston Basin
0
71
3,049
1,307
1,313
1,241
1,232
430 - Permian Basin
0
215
4,315
2,728
2,159
2,767
3,391
"Other" Basins
0
398
704
424
370
428
407
Total Emissions
0
787
8,677
5,112
4,644
5,092
6,100
Previous Estimate
0
786
8,678
5,110
4,638
5,028
NA
NA (Not Applicable)
Note: Totals may not sum due to independent rounding.
Transportation
Recalculations for the transportation segment have resulted in calculated CH4 and C02 emissions over
the time series from this segment that are lower (by less than 0.1 percent) than in the previous Inventory.
Refining
Recalculations due to resubmitted GHGRP data in the refining segment have resulted in an increase in
calculated CH4 emissions by an average of 3.7 percent across the time series and a decrease of 0.6
percent in 2022, compared to the previous Inventory.
Refining C02 emission estimates decreased by an average of 0.1 percent across the time series and
decreased by 1.8 percent in 2022, compared to the previous Inventory. This change in emissions is due
to GHGRP resubmissions and was largely due to a change in reported flaring C02 emissions.
Table 3-61: Refining National CH4 Emissions (Metric Tons CH4)
Source
1990
1
2005 | 2019
2020
2021
2022
2023
Refining
26,774
30,3891
30,779
25,794
25,299
24,529
23,579
Previous Estimate
25,742
29,218
30,814
25,861
25,366
24,685
NA
NA (Not Applicable)
Table 3-62: Refining National CO;
>
Emissions (kt C02)
Source
1990
2005
2019
2020
2021
2022
2023
Flares
3,023 I
3,431
3,512
2,840
2,969
2,784
2,829
Total Refining
3,174
3,602
3,560
2,874
3,001
2,820
2,862
Previous Estimate
3,174 |
3,602
3,571
2,893
3,021
2,872
NA
NA (Not Applicable)
Planned Improvements
Planned Improvements for 2025 Inventory
EPA updated oil and gas well counts and oil and gas production for this 2025 Inventory using Enverus
data. However, EPA did not update the number of completion events, due to significant changes in the
data across the time series. EPA will assess the underlying Enverus data to develop an appropriate
methodology to determine the number of completions for each year of the time series.
3-100 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Upcoming Data, and Additional Data that Could Inform the Inventory
EPA will assess new data received by the Greenhouse Gas Reporting Program and other relevant
programs on an ongoing basis, which may be used to confirm or improve existing estimates and
assumptions. In December 2024, EPA released a memorandum discussing updates under consideration
for a future Inventory to incorporate revised GHGRP subpart W emission factors and requested
stakeholder feedback (Inventory of U.S. Greenhouse Gas Emissions and Sinks: Updates Under
Consideration to Use Revised Subpart WEmission Factors).74 One commenter provided feedback on
the potential subpart W-based revisions. The commenter had concerns with using the revised subpart
W equipment leak emission factors though the commenter supported incorporating leaker survey data
into the Inventory's equipment leaks methodology.
EPA continues to track studies that contain data that may be used to update the Inventory. EPA will also
continue to assess studies that include and compare both top-down and bottom-up estimates, and
which could lead to improved understanding of unassigned high emitters (e.g., identification of emission
sources and information on frequency of high emitters) as recommended in previous stakeholder
comments.
3.7 Natural Gas Systems (Source Category
1B2b)
The U.S. natural gas system encompasses hundreds of thousands of wells, hundreds of processing
facilities, and over a million miles of transmission and distribution pipelines. This category (1 B2b) as
defined in the IPCC methodological guidance is for fugitive emissions from natural gas systems, which
per IPCC guidelines include emissions from leaks, venting, and flaring. Total greenhouse gas emissions
(CH4, C02, and N20) from natural gas systems in 2023 were 200.1 MMT C02 Eq., a decrease of 21
percent from 1990 and a decrease of 4 percent from 2022, both primarily due to decreases in CH4
emissions. From 2011, emissions decreased by 9 percent, primarily due to decreases in CH4 emissions.
National total dry gas production in the United States increased by 113 percent from 1990 to 2023,
increased by 4 percent from 2022 to 2023, and increased by 65 percent from 2011 to 2023. Of the overall
greenhouse gas emissions (200.1 MMT C02 Eq.), 81 percent are CH4 emissions (162.4 MMT C02 Eq.), 19
percent are C02 emissions (37.7 MMT), and less than 0.1 percent are N20 emissions (0.01 MMT C02
Eq.).
Overall, natural gas systems emitted 162.4 MMT C02 Eq. (5,802 kt CH4) of CH4 in 2023, a 26 percent
decrease compared to 1990 emissions, and 6 percent decrease compared to 2022 emissions (see Table
3-64 and Table 3-65). For non-combustion C02, a total of 37.7 MMT C02 Eq. (37,682 kt) was emitted in
2023, a 16 percent increase compared to 1990 emissions, and a 3 percent increase compared to 2022
levels. The 2023 N20 emissions were estimated to be 0.01 MMT C02 Eq. (0.03 kt N20), a 73 percent
increase compared to 1990 emissions, and a 55 percent decrease compared to 2022 levels.
74 The memo is available online: https://www.epa.gov/ghgemissions/stakeholder-process-natijral-gas-and-petroleum-
svstems-1990-?0?3-inventory
Energy 3-101
-------
The 1990 to 2023 emissions trend is not consistent across segments or gases. Overall, the 1990 to 2023
decrease in CH4 emissions is due primarily to the decrease in emissions from the following segments:
distribution (70 percent decrease), transmission and storage (42 percent decrease), processing (36
percent decrease), and exploration (98 percent decrease). Over the same time period, the production
segment saw increased CH4 emissions of 23 percent (with onshore production emissions increasing 1
percent, offshore production emissions decreasing 98 percent, and gathering and boosting [G&B]
emissions increasing 90 percent), and post-meter emissions increasing by 70 percent. The 1990 to 2023
increase in C02 emissions is primarily due to an increase in C02 emissions in the production segment,
where emissions from flaring have increased over time.
Methane and C02 emissions from natural gas systems include those resulting from normal operations,
routine maintenance, and system upsets. Emissions from normal operations include natural gas engine
and turbine uncombusted exhaust, flaring, and leak emissions from system components. Routine
maintenance emissions originate from pipelines, equipment, and wells during repair and maintenance
activities. Pressure surge relief systems and accidents can lead to system upset emissions. Emissions
of N20 from flaring activities are included in the Inventory, with most of the emissions occurring in the
processing and production segments. Note, C02 emissions exclude all combustion emissions (e.g.,
engine combustion) except for flaring C02 emissions. All combustion C02 emissions (except for flaring)
are accounted for in Section 3.1.
Each year, some estimates in the Inventory are recalculated with improved methods and/or data. These
improvements are implemented consistently across the previous Inventory's time series (i.e., 1990 to
2023) to ensure that the trend is representative of changes in emissions. Recalculations in natural gas
systems in this year's Inventory include:
• Updates to oil and gas well counts, oil and gas production volumes, and produced water
production volumes using the most recent data from Enverus.
• Methodological updates for offshore production in the Gulf of America.
• Recalculations due to Greenhouse Gas Reporting Program (GHGRP) submission revisions.
The Recalculations Discussion section below provides more details on the updated methods.
Below is a characterization of the six emission subcategories of natural gas systems: exploration,
production (including gathering and boosting), processing, transmission and storage, distribution, and
post-meter. Each of the segments is described and the different factors affecting CH4, C02, and N20
emissions are discussed.
Exploration. Exploration includes well drilling, testing, and completion. Emissions from exploration
accounted for 0.1 percent of CH4 emissions and 0.1 percent of C02 emissions from natural gas systems
in 2023. Well completions accounted for approximately 87 percent of CH4 emissions from the
exploration segment in 2023, with the rest resulting from well testing and drilling. Well completion flaring
emissions account for most of the C02 emissions. Methane emissions from exploration decreased by 98
percent from 1990 to 2023, with the largest decreases coming from hydraulically fractured gas well
completions without reduced emissions completions (RECs). Methane emissions from exploration
decreased 25 percent from 2022 to 2023 due to decreases in emissions from hydraulically fractured well
completions (both non-REC with flaring and REC with venting). Methane emissions from exploration
were highest from 2006 to 2008. Carbon dioxide emissions from exploration decreased by 94 percent
from 1990 to 2023 primarily due to decreases in hydraulically fractured gas well completions. Carbon
3-102 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
dioxide emissions from exploration decreased by 2 percent from 2022 to 2023 due to decreases in
emissions from hydraulically fractured gas well completions (REC with flaring) and non-hydraulically
fractured gas well completions (vented). Carbon dioxide emissions from exploration were highest from
2006 to 2008. Nitrous oxide emissions from exploration decreased 96 percent from 1990 to 2023 and
decreased 28 percent from 2022 to 2023.
Production (includinggathering and boosting). In the production segment, wells are used to withdraw raw
gas from underground formations. Emissions arise from the wells themselves, and from well-site
equipment and activities such as pneumatic controllers, tanks and separators, and liquids unloading.
Gathering and boosting emission sources are included within the production sector. The gathering and
boosting sources include gathering and boosting stations (with multiple emission sources on site) and
gathering pipelines. The gathering and boosting stations receive natural gas from production sites and
transfer it, via gathering pipelines, to transmission pipelines or processing facilities (custody transfer
points are typically used to segregate sources between each segment). Boosting processes include
compression, dehydration, and transport of gas to a processing facility or pipeline. Emissions from
production (including gathering and boosting) accounted for 50 percent of CH4 emissions and 26 percent
of C02 emissions from natural gas systems in 2023. Emissions from gathering and boosting and
pneumatic controllers in onshore production accounted for most of the production segment CH4
emissions in 2023. Within gathering and boosting, the largest sources of CH4 are compressor exhaust slip,
compressor venting and leaks, and tanks. Flaring emissions account for most of the C02 emissions from
production, with the highest emissions coming from flare stacks at gathering stations, miscellaneous
onshore production flaring, and tank flaring. Methane emissions from production increased by 23 percent
from 1990 to 2023, due primarily to increases in emissions from pneumatic controllers (due to an increase
in the number of controllers, particularly in the number of intermittent bleed controllers) and increases in
emissions from compressor exhaust slip in gathering and boosting. Methane emissions from production
decreased 10 percent from 2022 to 2023 due to decreases in emissions from well pad equipment leaks
(compressors) and pneumatic controllers. Carbon dioxide emissions from production increased by
approximately a factor of 2.9 from 1990 to 2023 due to increases in emissions at flare stacks in gathering
and boosting and miscellaneous onshore production flaring and increased 12 percent from 2022 to 2023
due primarily to increases in emissions at flare stacks at gathering and boosting stations and in
miscellaneous onshore production flaring and tank venting. Nitrous oxide emissions from production
decreased by 6 percent from 1990 to 2023 due to decreases in emissions from dehydrator units at
gathering and boosting stations and decreased 56 percent from 2022 to 2023 due to decreases in
emissions from flare stacks at gathering and boosting stations.
Processing. In the processing segment, natural gas liquids and various other constituents from the raw
gas are removed, resulting in "pipeline quality" gas, which is injected into the transmission system.
Methane emissions from compressors, including compressor seals, are the primary emission source
from this stage. Most of the C02 emissions come from acid gas removal (AGR) units, which are designed
to remove C02 from natural gas. Processing plants accounted for 9 percent of CH4 emissions and 71
percent of C02 emissions from natural gas systems. Methane emissions from processing decreased by
36 percent from 1990 to 2023 as emissions from compressors (leaks and venting) and equipment leaks
decreased; and increased 3 percent from 2022 to 2023 due to increased emissions from gas engines.
Carbon dioxide emissions from processing decreased by 5 percent from 1990 to 2023, due to a
decrease in AGR emissions, and increased 1 percent from 2022 to 2023 due to increased AGR
emissions. Nitrous oxide emissions decreased 53 percent from 2022 to 2023 due to decreased
emissions from flares at gas processing plants.
Energy 3-103
-------
Transmission and Storage. Natural gas transmission involves high pressure, large diameter pipelines
that transport gas long distances from field production and processing areas to distribution systems or
large volume customers such as power plants or chemical plants. Compressor station facilities are
used to move the gas throughout the U.S. transmission system. Leak CH4 emissions from these
compressor stations and venting from pneumatic controllers account for most of the emissions from
this stage. Uncombusted compressor engine exhaust and pipeline venting are also sources of CH4
emissions from transmission. Natural gas is also injected and stored in underground formations, or
liquefied and stored in above ground tanks, during periods of low demand (e.g., summer), and
withdrawn, processed, and distributed during periods of high demand (e.g., winter). Leak and venting
emissions from compressors are the primary contributors to CH4 emissions from storage. Emissions
from liquefied natural gas (LNG) stations and terminals are also calculated under the transmission and
storage segment. Methane emissions from the transmission and storage segment accounted for
approximately 23 percent of methane emissions from natural gas systems, while C02 emissions from
transmission and storage accounted for 3 percent of the C02 emissions from natural gas systems. CH4
emissions from this source decreased by 42 percent from 1990 to 2023 due to reduced pneumatic
device and compressor station emissions (including emissions from compressors and leaks) and
decreased 6 percent from 2022 to 2023 due to decreased emissions from pipeline venting and
transmission compressors. C02 emissions from transmission and storage were 6.4 times higher in 2023
than in 1990, due to increased emissions from LNG export terminals, and increased by 4 percent from
2022 to 2023, due to increased emissions from LNG stations. The quantity of LNG exported from the
United States increased by a factor of 83 from 1990 to 2023, and by 12 percent from 2022 to 2023. LNG
emissions are about 2 percent of CH4 and 86 percent of C02 emissions from transmission and storage in
year 2023. Nitrous oxide emissions from transmission and storage increased by 68 percent from 1990 to
2023 and decreased by 66 percent from 2022 to 2023.
Distribution. Distribution pipelines take the high-pressure gas from the transmission system at "city
gate" stations, reduce the pressure and distribute the gas through primarily underground mains and
service lines to individual end users. There were 1,364,070 miles of distribution mains in 2023, an
increase of 419,913 miles since 1990 (PHMSA 2024). Distribution system emissions, which accounted
for 9 percent of CH4 emissions from natural gas systems and less than 0.1 percent of C02 emissions
from natural gas systems, result mainly from leak emissions from pipelines and stations. An increased
use of plastic piping, which has lower emissions than other pipe materials, has reduced both CH4 and
C02 emissions from this stage, as have station upgrades at metering and regulating (M&R) stations.
Distribution system CH4 emissions in 2023 were 70 percent lower than 1990 levels and less than 1
percent lower than 2022 emissions. Distribution system C02 emissions in 2023 were 70 percent lower
than 1990 levels and less than 1 percent lower than 2022 emissions. Annual C02 emissions from this
segment are less than 0.1 MMT C02 Eq. across the time series.
Post-Meter. Post-meter includes leak emissions from residential and commercial appliances, industrial
facilities and power plants, and natural gas fueled vehicles. Leak emissions from residential appliances
and industrial facilities and power plants account for the majority of post-meter CH4 emissions. Methane
emissions from the post-meter segment accounted for approximately 8 percent of emissions from natural
gas systems in 2023. Post-meter CH4 emissions increased by 70 percent from 1990 to 2023 and increased
by 3 percent from 2022 to 2023, due to increases in the number of residential houses using natural gas and
increased natural gas consumption at industrial facilities and power plants. C02 emissions from post-
meter account for less than 0.01 percent of total C02 emissions from natural gas systems.
3-104 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Total greenhouse gas emissions from the six subcategories within natural gas systems are shown in
MMT C02 Eq. in Table 3-63. Total CH4 emissions for these same segments of natural gas systems are
shown in MMT C02 Eq. (Table 3-64) and kt (Table 3-65). Most emission estimates are calculated using a
net emission approach. However, a few sources are still calculated with a potential emission approach.
Reductions data are applied to those sources. In 2023, 2.6 MMT C02 Eq. CH4 is subtracted from
production segment emissions, 4.3 MMT C02 Eq. CH4 is subtracted from the transmission and storage
segment, and 0.1 MMT C02 Eq. CH4 is subtracted from the distribution segment to calculate net
emissions. More disaggregated information on potential emissions, net emissions, and reductions data
is available in Annex 3.6, Methodology for Estimating CH4 and C02 Emissions from Natural Gas Systems.
Table 3-63: Total Greenhouse Gas Emissions (CH4, C02, and N20) from Natural Gas
Systems (MMT C02 Eq.)
Segment
19901
2005
2019
2020
2021
2022
2023
Exploration
7.31
22.5
2.3
0.3
0.2
0.2
0.2
Production
69.31
98.3
114.9
105.4
101.4
98.3
90.5
Processing
52.21
31.8
40.4
39.5
39.7
41.4
42.0
Transmission and Storage
64.21
46.3
41.8
43.1
40.6
40.7
38.4
Distribution
51.01
28.5
15.5
15.5
15.3
15.3
15.3
Post-Meter
8.1 |
9.6
12.8
13.0
13.0
13.4
13.8
Total
252.11
237.0
227.7
216.9
210.4
209.3
200.1
Note: Totals may not sum due to independent rounding.
Table 3-64: CH4 Emissions from Natural Gas Systems (MMT C02
Eq.)
Segment
1990
2005
2019
2020
2021
2022
2023
Exploration
6.7
19.7
2.1
0.2
0.1
0.2
0.1
Production
65.9
93.7
103.9
96.3
92.1
89.7
80.8
Onshore Production
40.0
64.8
58.2
53.6
50.0
48.6
40.5
Gathering and Boosting
21.1
26.9
45.3
42.5
42.0
41.0
40.2
Offshore Production
4.8
2.0
0.5
0.1
0.1
0.1
0.1
Processing
23.9
13.0
14.2
14.0
14.2
14.8
15.2
Transmission and Storage
64.0
46.1
40.6
41.1
39.8
39.6
37.3
Distribution
50.9
28.5
15.5
15.5
15.3
15.2
15.2
Post-Meter
8.1
9.6
12.8
13.0
13.0
13.4
13.8
Total
219.6
210.7
189.0
180.1
174.6
172.8
162.4
Note: Totals may not sum due to independent rounding.
Table 3-65: CH4 Emissions from Natural Gas Systems (kt)
Segment
1990
2005
2019
2020
2021
2022
2023
Exploration
239
705
75
7
5
6
5
Production
2,354
3,348
3,711
3,438
3,289
3,203
2,886
Onshore Production
1,429
2,314
2,079
1,914
1,785
1,736
1,447
Gathering and Boosting
755
960
1,616
1,519
1,500
1,463
1,436
Offshore Production
170
73
16
5
4
3
3
Processing
853
463
506
501
508
529
544
Transmission and Storage
2,286
1,646
1,448
1,468
1,421
1,413
1,330
Distribution
1,819
1,018
554
553
547
544
544
Post-Meter
290
344
457
464
465
478
492
Energy 3-105
-------
Total
7,842
7,525
6,751
6,431
6,236
6,173
5,802
Note: Totals may not sum due to independent rounding.
Table 3-66: C02 Emissions from Natural Gas Systems (MMT)
Segment
19901
2005
2019
2020
2021
2022
2023
Exploration
0.6
2.7
0.2
0.1
0.0
+
+
Production
3.3
4.6
11.0
9.2
9.3
8.6
9.7
Processing
28.3
18.81
26.2
25.5
25.5
26.6
26.8
Transmission and Storage
0.2
0.2
1.2
2.0
0.9
1.1
1.2
Distribution
0.1
+1
+
+
+
+
+
Post-Meter
+
+
+
+
+
+
+
Total
32.51
26.31
38.7
36.8
35.7
36.4
37.7
+ Does not exceed 0.05 MMT C02 Eq.
Note: Totals may not sum due to independent rounding.
Table 3-67: C02 Emissions from Natural Gas Systems (kt)
Segment
19901
2005
2019
2020
2021
2022
2023
Exploration
619
2,708
249
97
38
37
36
Production
3,332
4,562
11,000
9,173
9,330
8,648
9,686
Processing
28,338
18,836
26,184
25,494
25,502
26,588
26,781
Transmission and Storage
182
189
1,244
2,028
857
1,118
1,160
Distribution
54
30
16
16
16
16
16
Post-Meter
1 I
1 I
2
2
2
2
2
Total
32,5251
26,3251
38,696
36,810
35,745
36,410
37,682
Note: Totals may not sum due to independent rounding.
Table 3-68: N2Q Emissions from Natural Gas Systems (Metric Tons C02
Eq.)
Segment
19901
2005
2019
2020
2021
2022
2023
Exploration
518
1
1,707
114
46
19
27
19
Production
3,983
5,204
5,098
3,737
3,955
8,385
3,729
Processing
NO
2,977
5,088
4,367
4,098
8,672
4,033
Transmission and Storage
229
280
563
941
399
1,142
384
Distribution
NO
NO
NO
NO
NO
NO
NO
Post-Meter
NO |
NO
NO
NO
NO
NO
NO
Total
4,7301
10,169
10,863
9,091
8,471
18,227
8,165
NO (Not Occurring)
Note: Totals may not sum due to independent rounding.
3-106 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Table 3-69: N20 Emissions from Natural Gas Systems (Metric Tons N2Q)
Segment
1990
2005
2019
2020
2021
2022
2023
Exploration
2.0
6.4
0.4
0.2
0.1
0.1
0.1
Production
15.0
19.6
19.2
14.1
14.9
31.6
14.1
Processing
NO
11.2
19.2
16.5
15.5
32.7
15.2
Transmission and Storage
0.9
1.1
2.1
3.6
1.5
4.3
1.4
Distribution
NO
NO
NO
NO
NO
NO
NO
Post-Meter
NO
NO
NO
NO
NO
NO
NO
Total
17.9
38.4
41.0
34.3
32.0
68.8
30.8
NO (Not Occurring)
Note: Totals may not sum due to independent rounding.
Methodology and Time-Series Consistency
See Annex 3.6 for the full time series of emissions data, activity data, and emission factors, and
additional information on methods and data sources—for example, the specific years of reporting data
from EPA's GHGRP that are used to develop certain factors.
This section provides a general overview of the methodology for natural gas system emission estimates
in the Inventory, which involves the calculation of CH4, C02, and N20 emissions for over 100 emissions
sources (i.e., equipment types or processes), and then the summation of emissions for each natural gas
segment.
The approach for calculating emissions for natural gas systems generally involves the application of
emission factors to activity data. For most sources, the approach uses technology-specific emission
factors or emission factors that vary over time and take into account changes to technologies and
practices, which are used to calculate net emissions directly. For others, the approach uses what are
considered "potential methane factors" and emission reduction data to calculate net emissions. The
estimates are developed with an IPCC Tier 2 approach; Tier 1 approaches are not used.
Emission Factors. Key references for emission factors for CH4 and C02 emissions from the U.S. natural
gas industry include a 1996 study published by the Gas Research Institute (GRI) and EPA (GRI/EPA
1996), EPA's GHGRP (EPA 2024), and others.
The 1996 GRI/EPA study developed over 80 CH4 emission factors to characterize emissions from the
various components within the operating segments of the U.S. natural gas system. The GRI/EPA study
was based on a combination of process engineering studies, collection of activity data, and
measurements at representative natural gas facilities conducted in the early 1990s. Year-specific
natural gas CH4 compositions are calculated using U.S. Department of Energy's Energy Information
Administration (EIA) annual gross production data for National Energy Modeling System (NEMS) oil and
gas supply module regions in conjunction with data from the Gas Technology Institute (GTI, formerly
GRI) Unconventional Natural Gas and Gas Composition Databases (GTI 2001). These year-specific CH4
compositions are applied to emission factors, which therefore may vary from year to year due to slight
changes in the CH4 composition of natural gas for each NEMS region.
GHGRP Subpart W data were used to develop CH4, C02, and N20 emission factors for many sources in
the Inventory. In the exploration and production segments, GHGRP data were used to develop emission
factors used for all years of the time series for well testing, gas well completions and workovers with and
Energy 3-107
-------
without hydraulic fracturing, pneumatic controllers and chemical injection pumps, condensate tanks,
liquids unloading, miscellaneous flaring, gathering and boosting pipelines, and certain sources at
gathering and boosting stations. In the processing segment, for recent years of the time series, GHGRP
data were used to develop emission factors for leaks, compressors, flares, dehydrators, and
blowdowns/venting. In the transmission and storage segment, GHGRP data were used to develop
factors for all years of the time series for LNG stations and terminals and transmission pipeline
blowdowns, and for pneumatic controllers for recent years of the time series.
Other data sources used for CH4emission factors include Zimmerle et al. (2015) for transmission and
storage station leaks and compressors, GTI (2009 and 2019) for commercial and industrial meters,
Lamb et al. (2015) for recent years for distribution pipelines and meter/regulator stations, Zimmerle et
al. (2019) for gathering and boosting stations, Bureau of Ocean Energy Management (BOEM) reports,
and Fischer et al. (2018) and IPCC (2019) for post-meter emissions.
For C02 emissions from sources in the exploration, production, and processing segments that use
emission factors not directly calculated from GHGRP data, data from the 1996 GRI/EPA study and the
2001 GTI publication were used to adapt the CH4 emission factors into related C02 emission factors. For
sources in the transmission and storage segment that use emission factors not directly calculated from
GHGRP data, and for sources in the distribution segment, data from the 1996 GRI/EPA study and a GTI
publication (1993) were used to adapt the CH4 emission factors into non-combustion related C02
emission factors. C02 emissions from post-meter sources (commercial, industrial and vehicles) were
estimated using default emission factors from IPCC (2019). Carbon dioxide emissions from post-meter
residential sources are included in fossil fuel combustion data.
Flaring N20 emissions were estimated for flaring sources using GHGRP data.
See Annex 3.6 for more detailed information on the methodology and data used to calculate CH4, C02,
and N20 emissions from natural gas systems.
Activity Data. Activity data were taken from various published data sets, as detailed in Annex 3.6. Key
activity data sources include data sets developed and maintained by EPA's GHGRP (EPA 2024); Enverus
(Enverus 2025); BOEM; Federal Energy Regulatory Commission (FERC) (FERC 2024); EIA; the Natural
Gas STAR and Methane Challenge Programs annual data; Oil and Gas Journal; and PHMSA.
For a few sources, recent direct activity data are not available. For these sources, either 2022 data were
used as a proxy for 2023 data, or a set of industry activity data drivers was developed and used to
calculate activity data over the time series. Drivers include statistics on gas production, number of
wells, system throughput, miles of various kinds of pipe, and other statistics that characterize the
changes in the U.S. natural gas system infrastructure and operations. More information on activity data
and drivers is available in Annex 3.6.
A complete list of references for emission factors and activity data by emission source is provided in
Annex 3.6.
Calculating Net Emissions. For most sources, net emissions are calculated directly by applying
emission factors to activity data. Emission factors used in net emission approaches reflect technology-
specific information and take into account regulatory and voluntary reductions. However, for
production, transmission and storage, and distribution, some sources are calculated using potential
emission factors, and CH4 that is not emitted is deducted from the total CH4 potential estimates. To
account for use of such technologies and practices that result in lower emissions but are not reflected
3-108 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
in "potential" emission factors, data are collected on both regulatory and voluntary reductions.
Regulatory actions addressed using this method include EPA National Emission Standards for
Hazardous Air Pollutants (NESHAP) regulations for dehydrator vents. Voluntary reductions included in
the Inventory are those reported to Natural Gas STAR and Methane Challenge for certain sources.
Methodological recalculations were applied to the entire time series to ensure time-series consistency
from 1990 through 2023. Available GHGRP data (beginning in 2011) and other recent data sources have
improved estimates of emissions from natural gas systems. To develop a consistent time series, for
sources with new data, EPA reviewed available information on factors that may have resulted in changes
over the time series (e.g., regulations, voluntary actions) and requested stakeholder feedback on trends
as well. For most sources, EPA developed annual data for 1993 through 2010 by interpolating activity
data or emission factors or both between 1992 and 2011 data points. Information on time-series
consistency for sources updated in this year's Inventory can be found in the Recalculations Discussion
below, with additional detail provided in supporting memos (relevant memos are cited in the
Recalculations Discussion). For detailed documentation of methodologies, please see Annex 3.5.
The notation key "IE" is used for C02 and CH4 emissions from venting and flaring in common data tables
category 1 .B.2. Disaggregating flaring and venting estimates across the Inventory would involve the
application of assumptions and could result in inconsistent reporting and, potentially, decreased
transparency. Data availability varies across segments within oil and gas activities systems, and
emission factor data available for activities that include flaring can include emissions from multiple
sources (flaring, venting and leaks).
Uncertainty
EPA has conducted a quantitative uncertainty analysis using the IPCC Approach 2 methodology
(Monte Carlo Simulation technique) to characterize the uncertainty for natural gas systems. For
more information on the approach, please seethe memoranda Inventory of U.S. Greenhouse Gas
Emissions and Sinks 1990-2016: Natural Gas and Petroleum Systems Uncertainty Estimates (2018
Uncertainty memo) and Inventory of U.S. Greenhouse Gas Emissions and Sinks 1990-2019: Update
for Natural Gas and Petroleum Systems C02 Uncertainty Estimates (2021 Uncertainty memo).75
EPA used Palisade's @RISK add-in tool for Microsoft Excel to estimate the 95 percent confidence
bound around CH4and C02 emissions from natural gas systems for the current Inventory. For the
CH4 uncertainty analysis, EPA focused on the 6 highest-emitting sources for the year 2023, which
together emitted 51 percent of methane from natural gas systems in 2023, and extrapolated the
estimated uncertainty for the remaining sources. For the C02 uncertainty analysis, EPA focused on
the highest-emitting source for the year 2023, which emitted 50 percent of C02 from natural gas
systems in 2023, and extrapolated the estimated uncertainty for the remaining sources. To estimate
uncertainty for N20, EPA applied the uncertainty bounds calculated for C02. EPA will seek to refine
this estimate in future Inventories. The @RISK add-in provides for the specification of probability
density functions (PDFs) for key variables within a computational structure that mirrors the
calculation of the inventory estimate. For emission factors that are derived from methane
emissions measurement studies, the PDFs are commonly determined to be lognormally distributed
(GRI/EPA1996; GTI 2001; GTI 2009; Lamb et al. 2015; Zimmerle et al. 2015; Fischer et al. 2018; GTI
75 See https://www.epa.gov/ghgemissions/natijral-gas-and-petroleum-systems.
Energy 3-109
-------
2019). For activity data that are derived from national datasets, the PDFs are set to a uniform
distribution (see 2018 and 2021 Uncertainty memos). Many emission factors and activity factors
are calculated using Subpart W data, and for these, the @RISK add-in determines the best fitting
PDF (e.g., lognormal, gaussian), based on bootstrapping of the underlying data (see 2018 and 2021
Uncertainty memos). The IPCC guidance notes that in using this Approach 2 method, "some
uncertainties that are not addressed by statistical means may exist, including those arising from
omissions or double counting, or other conceptual errors, or from incomplete understanding of the
processes that may lead to inaccuracies in estimates developed from models." The uncertainty
bounds reported below only reflect those uncertainties that EPA has been able to quantify and do
not incorporate considerations such as modeling uncertainty, data representativeness,
measurement errors, misreporting or misclassification. The understanding of the uncertainty of
emission estimates for this category evolves and improves as the underlying methodologies and
datasets improve.
The results presented below provide the 95 percent confidence bound within which actual
emissions from this source category are likely to fall for the year 2023, using the IPCC methodology.
The results of the Approach 2 uncertainty analysis are summarized in Table 3-70. Natural gas
systems CH4 emissions in 2023 were estimated to be between 146.4 and 179.9 MMT C02 Eq. at a 95
percent confidence level. Natural gas systems C02 emissions in 2023 were estimated to be
between 32.4 and 44.1 MMT C02 Eq. at a 95 percent confidence level. Natural gas systems N20
emissions in 2023 were estimated to be between 0.008 and 0.010 MMT C02 Eq. at a 95 percent
confidence level.
Uncertainty bounds for other years of the time series have not been calculated, but uncertainty is
expected to vary over the time series. For example, years where many emission sources are
calculated with interpolated data would likely have higher uncertainty than years with
predominantly year-specific data. In addition, the emission sources that contribute the most to CH4
and C02 emissions are different overthe time series, particularly when comparing recentyears to
early years in the time series. For example, venting emissions were higher and flaring emissions
were lower in earlyyears of the time series, compared to recentyears. Technologies also changed
overthe time series (e.g., liquids unloading with plunger lifts and reduced emissions completions
were not used early in the time series and cast iron distribution mains were more prevalent than
plastic mains in earlyyears). Transmission and gas processing compressor leak and vent emissions
were also higher in the earlyyears of the time series.
Table 3-70: Approach 2 Quantitative Uncertainty Estimates for CH4 and Non-
combustion C02 Emissions from Natural Gas Systems (MMT C02 Eq. and Percent)
Source
Gas
2023 Emission Estimate
Uncertainty Range Relative to Emission Estimate"
(MMT CO2 Eq.)b
(MMT CO2 Eq.)
(%)
Lower
Bound"
Upper
Bound"
Lower
Bound"
Upper
Bound"
Natural Gas Systems
CH4
162.4
146.1
179.9
-10%
+11%
Natural Gas Systems
CO2
37.7
32.4
44.1
-14%
+ 17%
Natural Gas Systems
N2O
0.008
0.007
0.010
-14%
+ 17%
3-110 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
a Range of emission estimates estimated by applying the 95 percent confidence intervals obtained from the Monte
Carlo simulation analysis conducted for the year 2023 ChU and CO2 emissions.
b All reported values are rounded after calculation. As a result, lower and upper bounds may not be duplicable from
other rounded values as shown in Table 3-61 and Table 3-62.
QA/QC and Verification Discussion
In order to ensure the quality of the emission estimates for natural gas systems, general (IPCC Tier 1)
Quality Assurance/Quality Control (QA/QC) procedures were implemented consistent with the U.S.
Inventory QA/QC plan outlined in Annex 8.
The natural gas systems emission estimates in the Inventory are continually being reviewed and
assessed to determine whether emission factors and activity factors accurately reflect current industry
practices. A QA/QC analysis was performed for data gathering and input, documentation, and
calculation. QA/QC checks are consistently conducted to minimize human error in the model
calculations. EPA performs a thorough review of information associated with new studies, GHGRP data,
regulations, public webcasts, and the Natural Gas STAR Program to assess whether the assumptions in
the Inventory are consistent with current industry practices. The EPA has a multi-step data verification
process for GHGRP data, including automatic checks during data-entry, statistical analyses on
completed reports, and staff review of the reported data. Based on the results of the verification
process, the EPA follows up with facilities to resolve mistakes that may have occurred.76
As in previous years, EPA conducted early engagement and communication with stakeholders on
updates prior to public review of the current Inventory. EPA held a stakeholder webinar in November
2024. EPA released a memo detailing updates under consideration and requesting stakeholder
feedback.
In recent years, several studies have measured emissions at the source level and at the national or
regional level and calculated emission estimates that may differ from the Inventory. There are a variety
of potential uses of data from new studies, including replacing a previous estimate or factor, verifying or
QA of an existing estimate or factor, and identifying areas for updates. In general, there are two major
types of studies related to oil and gas greenhouse gas data: studies that focus on measurement or
quantification of emissions from specific activities, processes and equipment, and studies that use
tools such as inverse modeling to estimate the level of overall emissions needed to account for
measured atmospheric concentrations of greenhouse gases at various scales. The first type of study
can lead to direct improvements to or verification of Inventory estimates. In the past few years, EPA has
reviewed and, in many cases, incorporated data from these data sources. The second type of study can
provide general indications of potential over- and under-estimates. In addition, in recent years
information from top-down studies has been directly incorporated to quantify emissions from well
blowouts.
A key challenge in using these types of studies to assess Inventory results is having a relevant basis for
comparison (e.g., the two data sets should have comparable time frames and geographic coverage, and
the independent study should assess data from the Inventory and not another data set, such as the
Emissions Database for Global Atmospheric Research, or "EDGAR"). In an effort to improve the ability to
compare the national-level Inventory with measurement results that may be at other spatial and
76 See https://www.epa.gov/sites/prodijction/files/9015-07/documents/ghgrp verification factsheet.pdf.
Energy 3-111
-------
temporal scales, EPA has developed a gridded inventory of U.S. anthropogenic methane emissions with
0.1 degree x 0.1 degree spatial resolution, monthly temporal resolution, and detailed scale-dependent
error characterization.77 The most recent version of the gridded methane inventory is designed to be
consistent with the U.S. EPA's Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2018
estimates for the years 2012 to 2018. The gridded inventory improves efforts to compare results of this
Inventory with atmospheric studies.
Recalculations Discussion
EPA received information and data related to the emission estimates through GHGRP reporting and
presented information to stakeholders regarding the updates under consideration. In December 2024,
EPA released a draft memorandum that discussed changes under consideration and requested
stakeholder feedback on those changes. EPA then released a final memorandum documenting the
methodology implemented in the current Inventory,78 The memorandum cited in the Recalculations
Discussion below is: Inventory of U.S. Greenhouse Gas Emissions and Sinks 1990-2023: Updates to Use
New Offshore Data (Offshore Production memo).
EPA evaluated relevant information available and made updates to the Inventory, including for offshore
production in the Gulf of America (GOA). General information for the source specific recalculations is
presented below and details are available in the Offshore Production memo.
In addition to the updates to the source mentioned above, for certain sources, CH4 and/or C02
emissions changed by greater than 0.05 MMT C02 Eq., comparing the previous estimate for 2022 to the
current (recalculated) estimate for 2022. The emissions changes were mostly due to GHGRP data
submission revisions and updated Enverus data. These sources are discussed below and include
pneumatic controllers, chemical injection pumps, liquids unloading, wellpad equipment leaks, kimray
pumps, produced water, and offshore production (in the production segment); gathering and boosting
(G&B) dehydrators, pneumatic controllers, blowdowns, and storage tanks; natural gas processing
blowdowns and acid gas removal (AGR); and LNG export sources.
The combined impact of revisions to 2022 natural gas systems CH4 emissions, compared to the
previous Inventory, is a decrease from 173.1 to 172.8 MMT C02 Eq. (0.3 MMT C02 Eq., or 0.2 percent).
The recalculations resulted in an average increase in the annual CH4 emission estimates across the
1990 through 2022 time series, compared to the previous Inventory, of 0.25 MMT C02 Eq., or about 0.1
percent.
The combined impact of revisions to 2022 natural gas systems C02 emissions, compared to the
previous Inventory, is a decrease from 36.5 MMT to 36.4 MMT (0.1 MMT or 0.2 percent). The
recalculations resulted in an average increase in emission estimates across the 1990 through 2022 time
series, compared to the previous Inventory, of less than 0.1 MMT C02 Eq., or less than 0.1 percent.
The combined impact of revisions to 2022 natural gas systems N20 emissions, compared to the
previous Inventory, is a decrease from 152.0 kt C02 Eq. to 18.2 kt C02 Eq., or 88 percent. This change for
2022 was due to a correction in the emission factor calculation for production storage tank flaring. The
77 See https://www.epa.gov/ghgemissionsAjs-gridded-methane-emissions.
78 Stakeholder materials including draft and final memoranda for the current (i.e., 1990 to 2023) Inventory are available at
https://www.epa.gov/ghgemissions/natural-gas-and-petroleum-svstems.
3-112 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
recalculations resulted in an average decrease in emission estimates across the 1990 through 2022
time series, compared to the previous Inventory, of 4.5 kt C02 Eq., or 7 percent.
In Table 3-71 and Table 3-72 below are categories in natural gas systems with recalculations resulting in
a change of greater than 0.05 MMT C02 Eq., comparing the previous estimate for 2022 to the current
(recalculated) estimate for 2022. No changes made to N20 estimates resulted in a change greater than
0.05 MMT C02 Eq. For more information, please see the Recalculations Discussion below.
Table 3-71: Recalculations of C02 in Natural Gas Systems (MMT C02)
Segment and Emission Sources with Changes of Greater than
0.05 MMT CO2 due to Recalculations
Previous
Estimate Year
2022,
2024 Inventory
Current
Estimate Year
2022,
2025 Inventory
Current
Estimate
Year 2023,
2025 Inventory
Exploration
+
+
+
Production
8.6
8.6
9.7
Offshore Production - GOA Federal Waters
+
+
+
Offshore Production - GOA State Waters
+
+
+
Processing
26.7
26.6
26.8
AGR Vents
18.1
18.0
18.7
Transmission and Storage
1.2
1.1
1.2
LNG Export Terminals (equipment leaks, compressors, flares)
1.0
0.9
0.9
Distribution
+
+
+
Post-Meter
+
+
+
Total
36.5
36.4
37.7
+ Does not exceed 0.05 MMT CO2 Eq.
Note: Totals may not sum due to independent rounding.
Table 3-72: Recalculations of CH4 in Natural Gas Systems (MMT C02 Eq.)
Segment and Emission Sources with
Changes of Greater than 0.05 MMT CO2 Eq.
due to Recalculations
Previous Estimate
Year2022,
2024 Inventory
Current Estimate
Year 2022,
2025 Inventory
Current Estimate
Year 2023,
2025 Inventory
Exploration
+
+
+
Production
89.7
89.7
80.8
Pneumatic Controllers
18.0
19.9
16.4
Chemical Injection Pumps
2.1
2.0
1.7
Liquids Unloading
2.4
2.3
1.4
Wellpad Equipment Leaks
10.8
11.5
8.0
Produced Water
4.0
3.9
4.1
Kimray Pumps
0.8
0.9
0.9
Offshore Production - GOA Federal Waters
0.4
0.05
0.04
Offshore Production - GOA State Waters
0.3
0.04
0.04
G&B Stations - Tanks
8.7
6.9
5.3
G&B Stations-Station Blowdowns
0.9
1.0
0.9
G&B Stations - Dehydrator Vents
1.1
1.2
1.1
G&B Stations - Pneumatic Controllers
4.8
4.6
3.9
Processing
15.1
14.8
15.2
Blowdowns/Venting
1.3
0.9
0.6
Energy 3-113
-------
Segment and Emission Sources with
Changes of Greater than 0.05 MMT CO2 Eq.
due to Recalculations
Previous Estimate
Year2022,
2024 Inventory
Current Estimate
Year 2022,
2025 Inventory
Current Estimate
Year 2023,
2025 Inventory
Transmission and Storage
39.6
39.6
37.3
Distribution
+
+
+
Post-Meter
13.4
13.4
13.8
Total
173.1
172.8
162.4
+ Does not exceed 0.05 MMT C02 Eq.
Note: Totals may not sum due to independent rounding.
Exploration
There were no methodological updates for exploration and recalculations resulted in an average
increase in CH4 emissions across the time series of 1 percent and an average increase in calculated C02
emissions across the time series of 0.5 percent, compared to the previous Inventory.
Production
Offshore Production in Gulf of America (Methodological Update)
EPA updated the calculation methodology for offshore production in the Gulf of America (GOA) to use
new emission factors calculated from year 2021 data from the Bureau of Ocean Energy Management
(BOEM) (BOEM 2023). Offshore production in the GOA occurs in two areas, federal waters and state
waters. State waters are closer to the shoreline and federal waters are beyond this. BOEM provides
periodic emission inventories which account for emissions specific to GOA federal waters production.
Previously, year 2017 BOEM data were the most recent that was incorporated into the Inventory's
calculation methodology. EPA previously applied emission source-specific emission factors calculated
from BOEM's 2017 dataset to calculate emissions for all years from 2016 - 2022 for GOA federal waters.
EPA then calculated GOA state waters emissions using the federal waters emissions, assuming the
emissions were equivalent on a production basis. With the release of the BOEM 2021 dataset, EPA
calculated new emission source-specific emission factors. EPA applied the same approach to calculate
emission factors from the 2021 BOEM dataset as it did for the prior BOEM datasets. EPA applied the
emission factors calculated from the BOEM 2021 dataset for years 2020 - 2023, maintained the
emission factors from the BOEM 2017 dataset for 2016 - 2018, and calculated emission factors that
average both BOEM datasets together for year 2019. This update impacts sources of vent and leak
emissions only, flaring emissions are not affected. Details for this update are available in the Offshore
Production memo.
As a result of this methodological update, CH4emissions estimates for offshore production in the GOA
are on average 76 percent lower for 2019 - 2022 compared to the previous Inventory. The 2022 CH4
emissions estimate is 87 percent lower than in the previous Inventory. The update resulted in C02
emissions estimates for offshore production in the GOA that are on average 82 percent lower for 2019 -
2022 compared to the previous Inventory. The 2022 C02 emissions estimate is 94 percent lower than in
the previous Inventory. The emission decreases are due to lower emission factors calculated from the
BOEM 2021 dataset compared to the BOEM 2017 dataset. The methodological update did not impact
emissions for years prior to 2019; differences in emissions compared to the previous Inventory for years
prior to 2019 are due to changes in underlying activity data (e.g., number of offshore complexes, oil and
gas production).
3-114 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Table 3-73: GOA Offshore Production Vent and Leak National CH4 Emissions (Metric
Tons CH4)
Source
19901
2005
2019
2020
2021
2022
2023
GOA Federal Waters - Major Complexes
134,807
43,272
8,327
1,946
1,243
1,271
1,042
GOA Federal Waters - Minor Complexes
19,354
17,685
1,142
543
337
344
284
GOA State Waters
14,2021
10,675
6,068
1,962
1,365
1,260
1,538
Total Emissions
168,3641
71,633
15,537
4,451
2,945
2,875
2,863
Previous Estimate
168,1511
71,526
27,136
31,148
21,533
22,712
NA
NA (Not Applicable)
Note: Totals may not sum due to independent rounding.
Table 3-74: GOA Offshore Production Vent and Leak National C02 Emissions (Metric
Tons C02)
Source
1990
2005
2019
2020
2021
2022
2023
GOA Federal Waters - Major Complexes
1,514
342
248
19
12
12
10
GOA Federal Waters - Minor Complexes
410
374
14
18
11
11
9
GOA State Waters
177
125
168
29
20
18
22
Total Emissions
2,100
842
431
65
43
42
42
Previous Estimate
2,098
840
802
920
638
673
NA
NA (Not Applicable)
Note: Totals may not sum due to independent rounding.
Chemical Injection Pumps (Recalculation with Updated Data)
Chemical injection pump CH4 emission estimates resulted in an average decrease of 0.1 percent
across the time series compared to the previous Inventory. The estimate for 2022 is 5 percent lower
than the previous Inventory. These changes were due to GHGRP submission revisions and updated
gas well counts.
Table 3-75: Chemical Injection Pumps National CH4 Emissions (Metric Tons CH4)
Source
1990 | 2005 | 2019
2020
2021
2022
2023
Chemical Injection Pumps
25,5631
80,288 I
I 111,734
87,266
76,698
72,604
62,414
Previous Estimate
25,587
80,213
111,631
87,227
76,893
76,407
NA
NA (Not Applicable)
Pneumatic Controllers (Recalculation with Updated Data)
Pneumatic controller CH4 emissions estimates are on average 1.4 percent higher across the time-
series than in the previous Inventory. The estimate for 2022 is 11 percent higher than in the previous
Inventory. These changes were due to GHGRP submission revisions and updated gas well counts.
Table 3-76: Pneumatic Controllers National CH4 Emissions (Metric Tons CH4)
Source
19901 20051 2019
2020
2021
2022
2023
Low Bleed Controllers
°|
22,6691
23,405
20,551
21,147
27,421
22,842
High Bleed Controllers
358,506
484,469
52,765
42,394
41,566
30,839
11,819
Intermittent Bleed Controllers
235,111 |
575,2071
887,978
772,701
752,638
654,157
549,430
Energy 3-115
-------
Total Emissions
593,6171
1,082,345
964,148
835,646
815,351
712,417
584,091
Previous Estimate
589,3321
1,067,997 J
958,943
817,727
747,391
643,721
NA
NO (Not Occurring)
NA (Not Applicable)
Liquids Unloading (Recalculation with Updated Data)
Liquids unloading CH4 emissions estimates decreased by an average of less than 0.1 percent
across the 1990 to 2022 time series compared with the previous Inventory. The 2022 estimate
decreased by 5 percent compared with the previous Inventory. These changes were due to GHGRP
submission revisions and updated gas well counts.
Table 3-77: Liquids Unloading National CH4 Emissions (Metric Tons CH4)
Source
19901
2005
2019
2020
2021
2022
2023
Liquids UnloadingWith Plunger Lifts
0
128,572
75,230
51,485
33,918
23,869
16,386
Liquids UnloadingWithout Plunger Lifts
77,8221
| 199,026
| 104,630
84,551
65,760
56,600
33,989
Total Emissions
77,8221
327,598
| 179,860
136,037
99,678
80,470
50,376
Previous Estimate
77,7671
| 327,0231
179,565
135,707
99,572
84,611
NA
NA (Not Applicable)
Note: Totals may not sum due to independent rounding.
Wellpad Equipment Leaks (Recalculation with Updated Data)
Wellpad equipment leak CH4 emissions estimates increased by an average of 0.2 percent across
the 1990 to 2022 time series compared with the previous Inventory. The 2022 estimate increased by
6 percent compared with the previous Inventory. These changes were due to GHGRP submission
revisions and updated gas well counts.
Table 3-78: Wellpad Equipment Leaks National CH4 Emissions (Metric Tons CH4)
Source
1990 | 2005 | 2019
2020
2021
2022
2023
Heaters
12,2821
18,4371
16,205
18,610
17,635
18,595
27,160
Separators
41,496
80,827
126,143
129,208
109,697
94,096
105,200
Dehydrators
12,898
11,394
3,656
3,070
4,081
3,111
3,361
Meters/Piping
42,964
63,842
84,850
154,043
130,602
76,543
75,192
Compressors
30,2401
61,781 |
65,518
61,041
74,234
217,523
75,396
Total Emissions
139,880
| 236,282
| 296,371
365,971
336,248
409,867
286,308
Previous Estimate
740,1501
| 236,0791
| 295,352
365,325
335,295
385,280
NA
NA (Not Applicable)
Note: Totals may not sum due to independent rounding.
Produced Water (Recalculation with Updated Data)
Produced water CH4 emissions estimates decreased by an average of 0.5 percent across the 1990
to 2022 time series compared with the previous Inventory. The 2022 estimate decreased by 3
percent compared with the previous Inventory. These changes were due to updated produced
water volumes.
3-116 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Table 3-79: Produced Water National CH4 Emissions (Metric Tons CH4)
Source
1990
2005
2019
2020
2021
2022
2023
Produced Water
121,669
152,809
158,918
140,054
130,117
138,837
147,634
Previous Estimate
121,867
153,081
159,525
140,299
140,299
143,132
NA
NA (Not Applicable)
Note: Totals may not sum due to independent rounding.
Kimray Pumps (Recalculation with Updated Data)
Kimray pump CH4 emissions estimates increased by an average of 0.6 percent across the 1990 to
2022 time series compared with the previous Inventory. The 2022 estimate increased by 10 percent
compared with the previous Inventory. These changes were due to updated gas well counts.
Table 3-80: Kimray Pumps National CH4 Emissions (Metric Tons CH4)
Source
1990 2005 2019
2020
2021
2022
2023
Kimray Pumps
149,3591
192,7191
134,876
33,634
31,690
31,479
30,594
Previous Estimate
149,192\
92,6691
34,630
33,353
31,100
28,709
NA
NA (Not Applicable)
Note: Totals may not sum due to independent rounding.
Gathering and Boosting-Storage Tanks (Recalculation with Updated Data)
Gathering and boosting (G&B) station storage tank CH4 emissions estimates are on average 9
percent lower across the 1990 to 2022 time series than in the previous Inventory. The 2022 estimate
is 21 percent lower than in the previous Inventory. These changes were due to GHGRP submission
revisions and a revised approach to determine the number of tanks reported under GHGRP.
Table 3-81: G&B Storage Tanks National Emissions (Metric Tons CH4)
Source
1990
2005
2019
2020
2021
2022
2023
G&B Storage Tanks
120,816
150,611
293,786
223,609
238,478
244,981
189,262
Previous Estimate
128,572
166,324
| 297,668
239,291
276,586
310,216
NA
NA (Not Applicable)
Note: Totals may not sum due to independent rounding.
Gathering and Boosting-Station Blowdowns (Recalculation with Updated Data)
G&B station blowdown CH4 emissions estimates are on average 44 percent higher across the 1990 to
2022 time series than in the previous Inventory. The 2022 estimate is 13 percent higher than in the
previous Inventory. These changes were due to GHGRP submission revisions and a revised approach to
incorporate blowdown emissions reported under GHGRP when facilities use flow meters to determine
emissions.
Table 3-82: G&B Station Blowdowns National Emissions (Metric Tons CH4)
Source
1990
2005
2019
2020
2021
2022
2023
G&B Station Blowdowns
30,979
38,619
59,948
51,211
40,462
36,269
32,672
Previous Estimate
20,218
26,155
39,059
40,519
35,161
32,036
NA
NA (Not Applicable)
Energy 3-117
-------
Gathering and Boosting - Pneumatic Controllers (Recalculation with Updated Data)
G&B pneumatic controllers CH4 emissions estimates are on average 0.2 percent higher across the 1990
to 2022 time series compared with the previous Inventory. The emissions estimate for 2022 is 4 percent
lower than in the previous Inventory, largely because of a decrease in emissions from high-bleed
pneumatic controllers. These changes were due to GHGRP submission revisions.
Table 3-83: G&B Pneumatic Controllers National Emissions (Metric Tons CH4)
Source
1990
2005
2019
2020
2021
2022
2023
High-Bleed Pneumatic Controllers
17,751
22,128
22,644
21,608
19,296
17,342
15,259
Intermittent Bleed Pneumatic Controllers
81,445
101,530
184,679
171,860
156,290
139,715
118,834
Low-Bleed Pneumatic Controllers
2,817
3,512
6,938
6,915
6,524
6,504
5,576
Total Emissions
102,013
127,170
214,261
200,383
182,110
163,562
139,670
Previous Estimate
98,229
127,072\
215,725
201,625
184,116
171,000
NA
NA (Not Applicable)
Note: Totals may not sum due to independent rounding.
Gathering and Boosting- Dehydrator Vents (Recalculation with Updated Data)
G&B dehydrator vent CH4 emissions estimates are on average 0.6 percent higher across the 1990 to
2022 time series compared with the previous Inventory. The emissions estimate for 2022 is 7 percent
higher than in the previous Inventory. These changes were due to GHGRP submission revisions.
Table 3-84: G&B Dehydrator Vent National Emissions (Metric Tons CH4)
Source
1990
2005
2019
2020
2021
2022
2023
G&B Dehydrator Vents
36,945
46,056
57,148
52,958
59,738
43,491
37,917
Previous Estimate
35,579
46,026
57,084
52,912
59,836
40,517
NA
NA (Not Applicable)
Processing
AGR (Recalculation with Updated Data)
Acid gas removal (AGR) C02 emission estimates are on average 0.02 percent lower across the time
series than in the previous Inventory. The C02 estimate for 2022 is 0.4 percent lower than in the previous
Inventory. These changes were due to GHGRP submission revisions.
Table 3-85: Processing Segment AGR National C02 Emissions (kt C02)
Source
1990 2005
2019
2020
2021
2022
2023
Flares
28,282 I
15,281
16,371
17,305
18,482
18,003
18,661
Previous Estimate
28,282
15,281
16,371
17,305
18,526
18,069
NA
NA (Not Applicable)
Blowdowns (Recalculation with Updated Data)
Processing blowdown CH4 emissions estimates are on average 0.4 percent lower across the time series
than in the previous Inventory. The emissions estimate for 2022 is 31 percent lower than in the previous
3-118 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Inventory. These changes were due to GHGRP submission revisions and a correction in the emission
factor calculation.
Table 3-86: Processing Blowdowns National CH4 Emissions (Metric Tons CH4)
Source
1990
2005 2019
2020
2021
2022
2023
Blowdowns
59,507
34,2441
44,076
50,286
47,458
32,013
20,638
Previous Estimate
59,507
34,244
44,581
44,197
45,370
46,188
NA
NA (Not Applicable)
Transmission and Storage
LNG Export Terminals (Recalculation with Updated Data)
LNG export terminal C02 emissions estimates are on average 0.3 percent lower across the time series
than in the previous Inventory. The 2022 estimate is 6 percent lower than in the previous Inventory.
These changes were due to updated data and GHGRP submission revisions.
Table 3-87: LNG Export Terminals National C02 Emissions (kt C02)
Source
1990
2005
2019
2020
2021
2022
2023
LNG Export Terminals (eq. leaks, compressors, flares)
0.02
0.02
1,007
1,767
693
940
883
Previous Estimate
0.02
0.021
979
1,767
707
1,005
NA
NA(Not Applicable)
Distribution
There were no methodological updates to the distribution segment and recalculations resulted in an
average increase in CH4 emissions across the time series of less than 0.1 percent and an average
increase in calculated C02 emissions across the time series of less than 0.1 percent, compared to the
previous Inventory.
Post-Meter
There were no methodological updates to post-meter emissions, and recalculations resulted in an
average increase in CH4 emissions across the time series of less than 0.1 percent and an average
increase in calculated C02 emissions across the time series of less than 0.1 percent, compared to the
previous Inventory.
Planned Improvements
Planned Improvements for 2026 Inventory
EPA updated oil and gas well counts and oil and gas production for this 2025 Inventory using Enverus
data. However, EPA did not update the number of completion events, due to significant changes in the
data across the time series. EPA will assess the underlying Enverus data to develop an appropriate
methodology to determine the number of completions for each year of the time series.
Energy 3-119
-------
Upcoming Data, and Additional Data that Could Inform the Inventory
EPA will assess new data received by EPA's Greenhouse Gas Reporting Program on an ongoing basis,
which may be used to validate or improve existing estimates and assumptions. In December 2024, EPA
released a memorandum discussing updates under consideration for a future Inventory to incorporate
revised GHGRP subpart W emission factors and requested stakeholder feedback (Inventory of U.S.
Greenhouse Gas Emissions and Sinks: Updates Under Consideration to Use Revised Subpart W
Emission Factors).79 One commenter provided feedback on the potential subpart W-based revisions.
The commenter had concerns with using the revised subpart W equipment leak emission factors though
the commenter supported incorporating leaker survey data into the Inventory's equipment leaks
methodology. EPA continues to track studies that contain data that may be used to update the
Inventory. EPA will also continue to assess studies that include and compare both top-down and
bottom-up emission estimates, which could lead to improved understanding of unassigned high
emitters (e.g., identification of emission sources and information on frequency of high emitters) as
recommended in previous stakeholder comments.
3.8 Abandoned Oil and Gas Wells (Source
Categories 1 B2a and 1 B2b)
The term "abandoned wells", as used in the Inventory, encompasses various types of oil and gas wells,
including orphaned wells and other non-producing wells:
• Wells with no recent production, and not plugged. Common terms (such as those used in state
databases) might include: inactive, temporarily abandoned, shut-in, dormant, and idle.
• Wells with no recent production and no responsible operator. Common terms might include:
orphaned, deserted, long-term idle, and abandoned.
• Wells that have been plugged to prevent migration of gas or fluids.
The U.S. population of abandoned oil and gas wells (including orphaned wells and other non-producing
wells) is around 3.9 million (with around 3.0 million abandoned oilwells and 0.9 million abandoned gas
wells). The methods to calculate emissions from abandoned wells involve calculating the total
populations of plugged and unplugged abandoned oil and gas wells in the United States and the
application of emission factors. An estimate of the number of orphaned wells within this population is
not developed as part of the methodology. Wells that are plugged have much lower average emissions
than wells that are unplugged (less than 1 kg CH4 per well per year, versus over 100 kg CH4 per well per
year). Around 43 percent of the abandoned well population in the United States is plugged. This fraction
has increased over the Inventory time series (from around 22 percent in 1990) as more wells fall under
regulations and programs requiring or promoting plugging of abandoned wells. Revised abandoned oil
and gas well counts from Enverus were not available for this version of the Inventory. This version of the
Inventory used 2022 activity data as proxy for 2023 (Enverus 2023).
79 The memo is available online: https://www.epa.gov/ghgemissions/stakeholder-process-natijral-gas-and-petroleum-
svstems-1990-?0?3-inventory
3-120 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Abandoned oil wells. Abandoned oil wells emitted 235 kt CH4 and 5 kt C02 in 2023. Emissions of both
gases increased by 3 percent from 1990, while the total population of abandoned oil wells increased 40
percent.
Abandoned gas wells. Abandoned gas wells emitted 68 kt CH4 and 3 kt C02 in 2023. Emissions of both
gases increased by 33 percent from 1990, while the total population of abandoned gas wells increased
83 percent.
Table 3-88: CH4 Emissions from Abandoned Oil and Gas Wells (MMT C02 Eq.)
Activity 1990
2005
2019
2020
2021
2022
2023
Abandoned Oil Wells 6.41
66
6.6
6.6
6.6
6.6
6.6
Abandoned Gas Wells 1.4
1.61
1.8
1.9
1.9
1.9
1.9
Total 7.8
8.21
8.5
8.5
8.6
8.5
8.5
Note: Totals may not sum due to independent rounding.
Table 3-89: CH4 Emissions from Abandoned Oil and Gas Wells (kt)
Activity 1990 I
20051 2019
2020
2021
2022
2023
Abandoned Oil Wells 2281
2361
237
237
237
235
235
Abandoned Gas Wells 51
581
65
66
69
68
68
Total 279
2941
302
303
306
303
303
Note: Totals may not sum due to independent rounding.
Table 3-90: C02 Emissions from Abandoned Oil and Gas Wells (MMT C02)
Activity 1990
2005
2019
2020
2021
2022
2023
Abandoned Oil Wells +|
*
*
+
+
+
+
Abandoned Gas Wells +
+ 1
+
+
+
+
+
Total +
I
+
+
+
+
+ Does not exceed 0.05 MMT C02 Eq.
Table 3-91: C02 Emissions from Abandoned Oil and Gas Wells (kt)
Activity 1990 |
20051 2019
2020
2021
2022
2023
Abandoned Oil Wells 5l
5I
5
5
5
5
5
Abandoned Gas Wells 2
3I
3
3
3
3
3
Total 7
71
8
8
8
8
8
Note: Totals may not sum due to independent rounding.
Methodology and Time-Series Consistency
EPA uses a Tier 2 method from IPCC (2019) to quantify emissions from abandoned oil and gas wells.
EPA's approach is based on the number of plugged and unplugged abandoned wells in the Appalachian
region and in the rest of the U.S., and emission factors for plugged and unplugged abandoned wells in
Appalachia and the rest of the U.S. Methods for abandoned wells are unavailable in IPCC (2006). The
details of this approach and of the data sources used are described in the memorandum Inventory of
U.S. Greenhouse Gas Emissions and Sinks 1990-2016: Abandoned Wells in Natural Gas and Petroleum
Systems (2018 Abandoned Wells Memo).
Energy 3-121
-------
EPA developed abandoned well CH4 emission factors using data from Kang et al. (2016) and Townsend-
Small et al. (2016). Plugged and unplugged abandoned well CH4 emission factors were developed at the
national-level (using emission data from Townsend-Small et al.) and for the Appalachia region (using
emission data from measurements in Pennsylvania and Ohio conducted by Kang et al. and Townsend-
Small et al., respectively). The Appalachia region emissions factors were applied to abandoned wells in
states in the Appalachian basin region, and the national-level emission factors were applied to
abandoned wells in all other states. EPA developed abandoned well C02 emission factors using the CH4
emission factors and an assumed ratio of C02-to-CH4 gas content, similar to the approach used to
calculate C02 emissions for many sources in Petroleum Systems and Natural Gas Systems. For
abandoned oil wells, EPA used the Petroleum Systems default production segment associated gas ratio
of 0.020 MT C02/MT CH4, which was derived through API TankCalc modeling runs. For abandoned gas
wells, EPA used the Natural Gas Systems default production segment CH4 and C02 gas content values
(GRI/EPA1996, GTI 2001) to develop a ratio of 0.044 MT C02/MT CH4. The same respective emission
factors are applied for each year of the time series.
EPA developed state-level annual counts of abandoned wells for 1990 through 2023 by summing
together an annual estimate of abandoned wells in the Enverus data set (Enverus 2023), and an
estimate of total abandoned wells not included the Enverus dataset (see 2018 Abandoned Wells Memo
for additional information on how the value was calculated) for each state. References reviewed to
develop the number of abandoned wells not included in the Enverus dataset include historical records
collected by state agencies and by USGS.
The state-level abandoned well population was then split into plugged and unplugged wells by applying
an assumption that all abandoned wells were unplugged in 1950 and using Enverus data to calculate the
fraction of plugged abandoned wells in 2023. Linear interpolation was applied between the 1950 value
and 2023 value to calculate the plugged fraction for intermediate years. See the memorandum Inventory
of U.S. Greenhouse Gas Emissions and Sinks 1990-2016: Abandoned Wells in Natural Gas and
Petroleum Systems (2018 Abandoned Wells Memo) for details.80
Abandoned Oil Wells
Table 3-92: Abandoned Oil Wells Activity Data, CH4 and C02 Emissions (kt)
Source
1990
2005
2019
2020
2021
2022
2023
Plugged abandoned oilwells
475,939
810,564
1,192,282
1,227,566
1,263,583
1,281,380
1,281,380
Unplugged abandoned oilwells
1,697,730
1,787,095
1,783,807
1,784,834
1,785,340
1,767,543
1,767,543
TotalAbandoned Oil Wells
2,173,669
2,597,659
2,976,089
3,012,400
3,048,923
3,048,923
3,048,923
Abandoned oilwells in Appalachia
22%
20%
19%
18%
18%
18%
18%
Abandoned oilwells outside of Appalachia
78%
80%
81%
82%
82%
82%
82%
ChUfrom plugged abandoned oilwells (kt)
0.17
0.25
0.35
0.36
0.36
0.37
0.37
ChUfrom unplugged abandoned oilwells(kt)
227.6
236.1
236.9
237.0
236.8
235.0
235.0
Total ChUfrom abandoned oil wells (kt)
227.7
236.4
237.2
237.3
237.2
235.4
235.4
Total CO2 from abandoned oil wells (kt)
4.6
4.8
4.8
4.8
4.8
4.8
4.8
Note: Totals may not sum due to independent rounding.
80 See https://www.epa.gov/ghgemissions/natijral-gas-and-petroleum-systems.
3-122 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Abandoned Gas Wells
Table 3-93: Abandoned Gas Wells Activity Data, CH4 and C02 Emissions (kt)
Source
1990
2005
2019
2020
2021
2022
2023
Plugged abandoned gas wells
110,089
210,902
359,018
372,605
389,745
395,236
395,236
Unplugged abandoned gas wells
355,620
404,960
448,504
453,988
463,119
457,628
457,628
TotalAbandoned Gas Wells
465,709
615,862
807,522
826,593
852,864
852,864
852,864
Abandoned gas wells in Appalachia
28%
25%
24%
24%
26%
26%
26%
Abandoned gas wells outside of Appalachia
72%
75%
76%
76%
74%
74%
74%
ChUfrom plugged abandoned gas wells (kt)
0.06
0.11
0.17
0.19
0.21
0.21
0.21
ChUfrom unplugged abandoned gas wells
(kt)
51.1
57.5
64.5
65.9
68.5
67.8
67.8
Total CH4 from abandoned gas wells (kt)
51.1
57.6
64.7
66.1
68.7
68.0
68.0
Total CO2 from abandoned gas wells (kt)
2.2
2.5
2.8
2.9
3.0
3.0
3.0
Note: Totals may not sum due to independent rounding.
Uncertainty
To characterize uncertainty surrounding estimates of abandoned well emissions, EPA conducted a
quantitative uncertainty analysis using the IPCC Approach 2 methodology (Monte Carlo simulation
technique). See the 2018 Abandoned Wells Memo for details of the uncertainty analysis methods. EPA
used Microsoft Excel's @RISK add-in tool to estimate the 95 percent confidence bound around total
methane emissions from abandoned oil and gas wells, then applied the calculated bounds to both CH4
and C02 emissions estimates for each population. The @RISK add-in provides for the specification of
probability density functions (PDFs) for key variables within a computational structure that mirrors the
calculation of the inventory estimate. EPA used measurement data from the Kang et al. (2016) and
Townsend-Small et al. (2016) studies to characterize the CH4 emission factor PDFs. For activity data
inputs (e.g., total count of abandoned wells, split between plugged and unplugged), EPA assigned
default uncertainty bounds of ± 10 percent based on expert judgment.
The IPCC guidance notes that in using this method, "some uncertainties that are not addressed by
statistical means may exist, including those arising from omissions or double counting, or other
conceptual errors, or from incomplete understanding of the processes that may lead to inaccuracies in
estimates developed from models." As a result, the understanding of the uncertainty of emission
estimates for this category evolves and improves as the underlying methodologies and datasets
improve. The uncertainty bounds reported below only reflect those uncertainties that EPA has been able
to quantify and do not incorporate considerations such as modeling uncertainty, data
representativeness, measurement errors, misreporting or misclassification.
The results presented below in Table 3-94 provide the 95 percent confidence bound within which actual
emissions from abandoned oil and gas wells are likely to fall for the year 2023, using the recommended
IPCC methodology. Abandoned oil well CH4 emissions in 2023 were estimated to be between 1.1 and
21.3 MMT C02 Eq., while abandoned gas well CH4 emissions were estimated to be between 0.3 and 6.8
MMT C02 Eq. at a 95 percent confidence level. Uncertainty bounds for other years of the time series
have not been calculated, but uncertainty is expected to vary over the time series.
Energy 3-123
-------
Table 3-94: Approach 2 Quantitative Uncertainty Estimates for CH4 and C02 Emissions
from Petroleum and Natural Gas Systems (MMT C02 Eq. and Percent)
Source
Gas
2023 Emission Estimate
(MMT CO2 Eq.)b
Uncertainty Range Relative to Emission Estimate"
(MMT CO2 Eq.) (%)
Lower
Upper
Lower
Upper
Bound
Bound
Bound
Bound
Abandoned OilWells
CH4
6.6
1.1
21.3
-83%
+223%
Abandoned Gas
Wells
ch4
1.9
0.3
6.8
-83%
+255%
Abandoned OilWells
co2
0.005
0.001
0.015
-83%
+223%
Abandoned Gas
Wells
co2
0.003
0.0005
0.011
-83%
+255%
a Range of emission estimates estimated by applying the 95 percent confidence intervals obtained from the Monte Carlo
simulation analysis conducted for total abandoned oil and gas well ChU emissions in year 2023.
b All reported values are rounded after calculation. As a result, lower and upper bounds may not be duplicable from other
rounded values as shown in the table.
QA/QC and Verification Discussion
The emission estimates in the Inventory are continually reviewed and assessed to determine whether
emission factors and activity factors accurately reflect current industry practices. In order to ensure the
quality of emission estimates for abandoned wells, general (IPCC Tier 1) quality assurance/quality
control (QA/QC) procedures were implemented consistent with the U.S. Inventory QA/QC plan outlined
in Annex 8. Additionally, EPA reviewed the current Enverus dataset and compared it with results from the
previous dataset to identify outliers and instances of significant changes to abandoned oil and gas well
counts.
EPA performs a thorough review of information associated with new studies to assess whether the
assumptions in the Inventory are consistent with industry practices and whether new data is available
that could be considered for updates to the estimates. As in previous years, EPA conducted early
engagement and communication with stakeholders on updates prior to public review. EPA held a
stakeholder webinar on greenhouse gas data for oil and gas in November of 2024.
Recalculations Discussion
Revised abandoned oil and gas well counts from Enverus were not available for this version of the
Inventory. This version of the Inventory used 2022 data as proxy for 2023 (Enverus 2023).
Planned Improvements
EPA will continue to assess new data and stakeholder feedback on considerations (such as potential
use of emission factor data from regions not included in the measurement studies on which current
emission factors are based) to improve the abandoned well count estimates and emission factors. In
future Inventories, EPA will assess data that become available from Department of Interior and
Department of Energy orphan well plugging programs. EPA will update the 2026 Inventory with revised
abandoned oil and gas well counts developed from Enverus data.
3-124 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
3.9 C02Transport, Injection, and
Geological Storage (Source Category
1C)
Emissions and reductions from C02 capture and sequestration are reported under the IPCC sector in
which capture takes place, as per the 2006 IPCC Guidelines for National Greenhouse Gas Inventories
(IPCC 2006). Fugitive emissions from the systems used to transport captured C02 from the source to the
injection site, fugitive emissions from activities and equipment at the injection site and those from the
end containment once the C02 is placed in storage are represented as part of C02 transport, injection,
and geological storage (TIGS) reporting. Figure 3-18 shows the flow and accounting of C02 emissions
across the C02 TIGS chain. Emissions from TIGS are shown in Table 3-95.
Figure 3-18: Flow of C02 Capture and Sequestration
Capture Facility
Net COs Emissions
ft
Capture Facility
CO2 Emissions from
Transport Leaks
C02 Emissions from
Equipment Leaks
11
C02 Transport
CO3 Injection
CO2 Emissions from
Surface Leaks
C02 Captured for Sequestration
Note: The Capture Facility Net C02 Emissions are the result of subtracting the amount of C02 Captured for Sequestration from
the Capture Facility C02 Emissions that would have occurred without C02 capture.
Table 3-95: Emission from TIGS (kt C02)
1990
2005
2019
2020
2021
2022
2023
Transport
NO
NO
2
2
2
2
2
Injection
NO
NO
16
13
37
28
31
Geological Storage
NO
NO
0
23
26
23
64
Total
NO
NO
18
39
65
53
98
NO (Not Occurring)
Note: Totals may not sum due to independent rounding.
Sequestered C02 is allocated across the different possible source categories as shown in Table 3-96.
The source categories are based on information from the Suppliers of C02 source category of EPA
Greenhouse Gas Reporting Program (GHGRP), 40 CFR Part 98, Subpart PP, also referred to as "Subpart
PP" (EPA 2024a). More information is provided in the Methodology section below.
Energy 3-125
-------
Table 3-96: Allocation of Sequestered C02 for Inventory Adjustment (kt C02)
19901
2005
2019
2020
2021
2022
2023
Inventory Adjustments Needed
Power Plants
NO
NO
0
0
0
0
360
Industrial Gas Plants
NO
NO
0
0
0
0
0
Chemical Plants
NO
NO
0
0
0
0
0
Synthetic Gas Production
NO
NO
0
0
0
0
0
Ammonia Plants
NO
NO
0
660
714
652
665
Ethanol Plants
NO
NO
520
522
444
603
903
Breweries
NO
NO
0
0
0
0
0
Distilleries
NO
NO
0
0
0
0
0
Paper Mills
NO |
NO
0
0
0
0
0
Total(lnvAdj)
NO
NO
520
1,182
1,158
1,255
1,928
Inventory Adjustments Not Needed
CO2 Domes
NO
NO
5,716
4,156
3,960
4,624
10,420
Petroleum Refineries
NO
NO
0
0
0
0
0
NG Processing
NO
NO
2,097
1,465
1,835
2,174
3,951
Total (No Adj)
NO |
NO
7,813
5,621
5,794
6,798
14,370
NO (Not Occurring)
Note: Totals may not sum due to independent rounding.
For the C02 sequestered sourced from natural domes, there is no adjustment needed to the Inventory
since it is considered a transfer from one sink to another. For the C02 from Natural gas processing and
Petroleum refining there is no need to further adjust the Inventory methodology since those emissions
are already netted out in the Inventory.
For the C02 from any other industrial process source the Inventory has been adjusted to subtract that
C02 capture from the source. This includes C02 captured from biogenic sources such as ethanol
facilities. Since fermentation emissions are biogenic C02 emissions, they are not included in the
national inventory (these are already included in national totals due to their treatment in the
Agricultural, Forestry and Other Land Use [AFOLU] sector). So, the subtraction of the amount of biogenic
C02 transferred to long-term storage may result in negative emissions. See Section 4.16 for more
information on this.
Methodology and Time-Series Consistency
The following section describes the methodology used to estimate C02 emissions from transport,
injection, and geological storage of C02. The allocation approach for determining the source of C02
capture for sequestration is also discussed.
Fugitive CO2 from Transport
To estimate C02 emissions from pipeline transport, EPA used the IPCC Tier 1 default factor for pipelines
as provided by the 2006 IPCC Guidelines for National Greenhouse Gas Inventories (IPCC 2006). In this
approach, the leakage emissions estimates from pipeline transport are assumed to be independent of
throughput and are based on distance (length) of pipeline. EPA estimated emissions associated with the
entire C02 pipeline network in the United States. This could potentially overestimate emissions, since
3-126 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
the amount of captured C02 subtraction at the source for the most part (except for NG Processing and
petroleum refining) is based on C02 received for sequestration, which would already account for any
pipeline losses. However, since that value is uncertain, and other sources of C02 (e.g., from natural
domes) are not being counted, using total C02 pipeline length to estimate transport emissions was
deemed appropriate.
The IPCC Tier 1 default fugitive C02 emissions rate from pipelines is 0.25-0.28 metric tons C02/km
pipeline, based on empirical data and analysis. Actual pipeline leakage rates depend on the type and
size of equipment installed in the pipeline systems, and are sourced from PHMSA (2024). In 2023, 5,331
miles (8,580 km) of C02 pipeline were in operation in the United States. This equates to an estimated
average leakage of 2,274 metric tons of C02 per year. Annual mileage is shown in Table 3-97.
Table 3-97: Pipeline Mileage (Miles)
1990 |
2005
2019
2020
2021
2022
2023
Miles
NO |
NO |
5,147
5,150
5,339
5,354
5,331
NO (Not Occurring)
Fugitive C02from Injection and Storage
GHGRP reporters provide an estimate of fugitive emissions from C02 injection, assumed to be reported
under C02 equipment leaks as part of the Geological Sequestration of Carbon Dioxide source category
of the GHGRP (40 CFR Part 98, Subpart RR, also referred to as "Subpart RR"), as shown in Table 3-98
(EPA 2024b). This information was used to estimate national emissions associated with C02 injection in
the Inventory. The GHGRP data include injection related emissions from the equipment between the
flow meter used to measure injection quantity and the injection wellhead which would be included in
the Inventory. Any fugitive C02 emission between the capture facility fence line and the injection point
would not be captured using this method, but would be captured as part of transport emissions
discussed above.
GHGRP reporters also provide an estimate of storage and any measured leakage of C02 from storage,
assumed to be under C02 surface leaks in subpart RR reporting as shown in Table 3-98, which has been
incorporated into the Inventory as well. GHGRP reporters report the annual mass of C02 that is emitted
by surface leakage as appropriate in accordance with their approved monitoring, reporting, and
verification (MRV) plan.81
As with transportation emissions, including equipment and surface leakage could potentially
overestimate emissions since the amount of captured C02 subtraction at the source for the most part
(except for NG Processing and petroleum refining) is based on C02 sequestered, which already
accounts for any equipment or surface losses. However, like for transport emissions, since that value is
uncertain, and other sources of C02 (e.g., from natural domes) are not being counted, including
equipment and surface leaks was deemed appropriate.
Under subpart RR, owners or operators of sequestration facilities submit a proposed MRV plan to EPA who reviews the
plan and issues a final MRV plan.
Energy 3-127
-------
Table 3-98: Emissions from Injection and Storage (kt C02)
1990
2005
2019
2020
2021
2022
2023
CO2 Injection Leaks
NO I
NO
16
13
37
28
31
CO2 Storage Leaks
NO
NO
0
23
26
23
64
NO (Not Occurring)
CO2 Sequestration and Capture
IPCC includes methodological guidance to estimate emissions from the capture, transport, injection,
and geological storage of C02. The methodology is based on the principle that the carbon capture and
storage system should be handled in a complete and consistent manner across the entire Energy sector.
The approach accounts for C02 captured at natural and industrial sites as well as emissions from
capture, transport, and use. For storage specifically, a Tier 3 methodology is outlined for estimating and
reporting emissions based on site-specific evaluations. However, IPCC (IPCC 2006) notes that if a
national regulatory process exists, emissions information available through that process may support
development of C02 emission estimates for geologic storage.
GHGRP reporters provide an estimate of C02 sequestered under Subpart RR, as shown in Table 3-99.
Subpart RR provides a mechanism for facilities to report the amount of C02 sequestered in geologic
formations on an annual basis to EPA. Facilities that conduct geologic sequestration of C02, and
facilities that inject C02 underground, are required to report greenhouse gas data annually to EPA
through its GHGRP. Facilities reporting geologic sequestration of C02 to the GHGRP develop and
implement an EPA-approved site-specific monitoring, reporting and verification plan, and report the
amount of C02 sequestered using a mass balance approach. Facilities measure and report data on the
amount of C02 received, data used to calculate the amount, and the source of the received C02 (if
known); various mass balance equation inputs (mass of C02 injected, recycled, emitted, produced,
equipment leaks, surface leakage, and entrained C02 in produced hydrocarbons), the amount of C02
sequestered, data used to calculate the inputs/amounts, and an annual monitoring report.
Table 3-99: Sequestered C02 (kt C02)
1990 |
2005
2019
2020
2021
2022
2023
CO2 Sequestered
NO |
NO |
8,332
6,802
6,952
8,053
16,299
NO (Not Occurring)
C02 sequestered is allocated to its source directly if known based on subpart RR. If the source is
unknown or if multiple sources are listed in Subpart RR, C02 sequestered is allocated across sources
based on subpart PP data. This mainly applies to splitting between natural domes and industrial
sources, and in particular natural gas processing. For facilities with annual C02 sourced from both C02
production wells and natural gas processing, C02 was split between the two sources based on subpart
PP enhanced oil recovery (EOR) data, as shown in Table 3-100 (EPA 2024a). The kt of C02 data is the
amount of C02 produced (natural domes) transferred to EOR and the amount of C02 captured
(industrial sources) transferred to EOR. Transfer to EOR is used since that is felt to best represent C02
supplied for sequestration.
Under subpart PP, EPA receives data from facilities with C02 production wells (natural C02 domes) and
other industrial facilities that extract or capture C02 streams. Importers and exporters of bulk C02 are
also required to report if total combined imports/exports of C02 and other greenhouse gases exceed
25,000 tons C02 Eq. per year. Reporters provide information on the mass of C02 captured or extracted,
3-128 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
data used to calculate that amount, and information on the amount of C02 that is supplied to various
end use categories. The amount of C02 captured by a specific facility is classified as confidential
business information (CBI) under the GHGRP and therefore only aggregated data is available for use
within the Inventory. Note that Subpart PP data does not include captured C02 if it is used on-site. Data
is available on the types and number of facilities that capture C02 and that was used to determine the
source categories of C02 capture as shown in Table 3-96.
For facilities with annual C02 sourced from C02 production wells, natural gas processing, and ethanol
plants, C02 was first split between natural domes (C02 production wells) and industrial capture (natural
gas processing and ethanol plants) based on Subpart PP EOR data. Then, industrial capture was spilt
evenly between natural gas processing and ethanol plants.
Table 3-100: Percentage of C02 (kt) Supplied to EOR from Different Sources
1990| 2005
2019
2020
2021
2022
2023
CO2 Extracted (domes) for EOR
kt of CO2
NO
NO
37,425
25,290
24,987
26,739
25,092
% of Total
NO
NO
72%
72%
71%
73%
74%
CO2 Captured (industrial) for EOR
kt of CO2
NO
NO
14,700
9,910
10,100
9,980
8,660
% of Total
NO
NO
28%
28%
29%
27%
26%
Total
NO
NO
52,125
35,200
35,087
36,719
33,752
NO (Not Occurring)
Note: Totals may not sum due to independent rounding.
Based on this methodology, sequestered C02 was allocated across the different possible source
categories, as shown in Table 3-96.
Treatment of EOR in the Inventory
The process of EOR can lead to incidental storage of C02 that is received for injection (i.e., storage is not
the main goal of EOR). In an EOR project, a portion of the injected C02 gets trapped in the reservoir in the
form of one or more C02 trapping mechanisms (stratigraphic trapping, dissolution in residual oil/brine,
residual trapping due to hysteresis, and mineral trapping). The remaining portion of the C02 is produced
along with hydrocarbons and brine through the production wells, which will be separated and re-
injected back into the reservoir along with newly received C02. Volumes of C02 that are recycled at the
last stage of the EOR project can be re-injected back into the reservoir as wells are shut-in or could be
transported to another EOR project.
For EOR C02, as noted in the 2006IPCC Guidelines, "At the Tier 1 or 2 methodology levels [EOR C02 is]
indistinguishable from fugitive greenhouse gas emissions by the associated oil and gas activities." In the
U.S. estimates for oil and gas fugitive emissions, the Tier 2 emission factors for C02 include C02 that
was originally injected and is emitted along with other gas from leak, venting, and flaring pathways, as
measurement data used to develop those factors would not be able to distinguish between C02 from
EOR and C02 occurring in the produced natural gas. Therefore, EOR C02 emitted through those
pathways is included in C02 estimates in 1B2.
More data on EOR may become available in the future through GHGRP subpart VV (see the Planned
Improvements section below).
Energy 3-129
-------
Uncertainty
A quantitative uncertainty analysis was conducted for C02 capture and sequestration using the IPCC-
recommended Approach 2 uncertainty estimation methodology. This analysis utilized the Monte Carlo
stochastic simulation software @Risk to estimate the 95 percent confidence bound around total C02
emissions.
There are uncertainties in pipeline emissions, equipment leakage, and surface leakage. A normal
distribution was assumed for all 13 input variables (two for pipeline emissions, seven for equipment
leakage, and four for surface leakage.) For these variables, the uncertainty ranges were assigned to the
input variables based on IPCC default uncertainty estimates (IPCC 2006) and expert opinion (ICF 2025).
The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 3-101. Total
emissions associated with CCS were estimated to be between 47.7 and 147.0 kt C02 Eq. at the 95
percent confidence level. This indicates a range of approximately 51 percent below to 51 percent above
the 2023 emission estimate of 97.6 kt C02 Eq.
Table 3-101: Approach 2 Quantitative Uncertainty Estimates for C02 Emissions from
TIGS (kt C02 Eq. and Percent)
Uncertainty Range Relative to Emission Estimate"
2023 Emission
(kt CO2 Eq.)
(%)
Source
Gas
Estimate
(kt CO2 Eq.)
Lower
Bound
Upper
Bound
Lower Upper
Bound Bound
Total Emissions from TIGS
C02
97.6
47.7
147.0
-51% +51%
a Range of emission estimates predicted by Monte Carlo stochastic simulation for a 95 percent confidence interval.
QA/QC and Verification
In order to ensure the quality of the C02 emission estimates from TIGS, general (IPCC Tier 1) and
category-specific (Tier 2) Quality Assurance/Quality Control (QA/QC) procedures were implemented
consistent with the U.S. Inventory QA/QC plan outlined in Annex 8. The Tier 2 procedures that were
implemented involved checks specifically focusing on the activity data and emission factor sources and
methodology used for estimating C02 from TIGS in the United States.
More details on the monitoring and QA/QC methods applicable to the GHGRP data used can be found
under the regulation (40 CFR Part 98).82 EPA verifies annual facility-level GHGRP reports through a multi-
step process (e.g., combination of electronic checks and manual reviews) to identify potential errors
and ensure that data submitted to EPA are accurate, complete, and consistent.83 Based on the results of
the verification process, EPA follows up with facilities to resolve mistakes that may have occurred. The
post-submittals checks are consistent with a number of general and category-specific QC procedures,
including range checks, statistical checks, algorithm checks, and year-to-year checks of reported data
and emissions.
82 See http://www.ecfr.gov/cgi-bin/text-idx7tpN/ecfrbrowse/Title40/40cfr98 main 0?.tpl.
83 See https://www.epa.gov/sites/prodijction/files/9015-07/documents/ghgrp verification factsheet.pdf.
3-130 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Recalculation Discussion
No recalculations were performed for the current Inventory.
Planned Improvements
EPA updated the GHGRP rules to add a new subpart VV (40 CFR Part 98 Subpart VV). Subpart VV creates
a reporting pathway for EOR operators who use the ISO 27916:2019 standard (ISO standard) to quantify
the C02 sequestered as a result of their operations. The ISO standard has requirements similar to the
site-specific monitoring, reporting and verification (MRV) plan required in order to report geologic
sequestration under subpart RR. When EOR facilities start to report using subpart VV facilities, that
would help to update treatment of C02 captured for EOR in the Inventory. Data on C02 sequestered
under subpart VV could be treated in the Inventory in the same way as the subpart RR data.
EPA also updated the GHGRP rules to add subpart PP data reporting requirements, that if a C02 stream
is captured from any facility subject to 40 CFR part 98 as well as supplying to RR or VV facilities , they
must:
1. Report the facility identification number associated with the annual greenhouse gas report for
the Subpart PP facility;
2. Report each facility identification number associated with the annual greenhouse gas reports
for each Subpart RR or VV facility to which C02 is transferred; and
3. Report the annual quantity of C02 in metric tons that is transferred to each Subpart RR or W
facility.
This provides a more direct link between carbon capture and sequestration in terms of Inventory
adjustments. This would include C02 captured at direct air capture (DAC) facilities in the future. To
prevent double counting, the updates also clarify that wells reported under subpart RR or W should not
also be counted under Subpart UU.
Furthermore, there could be additional existing GHGRP data available that could provide more input to
refine the allocation process. For example, subpart PP reporters track and report biogenic and fossil C02
separately. That information could be used to help allocate C02 from the different capture sources to
end uses based on assumptions about the biogenic content of captured C02. This data has not yet been
incorporated but could be used to help allocate capture and sequestration in the future.
Currently, there are no data included in this memo regarding C02 sequestered in years prior to 2010.
Alternate data sources could be explored, including reported quantities from the Regional Carbon
Sequestration Partnerships (RCSPs). Data would need to be available on an annual basis to consider for
conclusion.
Other possible updates include the treatment of exported C02. Exported C02 is currently accounted for
by adjusting down the amount of fuel combustion to net out emission results. Exported C02 could be
more explicitly accounted for in the Inventory through reporting C02 capture from the energy use
industrial sector and reporting the quantity of C02 export as part of the C02 TIGS accounting.
Energy 3-131
-------
3.10 International Bunker Fuels (Source
Category 1: Memo Items)
Emissions resulting from the combustion of fuels used for international transport activities, termed
international bunker fuels, are not included in national emission totals, but are reported separately
based upon location of fuel sales. The decision to report emissions from international bunker fuels
separately, instead of allocating them to a particular country, reflect the IPCC methodological guidance,
including IPCC (2006), in which countries are requested to report emissions from ships or aircraft that
depart from their ports with fuel purchased within national boundaries and are engaged in international
transport separately from national totals (IPCC 2006).
Two transport modes are addressed under the IPCC definition of international bunker fuels: aviation and
marine.84 Greenhouse gases emitted from the combustion of international bunker fuels, like other fossil
fuels, include C02, CH4 and N20 for marine transport modes, and C02 and N20 for aviation transport
modes. Emissions from ground transport activities—by road vehicles and trains—even when crossing
international borders are allocated to the country where the fuel was loaded into the vehicle and,
therefore, are not counted as bunker fuel emissions.
The 2006 IPCC Guidelines distinguish between three different modes of air traffic: civil aviation, military
aviation, and general aviation. Civil aviation comprises aircraft used for the commercial transport of
passengers and freight, military aviation comprises aircraft under the control of national armed forces,
and general aviation applies to recreational and small corporate aircraft. The 2006 IPCC Guidelines
further define international bunker fuel use from civil aviation as the fuel combusted for civil (e.g.,
commercial) aviation purposes by aircraft arriving or departing on international flight segments.
However, as mentioned above, and in keeping with the 2006 IPCC Guidelines, only the fuel purchased in
the United States and used by aircraft taking-off (i.e., departing) from the United States are reported
here. The standard fuel used for civil and military aviation is kerosene-type jet fuel, while the typical fuel
used for general aviation is aviation gasoline.85
Emissions of C02 from aircraft are essentially a function of fuel consumption. Nitrous oxide emissions
also depend upon engine characteristics, flight conditions, and flight phase (i.e., take-off, climb, cruise,
decent, and landing). Recent data suggest that little or no CH4 is emitted by modern engines (Anderson
et al. 2011), and as a result, CH4 emissions from this category are reported as zero. In jet engines, N20 is
primarily produced by the oxidation of atmospheric nitrogen, and the majority of emissions occur during
the cruise phase.
International marine bunkers comprise emissions from fuels burned by ocean-going ships of all flags
that are engaged in international transport. Ocean-going ships are generally classified as cargo and
passenger carrying, military (i.e., U.S. Navy), fishing, and miscellaneous support ships (e.g., tugboats).
For the purpose of estimating greenhouse gas emissions, international bunker fuels are solely related to
cargo and passenger carrying vessels, which is the largest of the four categories, and military vessels.
84 Most emission related international aviation and marine regulations are under the rubric of the International Civil Aviation
Organization (ICAO) or the International Maritime Organization (IMO), which develop international codes,
recommendations, and conventions, such as the International Convention of the Prevention of Pollution from Ships
(MARPOL).
85 Naphtha-type jet fuel was used in the past by the military in turbojet and turboprop aircraft engines.
3-132 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Two main types of fuels are used on sea-going vessels: distillate dieselfuel and residual fuel oil. Carbon
dioxide is the primary greenhouse gas emitted from marine shipping.
Overall, aggregate greenhouse gas emissions in 2023 from the combustion of international bunker fuels
from both aviation and marine activities were 97.0 MMT C02 Eq., or 7.2 percent below emissions in 1990
(see Table 3-102 and Table ). Emissions from international flights and international shipping voyages
departing from the United States have increased by 74.1 percent and decreased by 54.7 percent,
respectively, since 1990. The majority of these emissions were in the form of C02; however, small
amounts of CH4 (from marine transport modes) and N20 were also emitted.
For this Inventory, 2023 marine and military bunker fuel data were available, however civilian aviation
bunker fuel data were not available and were proxied based on 2022 values.
Table 3-102: C02, CH4, and N20 Emissions from International Bunker Fuels (MMT C02
Eq.)
Gas/Mode
1990
2005
2019
2020
2021
2022
2023
CO2
103.6
113.3
113.6
69.6
80.2
98.2
96.2
Aviation
38.2
60.2
78.3
39.8
50.8
66.6
66.5
Commercial
30.0
55.6
75.1
36.7
47.6
63.5
63.5
Military
8.2
4.6
3.2
3.1
3.2
3.1
3.0
Marine
65.4
53.1
35.4
29.9
29.4
31.6
29.6
CH4
0.2
0.1
0.1
0.1
0.1
0.1
0.1
Aviation
NO
NO
NO
NO
NO
NO
NO
Marine
0.2
0.1
0.1
0.1
0.1
0.1
0.1
N2O
0.8
0.9
0.9
0.5
0.6
0.8
0.8
Aviation
0.3
0.5
0.7
0.3
0.4
0.6
0.6
Marine
0.4
0.4
0.2
0.2
0.2
0.2
0.2
Total
104.6
114.3
114.6
70.3
80.9
99.1
97.0
NO (Not Occurring)
Notes: Totals may not sum due to independent rounding. Includes aircraft cruise altitude emissions. Civilian aviation bunker fuel
data were not available and were proxied based on 2022 values.
Table 3-103: C02, CH4, and N20 Emissions from International Bunker Fuels (kt)
Gas/Mode
1990
2005
2019
2020
2021
2022
2023
CO2
103,634
113,328
113,632
69,638
80,180
98,241
96,160
Aviation
38,205
60,221
78,280
39,781
50,812
66,646
66,526
Marine
65,429
53,107
35,351
29,857
29,369
31,595
29,634
CH4
7
5
4
3
3
3
3
Aviation
NO
NO
NO
NO
NO
NO
NO
Marine
7
5
4
3
3
3
3
N2O
3
3
3
2
2
3
3
Aviation
1
2
2
1
2
2
2
Marine
2 I
1
1
1
1
1
1
NO (Not Occurring)
Notes: Totals may not sum due to independent rounding. Includes aircraft cruise altitude emissions. Civilian aviation bunker fuel
data were not available and were proxied based on 2022 values.
Energy 3-133
-------
Methodology and Time-Series Consistency
Emissions of C02 were for the most part estimated by applying carbon content and fraction oxidized
factors to fuel consumption activity data. This approach is analogous to that described under Section
3.1. Carbon content and fraction oxidized factors for jet fuel (except for commercial aviation as per
below), distillate fuel oil, and residual fuel oil are the same as used for C02 from Fossil Fuel Combustion
and are presented in Annex 2.1, Annex 2.2, and Annex 3.8 of this Inventory. Density conversions were
taken from ASTM (1989) and USAF (1998). Heat content for distillate fuel oil and residual fuel oil were
taken from EIA (2025) and USAF (1998), and heat content for jet fuel was taken from EIA (2025). See
below for details on how emission estimates for commercial aviation were determined.
A complete description of the methodology and a listing of the various factors employed can be found in
Annex 2.1. See Annex 3.8 for a specific discussion on the methodology used for estimating emissions
from international bunker fuel use by the U.S. military.
Emission estimates for CH4 and N20 were calculated by multiplying emission factors by measures of
fuel consumption by fuel type and mode. Emission factors used in the calculations of CH4 and N20
emissions were obtained from the Revised 1996IPCC Guidelines (IPCC/UNEP/OECD/IEA1997), which
is also referenced in the 2006 IPCC Guidelines (IPCC 2006). For aircraft emissions, the following value,
in units of grams of pollutant per kilogram of fuel consumed (g/kg), was employed: 0.1 for N20 (IPCC
2006). For marine vessels consuming either distillate diesel or residual fuel oil the following values
(g/MJ), were employed: 0.315 for CH4 and 0.08 for N20. Activity data for aviation included solely jet fuel
consumption statistics, while the marine mode included both distillate diesel and residual fuel oil.
Activity data on domestic and international aircraft fuel consumption were developed by the U.S.
Federal Aviation Administration (FAA) using radar-informed data from the FAA Enhanced Traffic
Management System (ETMS) for 1990 and 2000 through 2022 as modeled with the Aviation
Environmental Design Tool (AEDT). This bottom-up approach is built from modeling dynamic aircraft
performance for each flight occurring within an individual calendar year. The analysis incorporates data
on the aircraft type, date, flight identifier, departure time, arrival time, departure airport, arrival airport,
ground delay at each airport, and real-world flight trajectories. To generate results for a given flight within
AEDT, the radar-informed aircraft data is correlated with engine and aircraft performance data to
calculate fuel burn and exhaust emissions. Information on exhaust emissions for in-production aircraft
engines comes from the International Civil Aviation Organization (ICAO) Aircraft Engine Emissions
Databank (EDB). This bottom-up approach is in accordance with the Tier 3B method from the 2006 IPCC
Guidelines (IPCC 2006).
International aviation C02 estimates for 1990 and 2000 through 2022 were obtained directly from FAA's
AEDT model (FAA 2024), data for 2023 was not yet available and has been proxied to 2022 in the current
Inventory. The radar-informed method that was used to estimate C02 emissions for commercial aircraft
for 1990 and 2000 through 2022 was not possible for 1991 through 1999 because the radar dataset was
not available for years prior to 2000. FAA developed Official Airline Guide (OAG) schedule-informed
inventories modeled with AEDT and great circle trajectories for 1990, 2000, and 2010. Because fuel
consumption and C02 emission estimates for years 1991 through 1999 are unavailable, consumption
estimates for these years were calculated using fuel consumption estimates from the Bureau of
Transportation Statistics (DOT 1991 through 2013), adjusted based on 2000 through 2005 data. See
Annex 3.3 for more information on the methodology for estimating emissions from commercial aircraft
jet fuel consumption.
3-134 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Data on U.S. Department of Defense (DoD) aviation bunker fuels and total jet fuel consumed by the U.S.
military was supplied by the Office of the Under Secretary of Defense (Installations and Environment),
DoD. Estimates of the percentage of each Service's total operations that were international operations
were developed by DoD. Military aviation bunkers included international operations, operations
conducted from naval vessels at sea, and operations conducted from U.S. installations principally over
international water in direct support of military operations at sea. Military aviation bunker fuel emissions
were estimated using military fuel and operations data synthesized from unpublished data from DoD's
Defense Logistics Agency Energy (DLA Energy 2025). Together, the data allow the quantity of fuel used in
military international operations to be estimated. Densities for each jet fuel type were obtained from a
report from the U.S. Air Force (USAF 1998). Final jet fuel consumption estimates are presented in Table .
See Annex 3.8 for additional discussion of military data.
Table 3-104: Aviation Jet Fuel Consumption for International Transport (TBtu)
Nationality
1990
2005
2019
2020
2021
2022
2023
U.S. and Foreign Carriers
426
791
1,068
521
677
902
902
U.S. Military
116
64
44
43
44
44
42
Total
542
854
1,112
564
721
946
944
Note: Totals may not sum due to independent rounding. Civilian aviation bunker fuel data were not available and were proxied
based on 2022 values.
In order to quantify the civilian international component of marine bunker fuels, activity data on
distillate diesel and residual fuel oil consumption by cargo or passenger carrying marine vessels
departing from U.S. ports were collected for individual shipping agents on a monthly basis by the U.S.
Customs and Border Protection. This information was then reported in unpublished data collected by
the Foreign Trade Division of the U.S. Department of Commerce's Bureau of the Census (DOC 1991
through 2023) for 1990 through 2001, 2007 through 2023, and the Department of Homeland Security's
Bunker Report for 2003 through 2006 (DHS 2008). Fuel consumption data for 2002 was interpolated due
to inconsistencies in reported fuel consumption data. Activity data on distillate diesel consumption by
military vessels departing from U.S. ports were provided by DLA Energy (2025). The total amount of fuel
provided to naval vessels was reduced by 21 percent to account for fuel used while the vessels were not
underway (i.e., in port). Data on the percentage of steaming hours underway versus not underway were
provided by the U.S. Navy. These fuel consumption estimates are presented in Table .
Table 3-105: Marine Fuel Consumption for International Transport (Million Gallons)
Fuel Type
1990
2005
2019
2020
2021
2022
2023
Residual Fuel Oil
4,781
3,881
2,246
1,964
1,953
2,172
2,016
Distillate Diesel Fuel & Other
617
444
702
461
437
435
423
U.S. Military Naval Fuels
5221
471
281
296
285
263
255
Total
5,920
4,796
3,229
2,721
2,674
2,870
2,694
Note: Totals may not sum due to independent rounding.
Methodological recalculations were applied to the entire time series to ensure time-series consistency
from 1990 through 2023.
Energy 3-135
-------
Uncertainty
Emission estimates related to the consumption of international bunker fuels are subject to the same
uncertainties as those from domestic aviation and marine mobile combustion emissions; however,
additional uncertainties result from the difficulty in collecting accurate fuel consumption activity data
for international transport activities separate from domestic transport activities.86 For example, smaller
aircraft on shorter routes often carry sufficient fuel to complete several flight segments without refueling
in order to minimize time spent at the airport gate or take advantage of lower fuel prices at particular
airports. This practice, called tankering, when done on international flights, complicates the use of fuel
sales data for estimating bunker fuel emissions. Tankering is less common with the type of large, long-
range aircraft that make many international flights from the United States, however. Similar practices
occur in the marine shipping industry where fuel costs represent a significant portion of overall
operating costs and fuel prices vary from port to port, leading to some tankering from ports with low fuel
costs.
Uncertainties exist with regard to the total fuel used by military aircraft and ships. Total aircraft and ship
fuel use estimates were developed from DoD records, which document fuel sold to the DoD
Components (e.g., Army, Department of Navy and Air Force) from the Defense Logistics Agency Energy.
These data may not include fuel used in aircraft and ships as a result of a Service procuring fuel from,
selling fuel to, trading fuel with, or giving fuel to other ships, aircraft, governments, or other entities.
Additionally, there are uncertainties in historical aircraft operations and training activity data. Estimates
for the quantity of fuel actually used in Navy and Air Force flying activities reported as bunker fuel
emissions had to be estimated based on a combination of available data and expert judgment.
Estimates of marine bunker fuel emissions were based on Navy vessel steaming hour data, which
reports fuel used while underway and fuel used while not underway. This approach does not capture
some voyages that would be classified as domestic for a commercial vessel. Conversely, emissions
from fuel used while not underway preceding an international voyage are reported as domestic rather
than international as would be done for a commercial vessel. There is uncertainty associated with
ground fuel estimates for 1997 through 2023, including estimates for the quantity of jet fuel allocated to
ground transportation. Small fuel quantities may have been used in vehicles or equipment other than
that which was assumed for each fuel type.
There are also uncertainties in fuel end-uses by fuel type, emissions factors, fuel densities, diesel fuel
sulfur content, aircraft and vessel engine characteristics and fuel efficiencies, and the methodology
used to back-calculate the data set to 1990 using the original set from 1995. The data were adjusted for
trends in fuel use based on a closely correlating, but not matching, data set. All assumptions used to
develop the estimate were based on process knowledge, DoD data, and expert judgments. The
magnitude of the potential errors related to the various uncertainties has not been calculated but is
believed to be small. The uncertainties associated with future military bunker fuel emission estimates
could be reduced through revalidation of assumptions based on data regarding current equipment and
operational tempo, however, it is doubtful data with more fidelity exist at this time.
Although aggregate fuel consumption data have been used to estimate emissions from aviation, the
recommended method for estimating emissions of gases other than C02 in the 2006IPCC Guidelines
(IPCC 2006) is to use data by specific aircraft type, number of individual flights and, ideally, movement
86 See uncertainty discussions under section 3.1 CCbfrom Fossil Fuel Combustion.
3-136 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
data to better differentiate between domestic and international aviation and to facilitate estimating the
effects of changes in technologies. The IPCC also recommends that cruise altitude emissions be
estimated separately using fuel consumption data, while landing and take-off (LTO) cycle data be used
to estimate near-ground level emissions of gases other than C02.87
There is also concern regarding the reliability of the existing DOC (1991 through 2024) data on marine
vessel fuel consumption reported at U.S. customs stations due to the significant degree of inter-annual
variation.
QA/QC and Verification
In order to ensure the quality of the emission estimates from international bunker fuels, General (IPCC
Tier 1) and category-specific (Tier 2) Quality Assurance/Quality Control (QA/QC) procedures were
implemented consistent with the U.S. Inventory QA/QC plan outlined in Annex 8 The Tier 2 procedures
that were implemented involved checks specifically focusing on the activity data and emission factor
sources and methodology used for estimating C02, CH4, and N20 emissions from international bunker
fuels in the United States. Emission totals for the different sectors and fuels were compared and trends
were investigated. No corrective actions were necessary.
Recalculations Discussion
No recalculations were performed for the current Inventory.
Planned Improvements
EPA will evaluate data availability to update the sources for densities, energy contents, and emission
factors applied to estimate emissions from aviation and marine fuels. Many are from sources from the
late 1990s, such as IPCC/UNEP/OECD/IEA (1997). Potential sources with more recent data include the
International Maritime Organization (IMO) greenhouse gas emission inventory, International Air
Transport Association (IATA)/ICAO greenhouse gas reporting system (CORSIA), and the EPA Greenhouse
Gas Reporting Program (GHGRP) Technical Support Document for Petroleum Products. Specifically, EPA
will evaluate data availability to support updating the heat contents and carbon contents of jet fuel with
input from EIA.
A longer-term effort is underway to consider the feasibility of including data from a broader range of
domestic and international sources for bunker fuels. Potential sources include the IMO greenhouse gas
emission inventory, data from the U.S. Coast Guard on vehicle operation currently used in criteria
pollutant modeling, data from the International Energy Agency (IEA), relevant updated FAA models to
improve aviation bunker fuel estimates, and researching newly available marine bunker data.
87 U.S. aviation emission estimates for CO, NOx, and NMVOCs are reported by EPA's National Emission Inventory (NEI) Air
Pollutant Emission Trends website, and reported under the Mobile Combustion section. It should be noted that these
estimates are based solely upon LTO cycles and consequently only capture near ground-level emissions, which are more
relevant for air quality evaluations. These estimates also include both domestic and international flights. Therefore,
estimates reported under the Mobile Combustion section overestimate IPCC-defined domestic CO, NOx, and NMVOC
emissions by including landing and take-off (LTO) cycles by aircraft on international flights, but underestimate because
they do not include emissions from aircraft on domestic flight segments at cruising altitudes.
Energy 3-137
-------
3.11 Biomass and Biofuels Consumption
(Source Category 1A)
The combustion of biomass—such as wood, charcoal, the biogenic portions of MSW, and wood waste
and biofuels such as ethanol, biogas, and biodiesel—generates C02 in addition to CH4 and N20 already
covered in this chapter. In line with the IPCC guidelines, C02 emissions from biomass and biofuel
combustion have been estimated separately from fossil fuel C02 emissions and are not directly
included in the energy sector contributions to U.S. totals. In accordance with IPCC methodological
guidelines, any such emissions are calculated by accounting for net carbon fluxes from changes in
biogenic carbon reservoirs in wooded or crop lands. For a more complete description of this
methodological approach, see the Land Use, Land-Use Change, and Forestry chapter (Chapter 6),
which accounts for the contribution of any resulting C02 emissions to U.S. totals within the Land Use,
Land-Use Change, and Forestry sector's approach.
Therefore, C02 emissions from biomass and biofuel consumption are not included specifically in
summing energy sector totals. However, they are presented here for informational purposes and to
provide detail on biomass and biofuels consumption.
In 2023, total C02 emissions from the burning of woody biomass in the industrial, residential,
commercial, and electric power sectors were approximately 187.7 MMT C02 Eq. (187,690 kt) (see Table
3-106 and Table 3-107). As the largest consumer of woody biomass, the industrial sector was
responsible for 61.8 percent of the C02 emissions from this source. The residential sector was the
second largest emitter, constituting 24.7 percent of the total, while the electric power and commercial
sectors accounted for the remainder.
Table 3-106: C02 Emissions from Wood Consumption by End-Use Sector (MMT C02
Eq.)
End-Use Sector
1990 | 2005 2019
2020
2021
2022
2023
Industrial
135.31
136.3 I
132.1
127.3
128.2
122.8
115.9
Residential
59.8
44.3
56.3
35.6
35.5
43.6
46.4
Commercial
6.8
7.2
7.7
7.5
7.5
7.5
7.4
Electric Power
13.3
19.1
20.7
19.1
20.3
20.4
17.9
Total
215.2
206.9
216.7
189.5
191.5
194.3
187.7
Note: Totals may not sum due to independent rounding.
Table 3-107: C02
Emissions from Wood Consumption by End-Use Sector (kt)
End-Use Sector
1990
2005
2019
2020
2021
2022
2023
Industrial
135,348
136,269
132,069
127,301
128,209
122,843
115,905
Residential
59,808
44,340
56,251
35,585
35,484
43,565
46,436
Commercial
6,779
7,218
7,654
7,515
7,490
7,525
7,399
Electric Power
13,252
19,074
20,677
19,115
20,288
20,385
17,950
Total
215,186
206,901
216,652
189,516
191,471
194,318
187,690
Note: Totals may not sum due to independent rounding.
3-138 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Carbon dioxide emissions from combustion of the biogenic components of MSW by the electric power
sector were an estimated 13.9 MMT C02 (13,936 kt) in 2023. Emissions across the time series are shown
in Table 3-108 and Table 3-109. As discussed in Section 3.3, MSW is combusted to produce electricity
and the C02 emissions from the fossil portion of the MSW (e.g., plastics, textiles, etc.) are included in
the energy sector FFC estimates. The MSW also includes biogenic components (e.g., food waste, yard
trimmings, natural fibers) and the C02 emissions associated with that biogenic portion is included here.
Table 3-108: C02 Emissions from Biogenic Components of MSW (MMT C02 Eq.)
End-Use Sector
1990 |
2005
2019
2020
2021
2022
2023
Electric Power
18.51
14.71
15.7
15.6
15.3
14.9
13.9
Table 3-109: C02 Emissions from Biogenic Components of MSW (kt)
End-Use Sector
1990
2005
2019
2020
2021
2022
2023
Electric Power
18,5341
14,7221
15,709
15,614
15,329
14,864
13,936
The transportation sector is responsible for most of the fuel ethanol consumption in the United States.
Ethanol used for fuel is currently produced primarily from corn grown in the Midwest, but it can be
produced from a variety of biomass feedstocks. Most ethanol for transportation use is blended with
gasoline to create a 90 percent gasoline, 10 percent by volume ethanol blend known as E-10 or gasohol,
In 2023, the United States transportation sector consumed an estimated 1,116.4 trillion Btu of ethanol
(95 percent of total), and as a result, produced approximately 76.4 MMT C02 Eq. (76,427 kt) (see Table
3-110 and Table 3-111) of C02 emissions. Smaller quantities of ethanol were also used in the industrial
and commercial sectors. Ethanol fuel production and consumption has grown significantly since 1990
due to the favorable economics of blending ethanol into gasoline and federal policies that have
encouraged use of renewable fuels.
Table 3-110: C02 Emissions from Ethanol Consumption (MMT C02 Eq.)
End-Use Sector
1990
2005
2019
2020
2021
2022
2023
Transportation®
4.1
21.6
78.7
68.1
75.4
75.0
76.4
Industrial
0.1
1.2
1.6
1.6
1.5
1.8
1.7
Commercial
0.1
0.2 |
2.2
2.2
2.1
2.8
2.6
Total
4.2
22.9
82.6
71.8
79.1
79.6
80.7
a See Annex 3.2, Table A-71 for additional information on transportation consumption of these fuels.
Note: Totals may not sum due to independent rounding.
Table 3-111: C02
Emissions from Ethanol Consumption (kt)
End-Use Sector
19901
2005
2019
2020
2021
2022
2023
Transportation®
4,059l
21,6161
78,739
68,085
75,417
74,953
76,427
Industrial
1051
1,176
1,610
1,582
1,509
1,790
1,652
Commercial
631
151 |
2,229
2,182
2,139
2,850
2,629
Total
4,2271
22,943
82,578
71,848
79,064
79,593
80,708
a See Annex 3.2, Table A-71 for additional information on transportation consumption of these fuels.
Note: Totals may not sum due to independent rounding.
Energy 3-139
-------
The transportation sector is assumed to be responsible for all of the biodiesel consumption in the
United States (EIA 2025). Biodiesel is currently produced primarily from soybean oil, but it can be
produced from a variety of biomass feedstocks including waste oils, fats, and greases. Biodiesel for
transportation use appears in low-level blends (less than 5 percent) with diesel fuel, high-level blends
(between 6 and 20 percent) with diesel fuel, and 100 percent biodiesel (EIA 2024).
In 2023, the United States consumed an estimated 246.3 trillion Btu of biodiesel, and as a result,
produced approximately 18.2 MMT C02 Eq. (18,185 kt) (see Table 3-112 and Table 3-113) of C02
emissions. Biodiesel production and consumption has grown significantly since 2001 due to the
favorable economics of blending biodiesel into diesel and federal policies that have encouraged use of
renewable fuels (EIA 2024). There was no measured biodiesel consumption prior to 2001 EIA (2025).
Table 3-112: C02 Emissions from Biodiesel Consumption (MMT C02 Eq.)
End-Use Sector 19901 20051 2019 2020 2021
2022
2023
Transportation® NO I
0.91
17.1 17.7 16.1
15.6
18.2
NO (Not Occurring)
a See Annex 3.2, Table A-71 for additional information on transportation consumption of these fuels.
Table 3-113: C02 Emissions from Biodiesel Consumption (kt)
End-Use Sector 1990
2005
2019 2020 2021
2022
2023
Transportation® NO I
8561
17,080 17,678 16,112
15,622
18,185
NO (Not Occurring)
a See Annex 3.2, Table A-71 for additional information on transportation consumption of these fuels.
Methodology and Time-Series Consistency
Woody biomass emissions were estimated by applying two gross heat contents from EIA (Lindstrom
2006) to U.S. consumption data (EIA 2025) (see Table 3-115), provided in energy units for the industrial,
residential, commercial, and electric power sectors. One heat content (16.95 MMBtu/MT wood and
wood waste) was applied to the industrial sector's consumption, while the other heat content (15.43
MMBtu/MT wood and wood waste) was applied to the consumption data for the other sectors. An EIA
emission factor of 0.434 MT C/MT wood (Lindstrom 2006) was then applied to the resulting quantities of
woody biomass to obtain C02 emission estimates. The woody biomass is assumed to contain black
liquor and other wood wastes, have a moisture content of 12 percent, and undergo complete
combustion to be converted into C02.
Data for total waste incinerated, excluding tires, from 1990 to 2023 was derived following the methodology
described in Section 3.3. Biogenic C02 emissions associated with MSW combustion were obtained from
EPA's GHGRP FLIGHT data for MSW combustion sources (EPA 2023). Dividing biogenic C02 emissions
from GHGRP FLIGHT data for MSW combustors by estimated MSW tonnage combusted yielded an annual
biogenic C02 emission factor. This approach follows the same approach used to develop the fossil C02
emissions from MSW combustion as discussed in Section 3.3. As this data was only available following
2011, all years prior use an average of the emission factors from 2011 through 2015.
Biogenic C02 emissions from MSW combustion were calculated by multiplying the annual tonnage
estimates, excluding tires, by the calculated emissions factor. Calculated biogenic C02 emission factors
are shown in Table 3-114.
3-140 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Table 3-114: Calculated Biogenic C02 Content per Ton Waste (kg C02/Short Ton
Combusted)
End-Use Sector
19901 2005 | 2019
2020
2021
2022
2023
CO2 Emission Factors
5561 5561 558
566
550
564
543
The amount of ethanol allocated across the transportation, industrial, and commercial sectors was
based on the sector allocations of ethanol-blended motor gasoline. The sector allocations of ethanol-
blended motor gasoline were determined using a bottom-up analysis conducted by EPA, as described in
the Methodology section of Fossil Fuel Combustion. Total U.S. ethanol consumption from EIA (2025)
was allocated to individual sectors using the same sector allocations as ethanol-blended motor
gasoline. The emissions from ethanol consumption were calculated by applying an emission factor of
18.67 MMT C/Qbtu (EPA 2010) to adjusted ethanol consumption estimates (see Table 3-116). The
emissions from biodiesel consumption were calculated by applying an emission factor of 20.1 MMT
C/Qbtu (EPA 2010) to U.S. biodiesel consumption estimates that were provided in energy units (EIA
2025) (see Table 3-117).88
Table 3-115: Woody Biomass Consumption by Sector (Trillion Btu)
End-Use Sector
1990
2005
2019
2020
2021
2022
2023
Industrial
1,441.91
1,451.71
1,407.0
1,356.2
1,365.9
1,308.7
1,234.8
Residential
580.0
430.0
545.5
345.1
344.1
422.5
450.3
Commercial
65.71
70.01
74.2
72.9
72.6
73.0
71.8
Electric Power
128.5
185.0
200.5
185.4
196.7
197.7
174.1
Total
2,216.2
2,136.7
2,227.2
1,959.5
1,979.4
2,001.8
1,930.9
Note: Totals may not sum due to independent rounding.
Table 3-116: Ethanol Consumption by Sector (Trillion Btu)
End-Use Sector
1990
2005
2019
2020
2021
2022
2023
Transportation
59.3
315.8
1,150.2
994.6
1,101.7
1,094.9
1,116.4
Industrial
1.5
17.2
23.5
23.1
22.0
26.2
24.1
Commercial
0.9
2.2
32.6
31.9
31.2
41.6
38.4
Total
61.7
335.1
1,206.3
1,049.5
1,155.0
1,162.7
1,179.0
Note: Totals may not sum due to independent rounding.
Table 3-117: Biodiesel Consumption by Sector (Trillion Btu)
End-Use Sector
19901
2005
2019
2020
2021
2022
2023
Transportation
NO |
I n.s|
231.3
239.4
218.2
211.6
246.3
NO (Not Occurring)
Methodological recalculations were applied to the entire time series to ensure time-series consistency
from 1990 through 2023.
88 CO2 emissions from biodiesel do not include emissions associated with the carbon in the fuel that is from the methanol
used in the process. Emissions from methanol use and combustion are assumed to be accounted for under Non-Energy
Use of Fuels. See Annex 2.3 - Methodology for Estimating Carbon Emitted from Non-Energy Uses of Fossil Fuels.
Energy 3-141
-------
Uncertainty
An uncertainty analysis was conducted for biomass and biofuel combustion using the IPCC-
recommended Approach 2 uncertainty estimation methodology. This analysis utilized the Monte Carlo
stochastic simulation software @RISK to estimate the 95 percent confidence bound around total
biomass and biofuel combustion emissions.
There are uncertainties in ethanol consumption, biodiesel consumption, woody biomass consumption,
and C02 emissions from waste incineration. It is assumed that the biodiesel and ethanol reported is 100
percent biodiesel rather than a blend. A normal distribution was assumed for all ethanol consumption,
wood consumption, and MSW input variables, while a uniform distribution was assumed for the
biodiesel emission factor. For these variables, the uncertainty ranges were assigned to the input
variables based on IPCC default uncertainty estimates (IPCC 2006) and expert opinion (ICF 2025).
The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 3-101. Biomass
and biofuel combustion C02 emissions in 2023 were estimated to be between 272.1 and 335.5 MMT C02
Eq. at a 95 percent confidence level. This indicates a range of 9 percent below to 12 percent above the
2023 emission estimate of 300.5 MMT C02 Eq.
Table 3-118: Approach 2 Quantitative Uncertainty Estimates for C02 Emissions from
Biomass and Biofuel Combustion (MMT C02 Eq. and Percent)
Uncertainty Range Relative to Emission Estimate"
2023 Emission
(MM CO2 Eq.)
(%)
Estimate
Lower
Upper
Lower
Upper
Source
Gas
(MMT CO2 Eq.)
Bound
Bound
Bound
Bound
Biomass and Biofuel Combustion
CO2
300.5
272.1
335.5
-9%
+ 12%
Recalculations Discussion
EIA (2025) updated electricity statistics which affected commercial sector wood consumption for the
years 2014 through 2022. This caused C02 emissions from commercial wood to decrease by an annual
average of 0.9 MMT C02 Eq. (11 percent) for the years 2014 and 2022, compared to estimates in the
previous Inventory.
EIA (2025) also updated ethanol consumed by all sectors in 2022, which caused C02 emissions from
industrial ethanol to decrease by 0.13 MMT C02 Eq. (6.7 percent), C02 emissions from transportation
ethanol to decrease by less than 0.05 MMT C02 Eq. (less than 0.05 percent), and C02 emissions from
commercial ethanol to increase by 0.13 MMT C02 Eq. (4.7 percent), compared to the previous Inventory.
Planned Improvements
Future research will investigate the availability of data on woody biomass heat contents and carbon
emission factors to see if there are newer, improved data sources available for these factors.
Currently, emission estimates from biomass and biomass-based fuels included in this Inventory are
limited to woody biomass, biogenic components of MSW, ethanol, and biodiesel. Additional forms of
biomass-based fuel consumption include biogas, renewable diesel and other biofuels. EPA will
investigate additional forms of biomass-based fuel consumption, research the availability of relevant
3-142 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
emissions factors, and integrate these into the Inventory as feasible. EPA will examine EIA data on
biogas and other biofuels to see if these fuel types can be included in future Inventories. EIA (2024a)
natural gas data already deducts biogas used in the natural gas supply, so no adjustments are needed to
the natural gas fuel consumption data to account for biogas. Distillate fuel statistics are adjusted in this
Inventory to remove renewable diesel fuels as well as biodiesel.
The availability of facility-level combustion emissions through EPA's GHGRP will be examined to help
better characterize the industrial sector's energy consumption in the United States and further classify
woody biomass consumption by business establishments according to industrial economic activity
type. Most methodologies used in EPA's GHGRP are consistent with IPCC, although for EPA's GHGRP,
facilities collect detailed information specific to their operations according to detailed measurement
standards, which may differ with the more aggregated data collected for the Inventory to estimate total,
national U.S. emissions. In addition, and unlike the reporting in this chapter, some facility-level fuel
combustion emissions reported under EPA's GHGRP may also include industrial process emissions.
In line with IPCC guidelines, fuel combustion emissions are included in this chapter, while process
emissions are included in the Industrial Processes and Product Use chapter of this report. In examining
data from EPA's GHGRP that would be useful to improve the emission estimates for the C02 from
biomass combustion category, particular attention will also be made to ensure time-series consistency,
as the facility-level reporting data from EPA's GHGRP are not available for all inventory years as reported
in this Inventory. Additionally, analyses will focus on aligning reported facility-level fuel types and IPCC
fuel types per the national energy statistics, ensuring C02 emissions from biomass are separated in the
facility-level reported data, and maintaining consistency with national energy statistics provided by EIA.
In implementing improvements and integration of data from EPA's GHGRP, the latest guidance from the
IPCC on the use of facility-level data in national inventories will be relied upon.89
Lastly, the C02 emission factor for wood developed by NESCAUM (2024) will be reviewed and potentially
incorporated based on this review.
3.12 Energy Sources of Precursor
Greenhouse Gases
In addition to the main greenhouse gases addressed above, energy-related activities are also sources of
greenhouse gas precursors. This section summarizes information on precursor emissions, which
include carbon monoxide (CO), nitrogen oxides (NOx), non-methane volatile organic compounds
(NMVOCs), ammonia (NH3), and sulfur dioxide (S02). These gases are not direct greenhouse gases, but
indirectly impact Earth's radiative balance by altering the concentrations of greenhouse gases (e.g.,
tropospheric ozone) and atmospheric aerosol (e.g., particulate sulfate). Total emissions of NOx, CO,
NMVOCs, NH3 ,and S02 from energy-related activities from 1990 to 2023 are reported in Table 3-119.
89 See http://www.ipcc-nggip.iges.or.jp/public/tb/TFI Technical Bulletin 1.pdf.
Energy 3-143
-------
Table 3-119: NOx, CO, NMVOC, NH3, and S02 Emissions from Energy-Related Activities
(kt)
Gas/Activity
1990
2005
2019
2020
2021
2022
2023
NOx
21,805
18,760
6,986
6,181
6,243
6,045
5,765
Fossil Fuel Combustion
21,678
18,188
6,496
5,626
5,557
5,405
5,125
Transportations
12,132
12,628
4,322
3,618
3,543
3,375
3,195
Industrial
2,475
1,486
800
753
720
727
730
Electric Power Sector
6,045
3,440
898
761
806
778
676
Commercial
451
288
187
192
188
206
206
Residential
575
346
290
300
300
318
318
Petroleum and Natural Gas
Systems
127
572
491
556
685
640
640
International Bunker Fuels
1,953
1,699
1,280
977
1,008
1,132
1,077
CO
124,583
64,319
30,258
28,316
28,704
27,889
27,360
Fossil Fuel Combustion
124,351
63,686
29,660
27,706
28,073
27,213
26,685
Transportations
119,478
59,540
25,621
23,546
23,889
23,003
22,526
Residential
3,620
2,393
2,860
2,968
2,950
2,960
2,960
Industrial
704
976
600
673
659
659
658
Electric Power Sector
329
582
428
361
423
422
371
Commercial
220
195
151
157
153
169
169
Petroleum and Natural Gas
Systems
232
632
599
610
630
676
676
International Bunker Fuels
102
131
150
83
101
128
127
NMVOCs
12,269
8,081
4,987
4,822
5,167
5,045
4,914
Fossil Fuel Combustion
11,793
6,079
2,593
2,391
2,454
2,329
2,198
Transportationa
10,932
5,608
2,072
1,846
1,912
1,786
1,655
Residential
693
322
397
431
429
429
429
Commercial
9
18
14
14
14
14
14
Industrial
117
87
81
74
73
74
74
Electric Power Sector
43
44
29
26
27
27
27
Petroleum and Natural Gas
Systems
476
2,002
2,394
2,431
2,713
2,716
2,716
International Bunker Fuels
57
54
45
32
34
40
38
NH3
190
218
177
178
265
269
266
Fossil Fuel Combustion
190
218
177
178
265
266
262
Transportation
169
153
96
84
171
172
168
Residential
4
17
52
63
63
66
66
Commercial
2
5
2
2
2
2
2
Industrial
15
20
12
13
13
12
12
Electric Power Sector
0
23
16
17
17
15
15
Petroleum and Natural Gas
Systems
+
+
+
+
+
4
4
International Bunker Fuels
NA
NA
NA
NA
NA
NA
NA
S02
21,638
13,331
1,509
1,289
1,423
1,422
1,230
Fossil Fuel Combustion
21,482
13,235
1,447
1,139
1,273
1,143
951
Electric Power Sector
14,432
9,436
921
758
898
819
627
Industrial
2,886
1,378
234
173
169
143
142
Transportations
793
724
40
23
24
26
26
Commercial
485
318
19
13
14
14
14
Residential
2,886
1,378
234
173
169
143
142
Petroleum and Natural Gas
Systems
156
96
61
150
150
279
279
International Bunker Fuels
NA
NA
NA
NA
NA
NA
NA
NA (Not Applicable)
aThe scope of the NEI for aircraft related precursor emissions included under the transportation is different from the Inventory
reporting scope. The NEI precursor estimate methodology does not exclude emissions that could be considered international
bunkers given local impacts from these emissions. The precursor estimates are modeled using FAA- and state-supplied landing
3-144 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
and take-off data for all aircraft types (including ground support equipment and auxiliary engines) used for public, private, and
military purposes.
Note: Totals may not sum due to independent rounding.
Source: (EPA 2023a). Emission categories from EPA (2023) are aggregated into sectors and categories reported as shown in Table
ES-3.
Methodology and Time-Series Consistency
Emission estimates for 1990 through 2023 were obtained from data published on the National
Emissions Inventory (NEI) Air Pollutant Emissions Trends Data website (EPA 2024). For Table 3-119, NEI
reported emissions of CO, NOx, NMVOCs, NH3, and S02 were recategorized from NEI Emissions
Inventory System (EIS) sectors to source categories more closely aligned with sectors and categories in
this report, based on discussions between the EPA Inventory and NEI staff (see crosswalk documented
in Annex 6.3).90 EIS sectors mapped to the energy sector categories in this report include: fuel
combustion for electric utilities, industrial, and other; petroleum and related industries; highway
vehicles; off-highway; and other mobile sources (e.g., commercial marine vessels and rail). As
described in the NEI Technical Support Documentation (TSD) (EPA 2023b), NEI emissions are estimated
through a combination of emissions data submitted directly to the EPA by state, local, and tribal air
agencies, as well as additional information added by the Agency from EPA emissions programs, such as
the emission trading program, Toxics Release Inventory (TRI), and data collected during rule
development or compliance testing.
Methodological approaches were applied to the entire time series to ensure time-series consistency
from 1990 through 2022, which are described in detail in the NEI's TSD and on EPA's Air Pollutant
Emission Trends website (EPA 2023b; EPA 2024). No quantitative estimates of uncertainty were
calculated for this source category.
90 The NEI estimates and reports emissions from six criteria air pollutants (CAPs) and 187 hazardous air pollutants (HAPs) in
support of National Ambient Air Quality Standards. EPA reported CAP emission trends are grouped into 60 sectors and 15
Tier 1 source categories, which broadly cover similar source categories to those presented in this chapter. For reporting
precursor emissions in the common data tables, EPA has mapped and regrouped emissions of greenhouse gas
precursors (CO, NOx, SO2, and NMVOCs) from NEI's EIS sectors to better align with NIR source categories, and to ensure
consistency and completeness to the extent possible. See Annex 6.3 for more information on this mapping.
Energy 3-145
-------
Industrial Processes
and Product Use |IPPU|
Emissions
-------
4 Industrial Processes and Product
Use
Industrial Processes and Product Use (IPPU) chapter includes greenhouse gas emissions occurring from
industrial processes and from the use of greenhouse gases in products. The industrial processes and
product use categories included in this chapter are presented in Figure 4-1 and Figure 4-2. Greenhouse
gas emissions from industrial processes can occur in two different ways. First, they may be generated
and emitted as the byproducts of various non-energy-related industrial activities. Second, they may be
emitted due to their use in manufacturing processes or by end-consumers. Combustion-related energy
use emissions from industry are reported in Chapter 3, Energy.
In the case of byproduct emissions, the emissions are generated by an industrial process itself and are
not directly a result of energy consumed during the process. For example, raw materials can be
chemically or physically transformed from one state to another. This transformation can result in the
release of greenhouse gases such as carbon dioxide (C02), methane (CH4), nitrous oxide (N20), and
fluorinated greenhouse gases (e.g., HFC-23). The greenhouse gas byproduct generating processes
included in this chapter include iron and steel production and metallurgical coke production, cement
production, petrochemical production, ammonia production, lime production, other process uses of
carbonates (e.g., flux stone, flue gas desulfurization, ceramics production, non-metallurgical magnesia
production, and soda ash consumption not associated with glass manufacturing), nitric acid
production, adipic acid production, urea consumption for non-agricultural purposes, aluminum
production, HCFC-22 production, other fluorochemical production, glass production, soda ash
production, ferroalloy production, titanium dioxide production, caprolactam production, zinc
production, phosphoric acid production, lead production, and silicon carbide production and
consumption.
Greenhouse gases that are used in manufacturing processes or by end-consumers include man-made
compounds such as hydrofluorocarbons (HFCs), perfluorocarbons (PFCs), sulfur hexafluoride (SF6), and
nitrogen trifluoride (NF3). The present contribution of HFCs, PFCs, SF6, and NF3 gases to the radiative
forcing effect of all anthropogenic greenhouse gases is small; however, because of their extremely long
lifetimes, many of them will continue to persist in the atmosphere long after they were first released. In
addition, many of these gases have high global warming potentials (GWPs); SF6 is the most potent
greenhouse gas the Intergovernmental Panel on Climate Change (IPCC) has evaluated. Use of HFCs
continues since they are the primary substitutes for ozone depleting substances (ODS), which are being
globally phased-out under the Montreal Protocol on Substances that Deplete the Ozone Layer; however,
production and consumption of HFCs are being phased down as well. Hydrofluorocarbons (HFCs),
PFCs, SF6, and NF3 are employed and emitted by a number of other industrial sources in the United
States, such as the electronics industry, electric power transmission and distribution, PFCs and SF6for
other product use, and magnesium metal production and processing. Carbon dioxide is also consumed
and emitted through various end-use applications. In addition, nitrous oxide is used in and emitted by
the electronics industry and anesthetic and aerosol applications.
4-2 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
In 2023, IPPU generated emissions of 385.7 million metric tons of C02 equivalent (MMT C02 Eq.), or 6,2
percent of total U.S. greenhouse gas emissions.1 Carbon dioxide emissions from all industrial
processes were 165.5 MMT C02 Eq. (165,533 kt C02) in 2023, or 3.4 percent of total U.S. C02 emissions.
Methane emissions from industrial processes resulted in emissions of approximately 0.04 MMT C02 Eq.
(1 kt CH4) in 2023, which was 0.01 percent of U.S. CH4 emissions. Nitrous oxide emissions from IPPU
were 14.9 MMT C02 Eq. (56 kt N20) in 2023, or 3.8 percent of total U.S. N20 emissions. In 2023
combined emissions of HFCs, PFCs, SF6, and NF3 totaled 205.3 MMT C02 Eq. Total emissions from IPPU
in 2023 were 4.6 percent more than 1990 emissions. Total emissions from IPPU remained relatively
constant between 2022 and 2023, decreasing by 1.0 percent due to offsetting trends within the sector.
More information on emissions of greenhouse gas precursors emissions that also result from IPPU are
presented in Section 4.29 of this chapter.
The largest source of IPPU-related emissions is the substitution of ozone depleting substances, which
accounted for 49.0 percent of sector emissions in 2023. These emissions have increased by 84.1
percent since 2005 and by 2.3 percent between 2022 and 2023. Iron and steel production and
metallurgical coke production was the second largest source of IPPU emissions in 2023, accounting for
12.0 percent of IPPU emissions in 2023. Cement production was the third largest source of IPPU
emissions, accounting for 10.5 percent of the sector total in 2023.
Figure 4-1: Industrial Processes and Product Use Sector Greenhouse Gas Sources
Substitution of Ozone Depleting Substances
Iron and Steel Production & Metallurgical Coke Production
Cement Production
Petrochemical Production
Ammonia Production
Lime Production
Nitric Acid Production
Other Process Uses of Carbonates
Urea Consumption for Non-Agricultural Purposes
Electrical Equipment
Fluorochemical Production
Electronics Industry
N2O from Product Uses
Non-EOR Carbon Dioxide Utilization
Glass Production
Soda Ash Production
Aluminum Production
Caprolactam, Glyoxal, and Glyoxylic Acid Production
Ferroalloy Production
Titanium Dioxide Production
Adipic Acid Production
Magnesium Production and Processing
Other Product Manufacture and Use1
Zinc Production
Phosphoric Acid Production
Lead Production
Carbide Production and Consumption
1189
Industrial Processes and Product
Use as a Portion of All Emissions
I Energy
Agriculture
I IPPU
i Waste
< 0.5
< 0.5
10
20
30
40
50
60
70
MMT CO2 Eq.
1 Other product manufacture and use includes SFe and PFCs from other product use, including the uses of HFCs and NF.s in those
applications. Other product manufacture and use categories are shown separately (e.g. electrical equipment, etc.).
The increase in overall IPPU emissions since 1990 reflects a range of emission trends among the
emission sources, as shown in Figure 4-2. Emissions resulting from most types of metal production have
1 Emissions reported in the IPPU chapter include those from all 50 states, including Hawaii and Alaska, as well as from
U.S. Territories.
Industrial Processes and Product Use 4-3
-------
declined significantly since 1990, largely due to production shifting to other countries, but also due to
transitions to less-emissive methods of production (in the case of iron and steel) and to improved
practices (in the case of PFC emissions from aluminum production). Carbon dioxide and CH4 emissions
from some chemical production sources (e.g., petrochemical production, urea consumption for non-
agricultural purposes) have increased since 1990, while emissions from other chemical production
sources (e.g., ammonia production, phosphoric acid production) have decreased. Emissions from
mineral sources have either increased (e.g., cement production) or not changed significantly (e.g., lime
production) since 1990 and largely follow economic cycles. HFC emissions from the substitution of ODS
have increased drastically since 1990 and are the largest source of IPPU emissions (49.0 percent in
2023), while the emissions of HFCs, PFCs, SF6, and NF3 from other sources have generally declined.
Nitrous oxide emissions from the production of nitric acid have decreased. Some emission sources
(e.g., adipic acid) exhibit varied interannual trends. Trends are explained further within each emission
source category throughout the chapter.
Figure 4-2: Trends in Industrial Processes and Product Use Sector Greenhouse Gas
Sources
500
450
I Electronics Industry
Other Product Manufacture and Use
I Mineral Industry
I Metal Industry
Chemical Industry
I Substitution of Ozone Depleting Substances
400 en
KD
350
300
250
200
150
100
50
0
. oo r~»
rr lo lti
£
cn
rsi
rN co h n co
r-v m r-»
m m m
-H O
CTi
Table 4-1 summarizes emissions for the IPPU chapter in MMT C02 Eq. using IPCC Fifth Assessment
Report (AR5) GWP values (IPCC 2013).Unweighted gas emissions in kt are also provided in Table 4-2.
The source descriptions that follow in the chapter are presented in the order consistent with national
inventory reporting guidelines, corresponding generally to: mineral industry, chemical industry, metal
industry, and emissions from the uses of HFCs, PFCs, SF6, and NF3.
Each year, some emission and sink estimates in the IPPU sector of the Inventory are recalculated and
revised with improved methods and/or data. In general, recalculations are made to the U.S. greenhouse
gas emission estimates either to incorporate new methodologies or, most commonly, to update recent
4-4 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
historical data. These improvements are implemented consistently across the previous Inventory's time
series (i.e., 1990 to 2022) to ensure that the trend is accurate. Key updates to this year's Inventory
include the incorporation of more complete activity data from the Greenhouse Gas Reporting Program
(GHGRP) within the non-EOR carbon dioxide utilization category; revisions to the method for estimating
emissions from production of fluorochemicals other than HCFC-22, specifically for emissions of gases
that are only reported by fluorinated GHG group from production and transformation processes; and
inclusion of minor uses of NF3 and HFCs under other product manufacture and use category to improve
completeness. Together, these methodological and other routine annual data updates increased IPPU
sector greenhouse gas emissions by an average 0.1 MMT C02 Eq. (less than 0.1 percent) across the time
series. For more information on specific methodological updates, please see the Recalculations
Discussion section for each category in this chapter.
Table 4-1: Emissions from Industrial Processes and Product Use (MMT C02 Eq.)
Gas/Source
1990
2005
2019
2020
2021
2022
2023
C02
213.7
195.9
169.5
161.0
171.4
169.0
165.5
Iron and Steel Production &
Metallurgical Coke Production
104.7
70.1
46.8
40.7
47.2
45.2
46.2
Iron and Steel Production
99.7
66.2
43.8
38.3
44.0
42.2
43.3
Metatturgicat Coke Production
5.6
3.9
3.0
2.3
3.2
3.0
3.0
Cement Production
33.5
46.2
40.9
40.7
41.3
41.9
40.6
Petrochemical Production
20.1
26.9
28.5
27.9
30.7
28.8
30.5
Ammonia Production
14.4
10.2
12.4
12.3
11.5
11.9
12.2
Lime Production
11.7
14.6
12.1
11.3
11.9
12.2
11.5
Other Process Uses of Carbonates
7.1
8.5
9.0
9.0
8.6
10.4
7.2
Urea Consumption for Non-
Agricultural Purposes
3.8
3.7
6.2
5.9
6.7
5.5
5.4
Non-EOR Carbon Dioxide Utilization
1.5
1.4
2.4
2.8
2.9
2.8
2.1
Glass Production
2.3
2.4
1.9
1.9
2.0
2.0
1.8
Soda Ash Production
1.4
1.7
1.8
1.5
1.7
1.7
1.7
Ferroalloy Production
2.2
1.4
1.6
1.4
1.4
1.3
1.2
Aluminum Production
6.8
4.1
1.9
1.7
1.5
1.4
1.2
Titanium Dioxide Production
1.2
1.8
1.3
1.3
1.5
1.5
1.2
Zinc Production
0.6
1.0
1.0
1.0
1.0
0.9
0.9
Phosphoric Acid Production
1.5
1.3
0.9
0.9
0.9
0.8
0.9
Lead Production
0.5
0.6
0.5
0.5
0.5
0.5
0.5
Carbide Production and
Consumption
0.2
0.2
0.2
0.2
0.2
0.2
0.2
Substitution of Ozone Depleting
Substances®
+
+
+
+
+
+
+
Magnesium Production and
Processing
0.1
+
+
+
+
+
+
CH4
0.1
+
+
+
+
+
+
Carbide Production and
Consumption
+
+
+
+
+
+
+
Ferroalloy Production
+
+
+
+
+
+
+
Iron and Steel Production &
Metallurgical Coke Production
+
+
+
+
+
+
+
Industrial Processes and Product Use 4-5
-------
Gas/Source
1990
2005
2019
2020
2021
2022
2023
Petrochemical Production
+
+
+
+
+
+
+
N20
29.6
22.2
18.7
20.8
19.7
16.1
14.9
Nitric Acid Production
10.8
10.1
8.9
8.3
7.9
8.6
8.3
N2O from Product Uses
3.8
3.8
3.8
3.8
3.8
3.8
3.8
Caprolactam, Glyoxal, and Glyoxylic
Acid Production
1.5
1.9
1.2
1.1
1.2
1.3
1.3
Adipic Acid Production
13.5
6.3
4.7
7.4
6.6
2.1
1.2
Electronics Industry
+
0.1
0.2
0.3
0.3
0.3
0.3
HFCs
47.8
125.0
175.8
177.8
184.3
189.5
191.0
Substitution of Ozone Depleting
Substances®
0.3
102.7
169.7
173.7
179.9
184.8
189.0
Fluorochemical Production
47.3
22.2
5.8
3.8
4.0
4.3
1.7
Electronics Industry
0.2
0.2
0.3
0.3
0.4
0.3
0.3
Magnesium Production and
Processing
0.0
0.0
0.1
0.1
+
+
+
Other Product Manufacture and Use0
0.0
0.0
0.0
+
0.0
0.0
0.0
PFCs
39.7
10.3
7.3
6.6
6.3
6.5
5.8
Fluorochemical Production
17.7
4.1
3.0
2.5
2.6
2.8
2.7
Electronics Industry
2.5
3.0
2.7
2.6
2.7
2.8
2.4
Aluminum Production
19.3
3.1
1.4
1.4
0.9
0.8
0.5
Other Product Manufacture and Use0
0.1
0.1
0.1
0.1
0.1
0.1
0.2
Substitution of Ozone Depleting
Substances®
0.0
+
+
+
+
+
+
Electrical Equipment
+
+
+
+
+
+
0.0
SF6
37.9
20.2
8.3
7.7
8.0
7.2
7.7
Electrical Equipment
24.6
11.8
6.0
5.5
5.5
4.9
5.1
Magnesium Production and
Processing
5.6
3.0
0.9
0.9
1.2
1.1
1.1
Other Product Manufacture and Use0
1.3
1.3
0.6
0.5
0.4
0.5
0.8
Electronics Industry
0.5
0.8
0.8
0.8
0.9
0.8
0.7
Fluorochemical Production
5.8
3.3
+
+
+
+
+
NF3
0.2
1.0
1.1
1.3
1.1
1.1
0.8
Electronics Industry
+
0.4
0.5
0.6
0.6
0.6
0.5
Fluorochemical Production
0.1
0.6
0.6
0.7
0.5
0.5
0.3
Other Product Manufacture and Use0
+
+
+
+
+
+
0.0
Totalb
368.9
374.7
380.8
375.3
390.9
389.6
385.7
+ Does not exceed 0.05 MMT C02 Eq.
a Small amounts of PFC emissions from this source are included under HFCs due to confidential business information.
b Total does not include other fluorinated gases, such as HFEs and PFPEs, which are reported separately in Section 4.24.
c Emissions included in section 4.27 of this chapter.
Notes: Totals may not sum due to independent rounding. Emissions of F-HTFs that are not HFCs, PFCs or SF6 are not included in
Inventory totals and are included for informational purposes only in Section 4.24. Emissions presented for informational
purposes include HFEs, PFPMIEs, perfluoroalkylmorpholines, and perfluorotrialkylamines.
4-6 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Table 4-2: Emissions from Industrial Processes and Product Use (kt)
Gas/Source
1990
2005
2019
2020
2021
2022
2023
CO2
213,681
195,930
169,519
160,995
171,433
169,037
165,533
Iron and Steel Production &
Metallurgical Coke Production
104,738
70,078
46,835
40,675
47,218
45,157
46,240
Iron and Steel Production
99,130
66,158
43,829
38,350
43,994
42,202
43,254
Metatturgicat Coke
Production
5,608
3,921
3,006
2,325
3,224
2,954
2,986
Cement Production
33,484
46,194
40,896
40,688
41,312
41,884
40,636
Petrochemical Production
20,075
26,882
28,483
27,926
30,656
28,788
30,540
Ammonia Production
14,404
10,234
12,388
12,335
11,458
11,945
12,211
Lime Production
11,700
14,552
12,112
11,299
11,870
12,208
11,548
Other Process Uses of
Carbonates
7,103
8,472
8,973
9,012
8,583
10,383
7,163
Urea Consumption for Non-
Agricultural Purposes
3,784
3,653
6,234
5,905
6,724
5,464
5,424
Non-EOR Carbon Dioxide
Utilization
1,472
1,375
2,415
2,842
2,889
2,812
2,150
Glass Production
2,263
2,402
1,940
1,858
1,969
1,956
1,774
Soda Ash Production
1,431
1,655
1,792
1,461
1,714
1,704
1,723
Ferroalloy Production
2,152
1,392
1,598
1,377
1,426
1,327
1,245
Aluminum Production
6,831
4,142
1,880
1,748
1,541
1,446
1,237
Titanium Dioxide Production
1,195
1,755
1,340
1,340
1,541
1,541
1,233
Zinc Production
632
1,030
1,026
977
1,007
947
920
Phosphoric Acid Production
1,529
1,342
909
901
874
804
850
Lead Production
516
553
518
491
473
455
450
Carbide Production and
Consumption
243
213
175
154
172
210
183
Substitution of Ozone Depleting
Substances®
+
1
3
4
4
4
4
Magnesium Production and
Processing
129
4
2
3
3
3
2
CH4
3
2
1
1
1
1
1
Carbide Production and
Consumption
1
+
+
+
+
+
+
Ferroalloy Production
1
+
+
+
+
+
+
Iron and Steel Production &
Metallurgical Coke Production
1
1
+
+
+
+
+
Petrochemical Production
+
+
+
+
+
+
+
N2O
111.7
83.7
70.8
78.6
74.4
60.7
56.1
Nitric Acid Production
41
38
34
31
30
33
32
N2O from Product Uses
14
14
14
14
14
14
14
Caprolactam, Glyoxal, and
Glyoxylic Acid Production
6
1
5
4
5
5
5
AdipicAcid Production
51
24
18
28
25
8
4
Electronics Industry
+
+
1
1
1
1
1
HFCs
M
M
M
M
M
M
M
Industrial Processes and Product Use 4-7
-------
Gas/Source
1990
2005
2019
2020
2021
2022
2023
Substitution of Ozone Depleting
Substances8
M
M
M
M
M
M
M
Fluorochemical Production
M
M
M
M
M
M
M
Electronics Industry
+
+
+
+
+
+
+
Magnesium Production and
Processing
0
0
+
+
0
+
0
Other Product Manufacture and
Useb
0
0
0
+
0
0
0
PFCs
M
M
M
M
M
M
M
Fluorochemical Production
2
+
+
+
+
+
+
Electronics Industry
+
+
+
+
+
+
+
Aluminum Production
M
M
M
M
M
M
M
Other Product Manufacture and
Useb
M
M
M
M
M
M
M
Substitution of Ozone Depleting
Substances8
0
+
+
+
+
+
+
Electrical Equipment
+
+
+
+
+
+
0
SF6
2
1
+
+
+
+
+
Electrical Equipment
1
1
+
+
+
+
+
Magnesium Production and
Processing
+
+
+
+
+
+
+
Other Product Manufacture and
Useb
+
+
+
+
+
+
+
Electronics Industry
+
+
+
+
+
+
+
Fluorochemical Production
+
+
+
+
+
+
+
NF3
+
+
+
+
+
+
+
Electronics Industry
+
+
+
+
+
+
+
Fluorochemical Production
+
+
+
+
+
+
+
Other Product Manufacture and
Useb
+
+
+
+
+
+
0
+ Does not exceed 0.5 kt.
M (Mixture of gases)
a Small amounts of PFC emissions from this source are included under HFCs due to confidential business information.
b Emissions included in section 4.27 of this chapter.
Note: Totals by gas may not sum due to independent rounding.
This chapter presents emission estimates calculated in accordance with the 2006IPCC Guidelines for
National Greenhouse Gas Inventories (2006 IPCC Guidelines) and its refinements. For additional detail
on IPPU sources that are not included in this Inventory report, please review Annex 5, Assessment of the
Sources and Sinks of Greenhouse Gas Emissions Not Included. These sources are not included due to
various national circumstances, such as emissions from a source may not currently occur in the United
States, data are not currently available for those emission sources (e.g., glyoxal and glyoxylic acid
production, CH4 from direct reduced iron production), emissions are included elsewhere within the
Inventory report, or data suggest that emissions are not significant (e.g., CH4 and N20 emissions from
petrochemical and carbon black production). In terms of geographic scope, emissions reported in the
IPPU chapter include those from all 50 states, including Hawaii and Alaska, as well as from District of
Columbia, U.S. Territories, and tribal lands to the extent to which industries are occurring. While most
IPPU sources do not occur in U.S. Territories (e.g., electronics manufacturing does not occur in U.S.
4-8 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Territories), they are estimated and accounted for where they are known to occur (e.g., cement
production, lime production, electrical equipment). EPA will review this on an ongoing basis to ensure
emission sources are included across all geographic areas if they occur. Information on planned
improvements for specific IPPU source categories can be found in the Planned Improvements section of
the individual source category.
In addition, as mentioned in the Energy chapter of this report (Box 3-5), fossil fuels consumed for non-
energy uses for primary purposes other than combustion for energy (including lubricants, paraffin
waxes, bitumen asphalt, and solvents) are reported in the Energy chapter. According to the 2006IPCC
Guidelines, these non-energy uses of fossil fuels are to be reported under the IPPU, rather than the
Energy sector; however, due to national circumstances regarding the allocation of energy statistics and
carbon balance data, the United States reports these non-energy uses in the Energy chapter of this
Inventory. Although emissions from these non-energy uses are reported in the Energy chapter, the
methodologies used to determine emissions are compatible with the 2006 IPCC Guidelines and are well
documented and scientifically based. The methodologies used are described in Section 3.2, Carbon
Emitted from Non-Energy Uses of Fossil Fuels and Annex 2.3, Methodology for Estimating Carbon
Emitted from Non-Energy Uses of Fossil Fuels. The emissions are reported under the Energy chapter to
improve transparency, report a more complete carbon balance, and avoid double counting. For
example, only the emissions from the first use of lubricants and waxes are to be reported underthe IPPU
sector, and emissions from use of lubricants in 2-stroke engines and emissions from secondary use of
lubricants and waxes in waste incineration with energy recovery are to be reported under the Energy
sector. Reporting non-energy use emissions from only first use of lubricants and waxes under IPPU
would involve making artificial adjustments to the non-energy use carbon balance and could potentially
result in double counting of emissions. These artificial adjustments would also be required for asphalt
and road oil and solvents (which are captured as part of petrochemical feedstock emissions) and could
also potentially result in double counting of emissions. For more information, see the Methodology
discussions in Section 3.1, C02 from Fossil Fuel Combustion, Section 3.2, Carbon Emitted from Non-
Energy Uses of Fossil Fuels and Annex 2.3, Methodology for Estimating Carbon Emitted from Non-
Energy Uses of Fossil Fuels.
Finally, as stated in the Energy chapter, portions of the fuel consumption data for seven fuel
categories—coking coal, distillate fuel, industrial other coal, petroleum coke, natural gas, residual fuel
oil, and other oil—are reallocated to the IPPU chapter, as they are consumed during non-energy related
industrial process activity. Emissions from the use of fossil fuels as feedstocks or reducing agents (e.g.,
petrochemical production, aluminum production, titanium dioxide, zinc production) are reported in the
IPPU chapter, unless otherwise noted due to specific national circumstances. This approach is
compatible with the 2006 IPCC Guidelines and is well documented and scientifically based. The
emissions from these feedstocks and reducing agents are reported under the IPPU chapter to improve
transparency and to avoid double counting of emissions under both the Energy and IPPU sectors. More
information on the methodology to adjust for these emissions within the Energy chapter is described in
the Methodology section of C02 from Fossil Fuel Combustion (3.1 Fossil Fuel Combustion [Source
Category 1 A]) and Annex 2.1 Methodology for Estimating Emissions of C02 from Fossil Fuel Combustion.
Additional information is listed within each IPPU emission source in which this approach applies.
Industrial Processes and Product Use 4-9
-------
Box 4-1: Uses of EPA's Greenhouse Gas Reporting Program Energy Data
EPA's Greenhouse Gas Reporting Program (GHGRP) dataset continues to be an important resource for
the Inventory, providing not only annual emissions information, but also other annual information, such
as activity data and emission factors that can improve and refine national emission estimates and
trends over time. GHGRP data also allow EPA to disaggregate national inventory estimates in new ways
that can highlight differences across regions and sub-categories of emissions, along with enhancing
application of QA/QC procedures and assessment of uncertainties.
EPA collects greenhouse gas emissions data from individual facilities and suppliers of certain fossil
fuels and industrial gases through its Greenhouse Gas Reporting Program (GHGRP). The GHGRP applies
to direct greenhouse gas emitters, fossil fuel suppliers, industrial gas suppliers, and facilities that inject
C02 underground for sequestration or other reasons and requires reporting by sources or suppliers in 41
industrial categories. Annual reporting is at the facility level, except for certain suppliers of fossil fuels
and industrial greenhouse gases. In general, the threshold for reporting is 25,000 metric tons or more of
C02 Eq. per year, but reporting is required for all facilities in some industries. Calendar year 2010 was
the first year for which data were collected for facilities subject to 40 CFR Part 98, though some source
categories first collected data for calendar year 2011.
EPA uses annual GHGRP data in a number of categories to improve the national estimates presented in
this Inventory, consistent with IPCC guidelines (e.g., minerals, chemicals, product uses). Methodologies
used in EPA's GHGRP are consistent with IPCC guidelines, including higher tier methods; however, it
should be noted that the coverage and definitions for source categories (e.g., allocation of energy and
IPPU emissions) in EPA's GHGRP may differ from those used in this Inventory and is an important
consideration when incorporating GHGRP data in the Inventory. This report is a comprehensive
accounting of all emissions from source categories identified in the 2006 IPCC Guidelines. EPA has paid
particular attention to ensuring both completeness and time-series consistency for major
recalculations that have occurred from the incorporation of GHGRP data into these categories,
consistent with 2006 IPCC Guidelines and the 2019 Refinement, Volume 1, Chapter 2, Section 2.3, Use
of Facility Data in Inventories.2 For certain source categories in this Inventory (e.g., nitric acid
production, lime production, cement production, petrochemical production, non-EOR carbon dioxide
utilization, ammonia production, and urea consumption for nonagricultural purposes), EPA has
integrated activity factors that have been derived from aggregated GHGRP data using criteria to confirm
that a given data aggregation shields underlying CBI from public disclosure and only publishing data
values that meet these aggregation criteria.3 Specific uses of aggregated facility-level data are described
in the respective methodological sections (e.g., including other sources using GHGRP data that is not
aggregated CBI, such as aluminum, electronics industry, electrical equipment, HCFC-22 production,
and magnesium production and processing). For other source categories in this chapter, as indicated in
the respective planned improvements sections, EPA is continuing to analyze how facility-level GHGRP
data may be used to improve the national estimates presented in this Inventory, giving particular
consideration to ensuring time-series consistency and completeness.
2 See https://www.ipcc-nggip.iges.or.jp/public/2019rf/pdf/1_Volume1/19R_V1_Ch02_DataCollection.pdf.
3 U.S. EPA Greenhouse Gas Reporting Program. Developments on Publication of Aggregated Greenhouse Gas Data,
November 25, 2014. See http://www.epa.gov/ghgreporting/confidential-business-information-ghg-reporting
4-10 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Additionally, EPA's GHGRP has and will continue to enhance QA/QC procedures and assessment of
uncertainties within the IPPU categories (see those categories for specific QA/QC details regarding the
use of GHGRP data).
4.1 Cement Production (Source Category
2A1)
Cement production is an energy- and raw material-intensive process that results in the generation of
carbon dioxide (C02) both from the energy consumed in making the clinker precursor to cement and
from the chemical process to make the clinker. This reporting category (2A1) includes emissions from
production of clinker and use of cement kiln dust. Per the IPCC methodological guidance, emissions
from fuels consumed for energy purposes during the production of cement are accounted for as part of
fossil fuel combustion in the industrial end-use sector reported under the Energy chapter.
During the clinker production process, the key reaction occurs when calcium carbonate (CaC03), in the
form of limestone or similar rocks or in the form of cement kiln dust (CKD), is heated in a cement kiln at
a temperature range of about 700 to 1,000 degrees Celsius (1,300 to 1,800 degrees Fahrenheit) to form
lime (i.e., calcium oxide, or CaO) and C02 in a process known as calcination or calcining. The quantity of
C02 emitted during clinker production is directly proportional to the lime content of the clinker. During
calcination, each mole of CaC03 heated in the clinker kiln forms one mole of CaO and one mole of C02.
The C02 is vented to the atmosphere as part of the kiln exhaust:
CaC03 + heat -> CaO +C02
Next, over a temperature range of 1,000 to 1,450 degrees Celsius, the CaO combines with alumina, iron
oxide and silica that are also present in the clinker raw material mix to form hydraulically reactive
compounds within white-hot semifused (sintered) nodules of clinker. These "sintering" reactions are
highly exothermic and produce few C02 process emissions. The clinker is then rapidly cooled to
maintain quality and then very finely ground with a small amount of gypsum and potentially other
materials (e.g., ground granulated blast furnace slag, etc.) to make portland and similar cements.
Masonry cement consists of plasticizers (e.g., ground limestone, lime, etc.) and portland cement, and
the amount of portland cement used accounts for approximately 3 percent of total clinker production
(USGS 2024b; 2024c). No additional emissions are associated with the production of masonry cement.
Carbon dioxide emissions that result from the production of lime used to produce portland and masonry
cement are included in Section 4.2.
Carbon dioxide emitted from the chemical process of cement production is the second largest source of
industrial C02 emissions in the United States. Cement is produced in 34 states and Puerto Rico. Texas,
Missouri, California, and Florida were the leading cement-producing states in 2023 and accounted for
approximately 43 percent of total U.S. production (USGS 2024b). In 2023, shipments of cement were
estimated to be equivalent to 2022 values (USGS 2024b).
In 2023, U.S. clinker production totaled 78,100 kilotons, which was a decrease of about 3 percent
compared to 2022 and an increase of 21 percent compared to 1990 (EPA 2024). The resulting C02
emissions were estimated to be 40.6 MMT C02 Eq. (40,636 kt) (see Table 4-3 and Table 4-4). Although
Industrial Processes and Product Use 4-11
-------
clinker production decreased between 2022 and 2023, imports of clinker and cement increased, and
exports remained steady. The total construction value increased by 5 percent during the first nine
months of 2023 compared to the same time period in 2022. Nonresidential construction spending
increased, but residential construction spending decreased. Growth was constrained by increased
costs, labor and production shortages, and ongoing supply chain disruptions (USGS 2024b). Cement
continues to be a critical component of the construction industry; therefore, the availability of public
and private construction funding, as well as overall economic conditions, have considerable impact on
the level of cement production.
Table 4-3: C02 Emissions from Cement Production (MMT C02 Eq.)
Year
1990
2005
2019
2020
2021
2022
2023
Cement Production Emissions
33.5
46.2
40.9
40.7
41.3
41.9
40.6
Table 4-4: C02 Emissions from Cement Production (kt C02)
Year
1990
2005
2019
2020
2021
2022
2023
Cement Production Emissions
33,484
46,194
40,896
40,688
41,312
41,884
40,636
Greenhouse gas emissions from cement production, which are primarily driven by production levels,
increased every year from 1991 through 2006 but decreased in the following years until 2009. Emissions
from cement production were at their highest levels in 2006 and at their lowest levels in 2009. Emissions
in 2009 were approximately 28 percent lower than 2008 emissions and 12 percent lower than 1990 due
to the economic recession and the associated decrease in demand for construction materials. Since
2009, emissions have increased by 37 percent due to increasing demand for cement.
Additionally, in 2022 several cement plants transitioned to portland-limestone blended cement (PLC), a
low-carbon cement, and the cement industry continued to announce increased use of alternative fuels
and alternative materials, carbon capture, utilization and storage projects, increased energy efficiency,
and shifting to renewable energy sources (USGS 2024b).
Methodology and Time-Series Consistency
Carbon dioxide emissions from cement production are estimated using the Tier 2 method from the 2006
IPCC Guidelines as this is a key category, in accordance with the IPCC methodological decision tree and
available data. The Tier 2 methodology was used because detailed and complete data (including
weights and composition) for carbonate(s) consumed in clinker production are not available,4 and thus
a rigorous Tier 3 approach is impractical. Tier 2 specifies the use of aggregated plant or national clinker
production data and an emission factor, which is the product of the average lime mass fraction for
clinker of 65 percent and a constant reflecting the mass of C02 released per unit of lime. The U.S.
Geological Survey (USGS) mineral commodity expert for cement has confirmed that this is a reasonable
assumption for the United States (Van Oss 2013a). This calculation yields an emission factor of 0.510
tons of C02 per ton of clinker produced, which was determined as follows:
As discussed further under "Planned Improvements," most cement-producing facilities that report their emissions to the
GHGRP use CEMS to monitor combined process and fuel combustion emissions for kilns, making it difficult to quantify
the process emissions on a facility-specific basis. By the end of 2022, the percentage of facilities not using CEMS was 1
percent.
4-12 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Equation 4-1:2006IPCC Guidelines Tier 1 Emission Factor for Clinker (precursor to
Equation 2.4)
Kg \ / g \i tons C02
44.01 —CO, ) h- (56.08—VCaO) = 0.510 ——
mole 2J V mole /J ton clinker
During clinker production, some of the raw materials, partially reacted raw materials, and clinker enters
the kiln line's exhaust system as non-calcinated, partially calcinated, or fully calcinated cement kiln
dust (CKD). To the degree that the CKD contains carbonate raw materials which are then calcined, there
are associated C02 emissions. At some plants, essentially all CKD is directly returned to the kiln,
becoming part of the raw material feed, or is likewise returned to the kiln after first being removed from
the exhaust. In either case, the returned CKD becomes a raw material, thus forming clinker, and the
associated C02 emissions are a component of those calculated for the clinker overall. At some plants,
however, the CKD cannot be returned to the kiln because it is chemically unsuitable as a raw material or
chemical issues limit the amount of CKD that can be so reused. Any clinker that cannot be returned to
the kiln is either used for other (non-clinker) purposes or is landfilled. The C02 emissions attributable to
the non-returned calcinated portion of the CKD are not accounted for by the clinker emission factor and
thus a CKD correction factor should be applied to account for those emissions. The USGS reports the
amount of CKD used to produce clinker, but no information is currently available on the total amount of
CKD produced annually.5 Because data are not currently available to derive a country-specific CKD
correction factor, a default correction factor of 1.02 (2 percent) was used to account for CKD C02
emissions, as recommended bythe IPCC (IPCC 2006).6 Total cement production emissions were
calculated by adding the emissions from clinker production and the emissions assigned to CKD.
Small amounts of impurities (i.e., not calcium carbonate) may exist in the raw limestone used to
produce clinker. The proportion of these impurities is generally minimal, although a small amount (1 to 2
percent) of magnesium oxide (MgO) may be desirable as a flux. Per the IPCC Tier 2 methodology, a
correction for MgO is not used, since the amount of MgO from carbonate is likely very small and the
assumption of a 100 percent carbonate source of CaO already yields an overestimation of emissions
(IPCC 2006).
The 1990 through 2012 activity data for clinker production were obtained from USGS (Van Oss 2013a),
which were based on U.S. Bureau of Mines data for 1990 through 1993 and USGS data for 1994 through
2012. Clinker production data for 2013 were also obtained from USGS (USGS 2014). USGS compiled the
data (to the nearest ton) through questionnaires sent to domestic clinker and cement manufacturing
plants, including facilities in Puerto Rico. Clinker production values in the current Inventory report utilize
GHGRP data for the years 2014 through 2023 (EPA 2024). Clinker production data are summarized in
Table 4-5. Details on how this GHGRP data compares to USGS reported data can be found in the section
on QA/QC and Verification.
5 The USGS Minerals Yearbook: Cement notes that CKD values used for clinker production are likely underreported.
6 As stated on p. 2.12 of the 2006 IPCC Guidelines, Vol. 3, Chapter 2: "...As data on the amount of CKD produced may be
scarce (except possibly for plant-level reporting), estimating emissions from lost CKD based on a default value can be
considered good practice. The amount of CO2 from lost CKD can vary but range typically from about 1.5 percent
(additional CO2 relative to that calculated for clinker) for a modern plant to about 20 percent for a plant losing a lot of
highly calcinated CKD (van Oss 2005). In the absence of data, the default CKD correction factor (CFckd) is 1.02 (i.e., add
2 percent to the CO2 calculated for clinker). If no calcined CKD is believed to be lost to the system, the CKD correction
factor will be 1.00 (van Oss 2005)..."
Industrial Processes and Product Use 4-13
-------
Table 4-5: Clinker Production (kt)
Year
1990
2005
2019
2020
2021
2022
2023
Clinker Production
64,355
88,783
78,600
78,200
79,400
80,500
78,100
Note: Clinker production from 1990 through 2023 includes Puerto Rico (relevant U.S. Territories).
Methodological approaches were applied to the entire time series to ensure time-series consistency
from 1990 through 2023. The methodology for cement production spliced activity data from two
different sources: USGS for 1990 through 2013 and GHGRP starting in 2014. Consistent with the 2006
IPCC Guidelines, the overlap technique was applied to compare the two data sets for years where there
was overlap, with findings that the data sets were consistent, and adjustments were not needed.
Uncertainty
The uncertainties contained in these estimates are primarily due to uncertainties in the lime content of
clinker and in the percentage of CKD recycled inside the cement kiln. Uncertainty is also associated
with the assumption that all calcium-containing raw materials are CaC03, when a small percentage
likely consists of other carbonate and non-carbonate raw materials. The lime content of clinker varies
from 60 to 67 percent; 65 percent is used as a representative value (Van Oss 2013b). This contributes to
the uncertainty surrounding the emission factor for clinker which has an uncertainty range of ±3 percent
with uniform densities (Van Oss 2013b). The amount of C02 from CKD loss can range from 1.5 to 8
percent depending upon plant specifications, and uncertainty was estimated at ±5 percent with uniform
densities (Van Oss 2013b). Additionally, some amount of C02 is reabsorbed when the cement is used for
construction. As cement reacts with water, alkaline substances such as calcium hydroxide are formed.
During this curing process, these compounds may react with C02 in the atmosphere to create calcium
carbonate. This reaction only occurs in roughly the outer 0.2 inches of the total thickness. Because the
amount of C02 reabsorbed is thought to be minimal, it was not estimated. EPA assigned uncertainty
bounds of ±3 percent and a normal probability density function for clinker production and uncertainty
bounds of ±5 percent and a uniform probability density function for the emission factor, based on expert
judgment (Van Oss 2013b).
The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 4-6. Based on
the uncertainties associated with total U.S. clinker production, the C02 emission factor for clinker
production, and the emission factor for additional C02 emissions from CKD, 2023 C02 emissions from
cement production were estimated to be between 38.9 and 42.5 MMT C02 Eq. at the 95 percent
confidence level. This confidence level indicates a range of approximately 4 percent below and 5
percent above the emission estimate of 40.6 MMT C02 Eq.
Table 4-6: Approach 2 Quantitative Uncertainty Estimates for C02 Emissions from
Cement Production (MMT C02 Eq. and Percent)
2023 Emission Estimate
Source Gas
(MMT CO2 Eq.)
Uncertainty Range Relative to Emission Estimatea
(MMT CO2 Eq.)
(%)
Lower Upper
Bound Bound
Lower Upper
Bound Bound
Cement
Production CO2 40.6
38.9 42.5
-4% +5%
4-14 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
a Range of emission estimates predicted by Monte Carlo stochastic simulation for a 95 percent confidence interval.
QA/QC and Verification
General quality assurance/quality control (QA/QC) procedures were applied consistent with the U.S.
Inventory QA/QC plan, which is in accordance with Volume 1, Chapter 6 of 2006IPCC Guidelines as
described in the introduction of the IPPU chapter (see Annex 8 for more details).
EPA relied upon the latest guidance from the IPCC on the use of facility-level data in national inventories
and applied a category-specific QC process to compare activity data from EPA's GHGRP with existing
data from USGS surveys. This was to ensure time-series consistency of the emission estimates
presented in the Inventory. Total U.S. clinker production is assumed to have low uncertainty because
facilities routinely measure this for economic reasons and because both USGS and GHGRP take
multiple steps to ensure that reported totals are accurate. EPA verifies annual facility-level GHGRP
reports through a multi-step process that is tailored to the reporting industry (e.g., combination of
electronic checks including range checks, statistical checks, algorithm checks, year-to-year
comparison checks, along with manual reviews involving outside data checks) to identify potential
errors and ensure that data submitted to EPA are accurate, complete, and consistent (EPA 2015). Based
on the results of the verification process, EPA follows up with facilities to resolve mistakes that may have
occurred.7 Facilities are also required to monitor and maintain records of monthly clinker production
per section 98.84 of the GHGRP regulation (40 CFR 98.84).
EPA's GHGRP requires all facilities producing portland cement to report greenhouse gas emissions,
including C02 process emissions from each kiln, C02 combustion emissions from each kiln, CH4 and
N20 combustion emissions from each kiln, and C02, CH4, and N20 emissions from each stationary
combustion unit other than kilns (40 CFR Part 98 Subpart H). Source-specific quality control measures
for the cement production category are included in section 98.84, Monitoring and QA/QC Requirements.
As mentioned above, EPA compares GHGRP clinker production data (EPA 2024) to the USGS clinker
production data (USGS 2024a; USGS 2024c). For the years 2014, 2020, and 2022, USGS and GHGRP
clinker production data showed a difference of approximately 1 percent. In 2018, the difference
between USGS and GHGRP clinker production data was approximately 3 percent, which resulted in a
difference in emissions of about 1.2 MMT C02 Eq. In 2015, 2016, 2017, 2019, and 2021, that difference
was less than 0.5 percent (less than 0.2 MMT C02 Eq.) between the two sets of activity data. For 2023,
the difference between USGS and GHGRP clinker production data was approximately 2 percent, which
resulted in a difference in emissions of about 0.9 MMT C02 Eq. The information collected by the USGS
National Minerals Information Center surveys continue to be an important data source.
Recalculations Discussion
No recalculations were performed for the 1990 through 2022 portion of the time series.
7 See GHGRP Verification Fact Sheet https://www.epa.gov/sites/production/files/7015-
07/rtocuments/ghgrp verification factsheet.pdf.
Industrial Processes and Product Use 4-15
-------
Planned Improvements
EPA is continuing to evaluate and analyze data reported under EPA's GHGRP that would be useful to
improve the emission estimates for the Cement Production source category. Most cement production
facilities reporting under EPA's GHGRP use Continuous Emission Monitoring Systems (CEMS) to monitor
and report C02 emissions, thus reporting combined process and combustion emissions from kilns. In
implementing further improvements and integration of data from EPA's GHGRP, the latest guidance from
the IPCC on the use of facility-level data in national inventories will be relied upon, in addition to
category-specific QC methods recommended by the 2006 IPCC Guidelines.8 EPA's long-term
improvement plan includes continued assessment of the feasibility of using additional GHGRP
information beyond aggregation of reported facility-level clinker data, in particular disaggregating the
combined process and combustion emissions reported using CEMS, to separately present national
process and combustion emissions streams consistent with IPCC guidelines. This long-term planned
analysis is still in development and has not been applied for this current Inventory.
EPA continues to review methods and data used to estimate C02 emissions from cement production in
order to account for organic material in the raw material and to discuss the carbonation that occurs
across the duration of the cement product. Work includes identifying data and studies on the average
carbon content for organic materials in kiln feed in the United States and on C02 reabsorption rates via
carbonation for various cement products. This information is not reported by facilities subject to GHGRP
reporting. This is a long-term improvement.
4.2 Lime Production (Source Category 2A2)
Lime is a manufactured product with many industrial, chemical, and environmental applications. This
reporting category (2A2) includes process emissions from the production of lime. Per the IPCC
methodological guidance, emissions from fuels consumed for energy purposes during the production of
lime are accounted for as part of fossil fuel combustion in the industrial end-use sector reported under
the Energy chapter.
Lime production involves three main processes: stone preparation, calcination, and hydration. Carbon
dioxide (C02) is generated during the calcination stage, when limestone—consisting of calcium
carbonate (CaC03) and/or magnesium carbonate (MgC03)—is roasted at high temperatures in a kiln to
produce calcium oxide (CaO) and C02. The C02 is given off as a gas and is normally emitted to the
atmosphere.
CctCO3 -> CaO -I- CO2
Some facilities, however, recover C02 generated during the production process for use in sugar refining
and precipitated calcium carbonate (PCC) production.9 PCC is used as a filler or coating in the paper,
See IPCC Technical Bulletin on Use of Facility-Specific Data in National Greenhouse Gas Inventories http://www.ipcc-
nggip.iges.or.ip/public/tb/TFI Technical Bulletin 1.pdf and the 2019 Refinement, Volume 1, Chapter 2, Section 2.3, Use
of Facility Data in Inventories at https://www.ipcc-
nggip.iges.or.jp/public/2019rf/pdf/1 Volume1/19R V1 Ch02 DataCollection.pdf.1
The amount of CO2 captured from lime production for sugar refining and PCC production is reported under Source
Category 2H3 "Other", but within this report, they are included in this chapter.
4-16 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
food, and plastic industries and is derived from reacting hydrated high-calcium quicklime with C02, a
production process that does not result in net emissions of C02 to the atmosphere.
For U.S. operations, the term "lime" refers to a variety of chemical compounds. These include CaO, or
high-calcium quicklime; calcium hydroxide (Ca(OH)2), or hydrated lime; dolomitic quicklime
([CaOMgO]); and dolomitic hydrate ([Ca(OH)2*MgO] or [Ca(OH)2»Mg(OH)2]).
The current lime market is approximately distributed across six end-use categories, as follows:
metallurgical uses, 36 percent; environmental uses, 26 percent; chemical and industrial uses, 21
percent; construction uses, 12 percent; miscellaneous uses, 4 percent; and refractory dolomite, 1
percent (USGS 2024a). The major uses are in steel making, chemical and industrial applications (such
as the manufacture of fertilizer, glass, paper and pulp, and precipitated calcium carbonate, and in sugar
refining), flue gas desulfurization (FGD) systems at coal-fired electric power plants, construction, and
water treatment, as well as uses in mining, pulp and paper and precipitated calcium carbonate
manufacturing (USGS 2024b). Lime is also used as a C02 scrubber, and there has been experimentation
on the use of lime to capture C02 from electric power plants. Both lime (CaO) and limestone (CaC03)
can be used as a sorbent for FGD systems. Emissions from limestone consumption for FGD systems are
reported under Section 4.4 Other Process Uses of Carbonate Production (Source Category 2A4).
Emissions from lime production have fluctuated over the time series depending on lime end-use
markets - primarily the steel making industry and FGD systems for utility and industrial plants - and also
energy costs. One significant change to lime end-use since 1990 has been the increase in demand for
lime for FGD at coal-fired electric power plants, which can be attributed to compliance with sulfur
dioxide (S02) emission regulations of the Clean Air Act Amendments of 1990. Phase I went into effect on
January 1,1995, followed by Phase II on January 1, 2000. To supply lime for the FGD market, the lime
industry installed more than 1.8 million tons per year of new capacity by the end of 1995 (USGS 1996).
The need for air pollution controls continued to drive the FGD lime market, which had doubled between
1990 and 2019 (USGS 1991 and 2020a).
The U.S. lime industry temporarily shut down some individual gas-fired kilns and, in some cases, entire
lime plants during 2000 and 2001, due to significant increases in the price of natural gas. Lime
production continued to decrease in 2001 and 2002, a result of lower demand from the steel making
industry, lime's largest end-use market, when domestic steel producers were affected by low priced
imports and slowing demand (USGS 2002).
Emissions from lime production peaked in 2006 at approximately 30.3 percent above 1990 levels, due to
strong demand from the steel and construction markets (road and highway construction projects),
before dropping to its second lowest level in 2009 at approximately 2.5 percent below 1990 emissions,
driven by the economic recession and downturn in major markets including construction, mining, and
steel (USGS 2007, 2008, 2010). In 2010, the lime industry began to recover as the steel, FGD, and
construction markets also recovered (USGS 2011 and 2012a). Fluctuation in lime production since 2015
has been driven largely by demand from the steel making industry (USGS 2018, 2019, 2020b). In 2020, a
significant decline in lime production occurred due to plants temporarily closing as a result of the global
COVID-19 pandemic (USGS 2021). This resulted in the lowest level of emissions in 2020 at
approximately 3.4 percent below 1990 emissions. Emissions increased slightly in 2021 and 2022, before
dropping again in 2023.
Lime production in the United States—including Puerto Rico—was reported to be 16,028 kilotons in
2023, a decrease of about 5.7 percent compared to 2022 levels (USGS 2024c). Compared to 1990, lime
Industrial Processes and Product Use 4-17
-------
production increased by about 1.2 percent. At year-end 2023, 73 primary lime plants were operating in
the United States, including Puerto Rico (USGS 2024c).10 Principal lime producing states were, in
alphabetical order, Alabama, Missouri, Ohio, and Texas (USGS 2024b).
U.S. lime production resulted in estimated net C02 emissions of 11.5 MMT C02 Eq.(11,548 kt) (see Table
4-7 and Table 4-8). Carbon dioxide emissions from lime production decreased by about 5.4 percent
compared to 2022 levels. Compared to 1990, C02 emissions have decreased by about 1.3 percent. The
trends in C02 emissions from lime production are directly proportional to trends in production, which
are described above.
Table 4-7: C02 Emissions from Lime Production (MMT C02 Eq.)
Year
1990
2005
2019
2020
2021
2022
2023
Lime Production
11.7
14.6
12.1
11.3
11.9
12.2
11.5
Table 4-8: Gross, Recovered, and Net C02 Emissions from Lime Production (kt C02)
Year
1990
2005
2019
2020
2021
2022
2023
Gross
Recovered®
11,959
259
15,074
522
12,676
564
11,875
576
12,586
716
12,750
542
12,043
495
Net Emissions
11,700
14,552
12,112
11,299
11,870
12,208
11,548
a For sugar refining and PCC production.
Note: Totals may not sum due to independent rounding.
Methodology and Time-Series Consistency
To calculate emissions, the amounts of high-calcium and dolomitic lime produced were multiplied by
their respective emission factors, consistent with Tier 2 methodology from the 2006IPCC Guidelines
and in accordance with the IPCC methodological decision tree and available data. The emission factor
is the product of the stoichiometric ratio between C02 and CaO, and the average CaO and MgO content
for lime. The CaO and MgO content for lime is assumed to be 95 percent for both high-calcium and
dolomitic lime (IPCC 2006). The emission factors were calculated as follows:
Equation 4-2:2006 IPCC Guidelines Tier 2 Emission Factor for Lime Production, High-
Calcium Lime (Equation 2.9)
EFH,8„-clc„i,„L,i„e = [(44.01^C02)+ (56.08^CaO)] x (o.950l>||) = 0.7455
Equation 4-3:2006 IPCC Guidelines Tier 2 Emission Factor for Lime Production,
Dolomitic Lime (Equation 2.9)
EFooiomitic Lime = [(s8.02 JjjCO,) + (96.39 JjjCaO . MgO)] x (0.9500 22±S2) =
0.8675
g lime
10 In 2023, 68 operating primary lime facilities in the United States reported to the EPA Greenhouse Gas Reporting
Program.
4-18 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Production was adjusted to remove the mass of chemically combined water found in hydrated lime,
determined according to the molecular weight ratios of H20 to (Ca(OH)2 and [Ca(OH)2*Mg(OH)2]) (IPCC
2006). These factors set the chemically combined water content to 27 percent for high-calcium
hydrated lime, and 30 percent for dolomitic hydrated lime.
The 2006 IPCC Guidelines (Tier 2 method) also recommends accounting for emissions from lime kiln
dust (LKD) through application of a correction factor. LKD is a byproduct of the lime manufacturing
process typically not recycled back to kilns. LKD is a very fine-grained material and is especially useful
for applications requiring very small particle size. Most common LKD applications include soil
reclamation and agriculture. Emissions from the application of lime for agricultural purposes are
reported in the Agriculture chapter under 5.5 Liming (Source Category 3G). Currently, data on annual
LKD production is not readily available to develop a country-specific correction factor. Lime emission
estimates were multiplied by a factor of 1.02 to account for emissions from LKD (IPCC 2006). See the
Planned Improvements section associated with efforts to improve uncertainty analysis and emission
estimates associated with LKD.
Lime emission estimates were further adjusted to account for the amount of C02 captured for use in on-
site processes. All the domestic lime facilities are required to report these data to EPA under its GHGRP.
The total national-level annual amount of C02 captured for on-site process use was obtained from EPA's
GHGRP (EPA 2024) based on reported facility-level data for years 2010 through 2023. The amount of C02
captured/recovered for non-marketed on-site process use is deducted from the total gross emissions
(i.e., from lime production and LKD). The net lime emissions are presented in Table 4-7 and Table 4-8.
GHGRP data on C02 removals (i.e., C02 captured/recovered) was available only for 2010 through 2023.
Since GHGRP data are not available for 1990 through 2009, IPCC "splicing" techniques were used as per
the 2006 IPCC Guidelines on time-series consistency (IPCC 2006, Volume 1, Chapter 5).
Lime production data (i.e., lime sold and non-marketed lime used by the producer) by type (i.e., high-
calcium and dolomitic quicklime, high-calcium and dolomitic hydrated lime, and dead-burned
dolomite) for 1990 through 2023 (see Table 4-9) were obtained from U.S. Geological Survey (USGS)
Minerals Yearbook (USGS 2024a) and are compiled by USGS to the nearest ton. Dead-burned dolomite
data are additionally rounded by USGS to no more than one significant digit to avoid disclosing company
proprietary data. Production data for the individual quicklime (i.e., high-calcium and dolomitic) and
hydrated lime (i.e., high-calcium and dolomitic) types were not provided prior to 1997. These were
calculated based on total quicklime and hydrated lime production data from 1990 through 1996 and the
three-year average ratio of the individual lime types from 1997 to 1999. Natural hydraulic lime, which is
produced from CaO and hydraulic calcium silicates, is not manufactured in the United States (USGS
2024b). Total lime production was adjusted to account for the water content of hydrated lime by
converting hydrate to oxide equivalent based on recommendations from the IPCC and using the water
content values for high-calcium hydrated lime and dolomitic hydrated lime mentioned above and is
presented in Table 4-10 (IPCC 2006). The CaO and CaO*MgO contents of lime, both 95 percent, were
obtained from the IPCC (IPCC 2006).
Table 4-9: High-Calcium- and Dolomitic-Quicklime, High-Calcium- and Dolomitic-
Hydrated, and Dead-Burned-Dolomite Lime Production (kt)
Year
1990
2005
2019
2020
2021
2022
2023
High-Calcium Quicklime
Dolomitic Quicklime
11,166
2,234
14,100
2,990
11,300
2,700
10,700
2,390
11,200
2,700
11,500
2,640
10,800
2,560
Industrial Processes and Product Use 4-19
-------
Year
1990
2005
2019
2020
2021
2022
2023
High-Calcium Hydrated
1,781
2,220
2,430
2,320
2,430
2,410
2,230
Dolomitic Hydrated
319
474
267
252
244
244
238
Dead-Burned Dolomite
342
200
200
200
200
200
200
Table 4-10: Adjusted Lime Production (kt)
Year
1990
2005
2019
2020
2021
2022
2023
High-Calcium
12,466
15,721
13,074
12,394
12,974
13,259
12,428
Dolomitic
2,800
3,522
3,087
2,766
3,071
3,011
2,927
Note: Minus water content of hydrated lime.
Methodological approaches were applied to the entire time series to ensure consistency in emissions
from 1990 through 2023.
Uncertainty
The uncertainties contained in these estimates can be attributed to slight differences in the chemical
composition of lime products and C02 recovery rates for on-site process use over the time series.
Although the methodology accounts for various formulations of lime, it does not account for the trace
impurities found in lime, such as iron oxide, alumina, and silica. Due to differences in the limestone
used as a raw material, a rigid specification of lime material is impossible. As a result, few plants
produce lime with exactly the same properties.
In addition, a portion of the C02 emitted during lime production will actually be reabsorbed when the
lime is consumed, especially at captive lime production facilities. As noted above, lime has many
different chemical, industrial, environmental, and construction applications. In many processes, C02
reacts with the lime to create calcium carbonate (e.g., water softening). Carbon dioxide reabsorption
rates vary, however, depending on the application. For example, 100 percent of the lime used to produce
precipitated calcium carbonate reacts with C02, whereas most of the lime used in steel making reacts
with impurities such as silica, sulfur, and aluminum compounds. Quantifying the amount of C02 that is
reabsorbed would require a detailed accounting of lime use in the United States and additional
information about the associated processes where both the lime and byproduct C02 are "reused."
Research conducted thus far has not yielded the necessary information to quantify C02 reabsorption
rates.11 Some additional information on the amount of C02 consumed on site at lime facilities, however,
has been obtained from EPA's GHGRP.
In some cases, lime is generated from calcium carbonate byproducts at pulp mills and water treatment
plants.12 The lime generated by these processes is included in the USGS data for commercial lime
consumption. In the pulping industry, mostly using the Kraft (sulfate) pulping process, lime is consumed
to causticize a process liquor (green liquor) composed of sodium carbonate and sodium sulfide. The
green liquor results from the dilution of the smelt created by combustion of the black liquor where
11 Representatives of the National Lime Association estimate that CO2 reabsorption that occurs from the use of lime may
offset as much as a quarter of the CO2 emissions from calcination (Males 2003).
12 Some carbide producers may also regenerate lime from their calcium hydroxide byproducts, which does not result in
emissions of CO2. In making calcium carbide, quicklime is mixed with coke and heated in electric furnaces. The
regeneration of lime in this process is done using a waste calcium hydroxide (hydrated lime) [CaC2 + 2H2O -> C2H2 +
Ca(OH)2], not calcium carbonate [CaCCh]. Thus, the calcium hydroxide is heated in the kiln to simply expel the water
[Ca(OH)2 + heat -> CaO + H2O], and no CO2 is released.
4-20 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
biogenic carbon (C) is present from the wood. Kraft mills recover the calcium carbonate "mud" after the
causticizing operation and calcine it back into lime—thereby generating C02—for reuse in the pulping
process. Although this re-generation of lime could be considered a lime manufacturing process, the C02
emitted during this process is mostly biogenic in origin and therefore is not included in the industrial
processes totals (Miner and Upton 2002). In accordance with IPCC methodological guidelines, any such
emissions are calculated by accounting for net carbon fluxes from changes in biogenic carbon
reservoirs in wooded or crop lands (see the Land Use, Land-Use Change, and Forestry chapter).
In the case of water treatment plants, lime is used in the softening process. Some large water treatment
plants may recover their waste calcium carbonate and calcine it into quicklime for reuse in the softening
process. Further research is necessary to determine the degree to which lime recycling is practiced by
water treatment plants in the United States.
Another uncertainty is the assumption that calcination emissions for LKD are around 2 percent. EPA
assigned uncertainty ranges of ±2 percent and a triangular probability density function for the LKD
correction factor based on expert judgment (RTI 2023). The National Lime Association (NLA) has
commented that the estimates of emissions from LKD in the United States could be closer to 6 percent.
They also note that additional emissions (approximately 2 percent) may also be generated through
production of other byproducts/wastes (off-spec lime that is not recycled, scrubber sludge) at lime
plants (Seeger 2013). Publicly available data on LKD generation rates, total quantities not used in
cement production, and types of other byproducts/wastes produced at lime facilities are limited. NLA
compiled and shared historical emissions information and quantities for some waste products reported
by member facilities associated with generation of total calcined byproducts and LKD, as well as
methodology and calculation worksheets that member facilities complete when reporting. There is
uncertainty regarding the availability of data across the time series needed to generate a representative
country-specific LKD factor. Uncertainty of the activity data is also a function of the reliability and
completeness of voluntarily reported plant-level production data. EPA assigned uncertainty ranges of ±1
percent for lime production and a normal probability density function, based on expert judgment (USGS
2012b). Further research, including discussion with NLA, and data is needed to improve understanding
of additional calcination emissions to consider revising the current assumptions that are based on the
2006 IPCC Guidelines. More information can be found in the Planned Improvements section below.
The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 4-11. Lime C02
emissions for 2023 were estimated to be between 11.3 and 11.8 MMT C02 Eq. at the 95 percent
confidence level. This confidence level indicates a range of approximately 2 percent below and 2
percent above the emission estimate of 11.5 MMT C02 Eq.
Table 4-11: Approach 2 Quantitative Uncertainty Estimates for C02 Emissions from
Lime Production (MMT C02 Eq. and Percent)
2023 Emission
Source
Gas
Estimate
Uncertainty Range Relative to Emission Estimate"
(MMT CO2 Eq.)
(MMT CO2 Eq.)
(%)
Lower Upper
Lower
Upper
Bound Bound
Bound
Bound
Lime Production
C02
11.5
11.3 11.8
-2%
+2%
a Range of emission estimates predicted by Monte Carlo stochastic simulation for a 95 percent confidence interval.
Industrial Processes and Product Use 4-21
-------
QA/QC and Verification
General quality assurance/quality control (QA/QC) procedures were applied consistent with the U.S.
Inventory QA/QC plan, which is in accordance with Volume 1, Chapter 6 of 2006IPCC Guidelines as
noted in the introduction of the IPPU chapter (see Annex 8 for more details).
More details on the greenhouse gas calculation, monitoring and QA/QC methods associated with
reporting on C02 captured for onsite use applicable to lime manufacturing facilities can be found under
Subpart S (lime manufacturing) of the GHGRP regulation (40 CFR Part 98).13 EPA verifies annual facility-
level GHGRP reports through a multi-step process (e.g., combination of electronic checks and manual
reviews) to identify potential errors and ensure that data submitted to EPA are accurate, complete, and
consistent (EPA 2024).14 Based on the results of the verification process, EPA follows up with facilities to
resolve mistakes that may have occurred. The post-submittals checks are consistent with a number of
general and category-specific QC procedures, including: range checks, statistical checks, algorithm
checks, and year-to-year checks of reported data and emissions.
Recalculations Discussion
No recalculations were performed for the 1990 through 2022 portion of the time series.
Planned Improvements
EPA plans to review GHGRP emissions and activity data reported to EPA under Subpart S of the GHGRP
regulation (40 CFR Part 98), and aggregated activity data on lime production by type in particular. In
addition, initial review of data has identified that several facilities use CEMS to report emissions. Under
Subpart S, if a facility is using a CEMS, they are required to report combined combustion emissions and
process emissions. EPA continues to review how best to incorporate GHGRP and notes that particular
attention will be made to also ensuring time-series consistency of the emissions estimates presented in
future Inventory reports. This is required because the facility-level reporting data from EPA's GHGRP,
with the program's initial requirements for reporting of emissions in calendar year 2010, are not
available for all inventory years (i.e., 1990 through 2009) as required for this Inventory. In implementing
improvements and integration of data from EPA's GHGRP, the latest guidance from the IPCC on the use
of facility-level data in national inventories will be relied upon.15
Future improvements involve improving and/or confirming the representativeness of current
assumptions associated with emissions from production of LKD and other byproducts/wastes as
discussed in the Uncertainty section, per comments from the NLA provided during a prior Public Review
comment period for a previous (1990 through 2018) Inventory EPA met with NLA in summer of 2020 for
clarification on data needs and available data and to discuss planned research into GHGRP data.
Previously, EPA met with NLA in spring of 2015 to outline specific information required to apply IPCC
methods to develop a country-specific correction factor to more accurately estimate emissions from
production of LKD. In 2016, NLA compiled and shared historical emissions information reported by
13 See http://www.ecfr.gov/cgi-bin/text-idx7tpN/ecfrbrowse/Title40/40cfr98 main 02.tpL
14 See https://www.epa.gov/sites/production/files/2015-07/documents/ghgrp verification factsheet.pdf.
15 See http://www.ipcc-nggip.iges.or.jp/public/tb/TFI Technical Bulletin 1 .pdf and the P019 Refinement. Volume 1,
Chapter 2, Section 2.3, Use of Facility Data in Inventories at https://www.ipcc-
nggip.iges.or.jp/public/2019rf/pdf/1 Volume1/19R V1 Ch02 DataCollection.pdf.
4-22 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
member facilities on an annual basis under voluntary reporting initiatives from 2002 through 2011
associated with generation of total calcined byproducts and LKD. Reporting of LKD was only
differentiated for the years 2010 and 2011. This emissions information was reported on a voluntary basis
consistent with NLA's facility-level reporting protocol, which was also provided to EPA. To reflect
information provided by NLA, EPA updated the qualitative description of uncertainty. At the time of this
Inventory, this planned improvement is in process and has not been incorporated into this current
Inventory report.
4.3 Glass Production (Source Category 2A3)
Glass production is an energy and raw-material intensive process that results in the generation of
carbon dioxide (C02) from both the energy consumed in making glass and the glass production process
itself. This reporting category (2A3) includes emissions from the production of glass. Emissions from
fuels consumed for energy purposes during the production of glass are accounted for as part of fossil
fuel combustion in the industrial end-use sector reported under the Energy chapter.
Glass production employs a variety of raw materials in a glass-batch. These include formers, fluxes,
stabilizers, and sometimes colorants. The major raw materials (i.e., fluxes and stabilizers) that emit
process-related C02 emissions during the glass melting process are limestone, dolomite, and soda ash.
The main former in all types of glass is silica (Si02). Other major formers in glass include feldspar and
boric acid (i.e., borax). Fluxes are added to lower the temperature at which the batch melts. Most
commonly used flux materials are soda ash (sodium carbonate, Na2C03) and potash (potassium
carbonate, K2C03). Stabilizers make glass more chemically stable and keep the finished glass from
dissolving and/or falling apart. Commonly used stabilizing agents in glass production are limestone
(CaC03), dolomite (CaC03MgC03), alumina (Al203), magnesia (MgO), barium carbonate (BaC03),
strontium carbonate (SrC03), lithium carbonate (Li2C03), and zirconia (Zr02) (DOE 2002). Glass makers
also use a certain amount of recycled scrap glass (cullet), which comes from in-house return of
glassware broken in the production process or other glass spillage or retention, such as recycling or
from cullet broker services.
The raw materials (primarily soda ash, limestone, and dolomite) release C02 emissions in a complex
high-temperature chemical reaction during the glass melting process. This process is not directly
comparable to the calcination process used in lime manufacturing, cement manufacturing, and
process uses of carbonates (i.e., limestone/dolomite use) but has the same net effect in terms of
generating process C02 emissions (IPCC 2006).
The U.S. glass industry can be divided into four main categories: containers, flat (window) glass, fiber
glass, and specialty glass. The majority of commercial glass produced is container and flat glass (EPA
2009). The United States is one of the major global exporters of glass. Domestically, demand comes
mainly from the construction, auto, bottling, and container industries. There are more than 1,700
facilities that manufacture glass in the United States, with the largest companies being Corning,
Guardian Industries, Owens-Illinois, and PPG Industries.16
16 Excerptfrom Glass & Glass Product Manufacturing Industry Profile, First Research. Available online at:
http://www.firstresearch.com/lndustry-Research/Glass-and-Glass-Product-Manufacturing.html.
Industrial Processes and Product Use 4-23
-------
The glass container sector is one of the leading soda ash consuming sectors in the United States. In
2023, glass production accounted for 46 percent of total domestic soda ash consumption (USGS 2024).
Emissions from soda ash production are reported in Section 4.12.
In 2023, 2,050 kilotons of soda ash, 1,252 kilotons of limestone, 824 kilotons of dolomite, and 1.7
kilotons of other carbonates were consumed for glass production (USGS 2024; EPA 2024). Use of soda
ash, limestone, dolomite, and other carbonates in glass production resulted in aggregate C02 emissions
of 1.8 MMT C02 Eq. (1,774 kt), which are summarized in Table 4-12 and Table 4-13. Overall, emissions
have decreased by 22 percent compared to 1990. Emissions decreased by 9 percent compared to 2022
levels.
Emissions from glass production have remained relatively consistent over the time series with some
fluctuations since 1990. In general, these fluctuations were related to the behavior of the export market
and the U.S. economy. Specifically, the extended downturn in residential and commercial construction
and automotive industries between 2008 and 2010 resulted in reduced consumption of glass products,
causing a drop in global demand for limestone, dolomite, and soda ash and resulting in lower emissions.
Some commercial food and beverage package manufacturers are shifting from glass containers towards
lighter and more cost-effective polyethylene terephthalate (PET) based containers, putting downward
pressure on domestic consumption of soda ash (USGS 1995 through 2015b). Glass production in 2023
decreased by as much as 7 percent between November to December and increased by as much as 6
percent from September to October (Federal Reserve 2024).
Table 4-12: C02 Emissions from Glass Production (MMT C02 Eq.)
Year 1990
2005
2019 2020
2021
2022
2023
Glass Production 2.3
2.4
1.9 1.9
2.0
2.0
1.8
Table 4-13: C02 Emissions from Glass Production (kt C02)
Year 1990
2005
2019 2020
2021
2022
2023
Glass Production 2,263
2,402
1,940 1,858
1,969
1,956
1,774
Methodology and Time-Series Consistency
Carbon dioxide emissions were calculated based on Tier 3 method from the 2006IPCC Guidelines, in
accordance with the IPCC methodological decision tree and available data, by multiplying the quantity
of input carbonates (i.e., limestone, dolomite, soda ash, and other carbonates) by the carbonate-based
emission factor (in metric tons C02/metric ton carbonate) and the average carbonate-based mineral
mass fraction.
2010 through 2023
The methodology for estimating C02 emissions from glass production for years 2010 through 2023 used
the quantities of limestone, dolomite, and a group of other carbonates (i.e., barium carbonate,
potassium carbonate, lithium carbonate, and strontium carbonate) used for glass production, obtained
from GHGRP (EPA 2024). USGS data on the quantity of soda ash used for glass production was used
because it was obtained directly from the soda ash producers and includes use by smaller artisanal
glass operations, which are excluded in the GHGRP data (USGS 2024).
4-24 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
GHGRP collects data from glass production facilities with greenhouse gas emissions greater than
25,000 metric tons C02 Eq. The reporting threshold is used to exclude artisanal glass operations that are
expected to have much lower greenhouse gas emissions than the threshold. These smaller facilities
have not been accounted for yet for this portion of the time series for limestone, dolomite, or other
carbonates due to limited data. Facilities report the total quantity of each type of carbonate used in
glass production each year to GHGRP, with data collection starting in 2010 (EPA 2024).
Using the total quantities of each carbonate, EPA calculated the metric tons of emissions resulting from
glass production by multiplying the quantity of input carbonates (i.e., limestone, dolomite, soda ash,
and other carbonates) by carbonate-based emission factors in metric tons C02/metric ton carbonate
(limestone, 0.43971; dolomite, 0.47732; soda ash, 0.41492; and other carbonates, 0.262), and by the
average carbonate-based mineral mass fraction for each year. IPCC default emission factors were used
for limestone, dolomite, and soda ash, and the emission factor for other carbonates is based on expert
judgment (RTI 2022).
1990 through 2009
Data from GHGRP on the quantity of limestone, dolomite, and other carbonates used in glass
production are not available for 1990 through 2009. Additionally, USGS does not collect data on the
quantity of other carbonates used for glass production.
To address time-series consistency, total emissions from 1990 to 2009 were calculated using the
Federal Reserve Industrial Production Index for glass production in the United States as a surrogate for
the total quantity of carbonates used in glass production. The production index measures real output
expressed as a percentage of real output in a base year, which is currently 2017 (Federal Reserve 2024).
Since January 1971, the Federal Reserve has released the monthly glass production index for NAICS
code 3272 (Glass and Glass Product Manufacturing) as part of release G.17, "Industrial Production and
Capacity Utilization" (Federal Reserve 2024). The monthly index values for each year were averaged to
calculate an average annual glass production index value. Total annual process emissions were
calculated by taking a ratio of the average annual glass production index for each year to the average
annual glass production index for base year 2017, and multiplying by the calculated 2017 emissions
(process-related) based on GHGRP data.
Emissions from limestone, dolomite, and other carbonate consumption were disaggregated from total
annual emissions, using the average percent contribution of each to annual emissions from these three
carbonates for 2010 through 2014 based on GHGRP data: 64.5 percent limestone, 35.5 percent
dolomite, and 0.1 percent other carbonates.
The methodology for estimating C02 emissions from the use of soda ash for glass production and data
sources for the amount of soda ash used in glass production are consistent with the methodology used
for 2010 through 2023. The average mineral mass fractions for soda ash are only available starting in
2010. The average carbonate-based mineral mass fractions from the GHGRP, averaged across 2010
through 2014, indicate that soda ash contained 98.7 percent sodium carbonate (Na2C03). This averaged
value is used to estimate emissions for 1990 through 2009. The years 2010 to 2014 were used to
determine the average carbonate-based mineral mass fractions because those years were deemed to
better represent historic glass production from 1990 to 2009.
Data on soda ash used for glass production for 1990 through 2023 were obtained from the U.S. Bureau
of Mines (1991 and 1993a), the USGS Minerals Yearbook: Soda Ash (USGS 1995 through 2015b), and
Industrial Processes and Product Use 4-25
-------
USGS Mineral Industry Surveys for Soda Ash (USGS 2017 through 2024). Data on limestone, dolomite,
and other carbonates used for glass production and on average carbonate-based mineral mass fraction
for 2010 through 2023 were obtained from GHGRP (EPA 2024). The quantities of limestone, dolomite,
and other carbonates were calculated for 1990 through 2009 using the Federal Reserve Industrial
Production Index (Federal Reserve 2024).
The amount of limestone, dolomite, soda ash, and other carbonates used in glass production each year
and the annual average Federal Reserve production indices for glass production are shown in Table 4-14.
Table 4-14: Limestone, Dolomite, Soda Ash, and Other Carbonates Used in Glass
Production (kt) and Average Annual Production Index for Glass and Glass Product
Manufacturing
Activity
1990
2005
2019
2020
2021
2022
2023
Limestone
1,409
1,690
1,370
1,334
1,397
1,370
1,252
Dolomite
714
857
883
824
893
925
824
Soda Ash
3,177
3,050
2,220
2,130
2,280
2,250
2,050
Other Carbonates
2
3
2
2
2
2
2
Total
5,302
5,599
4,475
4,289
4,572
4,547
4,127
Production Index®
94.3
113.1
99.8
92.4
88.3
86.6
85.2
a Average Annual Production Index uses 2017 as the base year.
Note: Totals may not sum due to independent rounding.
As discussed above, methodological approaches were applied to the entire time series to ensure
consistency in emissions from 1990 through 2023. Consistent with the 2006IPCC Guidelines, the
overlap technique was applied to compare USGS and GHGRP data sets for 2010 through 2023. To
address the inconsistencies, adjustments were made as described above.
Uncertainty
The methodology in this Inventory report uses GHGRP data for the average mass fraction of each
mineral used in glass production. These minerals are limestone, dolomite, soda ash, and other
carbonates (barium carbonate, potassium carbonate, lithium carbonate, and strontium carbonate). The
mass fractions are reported directly by the glass manufacturers, for each year from 2010 to 2023.
The methodology uses the quantities of limestone, dolomite, and other carbonates used in glass
manufacturing which is reported directly by the glass manufacturers for years 2010 through 2023 and
the amount of soda ash used in glass manufacturing which is reported by soda ash producers for the full
time series. EPA assigned an uncertainty range of ±5 percent and a normal probability density function
for all carbonate quantities and the Federal Reserve Industrial Production Index for glass production,
and using this suggested uncertainty provided in Section 2.4.2.2 of the 2006 IPCC Guidelines is
appropriate based on expert judgment (RTI 2023). EPA assigned an uncertainty range of ±2 percent for
the carbonate emission factors and ±1 percent for the calcination fraction, and using this suggested
uncertainty provided in Section 2.4.2.1 of the 2006 IPCC Guidelines is appropriate based on expert
judgment (RTI 2023). Per this expert judgment, a triangular probability density function was assigned for
emission factors, mineral mass fractions, and calcination fraction.
EPA assigned an uncertainty range of ±1 percent for the mineral mass fractions; ±6 percent for the
carbonate-based C02 emission factor for other carbonates; ±10 percent for the carbonate consumption
4-26 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
quantity of limestone; ±3 percent for the carbonate consumption quantity of dolomite; and ±6 percent
for the carbonate consumption quantity of all other carbonates (RTI 2025).
The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 4-15. In 2023,
glass production C02 emissions were estimated to be between 1.7 and 1.8 MMT C02 Eq. at the 95
percent confidence level. This indicates a range of approximately 3 percent below and 3 percent above
the emission estimate of 1.8 MMT C02 Eq.
Table 4-15: Approach 2 Quantitative Uncertainty Estimates for C02 Emissions from
Glass Production (MMT C02 Eq. and Percent)
Uncertainty Range Relative to Emission Estimate"
Source
Gas
2023 Emission Estimate
(MMT CO2 Eq.)
(MMT CO2 Eq.)
Lower Upper
Bound Bound
(%)
Lower
Bound
Upper
Bound
Glass Production
C02
1.8
1.7 1.8
-3%
+3%
a Range of emission estimates predicted by Monte Carlo stochastic simulation for a 95 percent confidence interval.
QA/QC and Verification
General quality assurance/quality control (QA/QC) procedures were applied consistent with the U.S.
Inventory QA/QC plan, which is in accordance with Volume 1, Chapter 6 of 2006IPCC Guidelines as
described in the introduction of the IPPU chapter (see Annex 8 for more details). For the GHGRP data,
EPA verifies annual facility-level reports through a multi-step process (e.g., including a combination of
pre-and post-submittal electronic checks and manual reviews by staff) to identify potential errors and
ensure that data submitted to EPA are accurate, complete, and consistent (EPA 2015).17 Based on the
results of the verification process, EPA follows up with facilities to resolve mistakes that may have
occurred. The post-submittals checks are consistent with a number of general and category-specific QC
procedures, including: range checks, statistical checks, algorithm checks, and year-to-year checks of
reported data and emissions.
Recalculations Discussion
No recalculations were performed for the 1990 through 2022 portion of the time series.
Planned Improvements
EPA plans to evaluate updates to uncertainty levels for the activity data and mineral mass fraction
values from EPA's GHGRP. This is a near-term planned improvement that is anticipated for inclusion in
2025 report.
Some glass producing facilities in the United States do not report to EPA's GHGRP because they fall
below the reporting threshold for this industry. EPA will continue ongoing research on the availability of
data to better assess the completeness of emission estimates from glass production and how to refine
the methodology to ensure complete national coverage of this category. When reporting began in 2010,
EPA received data from more facilities that were above the reporting threshold than expected, and total
17 GHGRP Report Verification Factsheet. See https://www.epa.gov/sites/production/files/7015-
07/rtocuments/ghgrp verification factsheet.pdf.
Industrial Processes and Product Use 4-27
-------
emissions for these reporting facilities were higher than expected for all glass production facilities in the
United States (EPA 2009). Research will include reassessing previous assessments of GHGRP industry
coverage using the reporting threshold of 25,000 metric tons C02 Eq. This is a medium-term planned
improvement.
4.4 Other Process Uses of Carbonates
(Source Category 2A4)
Limestone (CaC03), dolomite (CaC03MgC03),18 and other carbonates such as soda ash, magnesite,
and siderite are basic materials used by a wide variety of industries, including construction, agriculture,
chemical, metallurgy (i.e., iron and steel production, ferroalloy production, and magnesium production),
glass production, environmental pollution control, ceramics production, and non-metallurgical
magnesia production. This reporting category (2A4) includes emissions from other uses of limestone,
dolomite, and other carbonates not included in other categories; the production of ceramics; other uses
of soda ash not included elsewhere; and the production of non-metallurgical magnesia. This section
addresses mineral industry use of these carbonates: limestone, dolomite, soda ash, and magnesite.
Emissions from the use of these carbonates are organized into four subcategories: other process uses of
carbonates (i.e., limestone and dolomite consumption), ceramics production, other uses of soda ash,
and non-metallurgical magnesia production.
For industrial applications, carbonates are heated sufficiently enough to calcine the material and
generate C02 as a byproduct.
CctCO3 -> CaO -I- CO2
MgCOs MgO -I- CO2
Examples of such applications include limestone used as a flux or purifier in metallurgical furnaces, as a
sorbent in flue gas desulfurization (FGD) systems for utility and industrial plants, and as a raw material
for the production of glass, lime, and cement.
Emissions from limestone and dolomite used in the production of cement, lime, glass, and iron and
steel are excluded from the other process uses of carbonates category and reported under their
respective source categories (e.g., Section 4.3, Glass Production). Emissions from soda ash production
are reported under Section 4.12, Soda Ash Production (Source Category 2B7). Emissions from soda ash
consumption associated with glass manufacturing are reported under Section 4.3, Glass Production
(Source Category 2A3). Emissions from the use of limestone and dolomite in liming of agricultural soils
are included in the Agriculture chapter under Section 5.5, Liming (Source Category 3G). Emissions from
limestone and dolomite used in the production of iron and steel and magnesium production are
reported under Section 4.18, Iron and Steel Production (Source Category 2C1). Emissions from dolomite
used in the production of magnesium are reported under Section 4.21, Magnesium Production and
Processing (Source Category 2C4). As noted in Section 4.19, Ferroalloy Production (Source Category
2C2), emissions from the production of ferromanganese are not included in this Inventory because of
18 Limestone and dolomite are collectively referred to as limestone by the industry, and intermediate varieties are seldom
distinguished.
4-28 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
the small number of manufacturers of these materials in the United States. Government information
disclosure rules prevent the publication of production data for these production facilities. Emissions
from fuels consumed for energy purposes during these processes are accounted for as part of fossil fuel
combustion in the industrial end-use sector reported under the Energy chapter in Section 3.1, Fossil
Fuel Combustion (Source Category 1 A). Both lime (CaO) and limestone (CaC03) can be used as a
sorbent for FGD systems. Emissions from lime consumption for FGD systems and from sugar refining
are reported under Section 4.3, Lime Production (Source Category 2A2). Emissions from the use of
dolomite in primary magnesium metal production are reported under Section 4.21, Magnesium
Production and Processing (Source Category 2C4).
Limestone and dolomite are widely distributed throughout the world in deposits of varying sizes and
degrees of purity. Large deposits of limestone occur in nearly every state in the United States, and
significant quantities are extracted for industrial applications. In 2018, the leading limestone producing
states were Texas, Florida, Ohio, Missouri, and Pennsylvania, which contributed 46 percent of the total
U.S. output (USGS 2022a). Dolomite deposits are found in the United States, Canada, Mexico, Europe,
Africa, and Brazil. In the United States, the leading dolomite producing states are Pennsylvania, New
York, and Utah which currently contribute more than a third of the total U.S. output (USGS 2022a).
Ceramics include the production of bricks and roof tiles, vitrified clay pipes, refractory products,
expanded clay products, wall and floor tiles, table and ornamental ware (i.e., household ceramics),
sanitary ware, technical ceramics (e.g., aerospace, automotive, electronic, or biomedical applications),
and inorganic bonded abrasives. Most ceramic products are made from one or more different types of
clay (e.g., shales, fire clay, and ball clay) with varying carbonate contents. The process of manufacturing
ceramic products, regardless of the product type or scale, is essentially the same. This process consists
of raw material processing (grinding, calcining, and drying), forming (wet or dry process), firing (single or
multiple stage firing process), and final processing. Process C02 emissions are produced during the
calcination process in the kiln or dryer, where carbonates are heated to high temperatures which results
in metal oxides and C02. In 2018, the leading clay producing states were Georgia, Wyoming, Texas,
Alabama, and North Carolina, which contributed 60 percent of the total U.S. output (USGS 2022f).
Other uses of soda ash include the consumption of soda ash for non-glass purposes. Excluding glass
production, soda ash consumption by end use in 2023 included chemicals, 54 percent, soap and
detergent manufacturing, 9 percent; distributers, 11 percent; flue gas desulfurization, 6 percent; other
uses, 17 percent; pulp and paper production, 2 percent; and water treatment, 2 percent (USGS 2024).
Chemicals produced using soda ash include sodium-based inorganic chemicals such as sodium
bicarbonate, sodium chromates, sodium phosphates, and sodium silicates. (USGS 2022g).
Internationally, two types of soda ash are produced: natural and synthetic. In 2019, 93 percent of the
global soda ash production came from China, the United States, Russia, Germany, India, Turkey, Poland,
and France. The United States only produces natural soda ash and only in two states: Wyoming and
California (USGS 2021a).
Non-metallurgical magnesia production comprises three categories of magnesia products: calcined
magnesia, deadburned magnesia, and fused magnesia. Magnesia is produced by calcining magnesite
(MgC03) which results in the release of C02. Non-metallurgical magnesia is used in agricultural,
industrial, refractory, and electrical insulating applications. Specific applications include fertilizers,
construction materials, plastics, and flue gas desulphurization. China, Russia, and Turkey account for
83 percent of global production capacity of magnesia from magnesite (USGS 2022e). In the United
Industrial Processes and Product Use 4-29
-------
States, only one facility located in Nevada produces non-metallurgical magnesia using magnesite as the
raw material.
In 2023,12,668 kilotons (kt) of limestone, 682 kt of dolomite, 2,408 kt of soda ash, and 513 kt of
magnesite were consumed for these emissive applications, which excludes consumption for the
production of cement, lime, glass, and iron and steel (Willett 2024; USGS 2022c). Usage of limestone,
dolomite, soda ash, and magnesite resulted in aggregate C02 emissions of 7.2 MMT C02 Eq. (7,163 kt)
(see Table 4-16 and Table 4-17). The 2023 emissions decreased 31 percent compared to 2022, primarily
as a result of decreased limestone and dolomite consumption attributed to flux stone and decreased
limestone consumption attributed to sulfur oxide removal. Overall emissions for 2023 have increased
0.9 percent compared to 1990.
Table 4-16: C02 Emissions from Other Process Uses of Carbonates (MMT C02 Eq.)
Year
1990
2005
2019
2020
2021
2022
2023
Other Uses of Carbonates
4.8
6.2
7.4
7.4
7.0
8.8
5.5
Ceramics Production
0.8
0.8
0.4
0.4
0.4
0.4
0.4
Other Uses of Soda Asha
1.4
1.3
1.0
1.0
1.0
1.0
1.0
Non-Metallurgical Magnesia Production
0.1
0.2
0.2
0.2
0.2
0.2
0.3
Total
7.1
8.5
9.0
9.0
8.6
10.4
7.2
a Soda ash consumption not associated with glass manufacturing.
Note: Totals may not sum due to independent rounding.
Table 4-17: C02 Emissions from Other Process Uses of Carbonates (kt C02)
Year
1990
2005
2019
2020
2021
2022
2023
Other Uses of Carbonates
4,843
6,155
7,386
7,441
6,972
8,780
5,492
Ceramics Production
757
822
399
397
400
407
401
Other Uses of Soda Asha
1,390
1,305
1,036
958
979
992
999
Non-Metallurgical Magnesia Production
113
191
152
216
231
204
270
Total
7,103
8,472
8,973
9,012
8,583
10,383
7,163
a Soda ash consumption not associated with glass manufacturing.
Note: Totals may not sum due to independent rounding.
Methodology and Time-Series Consistency
Other Uses of Carbonates (Limestone and Dolomite Consumption)
Carbon dioxide emissions from other uses of carbonates, specifically limestone and dolomite
consumption, were calculated using a Tier 2 method from the 2006IPCC Guidelines, in accordance with
the IPCC methodological decision tree and available data, by multiplying the quantity of limestone or
dolomite consumed by the emission factor for limestone or dolomite calcination, respectively: 0.43971
metric ton C02/metric ton carbonate for limestone and 0.47732 metric ton C02/metric ton carbonate for
dolomite.19 This methodology was used for limestone and dolomite used for flux stone, flue gas
desulfurization systems, chemical stone, mine dusting or acid water treatment, and acid neutralization.
Flux stone used during the production of iron and steel was deducted from the other uses of carbonates
19 2006 IPCC Guidelines, Volume 3: Chapter 2, Table 2.1.
4-30 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
source category estimate and attributed to the iron and steel production source category estimate.
Similarly, limestone and dolomite consumption for glass manufacturing, cement, and lime
manufacturing are excluded from this category and attributed to their respective categories.
Consumption data for 1990 through 2023 of limestone and dolomite used for flux stone, flue gas
desulfurization systems, chemical stone, mine dusting or acid water treatment, and acid neutralization
(see Table 4-18) were obtained from the U.S. Geological Survey (USGS) Minerals Yearbook: Crushed
Stone Annual Report (1995a through 2023b), preliminary data for 2022 and 2023 from USGS Crushed
Stone Commodity Expert (Willett 2023, Willett 2024), American Iron and Steel Institute limestone and
dolomite consumption data (AISI 2018 through 2021), and the U.S. Bureau of Mines (1991 and 1993a),
which are reported to the nearest ton. In addition, the estimated values for limestone and dolomite
consumption for flux stone used during the production of iron and steel were adjusted using emissions
data from the EPA's Greenhouse Gas Reporting Program (GHGRP) Subpart Q for the iron and steel sector
for 2020 through 2023. Iron and steel GHGRP process emissions data increased by approximately 5
percent from 2022 to 2023 (EPA 2024). This adjustment method is consistent with the method used in
Section 4.18, Iron and Steel Production (Source Category 2C1).
During 1990 and 1992, the USGS did not conduct a detailed survey of limestone and dolomite
consumption by end-use; therefore, data on consumption by end use for 1990 was estimated by
applying the 1991 ratios of total limestone and dolomite consumption by end use to total 1990
limestone and dolomite consumption values. Similarly, the 1992 consumption figures were
approximated by applying an average of the 1991 and 1993 ratios of total limestone and dolomite use by
end uses to the 1992 total values.
In 1991, the U.S. Bureau of Mines, now known as the USGS, began compiling production and end use
information through surveys of crushed stone manufacturers. Manufacturers provided different levels of
detail in survey responses, so information was divided into three categories: (1) production by end-use,
as reported by manufacturers (i.e., "specified" production); (2) production reported by manufacturers
without end-uses specified (i.e., "unspecified-reported" production); and (3) estimated additional
production by manufacturers who did not respond to the survey (i.e., "unspecified-estimated"
production). Additionally, each year the USGS withholds data on certain limestone and dolomite end-
uses due to confidentiality agreements regarding company proprietary data. For the purposes of this
analysis, emissive end-uses that contained withheld data were estimated using one of the following
techniques: (1) the value for all the withheld data points for limestone or dolomite use was distributed
evenly to all withheld end-uses; (2) the average percent of total limestone or dolomite for the withheld
end-use in the preceding and succeeding years; or (3) the average fraction of total limestone or dolomite
for the end-use over the entire time period.
A large quantity of crushed stone was reported to the USGS under the category "unspecified uses." A
portion of this consumption is believed to be limestone or dolomite used for emissive end uses. The
quantity listed for "unspecified uses" was, therefore, allocated to all other reported end-uses according
to each end-use's fraction of total consumption in that year.20
Table 4-18: Limestone and Dolomite Consumption from Other Uses of Carbonates (kt)
Activity
1990
2005
2019
2020
2021
2022
2023
Limestone
10,016
10,465
15,146
13,707
12,788
17,891
11,897
20 This approach was recommended by USGS, the data collection agency.
Industrial Processes and Product Use 4-31
-------
Activity
1990
2005
2019
2020
2021
2022
2023
Dolomite
919
3,254
1,520
2,962
2,826
1,914
547
Total
10,935
13,719
16,667
16,669
15,614
19,805
12,444
Note: Totals may not sum due to independent rounding.
Ceramics Production
Carbon dioxide emissions from ceramics production were calculated using a Tier 1 method from the
2006IPCC Guidelines, in accordance with the IPCC methodological decision tree and available data, by
multiplying the quantity of clay consumed for emissive purposes by a carbonate content value for clay
of 10 percent, limestone fraction of 85 percent and dolomite fraction of 15 percent, and by the emission
factor for limestone or dolomite calcination, respectively: 0.43971 metric ton C02/metric ton of
limestone and 0.47732 metric ton C02/metric ton of dolomite.21 To estimate annual process C02
emissions, EPA evaluated the end-uses of each type of clay published by USGS to identify the emissive
end-uses that fall into the ceramics production subcategory. The emissive end-uses were organized into
three groups: ceramics, glass, and floor & tile; refractories; and heavy clay products. The total limestone
and dolomite consumption from the three emissive groupings for ceramics production for 1990 through
2023 (see Table 4-19) were obtained from USGS (Simmons 2025).
Table 4-19: Limestone and Dolomite Consumption from Ceramics Production (kt)
Activity
1990
2005
2019
2020
2021
2022
2023
Limestone
1,444
1,569
762
758
764
776
766
Dolomite
255
277
135
134
135
137
135
Total
1,699
1,846
897
892
899
913
901
Note: Totals may not sum due to independent rounding.
Other Uses of Soda Ash
Carbon dioxide emissions from soda ash consumption were calculated using a Tier 1 method from the
2006 IPCC Guidelines, in accordance with the IPCC methodological decision tree and available data.
Excluding glass manufacturing which is reported under Section 4.3 Glass Production (Source Category
2A3), most soda ash is consumed in chemical production, with smaller amounts used in soap
production, pulp and paper, flue gas desulfurization, and water treatment. In these applications, it is
assumed that one mole of carbon is released for every mole of soda ash used. Thus, approximately
0.113 metric tons of carbon (or 0.415 metric tons of C02) are released for every metric ton of soda ash
consumed. The activity data for soda ash consumption for 1990 to 2023 (see Table 4-20) were obtained
from the U.S. Geological Survey (USGS) Minerals Yearbook for Soda Ash (1994 through 2015b) and USGS
Mineral Industry Surveys for Soda Ash (USGS 2017a, 2018, 2019, 2020b, 2021 b, 2022b, 2023a, 2024).
Soda ash consumption data were collected by the USGS from voluntary surveys of the U.S. soda ash
industry.
21 2006 IPCC Guidelines, Volume 3: Chapter 2, Table 2.1.
4-32 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Table 4-20: Other Uses of Soda Ash Consumption Not Associated with Glass
Manufacturing (kt)
Activity
1990
2005
2019
2020
2021
2022
2023
Soda Asha
3,351
3,144
2,497
2,310
2,360
2,391
2,408
a Soda ash consumption is sales reported by producers which exclude imports. Historically, imported soda ash is less than 1
percent of the total U.S. consumption (Kostick2012).
Non-Metallurgical Magnesia Production
Carbon dioxide emissions from non-metallurgical magnesia production were calculated using a Tier 1
method from the 2006IPCC Guidelines, in accordance with the IPCC methodological decision tree and
available data, by multiplying the quantity of magnesium ore extracted from the mine and processed at
the facility by the carbonate content for magnesite or limestone, respectively, and by the emission factor
for magnesite or limestone calcination, respectively: 0.52197 metric ton C02/metric ton carbonate for
magnesite and 0.43971 metric ton C02/metric ton carbonate for limestone.22 A USGS report on
magnesite deposits at Gabbs, Nevada lists the carbonate content of magnesite as 98 percent
magnesite and 1 percent limestone (USGS 1948). In the absence of other data, all magnesium ore
extracted from the mine is assumed to be used for non-metallurgical magnesium production.
Magnesium ore extracted from the mine and processed at the facility for non-metallurgical magnesia
production for 2002 through 2023 (see Table 4-21) was obtained from the Nevada Department of
Environmental Quality (McNeece 2023, McNeece 2024). This data was not available for 1990 through
2001. To address this gap in data availability and time-series consistency, carbonate consumption for
1990 through 2001 were estimated by multiplying the average ratio of magnesium ore consumption to
production capacity for 2002 to 2004 bythe production capacity of the facility in Nevada. Production
capacity for 1990 through 2001 was obtained from the USGS Minerals Yearbook for Magnesium
Compounds (USGS 1990 through 2002).
Table 4-21: Magnesite and Limestone Consumption from Non-Metallurgical Magnesia
Production (kt)
Activity
1990
2005
2019
2020
2021
2022
2023
Magnesite
214
363
289
410
439
388
513
Limestone
2
4
3
4
4
4
5
Total
216
367
292
414
443
392
518
Note: Totals may not sum due to independent rounding.
Methodological approaches were applied to the entire time series to ensure consistency in emissions
from 1990 through 2023. Consistent with the 2006 IPCC Guidelines, the overlap technique was applied
for non-metallurgical magnesia production to compare the magnesium ore consumption data to
production capacity data for years where there was overlap. To address inconsistencies, adjustments
were made, as described above.
Uncertainty
The uncertainty levels presented in this section account for uncertainty associated with activity data.
Data on limestone and dolomite consumption are collected by USGS through voluntary national
22 2006 IPCC Guidelines, Volume 3: Chapter 2, Table 2.1.
Industrial Processes and Product Use 4-33
-------
surveys. USGS contacts the mines (i.e., producers of various types of crushed stone) for annual sales
data. Data on other carbonate consumption are not readily available. The producers report the annual
quantity sold to various end-users and industry types. USGS estimates the historical response rate for
the crushed stone survey to be approximately 70 percent, and the rest is estimated by USGS. Large
fluctuations in reported consumption exist, reflecting year-to-year changes in the number of survey
responders. The uncertainty resulting from a shifting survey population is exacerbated by the gaps in the
time series of reports. The accuracy of distribution by end use is also uncertain because this value is
reported by the producer/mines and not the end user. Additionally, there is significant inherent
uncertainty associated with estimating withheld data points for specific end uses of limestone and
dolomite. Lastly, much of the limestone consumed in the United States is reported as "other unspecified
uses;" therefore, it is difficult to accurately allocate this unspecified quantity to the correct end-uses.
EPA contacted the USGS National Minerals Information Center Crushed Stone commodity expert to
assess the current uncertainty ranges associated with the limestone and dolomite consumption data
compiled and published by USGS. During this discussion, the expert confirmed that EPA's range of
uncertainty was still reasonable (Willett 2017). EPA assigned an uncertainty range of ±10 percent for
limestone and dolomite consumption, based on expert judgement (Willett 2017). EPA assigned an
uncertainty range of ±5 percent for soda ash consumption, and using this suggested uncertainty
provided in Volume 3, Chapter 2, Section 2.4.2.2 of the 2006IPCC Guidelines is appropriate based on
expert judgment (RTI 2023).
Uncertainty in the estimates also arises in part due to variations in the chemical composition of
limestone. In addition to calcium carbonate, limestone may contain smaller amounts of magnesia,
silica, and sulfur, among other minerals. The exact specifications for limestone or dolomite used as flux
stone vary with the pyrometallurgical process and the kind of ore processed. EPA assigned an
uncertainty range of ±3 percent for the C02 emission factors for limestone and dolomite consumption,
and using this suggested uncertainty provided in Volume 3, Chapter 2, Section 2.5.2.1 of the 2006 IPCC
Guidelines is appropriate based on expert judgment (RTI 2023).
For emissions from ceramics production, data on clay consumption are collected by USGS through
voluntary national surveys. Large fluctuations in reported consumption exist, reflecting year-to-year
changes in the number of survey responders. The accuracy of distribution by end use is also uncertain
because this value is reported by the producer and not the end user. Uncertainty in the estimates also
arises in part due to the variations in the carbonate content of the various clays used for the various
types of ceramics. As discussed above, as no information is available on the carbonate content for each
clay, fractions of limestone and dolomite consumed and a carbonate content for clay from the 2006
IPCC Guidelines are used. EPA assigned an uncertainty range of ±10 percent for the activity data and ±3
percent for the emission factors, consistent with uncertainty ranges for limestone and dolomite activity
data and emission factors for other process uses of carbonates, respectively.
For emissions from soda ash consumption, the primary source of uncertainty results from the fact that
these emissions are dependent upon the type of processing employed by each end-use. Specific
emission factors for each end-use are not available, so a Tier 1 default emission factor is used for all
end-uses. Therefore, there is uncertainty surrounding the emission factors from the consumption of
soda ash. Additional uncertainty comes from the reported consumption and allocation of consumption
within sectors that is collected on a quarterly basis by the USGS. Efforts have been made to categorize
company sales within the correct end-use sector. EPA assigned an uncertainty range of ±2 percent for
the C02 emission factor for soda ash consumption. The uncertainty range is derived from the default
4-34 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
ranges for soda ash consumption for glass production in Volume 3, Chapter 2, Section 2.4.2.1 of the
2006IPCC Guidelines which is representative of soda ash consumption not associated with glass
production, based on expert judgment (RTI 2023).
For non-metallurgical magnesia production, uncertainties arise due to variations in the chemical
composition of the carbonates used in production of caustic-calcined magnesia production. As noted,
minor quantities of other carbonates beyond limestone and magnesite are also used but unknown.
These other carbonates are likely small and have a minimal impact on the derived emission factor. EPA
assigned an uncertainty range of ±10 percent for the activity data and ±3 percent for the emission
factors, consistent with uncertainty ranges for limestone and dolomite activity data and emission
factors for other process uses of carbonates, respectively. The results of the Approach 2 quantitative
uncertainty analysis are summarized in Table 4-22.
A normal probability density function was assigned for all activity data, and a triangular probability
density function was assigned for all emission factors (RTI 2023). Carbon dioxide emissions from other
process uses of carbonates in 2023 were estimated to be between 6.4 and 8.3 MMT C02 Eq. at the 95
percent confidence level. This indicates a range of approximately 11 percent below and 15 percent
above the emission estimate of 7.2 MMT C02 Eq.
Table 4-22: Approach 2 Quantitative Uncertainty Estimates for C02 Emissions from
Other Process Uses of Carbonates (MMT C02 Eq. and Percent)
2023 Emission Estimate
Uncertainty Range Relative to Emission Estimate"
Source
Gas
(MMT CO2 Eq.)
(MMT CO2 Eq.)
(MMT CO2 Eq.)
Lower
Bound
Upper
Bound
Lower Upper
Bound Bound
Other Process Use
of Carbonates
C02
7.2
6.4
8.3
-11% +15%
a Range of emission estimates predicted by Monte Carlo stochastic simulation for a 95 percent confidence interval.
QA/QC and Verification
General quality assurance/quality control (QA/QC) procedures were applied consistent with the U.S.
Inventory QA/QC plan, which is in accordance with Volume 1, Chapter 6 of 2006 IPCC Guidelines as
described in the introduction of the IPPU chapter (see Annex 8 for more details).
Recalculations Discussion
No recalculations were performed for the 1990 through 2022 portion of the time series.
Planned Improvements
EPA plans to review the uncertainty ranges assigned to activity data. This planned improvement is
currently planned as a medium-term improvement.
Industrial Processes and Product Use 4-35
-------
4.5 Ammonia Production (Source Category
2B1)
Emissions of carbon dioxide (C02) occur during the production of synthetic ammonia (NH3), primarily
through the use of natural gas, petroleum coke, or naphtha as a feedstock. The natural gas-, naphtha-,
and petroleum coke-based processes produce C02 and hydrogen (H2), the latter of which is used in the
production of ammonia. The brine electrolysis process for production of ammonia does not lead to
process-based C02 emissions. This reporting category (2B1) includes emissions from the production of
ammonia. Due to national circumstances, emissions from fuels consumed for energy purposes during
the production of ammonia are accounted for as part of fossil fuel combustion in the industrial end-use
sector reported under the Energy chapter. More information on this approach can be found in the
Methodology section below.
Ammonia production requires a source of nitrogen (N) and hydrogen (H). Nitrogen is obtained from air
through liquid air distillation or an oxidative process where air is burnt and the residual nitrogen is
recovered. In the United States, the majority of ammonia is produced using a natural gas feedstock as
the hydrogen source. One synthetic ammonia production plant located in Kansas is producing ammonia
from petroleum coke feedstock. In some U.S. plants, some of the C02 produced by the process is
captured and used to produce urea rather than being emitted to the atmosphere. In 2023,17 companies
operated 36 ammonia producing facilities in 17 states. Approximately 60 percent of domestic ammonia
production capacity is concentrated in Louisiana, Oklahoma, and Texas (USGS 2024).
Synthetic ammonia production from natural gas feedstock consists of five principal process steps. The
primary reforming step converts methane (CH4) to C02, carbon monoxide (CO), and hydrogen (H2) in the
presence of a catalyst. Only 30 to 40 percent of the CH4 feedstock to the primary reformer is converted
to CO and C02 in this step of the process. The secondary reforming step converts the remaining CH4
feedstock to CO and C02. In the shift conversion step, the CO in the process gas from the secondary
reforming step (representing approximately 15 percent of the process gas) is converted to C02 in the
presence of a catalyst, water, and air. Carbon dioxide is removed from the process gas by the shift
conversion process, and the H2is combined with the nitrogen (N2) gas in the process gas during the
ammonia synthesis step to produce ammonia. The C02 is included in a waste gas stream with other
process impurities and is absorbed by a scrubber solution. In regenerating the scrubber solution, C02 is
released from the solution.
The conversion process for conventional steam reforming of CH4, including the primary and secondary
reforming and the shift conversion processes, is approximately as follows:
0.88CH4 + 1.26Air + 1.24H20 ^ 0.88C02 + N2 + 3H2
N2 + 3 H2 ^ 2 NH3
To produce synthetic ammonia from petroleum coke, the petroleum coke is gasified and converted to
C02 and H2. These gases are separated, and the H2 is used as a feedstock to the ammonia production
process, where it is reacted with N2 to form ammonia.
Not all of the C02 produced during the production of ammonia is emitted directly to the atmosphere.
Some of the ammonia and some of the C02 produced by the synthetic ammonia process are used as
4-36 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
raw materials in the production of urea [CO(NH2)2], which has a variety of agricultural and industrial
applications.
The chemical reaction that produces urea is:
2NH3+ C02 -> NH2C00NH4 -> CO(NH2)2 +h2o
Only the C02 emitted directly to the atmosphere from the synthetic ammonia production process is
accounted for in determining emissions from ammonia production. The C02 that is captured during the
ammonia production process and used to produce urea does not contribute to the C02 emission
estimates for ammonia production presented in this section. Instead, C02 emissions resulting from the
consumption of urea are attributed to the urea consumption or urea application source category (under
the assumption that the carbon stored in the urea during its manufacture is released into the
environment during its consumption or application). Emissions of C02 resulting from agricultural
applications of urea are accounted for in Section 5.6. Emissions of C02 resulting from non-agricultural
applications of urea (e.g., use as a feedstock in chemical production processes) are accounted for in
Section 4.6.
Another consideration in calculating emissions from ammonia production is C02 that is geologically
sequestered. There is one C02 sequestration facility associated with ammonia production in the United
States that reports to GHGRP Subpart RR (The Geologic Sequestration of Carbon Dioxide). The North
Burbank Unit has received C02 produced via gasification operations at the Coffeyville Resources
ammonia production facility since 2020. The C02 that is captured from the ammonia production
process and sequestered does not contribute to the C02 emission estimates for ammonia production.
This C02 is subtracted from the overall emissions from ammonia production. See Section 3.9 for more
detail on including C02 sequestration in the Inventory.
Emissions from fuel used for energy at ammonia plants are accounted for as part of fossil fuel
combustion in the industrial end-use sector reported under the Energy chapter. The consumption of
natural gas and petroleum coke as fossil fuel feedstocks for NH3 production are adjusted for within the
Energy chapter as these fuels were consumed during non-energy related activities. More information on
this methodology is described in Annex 2.1, Methodology for Estimating Emissions of C02from Fossil
Fuel Combustion.
Total net emissions of C02 from ammonia production in 2023 were 12.2 MMT C02 Eq. (12,211 kt) and are
summarized in Table 4-23 and Table 4-24. Ammonia production relies on natural gas as both a feedstock
and a fuel, and as such, market fluctuations and volatility in natural gas prices affect the production of
ammonia. Since 1990, emissions from ammonia production have decreased by 15 percent. Emissions
in 2023 increased by about 2 percent from the 2022 levels. One facility in Kansas produces ammonia
from petroleum coke and began operations in 2000. All other facilities use natural gas as feedstock.
Emissions from ammonia production increased steadily from 2015 to 2018, due to the addition of new
ammonia production facilities and new production units at existing facilities in 2016, 2017, and 2018.
Agriculture continues to drive demand for nitrogen fertilizers, accounting for approximately 88 percent
of domestic ammonia consumption (USGS 2024).
Table 4-23: C02 Emissions from Ammonia Production (MMT C02 Eq.)
Source
1990
2005
2019
2020
2021
2022
2023
Ammonia Production
14.4
10.2
12.4
12.3
11.5
11.9
12.2
Industrial Processes and Product Use 4-37
-------
Table 4-24: C02 Emissions from Ammonia Production (kt C02)
Source
1990
2005
2019
2020
2021
2022
2023
Ammonia Production
14,404
10,234
12,388
12,335
11,458
11,945
12,211
Methodology and Time-Series Consistency
Estimates of C02 emissions from the production of synthetic ammonia for 2010 through 2023 are
estimated using a country-specific approach consistent with a Tier 3 method from the 2006IPCC
Guidelines, in accordance with the IPCC methodological decision tree and available data (IPCC 2006).
The methodology for 2010 to 2023 directly uses the process C02 emissions reported to Subpart G of the
U.S. EPA Greenhouse Gas Reporting Program (GHGRP) (EPA 2018; EPA 2024a). Estimates for 1990 to
2009 emissions are based on reported and calculated data on natural gas and petroleum coke
feedstock used for ammonia production, consistent with IPCC Tier 2 methods and in accordance with
the IPCC methodological decision tree and available data.
Emissions from fuel used for energy at ammonia plants are accounted for in the Energy chapter. This
approach differs slightly from the 2006 IPCC Guidelines which indicates that "in the case of ammonia
production no distinction is made between fuel and feedstock emissions with all emissions accounted
for in the IPPU Sector." Disaggregated data on fuel used for ammonia feedstock and fuel used for energy
for ammonia production are not available in the United States. The Energy Information Administration
(EIA), where energy use data are obtained for the Inventory (see the Energy chapter), does not provide
data broken out by industrial category. EIA data are only available at the broad industry sector level.
Furthermore, the GHGRP data used to estimate emissions are based on feedstock use and not fuel use.
The method uses the same science informing the 2006 IPCC Guidelines and is consistent with avoiding
double counting in the reporting of fuel use emissions under Energy and IPPU reporting. See more
information in introduction to this Chapter.
Petroleum Coke Feedstock
Since 2000, one facility in the United States has produced ammonia using petroleum coke as a
feedstock. For 2010 to 2023, C02 emissions from the production of synthetic ammonia from petroleum
coke feedstock were estimated using C02 emissions reported by the facility to GHGRP (EPA 2018; EPA
2024a).
For 2006 to 2009, C02 emissions from the production of synthetic ammonia from petroleum coke
feedstock were estimated by multiplying the following: quantity of petroleum coke feedstock reported
by the facility (CVR 2008 through 2023); the Inventory heating content value for petroleum coke
(consistent with values used in the Energy chapter); the petroleum coke carbon content; and a
stoichiometric C02/C factor of 44/12.
For 2000 to 2005, the quantity of petroleum coke feedstock was not available and was estimated by
multiplying the average ratio of petroleum coke feedstock quantity to ammonia production quantity
produced from petroleum coke from 2006 through 2010 by total ammonia production for 2000 to 2005
(ACC 2024). The years 2006 to 2010 were used to determine the average ratio of petroleum coke
feedstock quantity to the ammonia quantity produced from petroleum coke because that period was
deemed to better represent historic ammonia production from petroleum coke for the period from 2000
to 2005.
4-38 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
For 2000 to 2005, C02 emissions from the production of synthetic ammonia from petroleum coke
feedstock were estimated by multiplying the following: the average ratio of petroleum coke feedstock
quantity to ammonia production quantity; total ammonia production quantity (ACC 2024); the Inventory
heating content value for petroleum coke (consistent with values used in the Energy chapter); the
petroleum coke carbon content; and the stoichiometric ratio of C02 to C (44/12).
Natural Gas Feedstock
For 2017 through 2023, facilities directly reported to GHGRP the quantity of natural gas feedstock used
for ammonia production along with the carbon content of the natural gas feedstock (EPA 2018; EPA
2024a).
For 2010 through 2016, the quantity of natural gas feedstock was calculated using GHGRP process C02
emissions for 2010 through 2016, average molecular weight of the feedstock from 2017 through 2021,
and average carbon content from 2017 through 2021. Data from years 2017 to 2021 were used to
determine the average molecular weight and the average carbon content because that period better
represents historic ammonia production from 2010 to 2016. Using all available data from 2017 to 2021
allowed for the maximum number of data points available at the time of adopting this methodology to
ensure that the average was representative. The averages were not updated using later data to exclude
any new facilities that might not be representative of facilities that were operating during the earlier
years of the GHGRP.
For 2010 to 2023, C02 emissions from the production of synthetic ammonia from natural gas feedstock
were estimated using the C02 emissions reported to the GHGRP (EPA 2018; EPA 2024a) and subtracting
the C02 emissions from the production of synthetic ammonia from petroleum coke feedstock as
determined in the Petroleum Coke Feedstock section above.
For 1990 to 2009, the quantity of natural gas feedstock was not available and was estimated by
multiplying the average ratio of natural gas feedstock quantity to ammonia production quantity from
2010 through 2014 by total ammonia production for each year for 1990 to 2009 (ACC 2024). The years
2010 to 2014 were used to determine the average ratio of natural gas feedstock quantity to ammonia
production because that period better represents historic ammonia production from 1990 to 2009.23 For
1990 to 2009, C02 emissions from the production of synthetic ammonia from natural gas feedstock
were estimated using the natural gas feedstock quantity as determined above and the Inventory C02
emissions factor and heating content value for natural gas (consistent with values used in the Energy
chapter).
Urea Production and Sequestered C02 Adjustments
Emissions of C02 from ammonia production from both feedstocks and for all years from 1990 to 2023
were adjusted to account for the use of some C02 emissions resulting from ammonia production as a
raw material in the production of urea and the capture and sequestration of some C02 emissions from
ammonia production. For urea, the C02 emissions reported for ammonia production are reduced by a
factor of 0.733, which corresponds to a stoichiometric C02/urea factor of 44/60, assuming complete
23 The number of facilities reporting to GHGRP has increased since 2010: 22 facilities reported from 2010 to 2012; 23 from
2013 to 2015; 26 in 2016; and 29 from 2017 to 2023. Using data from 2010 to 2014 excludes the newer facilities that
might not be representative of facilities in earlieryears.
Industrial Processes and Product Use 4-39
-------
conversion of ammonia (NH3) and C02to urea (IPCC 2006; EFMA2000) and multiplied by total annual
domestic urea production.
All synthetic ammonia production and subsequent urea production are assumed to be from the same
process—conventional catalytic reforming of natural gas feedstock, with the exception of ammonia
production from petroleum coke feedstock at the one facility located in Kansas.
Table 4-25: Total Ammonia Production, Total Urea Production, Recovered CO2 Consumed for
Urea Production, and Sequestered CO2 (kt)
Year
1990
2005
2019
2020
2021
2022
2023
Total Ammonia Production
15,425
10,143
16,410
17,020
15,420
16,800
17,800
Total Urea Production
7,450
5,270
11,400
11,500
10,521
11,272
11,306
Recovered CO2 Consumed for Urea Production
5,463
3,865
8,360
8,433
7,715
8,266
8,291
Sequestered CO2
-
-
660
714
652
665
Total ammonia production, total urea production, recovered C02 consumed for urea production, and
sequestered C02 are shown in Table 4-25. Total ammonia production data for 2011 through 2023 were
obtained from American Chemistry Council (ACC 2024). For years 1990 through 2010, ammonia
production data were obtained from the Census Bureau of the U.S. Department of Commerce (U.S.
Census Bureau 1991 through 1994,1998 through 2011) as reported in Current Industrial Reports
Fertilizer Materials and Related Products annual and quarterly reports. Data on facility-level process
emissions for 2010 through 2023 and data on natural gas feedstock used and carbon content of the
natural gas feedstock starting in 2017 were obtained from GHGRP (EPA 2018; EPA 2024a). Natural gas
and petroleum coke heating values come from national-level data (EIA 2023), and natural gas and
petroleum coke carbon contents are the same as used in the Energy chapter calculations.
Data on urea production for 2010 through 2023 and sequestered C02 for 2020 through 2023 were
obtained from GHGRP (EPA 2018, EPA 2024b, EPA 2024c). Urea production data for 2009 through 2010
were obtained from the U.S. Census Bureau (U.S. Census Bureau 2010 and 2011). Urea production data
for 1990 through 2008 were obtained from the USGS Minerals Yearbook: Nitrogen (USGS 1994-2009).
The U.S. Census Bureau ceased collection of urea production statistics in 2011.
Methodological approaches were applied to the entire time series to ensure consistency in emissions
from 1990 through 2023. The methodology for ammonia production spliced activity data from different
sources: U. S. Census Bureau data for 1990 through 2010, ACC data beginning in 2011, and GHGRP data
beginning in 2010 and 2017. Consistent with the 2006 IPCC Guidelines, the overlap technique was
applied to compare the two data sets for years where there was overlap, with findings that the data sets
were consistent and adjustments were not needed.
Uncertainty
The uncertainties presented in this section are primarily due to how accurately the emission factor used
represents an average across all ammonia plants using natural gas feedstock. Uncertainty in the back
calculation of natural gas feedstock used for 1990 through 2009 also exists. Using the average ratio of
natural gas feedstock quantity to ammonia production, determined using GHGRP data from 2010 to
2014, does not account for efficiency gains in ammonia production since 1990 (e.g., potential
decreases in gas usage per ton of ammonia, manufacturing shift from steam-driven turbines to
4-40 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
electrical-drive turbines). Uncertainties are also associated with ammonia production estimates and
the assumption that all ammonia production and subsequent urea production was from the same
process—conventional catalytic reforming of natural gas feedstock, with the exception of one ammonia
production plant located in Kansas that is manufacturing ammonia from petroleum coke feedstock.
Uncertainty is also associated with the representativeness of the emission factor used for the petroleum
coke-based ammonia process. It is also assumed that ammonia and urea are produced at co-located
plants from the same natural gas raw material. The uncertainty of the total urea production activity data,
based on USGS Minerals Yearbook: Nitrogen data, is a function of the reliability of reported production
data and is influenced by the completeness of the survey responses. EPA assigned an uncertainty range
of ±2 percent for urea production, natural gas feedstock quantity, petroleum coke feedstock quantity,
and carbon content of natural gas feedstock, and using the suggested uncertainty provided in Section
3.2.3.2 of the 2006IPCC Guidelines is appropriate based on expert judgment (RTI 2023). EPA assigned
an uncertainty range of ±7 percent for ammonia production (RTI 2025). Per these expert judgements, a
normal probability density function was assigned for all variables.
Recovery of C02 from ammonia production plants for purposes other than urea production (e.g.,
commercial sale, etc.) has not been considered in estimating the C02 emissions from ammonia
production, as data concerning the disposition of recovered C02 are not available. Such recovery may or
may not affect the overall estimate of C02 emissions depending upon the end use to which the
recovered C02 is applied. Further research is required to determine whether byproduct C02 is being
recovered from other ammonia production plants for application to end uses that are not accounted for
elsewhere; however, for reporting purposes, C02 consumption for urea production is provided in this
chapter.
The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 4-26. Carbon
dioxide emissions from ammonia production in 2023 were estimated to be between 11.8 and 12.7 MMT
C02 Eq. at the 95 percent confidence level. This indicates a range of approximately 4 percent below and
4 percent above the emission estimate of 12.2 MMT C02 Eq.
Table 4-26: Approach 2 Quantitative Uncertainty Estimates for C02 Emissions from
Ammonia Production (MMT C02 Eq. and Percent)
2023 Emission
Source
Gas
Estimate
Uncertainty Range Relative to Emission Estimate"
(MMT CO2 Eq.)
(MMT CO2 Eq.)
(%)
Lower Upper
Lower Upper
Bound Bound
Bound Bound
Ammonia Production
C02
12.2
11.8 12.7
-4% +4%
a Range of emission estimates predicted by Monte Carlo stochastic simulation for a 95 percent confidence interval.
QA/QC and Verification
General quality assurance/quality control (QA/QC) procedures were applied to ammonia production
emission estimates consistent with the U.S. Inventory QA/QC plan, which is in accordance with Volume
1, Chapter 6 of 2006 IPCC Guidelines as described in the introduction of the IPPU chapter (see Annex 8
for more details). More details on the greenhouse gas calculation, monitoring and QA/QC methods
applicable to ammonia facilities can be found under Subpart G (Ammonia Production) of the regulation
Industrial Processes and Product Use 4-41
-------
(40 CFR Part 98).24 EPA verifies annual facility-level GHGRP reports through a multi-step process (e.g.,
combination of electronic checks and manual reviews) to identify potential errors and ensure that data
submitted to EPA are accurate, complete, and consistent.25 Based on the results of the verification
process, EPA follows up with facilities to resolve mistakes that may have occurred. The post-submittals
checks are consistent with a number of general and category-specific QC procedures, including range
checks, statistical checks, algorithm checks, and year-to-year checks of reported data and emissions.
More details on the greenhouse gas calculation, monitoring, and QA/QC methods applicable to
reporting of urea produced at ammonia production facilities can be found under Section 4.6 Urea
Consumption for Non-Agricultural Purposes.
Recalculations Discussion
For 2020 to 2023, sequestered C02 was incorporated into the emission calculations. In addition, the
GHGRP data for ammonia production for years 2019 to 2022 was adjusted according to the latest data.
As a result, recalculations were performed for emissions from ammonia for 2019 to 2022. Compared to
the previous Inventory, total C02 emissions from the production of ammonia production (from natural
gas and petroleum coke feedstocks) decreased by less than 1 percent (13 kt C02) in 2019 and an
average of 5.5 percent (690 kt C02) per year for 2020 to 2022.
Planned Improvements
Currently the Inventory does not separately track fuel energy use for ammonia production. To be more
consistent with 2006IPCC Guidelines, EPA is considering whether to include natural gas fuel use as part
of ammonia production emissions as a future improvement. The data are still being evaluated as part of
EPA's efforts to disaggregate other industrial sector categories' energy use in the Energy chapter of the
Inventory. If possible, this will be incorporated in future Inventory reports. If incorporated, the fuel energy
use and emissions will be removed from current reporting under Energy to avoid double counting.
4.6 Urea Consumption for Non-Agricultural
Purposes (Source Category 2B10)
Urea is produced using ammonia (NH3) and carbon dioxide (C02) as raw materials. All urea produced in
the United States is assumed to be produced at ammonia production facilities where both ammonia
and C02 are generated. There were 36 plants producing ammonia in the United States in 2023, with two
additional plants sitting idle for the entire year (USGS 2024).
The chemical reaction that produces urea is:
2NH3+ C02 -> NH2C00NH4 -> CO(NH2)2 +h2o
This section accounts for C02 emissions associated with urea consumed exclusively for non-
agricultural purposes. This reporting category (2B10) includes emissions from IPCC assessment reports
24 See http://www.ecfr.gov/cgi-bin/text-idx7tpN/ecfrbrowse/Title40/40cfr98 main 0?.tpl.
25 See https://www.epa.gov/sites/prodijction/files/9015-07/documents/ghgrp verification factsheet.pdf.
4-42 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
that do not fall within any other source category, which includes emissions from urea consumption for
non-agricultural purposes. Emissions of C02 resulting from agricultural applications of urea are
accounted for in Section 5.6 of the Agriculture chapter.
The industrial applications of urea include its use in adhesives, binders, sealants, resins, fillers,
analytical reagents, catalysts, intermediates, solvents, dyestuffs, fragrances, deodorizers, flavoring
agents, humectants and dehydrating agents, formulation components, monomers, paint and coating
additives, photosensitive agents, and surface treatments agents. In addition, urea is used for abating
nitrogen oxide (NOx) emissions from coal-fired power plants and diesel transportation motors.
Emissions of C02 from urea consumed for non-agricultural purposes in 2023 were estimated to be 5.4
MMT C02 Eq. (5,424 kt) and are summarized in Table 4-27 and Table 4-28. Net C02 emissions from urea
consumption for non-agricultural purposes have increased by approximately 43 percent from 1990 to
2023 and decreased by approximately 1 percent from 2022 to 2023.
Table 4-27: C02 Emissions from Urea Consumption for Non-Agricultural Purposes
(MMT C02 Eq.)
Source 1990
2005
2019 2020 2021 2022 2023
Urea Consumption 3.8
3.7
6.2 5.9 6.7 5.5 5.4
Table 4-28: C02 Emissions from Urea Consumption for Non-Agricultural Purposes (kt
co2)
Source 1990
2005
2019 2020 2021 2022 2023
Urea Consumption 3,784
3,653
6,234 5,905 6,724 5,464 5,424
Methodology and Time-Series Consistency
Emissions of C02 resulting from urea consumption for non-agricultural purposes are estimated using a
country-specific method consistent with the Tier 1 method used to estimate emissions from ammonia
production in the 2006IPCC Guidelines which states that the "C02 recovered [from ammonia
production] for downstream use can be estimated from the quantity of urea produced where C02 is
estimated by multiplying urea production by 44/60, the stoichiometric ratio of C02 to urea" (IPCC 2006).
The amount of urea consumed in the United States for non-agricultural purposes is multiplied by a
factor representing the amount of C02 used as a raw material to produce the urea. This method is based
on the assumption that all of the carbon in urea is released into the environment as C02 during use.
The amount of urea consumed for non-agricultural purposes in the United States is estimated by
deducting the quantity of urea fertilizer applied to agricultural lands, which is obtained directly from the
Agriculture chapter (see Table 5-25), from the total domestic supply of urea as reported in Table 4-29.
The domestic supply of urea is estimated based on the amount of urea produced plus urea imports and
minus urea exports. A factor of 0.733 tons of C02 per ton of urea consumed is then applied to the
resulting supply of urea for non-agricultural purposes to estimate C02 emissions from the amount of
urea consumed for non-agricultural purposes. The 0.733 tons of C02 per ton of urea emission factor is
based on the stoichiometry of carbon in urea. This corresponds to a stoichiometric ratio of C02 to urea
of 44/60, assuming complete conversion of carbon in urea to C02(IPCC 2006; EFMA2000).
Industrial Processes and Product Use 4-43
-------
Urea production data for 1990 through 2008 were obtained from the U.S. Geological Survey (USGS)
Minerals Yearbook: Nitrogen (USGS 1994 through 2019). Urea production data for 2009 through 2010
were obtained from the U.S. Census Bureau (2011). The U.S. Census Bureau ceased collection of urea
production statistics in 2011. Urea production data for 2011 through 2023 were obtained from GHGRP
(EPA 2018; EPA 2024a; EPA 2024b).
Urea import data for 2023 were not available at the time of publication and were estimated using 2022
values. Urea import data for 2013 to 2022 were obtained from the USGS Minerals Yearbook: Nitrogen
(USGS 1994 through 2019; USGS 2022; USGS 2023; USGS 2024a). Urea import data for 2011 and 2012
were taken from U.S. Fertilizer Import/Exports from the United States Department of Agriculture (USDA)
Economic Research Service Data Sets (U.S. Department of Agriculture 2012). USDA suspended updates
to this data after 2012. Urea import data for the previous years were obtained from the U.S. Census
Bureau Current Industrial Reports Fertilizer Materials and Related Products annual and quarterly reports
for 1997 through 2010 (U.S. Census Bureau 2001 through 2011), The Fertilizer Institute (TFI 2002) for
1993 through 1996, and the United States International Trade Commission Interactive Tariff and Trade
DataWeb (U.S. ITC 2002) for 1990 through 1992 (see Table 4-29).
Urea export data for 2023 were not available at the time of publication and were estimated using 2022
values. Urea export data for 2013 to 2022 were obtained from the USGS Minerals Yearbook: Nitrogen
(USGS 1994 through 2019; USGS 2022; USGS 2023; USGS 2024a). Urea export data for 1990 through
2012 were taken from U.S. Fertilizer Import/Exports from USDA Economic Research Service Data Sets
(U.S. Department of Agriculture 2012). USDA suspended updates to this data after 2012.
Table 4-29: Urea Production, Urea Applied as Fertilizer, Urea Imports, and Urea
Exports (kt)
Year
1990
2005
2019
2020
2021
2022
2023
Urea Production
7,450
5,270
11,400
11,500
10,521
11,272
11,306
Urea Applied as Fertilizer
3,296
4,779
6,750
6,860
6,962
7,081
7,169
Urea Imports
1,860
5,026
4,410
4,190
5,880
4,570
4,570
Urea Exports
854
536
559
111
270
1,310
1,310
Urea Consumed for Non-Agricultural Purposes
5,160
4,981
8,501
8,053
9,170
7,450
7,396
Methodological approaches were applied to the entire time series to ensure consistency in emissions
from 1990 through 2023. The methodology for urea consumption for non-agricultural purposes spliced
activity data from different sources: USGS data for 1990 through 2008, U. S. Census Bureau data for
2009 and 2010, and GHGRP data beginning in 2011. Consistent with the 2006IPCC Guidelines, the
overlap technique was applied to compare the data sets for years where there was overlap, with findings
that the data sets were consistent and adjustments were not needed.
Uncertainty
There is limited publicly available data on the quantities of urea produced and consumed for non-
agricultural purposes. Therefore, the amount of urea used for non-agricultural purposes is estimated
based on a balance that relies on estimates of urea production, urea imports, urea exports, and the
amount of urea used as fertilizer. EPA uses an uncertainty range of ±10 percent for urea production and
±5 percent for urea imports and urea exports, consistent with the ranges for activity data that are not
obtained directly from plants, and using this suggested uncertainty provided in Section 3.2.3.2 of the
4-44 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
2006IPCC Guidelines is appropriate based on expert judgment (RTI 2023). Per this expert judgment, a
normal probability density function was assigned for all activity data. The primary uncertainties
associated with this source category are associated with the accuracy of these estimates as well as the
fact that each estimate is obtained from a different data source. Because urea production estimates are
no longer available from the USGS, there is additional uncertainty associated with urea produced
beginning in 2011. There is also uncertainty associated with the assumption that all of the carbon in
urea is released into the environment as C02 during use.
The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 4-30. Carbon
dioxide emissions associated with urea consumption for non-agricultural purposes during 2023 were
estimated to be between 5.2 and 5.7 MMT C02 Eq. at the 95 percent confidence level. This indicates a
range of approximately 5 percent below and 4 percent above the emission estimate of 5.4 MMT C02 Eq.
Table 4-30: Approach 2 Quantitative Uncertainty Estimates for C02 Emissions from
Urea Consumption for Non-Agricultural Purposes (MMT C02 Eq. and Percent)
Source
Gas
2023 Emission Estimate
Uncertainty Range Relative to Emission
Estimate3
(MMT CO2 Eq.)
(MMT CO2 Eq.)
(%)
Lower
Upper
Lower
Upper
Bound
Bound
Bound
Bound
Urea Consumption
for Non-Agricultural
C02
5.4
5.2
5.7
-5%
+4%
Purposes
a Range of emission estimates predicted by Monte Carlo stochastic simulation for a 95 percent confidence interval.
QA/QC and Verification
General quality assurance/quality control (QA/QC) procedures were applied consistent with the U.S.
Inventory QA/QC plan, which is in accordance with Volume 1, Chapter 6 of 2006 IPCC Guidelines as
described in the introduction of the IPPU chapter (see Annex 8 for more details).
More details on the greenhouse gas calculation, monitoring and QA/QC methods applicable to reporting
of urea production occurring at ammonia facilities can be found under Subpart G (Ammonia
Manufacturing) of the regulation (40 CFR Part 98).26 EPA verifies annual facility-level GHGRP reports
through a multi-step process (e.g., combination of electronic checks and manual reviews) to identify
potential errors and ensure that data submitted to EPA are accurate, complete, and consistent.27 Based
on the results of the verification process, EPA follows up with facilities to resolve mistakes that may have
occurred. The post-submittals checks are consistent with a number of general and category-specific QC
procedures, including range checks, statistical checks, algorithm checks, and year-to-year checks of
reported data and emissions. EPA also conducts QA checks of GHGRP reported urea production data
against external datasets including the USGS Minerals Yearbook data. The comparison shows
consistent trends in urea production over time.
26 See http://www.ecfr.gov/cgi-bin/text-idx7tpN/ecfrbrowse/Title40/40cfr98 main 0?.tpl.
27 See https://www.epa.gov/sites/prodijction/files/9015-07/documents/ghgrp verification factsheet.pdf.
Industrial Processes and Product Use 4-45
-------
Recalculations Discussion
Based on updated quantities of urea applied for agricultural uses for 2016 through 2022, updated urea
imports from USGS for 2022, and updated urea exports from USGS for 2022, recalculations were
performed for 2016 through 2022. Compared to the previous Inventory, C02 emissions from urea
consumption for non-agricultural purposes increased by less than 1 percent for 2016 (2 kt C02), 2017 (9
kt C02), and 2018 (52 kt C02), increased by less than 2 percent for 2019 (84 kt C02), 2020 (101 kt C02),
and 2021 (124 kt C02), and decreased by 23 percent for 2022 (1,589 kt C02).
Planned Improvements
At this time, there are no specific planned improvements for estimating C02 emissions from urea
consumption for non-agricultural purposes.
4.7 Nitric Acid Production (Source Category
2B2)
Nitrous oxide (N20) is emitted during the production of nitric acid (HN03), an inorganic compound used
primarily to make synthetic commercial fertilizers. Nitric acid is also a major component in the
production of adipic acid—a feedstock for nylon—and explosives. This reporting category (2B2) includes
emissions from production of nitric acid. Emissions from fuels consumed for energy purposes during
the production of nitric acid are accounted for as part of fossil fuel combustion in the industrial end-use
sector reported under the Energy chapter.
There are two types of nitric acid: weak nitric acid and high-strength nitric acid. The weak nitric acid
production method utilizes oxidation, condensation, and absorption to produce nitric acid at
concentrations between 30 and 70 percent nitric acid. High-strength nitric acid (90 percent or greater
nitric acid) can be produced by two methods: (1) through the dehydration, bleaching, condensing, and
absorption of the weak nitric acid or (2) through the oxidation of ammonia into nitric oxide, which is
oxidized and cooled into dinitrogen tetroxide and then pressurized and oxidized into high-strength nitric
acid. Most U.S. plants were built between 1960 and 2000. As of 2023, there were 30 active nitric acid
production plants that produce weak nitric acid in the United States (EPA 2024). One plant produces
both weak and high-strength nitric acid (EPA 2010).
The basic process technology for producing nitric acid has not changed significantly over time. During
this process, N20 is formed as a byproduct and released from reactor vents into the atmosphere.
Nitric acid is made from the reaction of ammonia (NH3) with oxygen (02) in two stages. The overall
reaction is:
4NH3 + 802 -> 4HN03 + 4H2
Currently, the nitric acid industry in the United States controls emissions of NO and N02 (i.e., NOx),
using a combination of non-selective catalytic reduction (NSCR) and selective catalytic reduction (SCR)
technologies. In the process of destroying NOx, NSCR systems are also very effective at destroying N20.
Five nitric acid plants had NSCR systems installed between 1964 and 1977, over half due to the
4-46 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
finalization of the Nitric Acid Plant New Source Performance Standards (NSPS) which went into effect in
1971. Four additional nitric acid plants had NSCR systems installed between 2016 and 2018, as a result
of EPA Consent Decrees to control NOx emissions more effectively. NSCR systems are used in
approximately one-third of the weak acid production plants. For N20 abatement, U.S. facilities are using
both tertiary (i.e., NSCR and SCR) and secondary controls (i.e., catalysts added to the ammonia reactor
to lessen potential N20 production).
Emissions from the production of nitric acid are calculated as the product of the total annual production
and plant-specific emission factors. Generally, an increase/decrease in the annual amount of nitric acid
produced from year to year leads to an increase/decrease in the N20 emissions from year to year, with
some exceptions. For example, in 2015 and 2019, nitric acid production decreased and emissions
increased compared to the respective preceding years; in 2016, nitric acid production increased and
emissions decreased compared to 2015. N20 emissions for those years are calculated based on data
from the GHGRP as discussed in the Methodology section below. The data from plants reporting to
GHGRP indicate that plant-specific operations can affect the emission rate or factor, including: (1) site-
specific fluctuations in ambient temperature and humidity, (2) catalyst age and condition, (3) process
changes, such as fluctuations in process pressure or temperature and replacing the ammonia catalyst,
(4) the addition, removal, maintenance, and utilization of abatement technologies, and (5) the number
of nitric acid trains, which are reaction vessels where ammonia is oxidized to form nitric acid. Changes
in those operating conditions for the years in question (2015, 2016, and 2019) caused changes in the
emission rate or factor used, which resulted in the exceptions noted above.
Nitrous oxide emissions from this source were estimated to be 8.3 MMT C02 Eq. (32 kt of N20) in 2023
and are summarized in Table 4-31 and Table 4-32. Emissions from nitric acid production have decreased
by 23 percent since 1990, while production has increased by 8.6 percent over the same time period (see
Table 4-33). Emissions have decreased by 35 percent since 1997, the highest year of production in the
time series. From 2022 to 2023, nitric acid production decreased by 1 percent, while overall emissions
from nitric acid production decreased by 3.1 percent from 2022 to 2023.
Table 4-31: N20 Emissions from Nitric Acid Production (MMT C02 Eq.)
Year 1990
2005
2019 2020
2021
2022
2023
Nitric Acid Production 10.8
10.1
CO
CD
CO
CO
7.9
8.6
8.3
Table 4-32: N20 Emissions from Nitric Acid Production (kt N20)
Year 1990
2005
2019 2020
2021
2022
2023
Nitric Acid Production 41
38
34 31
30
33
32
Methodology and Time-Series Consistency
Emissions of N20 from nitric acid production are estimated using methods provided by the 2006IPCC
Guidelines, in accordance with the IPCC methodological decision tree and available data. For 2010
through 2023, a Tier 3 method was used to estimate emissions based on GHGRP data. For 1990 through
2009, a Tier 2 method was used to estimate emissions from nitric acid production based on U.S. Census
Bureau data.
Industrial Processes and Product Use 4-47
-------
2010 through 2023
Process N20 emissions and nitric acid production data were obtained directly from EPA's GHGRP for
2010 through 2023 by aggregating reported facility-level data (EPA 2018; EPA 2024).28
Since 2010, in the United States, all nitric acid facilities that produce weak nitric acid (30 to 70 percent)
have been required to report annual greenhouse gas emissions data to EPA as per the requirements of
the GHGRP (Subpart V). Beginning with 2018, the rule was changed to include facilities that produce
nitric acid of any strength. The only facility that produces high-strength nitric acid also produces weak
nitric acid. All N20 emissions from nitric acid production originate from the production of weak nitric
acid.
Process emissions and nitric acid production reported to the GHGRP provide complete estimates of
greenhouse gas emissions for the United States because there are no reporting thresholds. While
facilities are allowed to stop reporting to the GHGRP if the total reported emissions from nitric acid
production are less than 25,000 metric tons C02 Eq. per year for five consecutive years or less than
15,000 metric tons C02 Eq. per year for three consecutive years, no facilities have stopped reporting as
a result of these provisions.29 All nitric acid facilities are required to either calculate process N20
emissions using a site-specific emission factor that is the average of the emission factor determined
through annual performance tests for each nitric acid train under typical operating conditions or directly
measure process N20 emissions using monitoring equipment.30
Emissions from facilities vary from year to year, depending on the amount of nitric acid produced with
and without abatement technologies and other conditions affecting the site-specific emission factor. To
maintain consistency across the time series and with the rounding approaches taken by other data sets,
GHGRP nitric acid data are rounded and are shown in Table 4-33.
1990 through 2009
Using GHGRP data for 2010, country-specific N20 emission factors were calculated for nitric acid
production with abatement and without abatement (i.e., controlled and uncontrolled emission factors).
The following 2010 emission factors were derived for production with abatement and without
abatement: 3.3 kg N20/metric ton HN03 produced at plants using abatement technologies (e.g., tertiary
systems such as NSCR systems) and 5.99 kg N20/metric ton HN03 produced at plants not equipped
with abatement technology. Country-specific weighted emission factors were derived by weighting
these emission factors by percent production with abatement and without abatement over time periods
1990 through 2008 and 2009. These weighted emission factors were used to estimate N20 emissions
from nitric acid production for years prior to the availability of GHGRP data (i.e., 1990 through 2008 and
2009). A separate weighted emission factor is included for 2009 due to data availability for that year.
EPA verified the installation dates of N20 abatement technologies for all facilities based on GHGRP
facility-level information and confirmed that all abatement technologies were accounted for in the
28 National N2O process emissions, national production, and national share of nitric acid production with abatement and
without abatement technology were aggregated from the GHGRP facility-level data for 2010 to 2023 (i.e., percent
production with and without abatement).
29 See 40 CFR 98.2(i)(1) and 40 CFR 98.2(i)(2) for more information about these provisions.
30 Facilities must use standard methods - either EPA Method 320 or ASTM D6348-03 for annual performance tests—and
must follow associated QA/QC procedures consistent with category-specific QC of direct emission measurements
during these performance tests.
4-48 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
derived emission factors (Icenhour 2020). Due to the lack of information on abatement equipment
utilization, it is assumed that once abatement technology was installed in facilities, the equipment was
consistently operational for the duration of the time series considered in this report (especially NSCRs).
The country-specific weighted N20 emission factors were used in conjunction with annual production to
estimate N20 emissions for 1990 through 2009, using the following equations:
Equation 4-4:2006IPCC Guidelines Tier 3: N20 Emissions From Nitric Acid Production
(Equation 3.6)
Ei ~ Pi x EFWeighted,i
EFweighted,i = X EFc) + (%PUnc,i X EFunc)\
where,
Ei = Annual N20 Emissions for year i (kg/yr)
Pi = Annual nitric acid production foryear i (metric tons HN03)
EFweighted,i= Weighted N20 emission factor for year i (kg N20/metric ton HN03)
%Pc,i = Percent national production of HN03 with N20 abatement technology (%)
EFC = N20 emission factor, with abatement technology (kg N20/metric ton HN03)
%Puno,i = Percent national production of HN03 without N20 abatement technology (%)
EFunc = N20 emission factor, without abatement technology (kg N20/metric ton HN03)
i = year from 1990 through 2009
• For 2009: Weighted N20 emission factor = 5.46 kg N20/metric ton HN03.
• For 1990 through 2008: Weighted N20 emission factor = 5.66 kg N20/metric ton HN03.
Nitric acid production data for the United States for 1990 through 2009 were obtained from the U.S.
Census Bureau (U.S. Census Bureau 2008, 2009, 2010a, 2010b) (see Table 4-33). EPA used GHGRP
facility-level information to verify that all reported N20 abatement equipment were incorporated into the
estimation of N20 emissions from nitric acid production over the full time series (EPA 2024).
Table 4-33: Nitric Acid Production (kt)
Year
1990
2005
2019
2020
2021
2022
2023
Production
7,200
6,710
8,080
7,970
7,800
7,860
7,811
Methodological approaches were applied to the entire time series to ensure consistency in emissions
from 1990 through 2023. The methodology for nitric acid production spliced activity data from two
different sources: U.S. Census Bureau production data for 1990 through 2009 and GHGRP production
data starting in 2010. Consistent with the 2006 IPCC Guidelines, the overlap technique was applied to
compare the two data sets for years where there was overlap, with findings that the data sets were
consistent and adjustments were not needed.
Uncertainty
Uncertainty associated with the parameters used to estimate N20 emissions including, the share of U.S.
nitric acid production attributable to each emission abatement technology (i.e., utilization) over the time
series (especially prior to 2010), and the associated emission factors applied to each abatement
technology type. While some information has been obtained through outreach with industry
Industrial Processes and Product Use 4-49
-------
associations, limited information is available over the time series (especially prior to 2010) for a variety
of facility level variables, including plant-specific production levels, plant production technology (e.g.,
low or high pressure, etc.), and abatement technology destruction and removal efficiency rates.
Production data prior to 2010 were obtained from National Census Bureau, which does not provide
uncertainty estimates with their data. Facilities reporting to EPA's GHGRP must measure production
using equipment and practices used for accounting purposes. While emissions are often directly
proportional to production, the emission factor for individual facilities can vary significantly from year to
year due to site-specific fluctuations in ambient temperature and humidity, catalyst age and condition,
nitric acid production process changes, the addition or removal of abatement technologies, and the
number of nitric acid trains at the facility. At this time, EPA does not estimate uncertainty of the
aggregated facility-level information. As noted in the QA/QC and verification section below, EPA verifies
annual facility-level reports through a multi-step process (e.g., combination of electronic checks and
manual reviews by staff) to identify potential errors and ensure that data submitted to EPA are accurate,
complete, and consistent. The annual production reported by each nitric acid facility under EPA's
GHGRP and then aggregated to estimate national N20 emissions is assumed to have low uncertainty.
EPA assigned an uncertainty range of ±6 percent for facility-reported N20 emissions and EPA assigned
an uncertainty range of ±4 percent for nitric acid production based on expert judgment (RTI 2025). Per
this expert judgment, a normal probability density function was assigned for facility-reported N20
emissions and nitric acid production.
The results of this Approach 2 quantitative uncertainty analysis are summarized in Table 4-34. Nitrous
oxide emissions from nitric acid production were estimated to be between 7.8 and 8.9 MMT C02 Eq. at
the 95 percent confidence level. This indicates a range of approximately 6 percent below to 6 percent
above the 2023 emissions estimate of 8.3 MMT C02 Eq.
Table 4-34: Approach 2 Quantitative Uncertainty Estimates for N20 Emissions from
Nitric Acid Production (MMT C02 Eq. and Percent)
2023 Emission
Source
Gas
Estimate
Uncertainty Range Relative to Emission Estimate"
(MMT CO2 Eq.)
(MMT CO2 Eq.) (%)
Lower Upper Lower Upper
Bound Bound Bound Bound
Nitric Acid Production
n2o
8.3
7.8 8.9 -6% +6%
a Range of emission estimates predicted by Monte Carlo stochastic simulation for a 95 percent confidence interval.
QA/QC and Verification
General quality assurance/quality control (QA/QC) procedures were applied consistent with the U.S.
Inventory QA/QC plan, which is in accordance with Volume 1, Chapter 6 of the 2006IPCC Guidelines as
described in the introduction of the IPPU chapter (see Annex 8 for more details). More details on the
greenhouse gas calculation, monitoring and QA/QC methods applicable to nitric acid facilities can be
found under Subpart V: Nitric Acid Production of the GHGRP regulation (40 CFR Part 98).31
The main QA/QC activities are related to annual performance testing, which must follow either EPA
Method 320 or ASTM D6348-03. EPA verifies annual facility-level GHGRP reports through a multi-step
31 See Subpart V monitoring and reporting regulation http://www.e cf r. go v/c g i - b i n /text-
idx?tpl=/ecfrbrowse/Title40/40cfr98 main 0?.tpl.
4-50 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
process that is tailored to the Subpart (e.g., combination of electronic checks including range checks,
statistical checks, algorithm checks, year-to-year comparison checks, along with manual reviews) to
identify potential errors and ensure that data submitted to EPA are accurate, complete, and consistent.
Based on the results of the verification process, EPA follows up with facilities to resolve mistakes that
may have occurred (EPA 20 1 5).32 EPA's review of observed trends noted that while emissions have
generally mirrored production, in 2015 and 2019 nitric acid production decreased compared to the
previous year and emissions increased. While review is ongoing, based on feedback from the
verification process to date, these changes are due to facility-specific changes (e.g., in the nitric
production process and management of abatement equipment).
Recalculations Discussion
No recalculations were performed for the 1990 through 2022 portion of the time series.
Planned Improvements
Pending resources, EPA is considering a near-term improvement to both review and refine quantitative
uncertainty estimates and the associated qualitative discussion.
4.8 Adipic Acid Production (Source Category
2B3)
Adipic acid is a white crystalline solid used in the manufacture of synthetic fibers, plastics, coatings,
urethane foams, elastomers, and synthetic lubricants. This reporting category (2B3) includes emissions
from the production of adipic acid. Emissions from fuels consumed for energy purposes during the
production of adipic acid are accounted for as part of fossil fuel combustion in the industrial end-use
sector reported under the Energy chapter.
Adipic acid is produced through a two-stage process during which nitrous oxide (N20) is generated in the
second stage. The first stage of manufacturing usually involves the oxidation of cyclohexane to form a
cyclohexanone/cyclohexanol mixture. The second stage involves oxidizing this mixture with nitric acid to
produce adipic acid. Nitrous oxide is generated as a byproduct of the nitric acid oxidation stage and is
emitted in the waste gas stream (Thiemens and Trogler 1991). The second stage is represented by the
following chemical reaction:
(CH2)5CO(cyclohexanone) + (C H2)5CHOH (cyclohexanol) + wHN03
-> H00C(CH2)4C00H(adipic acid) + xN20 +yH20
Process emissions from the production of adipic acid vary with the types of technologies and level of
emission controls employed by a facility. In 1990, two major adipic acid-producing plants had N20
abatement technologies in place and, as of 1998, three major adipic acid production facilities had
32 See GHGRP Verification Factsheet https://www.epa.gov/sites/production/files/7015-
07/rtocuments/ghgrp verification factsheet.pdf.
Industrial Processes and Product Use 4-51
-------
control systems in place (Reimer et al. 1999). In 2023, thermal reduction was applied as an N20
abatement measure at one adipic acid facility (EPA 2024).
Worldwide, only a few adipic acid plants exist. The United States, Europe, and China are the major
producers, with the United States accounting for the largest share of global adipic acid production
capacity in recent years. In 2023, the United States had two companies with a total of two adipic acid
production facilities (one in Texas and one in Florida), following the ceased operations of a third major
production facility at the end of 2015 (EPA 2024).
Commercially, adipic acid is the most important of the aliphatic dicarboxylic acids, which are used to
manufacture polyesters. Eighty-four percent of all adipic acid produced in the United States is used in
the production of nylon 6,6; 9 percent is used in the production of polyester polyols; 4 percent is used in
the production of plasticizers; and the remaining 4 percent is accounted for by other uses, including
unsaturated polyester resins and food applications (ICIS 2007). Food grade adipic acid is used to
provide some foods with a "tangy" flavor (Thiemens and Trogler 1991).
Compared to 1990, national adipic acid production in 2023 has increased by 6 percent to approximately
800,000 metric tons (ACC 2024). Nitrous oxide emissions from adipic acid production were estimated to
be 1.2 MMT C02 Eq. (4 kt N20) in 2023 and are summarized in Table 4-35 and Table 4-36. Over the period
1990 through 2023, facilities have reduced emissions by 91.5 percent due to the widespread installation
of pollution control measures in the late 1990s. The main reason for the 45 percent decrease in N20
emissions from adipic acid production between 2022 and 2023 is increased utilization of N20
abatement equipment at one adipic acid production facility.
EPA reviewed GHGRP facility reported information on the date of abatement technology installation in
order to better reflect trends and changes in emissions abatement within the industry across the time
series. The facility using the facility-specific emission factor developed through annual performance
testing has reported no installation and no utilization of N20 abatement technology. The facility using
direct measurement of N20 emissions has reported the use of two thermal reduction units as N20
abatement technologies; the first unit began operation in 1980, and the second unit began operation in
2023 (Ard 2024; Ascend 2023).
Significant changes in the amount of time that the N20 abatement device at one facility was in operation
has been the main cause of fluctuating emissions in recent years. These fluctuations are most evident
for years where trends in emissions and adipic acid production were not directly proportional: (1)
between 2016 and 2017, (2) between 2017 and 2018, (3) between 2019 and 2020, (4) between 2020 and
2021, and (5) between 2021 and 2022. As noted above, changes in control measures and abatement
technologies at adipic acid production facilities, including maintenance of equipment, can result in
annual emission fluctuations. Little additional information is available on drivers of trends, and the
amount of adipic acid produced is not reported under EPA's GHGRP.
Table 4-35: N20 Emissions from Adipic Acid Production (MMT C02 Eq.)
Year 1990
2005
2019 2020
2021
2022
2023
Adipic Acid Production 13.5
6.3
4.7 7.4
6.6
2.1
1.2
Table 4-36: N20 Emissions from Adipic Acid Production (kt N20)
Year 1990
2005
2019 2020
2021
2022
2023
Adipic Acid Production 51
24
18 28
25
8
4
4-52 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Methodology and Time-Series Consistency
Emissions of N20 from adipic acid production are estimated using methods provided by the 2006IPCC
Guidelines, in accordance with the IPCC methodological decision tree and available data. For 2010
through 2023, a Tier 3 method was used to estimate emissions. For 1990 through 2009, emissions are
estimated using both Tier 2 and Tier 3 methods. Due to confidential business information (CBI), plant
names are not provided in this section; therefore, the four adipic acid-producing facilities that have
operated over the time series will be referred to as Plants 1 through 4. As noted above, one currently
operating facility uses thermal reduction as an N20 abatement technology.
2010 through 2023
All emission estimates for 2010 through 2023 were obtained through analysis of GHGRP data (EPA 2010
through 2024). Facility-level greenhouse gas emissions data were obtained from EPA's GHGRP for the
years 2010 through 2023 (EPA 2010 through 2024) and aggregated to national N20 emissions.
Consistent with IPCC Tier 3 methods, all adipic acid production facilities are required to either calculate
N20 emissions using a facility-specific emission factor developed through annual performance testing
under typical operating conditions or directly measure N20 emissions using monitoring equipment.33
1990 through 2009
For years 1990 through 2009, which were prior to EPA's GHGRP reporting, for both Plants 1 and 2,
emission estimates were obtained directly from the plant engineers and account for reductions due to
control systems in place at these plants during the time series. These prior estimates are considered
CBI and hence are not published (Desai 2010, 2011). These estimates were based on continuous
process monitoring equipment installed at the two facilities.
For Plant 4, 1990 through 2009 N20 emissions were estimated using the following Tier 2 equation from
the 2006 IPCC Guidelines:
Equation 4-5:2006 IPCC Guidelines Tier 2: N20 Emissions From Adipic Acid Production
(Equation 3.8)
Eaa = Qaa X EFaa X (1 ~ \DF X UF])
where,
Eaa
N20 emissions from adipic acid production, metric tons
Qaa =
Quantity of adipic acid produced, metric tons
EFaa =
Emission factor, metric ton N20/metric ton adipic acid produced
DF
N20 destruction factor
UF
Abatement system utility factor
The adipic acid production is multiplied by an emission factor (i.e., N20 emitted per unit of adipic acid
produced), which has been estimated to be approximately 0.3 metric tons of N20 per metric ton of
product (IPCC 2006). The "N20 destruction factor" in the equation represents the percentage of N20
emissions that are destroyed by the installed abatement technology. The "abatement system utility
33 Facilities must use standard methods, either EPA Method 320 or ASTM D6348-03 for annual performance testing, and
must follow associated QA/QC procedures during these performance tests consistent with category-specific QC of
direct emission measurements.
Industrial Processes and Product Use 4-53
-------
factor" represents the percentage of time that the abatement equipment operates during the annual
production period. Plant-specific production data for Plant 4 were obtained across the time series
through personal communications (Desai 2010, 2011). The plant-specific production data were then
used for calculating emissions as described above.
For Plant 3, 2005 through 2009 emissions were obtained directly from the plant (Desai 2010, 2011). For
1990 through 2004, emissions were estimated using plant-specific production data and the IPCC
factors as described above for Plant 4. Plant-level adipic acid production for 1990 through 2003 was
estimated by allocating national adipic acid production data to the plant level using the ratio of known
plant capacity to total national capacity for all U.S. plants (ACC 2023; CMR 2001, 1998; CW 1999; C&EN
1992 through 1995). For 2004, actual plant production data were obtained and used for emission
calculations (CW 2005).
Plant capacities for 1990 through 1994 were obtained from Chemical & Engineering News, "Facts and
Figures" and "Production of Top 50 Chemicals" (C&EN 1992 through 1995). Plant capacities for 1995 and
1996 were kept the same as 1994 data. The 1997 plant capacities were taken from Chemical Market
Reporter, "Chemical Profile: Adipic Acid" (CMR 1998). The 1998 plant capacities for all four plants and
1999 plant capacities for three of the plants were obtained from Chemical Week, Product Focus: Adipic
Acid/Adiponitrile (CW 1999). Plant capacities for the year 2000 for three of the plants were updated
using Chemical Market Reporter, "Chemical Profile: Adipic Acid" (CMR 2001). For 2001 through 2003,
the plant capacities for three plants were held constant at year 2000 capacities. Plant capacity for 1999
to 2003 for the one remaining plant was kept the same as 1998.
National adipic acid production data (see Table 4-37) from 1990 through 2023 were obtained from the
American Chemistry Council (ACC 2024).
Table 4-37: Adipic Acid Production (kt)
Year
1990
2005
2019
2020
2021
2022
2023
Production
755
865
810
710
760
760
800
Methodological approaches were applied to the entire time series to ensure consistency in emissions
from 1990 through 2023. The methodology for adipic acid production spliced activity data from multiple
sources: plant-specific emissions data and publicly available plant capacity data for 1990 through 2009
and GHGRP emission data starting in 2010. Consistent with the 2006 IPCC Guidelines, the overlap
technique was applied to compare the two data sets for years where there was overlap, with findings
that the data sets were consistent and adjustments were not needed.
Uncertainty
Uncertainty associated with N20 emission estimates includes the methods used by companies to
monitor and estimate emissions. While some information has been obtained through outreach with
facilities, limited information is available over the time series on these methods, abatement technology
destruction and removal efficiency rates, and plant-specific production levels. EPA assigned an
uncertainty range of ±5 percent and a normal probability density function for facility-reported N20
emissions, and using this suggested uncertainty provided in section 3.4.3.2 of the 2006 IPCC Guidelines
is appropriate based on expert judgment (RTI 2023).
4-54 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
The results of this Approach 2 quantitative uncertainty analysis are summarized in Table 4-38. Nitrous
oxide emissions from adipic acid production for 2023 were estimated to be between 1.1 and 1.2 MMT
C02 Eq. at the 95 percent confidence level. These values indicate a range of approximately 4 percent
below to 4 percent above the 2023 emission estimate of 1.2 MMT C02 Eq.
Table 4-38: Approach 2 Quantitative Uncertainty Estimates for N20 Emissions from
Adipic Acid Production (MMT C02 Eq. and Percent)
2023 Emission
Uncertainty Range Relative to Emission
Source
Gas
Estimate
Estimate3
(MMTCO2 Eq.)
(MMT CO2 Eq.)
(%)
Lower Upper
Lower Upper
Bound Bound
Bound Bound
Adipic Acid Production
n2o
1.2
1.1 1.2
-4% +4%
a Range of emission estimates predicted by Monte Carlo stochastic simulation for a 95 percent confidence interval.
QA/QC and Verification
General quality assurance/quality control (QA/QC) procedures were applied consistent with the U.S.
Inventory QA/QC plan, which is in accordance with Volume 1, Chapter 6 of the 2006IPCC Guidelines as
described in the introduction of the IPPU chapter (see Annex 8 for more details).
More details on the greenhouse gas calculation, monitoring and QA/QC methods applicable to adipic
acid facilities can be found under Subpart E (Adipic Acid Production) of the GHGRP regulation (40 CFR
Part 98).34 The main QA/QC activities are related to annual performance testing, which must follow
either EPA Method 320 or ASTM D6348-03. EPA verifies annual facility-level GHGRP reports through a
multi-step process (e.g., combination of electronic checks and manual reviews) to identify potential
errors and ensure that data submitted to EPA are accurate, complete, and consistent (EPA 2015).35
Based on the results of the verification process, EPA follows up with facilities to resolve mistakes that
may have occurred. The post-submittals checks are consistent with a number of general and category-
specific QC procedures, including range checks, statistical checks, algorithm checks, and year-to-year
comparisons of reported data.
Recalculations Discussion
No recalculations were performed for the 1990 through 2022 portion of the time series. While not used
in emissions calculations, the 2022 value for adipic acid production was updated and included for
informational purposes (ACC 2024) in Table 4-37 above.
Planned Improvements
Pending resources, EPA is considering a near-term improvement to both review and refine quantitative
uncertainty estimates and the associated qualitative discussion.
34 See http://www.ecfr.gov/cgi-bin/text-idx7tpN/ecfrbrowse/Title40/40cfr98 main 0?.tpl.
35 See https://www.epa.gov/sites/prodijction/files/9015-07/documents/ghgrp verification factsheet.pdf.
Industrial Processes and Product Use 4-55
-------
4.9 Caprolactam, Glyoxal and Glyoxylic Acid
Production (Source Category 2B4)
This reporting category (2B4) includes emissions from the production of caprolactam, glyoxal
(ethanedial), and glyoxylic acid. Emissions from fuels consumed for energy purposes during the
production of caprolactam, glyoxal, and glyoxylic acid are accounted for as part of fossil fuel
combustion in the industrial end-use sector reported under the Energy chapter.
Caprolactam
Caprolactam (C6HnNO) is a colorless monomer produced for nylon-6 fibers and plastics. A substantial
proportion of the fiber is used in carpet manufacturing. Most commercial processes used for the
manufacture of caprolactam begin with benzene, but toluene can also be used. The production of
caprolactam can give rise to emissions of nitrous oxide (N20).
During the production of caprolactam, emissions of N20 can occur from the ammonia oxidation step,
emissions of carbon dioxide (C02) from the ammonium carbonate step, emissions of sulfur dioxide
(S02) from the ammonium bisulfite step, and emissions of non-methane volatile organic compounds
(NMVOCs). Emissions of C02, S02 and NMVOCs from the conventional process are unlikely to be
significant in well-managed plants. Modified caprolactam production processes are primarily
concerned with elimination of the high volumes of ammonium sulfate that are produced as a byproduct
of the conventional process (IPCC 2006).
In the most commonly used process where caprolactam is produced from benzene, benzene is
hydrogenated to cyclohexane which is then oxidized to produce cyclohexanone (C6Hi0O). The classical
route (Raschig process) and basic reaction equations for production of caprolactam from
cyclohexanone are (IPCC 2006):
NO
Oxidation of NH-, to
1 3 N02
i
co2
NH3 reacted with——to yield ammonium carbonate (NH4)2C03
H20
i
NO
(NH4)2C03 reacted with (from NH3 oxidation)to yield ammonium nitrite (NH4N02)
N02
i
so2
NH3 reacted with——to yield ammonium bisulphite (NH4HS03)
H20
i
NH4N02 and (NH4HS03)reacted to yield hydroxylamine disulphonate (N0H(S03NH4)2)
I
4-56 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
(N0H(S03NH4)2) hydrolised to yield hydroxylamine sulphate ((NH2OH)2¦ H2S04) and
ammonium sulphate ((NH4)2S04)
I
Cylohexanone reaction:
1
C6HwO + ~(NH20H)2.H2S04(+NH3 and H2S04) -> C6H10NOH + (NH4)2S04 + H20
I
Beckmann rearrangement:
C6H10NOH (+H2S04 and S02) -> C6HltNO. H2S04 (+4JVH3 and Hz0) -> C6HltNO + 2(NH4-)2S04
In 2004, three facilities produced caprolactam in the United States (ICIS 2004). Another facility,
Evergreen Recycling, was in operation from 2000 to 2001 (ICIS 2004; Textile World 2000) and from 2007
through 2015 (Shaw 2015). Caprolactam production at Fibrant LLC (formerly DSM Chemicals) in Georgia
ceased in 2018 (Cline 2019). As of 2023, two companies in the United States produced caprolactam at
two facilities: AdvanSix (formerly Honeywell) in Virginia (AdvanSix 2024) and BASF in Texas (BASF 2024).
Nitrous oxide emissions from caprolactam production in the United States were estimated to be 1.3
MMT C02 Eq. (5 kt N20) in 2023 and are summarized in Table 4-39 and Table 4-40. National emissions
from caprolactam production decreased by approximately 10.5 percent over the period of 1990 through
2023. Emissions in 2023 are identical to 2022 emissions. The values in 2022 and 2023 indicate that
caprolactam production is consistent with 2017 levels, prior to the COVID-19 pandemic, but still below
annual average production from 1990-2016.
Table 4-39: N20 Emissions from Caprolactam Production (MMT C02 Eq.)
Year 1990
2005
2019 2020 2021
2022
2023
Caprolactam Production 1.5
1.9
1.2 1.1 1.2
1.3
1.3
Table 4-40: N2Q Emissions from Caprolactam Production (kt N20)
Year 1990
2005
2019 2020 2021
2022
2023
Caprolactam Production 6
7
5 4 5
5
5
Glyoxal and Glyoxylic Acid
Glyoxal (ethanedial) (C2H202) is mainly used as a crosslinking agent for vinyl acetate/acrylic resins,
disinfectant, gelatin hardening agent, textile finishing agent (permanent-press cotton, rayon fabrics),
and wet-resistance additive (paper coatings) (IPCC 2006). It is also used for enhanced oil-recovery. It is
produced from oxidation of acetaldehyde (ethanal) (C2H40) with concentrated nitric acid (HN03).
Glyoxal can also be produced from catalytic oxidation of ethylene glycol (ethanediol) (CH2OHCH2OH).
Glyoxylic acid (C2H203) is produced by nitric acid oxidation of glyoxal. Glyoxylic acid is used for the
production of synthetic aromas, agrochemicals, and pharmaceutical intermediates (IPCC 2006).
Preliminary data suggests that glyoxal and glyoxylic acid may be produced in small quantities
domestically but are largely imported to the United States. EPA does not currently estimate the
Industrial Processes and Product Use 4-57
-------
emissions associated with the production ofglyoxaland glyoxylic acid because activity data are not
available. See planned improvements below and Annex 5 of this report for more information.
Methodology and Time-Series Consistency
Emissions of N20 from the production of caprolactam are calculated using the Tier 1 methodology from
the 2006IPCC Guidelines, in accordance with the IPCC methodological decision tree and available
data. The Tier 1 equation is as follows:
Equation 4-6:2006 IPCC Guidelines Tier 1: N20 Emissions From Caprolactam
Production (Equation 3.9)
En2o = EF X CP
where,
EN2o = Annual N20 Emissions (kg)
EF = N20 emission factor (default) (kg N20/metric ton caprolactam produced)
CP = Caprolactam production (metric tons)
During the caprolactam production process, N20 is generated as a byproduct of the high temperature
catalytic oxidation of ammonia (NH3), which is the first reaction in the series of reactions to produce
caprolactam. The amount of N20 emissions can be estimated based on the chemical reaction shown
above. Based on this formula, which is consistent with an IPCC Tier 1 approach, approximately 111.1
metric tons of caprolactam are required to generate one metric ton of N20, resulting in an emission
factor of 9.0 kg N20 per metric ton of caprolactam (IPCC 2006). When applying the Tier 1 method, the
2006 IPCC Guidelines state that it is good practice to assume that there is no abatement of N20
emissions and to use the highest default emission factor available in the guidelines. In addition, EPA did
not find support for the use of secondary catalysts to reduce N20 emissions, such as those employed at
nitric acid plants.
The activity data for caprolactam production (see Table 4-41) from 1990 to 2023 were obtained from the
American Chemistry Council's Guide to the Business of Chemistry (ACC 2024). EPA will continue to
analyze and assess alternative sources of production data as a quality control measure.
Table 4-41: Caprolactam Production (kt)
Year
1990
2005
2019
2020
2021
2022
2023
Production (kt)
626
795
515
480
510
560
560
Carbon dioxide and methane (CH4) emissions may also occur from the production of caprolactam, but
currently the IPCC does not have methodologies for calculating these emissions associated with
caprolactam production (EPA 2023).
Methodological approaches, consistent with the 2006 IPCC Guidelines, have been applied to the entire
time series to ensure consistency in emissions from 1990 through 2023.
Uncertainty
Estimation of emissions of N20 from caprolactam production can be treated as analogous to estimation
of emissions of N20 from nitric acid production. Both production processes involve an initial step of NH3
4-58 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
oxidation, which is the source of N20 formation and emissions (IPCC 2006). Therefore, uncertainties for
the default emission factor values in the 2006 IPCC Guidelines are an estimate based on default values
for nitric acid plants. In general, default emission factors for gaseous substances have higher
uncertainties because mass values for gaseous substances are influenced by temperature and pressure
variations and gases are more easily lost through process leaks. The default values for caprolactam
production have a relatively high level of uncertainty due to the limited information available (IPCC
2006). EPA assigned uncertainty bounds of ±5 percent for caprolactam production, based on expert
judgment. EPA assigned an uncertainty range of ±40 percent for the N20 emission factor, and using this
suggested uncertainty provided in Section 3.5.2.1 of the 2006 IPCC Guidelines is appropriate based on
expert judgment (RTI 2023). Per this expert judgment, a normal probability density function was
assigned for activity data, and a triangular probably density function was assigned for the emission
factor.
The results of this Approach 2 quantitative uncertainty analysis are summarized in Table 4-42. Nitrous
oxide emissions from caprolactam, glyoxal and glyoxylic acid production for 2023 were estimated to be
between 0.9 and 1.8 MMT C02 Eq. at the 95 percent confidence level. These values indicate a range of
approximately 31 percent below to 31 percent above the 2023 emission estimate of 1.3 MMT C02 Eq.
Table 4-42: Approach 2 Quantitative Uncertainty Estimates for N20 Emissions from
Caprolactam, Glyoxal and Glyoxylic Acid Production (MMT C02 Eq. and Percent)
2023 Emission
Uncertainty Range Relative to Emission
Source
Gas
Estimate
Estimate3
(MMT CO2 Eq.)
(MMT CO2 Eq.)
(%)
Lower Upper
Lower Upper
Bound Bound
Bound Bound
Caprolactam Production
n2o
1.3
0.9 1.8
-31% +31%
a Range of emission estimates predicted by Monte Carlo stochastic simulation for a 95 percent confidence interval.
QA/QC and Verification
General quality assurance/quality control (QA/QC) procedures were applied consistent with the U.S.
Inventory QA/QC plan, which is in accordance with Volume 1, Chapter 6 of the 2006 IPCC Guidelines as
described in the introduction of the IPPU chapter (see Annex 8 for more details).
Recalculations Discussion
No recalculations were performed for the 1990 through 2022 portion of the time series.
Planned Improvements
EPA's GHGRP has been amended to include reporting from these industries and annual reporting will
begin in 2025 if production is occurring. Data would be publicly available in early fall 2026. Preliminary
data suggests that glyoxal and glyoxylic acid may be produced in small quantities domestically but are
largely imported to the United States (EPA 2023). To elaborate, it is also possible that there are other
facilities in the U.S. that do not have to report under TSCA because their total production volume is less
than 25,000 pounds per year or they are exempt from reporting because they are a small manufacturer
based on their total company sales revenue. See Annex 5 of this report for more information. This
Industrial Processes and Product Use 4-59
-------
planned improvement is subject to data availability and will be implemented in the medium- to long-
term.
4.10 Carbide Production and Consumption
(Source Category 2B5 & 2B10)
Carbon dioxide (C02) and methane (CH4) are emitted from the production of silicon carbide (SiC), a
material used for industrial abrasive, metallurgical, and other non-abrasive applications in the United
States, and C02 is emitted from the consumption of SiC. Per the IPCC methodological guidance,
emissions from fuels consumed for energy purposes during the production of silicon carbide are
accounted for in the industrial end-use sector reported under the Energy chapter. Additionally, some
metallurgical and non-abrasive applications of SiC are emissive at high temperatures due to the SiC
oxidation temperature (Biscay 2021). While emissions should be accounted for where they occur based
on 2006 IPCC Guidelines, emissions from SiC consumption are accounted for here until additional data
on SiC consumption by end-use are available. The reporting category (2B5) includes emissions from the
production of SiC, and the reporting category (2B10) includes emissions from the consumption of SiC.
To produce SiC, silica sand or quartz (Si02) is reacted with carbon (C) in the form of petroleum coke. A
portion (about 35 percent) of the carbon contained in the petroleum coke is retained in the SiC. The
remaining carbon is emitted as C02, CH4, or carbon monoxide (CO). The overall reaction is shown
below, but in practice, it does not proceed according to stoichiometry:
Si02 "I- 3C SiC + 2C0 (+ 02 2.CO2)
Carbon dioxide and CH4 are also emitted during the production of calcium carbide, a chemical used to
produce acetylene. Carbon dioxide is implicitly accounted for in the storage factor calculation for the
non-energy use of petroleum coke in the Energy chapter, using a country-specific approach given
calcium carbide production data.36
Markets for manufactured abrasives, including SiC, are heavily influenced by activity in the U.S.
manufacturing sector, especially in the aerospace, automotive, furniture, housing, and steel
manufacturing sectors. Specific applications of abrasive-grade SiC in 2018 included antislip abrasives,
blasting abrasives, bonded abrasives, coated abrasives, polishing and buffing compounds, tumbling
media, and wire-sawing abrasives (USGS 2021). Approximately 50 percent of SiC is used in metallurgical
applications, which include primarily iron and steel production, and other non-abrasive applications,
which include use in advanced or technical ceramics and refractories (USGS 2023a; Washington Mills
2023).
As a result of the economic downturn in 2008 and 2009, demand for SiC decreased in those years. Low-
cost imports, particularly from China, combined with high relative operating costs for domestic
36 The United States applies a country-specific approach for estimating CO2 emissions from production of calcium carbide
because currently there is no way to disaggregate and report emissions specifically associated with petroleum coke
used in calcium carbide production (as is done for silicon carbide) since production data are not available for calcium
carbide. Table A-42 in Annex 2 indicates a storage factor of 30 percent for petroleum coke used in non-energy uses. This
indicates effectively that 70 percent of any CO2 emissions associated with petroleum coke used in calcium carbide
production is released and accounted for under NEU emissions in the Inventory.
4-60 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
producers, continue to put downward pressure on the production of SiC in the United States.
Consumption of SiC in the United States has recovered somewhat from its low in 2009 to 2020; 2021
and 2022 consumption data was withheld to avoid disclosing company proprietary data (USGS 1991 b
through 2021), and 2023 USGS data has not yet been released.
Silicon carbide was manufactured by two facilities in the United States, one of which produced primarily
non-abrasive SiC (USGS 2021). USGS production values for the United States consists of SiC used for
abrasives and for metallurgical and other non-abrasive applications (USGS 2021). In 2023, production
remained consistent, and imports and exports decreased due to foreign competition (USGS 2024). Total
consumption of SiC decreased by approximately 25 percent from 2022 to 2023 (U.S. Census Bureau
2005 through 2023).
Carbon dioxide emissions from SiC production and consumption in 2023 were 0.2 MMT C02 Eq. (183 kt
C02), which are about 25 percent lower than emissions in 1990 (see Table 4-43 and Table 4-44).
Approximately 50 percent of these emissions resulted from SiC production, while the remainder
resulted from SiC consumption. Methane emissions from SiC production in 2023 were 0.01 MMT C02
Eq. (0.5 kt CH4) (see Table 4-43 and Table 4-44). These tables indicate minor changes in emissions in
recent years.
Table 4-43: C02 and CH4 Emissions from Silicon Carbide Production and Consumption
(MMT C02 Eq.)
Year
1990
2005
2019
2020
2021
2022
2023
SiC Production
C02
0.2
0.1
0.1
0.1
0.1
0.1
0.1
cm
+
+
+
+
+
+
+
SiC Consumption
CO2
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Total
0.2
0.2
0.2
0.2
0.2
0.2
0.2
Note: Totals may not sum due to independent rounding.
Table 4-44: C02 and CH4 Emissions from Silicon Carbide Production and Consumption
(kt)
Year
1990
2005
2019
2020
2021
2022
2023
SiC Production
CO2
170
92
92
92
92
105
105
CH4
1
+
+
+
+
+
+
SiC Consumption
CO2
73
121
84
62
80
105
78
Note: Totals may not sum due to independent rounding.
Methodology and Time-Series Consistency
Emissions of C02 and CH4 from the production of SiC are calculated using the Tier 1 method from the
2006IPCC Guidelines, in accordance with the IPCC methodological decision tree and available data.
Emissions of C02 from the consumption of SiC are a country-specific source calculated using a
country-specific methodology based on available data. The 2006 IPCC Guidelines do not provide
Industrial Processes and Product Use 4-61
-------
guidance for estimating emissions from use of SiC or SiC consumption, but the country-specific
methodology used is based on the stoichiometry of SiC consumption and is compatible with the 2006
IPCC Guidelines and consistent with a Tier 1 approach.
Annual estimates of SiC production were multiplied by the default emission factors, as shown below:
Equation 4-7:2006 IPCC Guidelines Tier 1: Emissions from Carbide Production
(Equation 3.11)
Emission factors were taken from the 2006 IPCC Guidelines:
• 2.62 metric tons C02/metric ton SiC
• 11.6 kg CH4/metric ton SiC
Production data includes silicon carbide manufactured for abrasive applications as well as for
metallurgical and other non-abrasive applications (USGS 2021).
Silicon carbide industrial abrasives production data for 1990 through 2023 were obtained from the U.S.
Geological Survey (USGS 1991 a through 2021; USGS 2023a, USGS 2024). Silicon carbide production
data published by USGS have been rounded to the nearest 5,000 metric tons to avoid disclosing
company proprietary data. For the period 1990 through 2001, reported USGS production data include
production from two facilities located in Canada that ceased operations in 1995 and 2001. Using SiC
production data from Canada (ECCC 2022), U.S. SiC production for 1990 through 2001 was adjusted to
reflect only U.S. production.
Emissions from SiC consumption are calculated by multiplying the annual SiC consumption for
metallurgical and other non-abrasive uses by the carbon content of SiC (about 30.0 percent), which is
based on the molecular weight of SiC, and converted to C02. This conversion calculation equates to
1.10 and is consistent with the IPCC default emission factor to calculate C02 emissions from the
consumption of acetylene, a calcium carbide product, and demonstrates a methodology consistent
with the 2006 IPCC Guidelines. The amount of SiC used by other non-abrasive applications is
determined by multiplying the annual SiC consumption by 50 percent (the percentage that the USGS
allocates as usage by metallurgical and other non-abrasive applications) and then subtracting the
amount of SiC used for metallurgical applications (USGS 1991 a through 2021; USGS 2023a).
SiC consumption data are estimated for the entire time series using USGS consumption data (USGS
1991 b through 2022) and data from the U.S. International Trade Commission (USITC) database on net
imports and exports of SiC (U.S. Census Bureau 2005 through 2023) (Table 4-45). Total annual SiC
consumption (utilization) was estimated by subtracting annual exports of SiC from the total of annual
Esc,C02 — EFsc,C02 X Qsc
1 metric ton
1000 kg
where,
C02 emissions from production of SiC, metric tons
Emission factor for production of SiC, metric ton C02/metric ton SiC
Quantity of SiC produced, metric tons
CH4 emissions from production of SiC, metric tons
Emission factor for production of SiC, kilogram CH4/metric ton SiC
4-62 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
national SiC production and annual imports. Data on the annual consumption of SiC for metallurgical
uses were obtained from USGS Minerals Yearbook: Silicon (USGS 1991 b-2021; USGS 2023b). USGS
withheld consumption data for metallurgical uses from publication for 2017, 2018, 2021, and 2022, due
to concerns of disclosing company-specific sensitive information. SiC consumption for 2017 and 2018
were estimated using 2016 values and SiC consumption for 2021 and 2022 were estimated using the
2020 value (USGS 2023b). Additionally, as the USGS has not yet released the 2023 data, SiC
consumption for 2023 was estimated using the 2020 value.
The petroleum coke portion of the total C02 process emissions from silicon carbide production is
adjusted for within the Energy chapter to avoid double counting emissions, as these fuels were
consumed during non-energy related activities. Additional information on the adjustments made within
the Energy sector for non-energy use of fuels is described in both the Methodology section of C02 from
Fossil Fuel Combustion (Section 3.1) and Annex 2.1, Methodology for Estimating Emissions of C02 from
Fossil Fuel Combustion.
Table 4-45: Production and Consumption of Silicon Carbide (Metric Tons)
Year
1990
2005
2019
2020
2021
2022
2023
SiC Production
65,000
35,000
35,000
35,000
35,000
40,000
40,000
SiC Consumption
132,465
220,149
152,412
113,756
146,312
191,133
142,569
Methodological approaches were applied to the entire time series to ensure consistency in emissions
from 1990 through 2023.
Uncertainty
Silicon carbide production data published by the USGS is rounded to the nearest 5,000 tons and has
been consistently reported at 35,000 tons since 2003 to avoid disclosure of company proprietary data.
This translates to an uncertainty range of ±7 percent and a normal probability density function for SiC
production (USGS 2021). There is uncertainty associated with the emission factors used because they
are based on stoichiometry as opposed to monitoring of actual SiC production plants. An alternative is
to calculate emissions based on the quantity of petroleum coke used during the production process
rather than on the amount of silicon carbide produced; however, these data were not available. For CH4,
there is also uncertainty associated with the hydrogen-containing volatile compounds in the petroleum
coke (IPCC 2006). EPA assigned an uncertainty of ±10 percent for the Tier 1 C02 and CH4 emission
factors for the SiC production processes, and using this suggested uncertainty provided in Section
3.6.3.1 of the 2006 IPCC Guidelines is appropriate based on expert judgment (RTI 2023). Per this expert
judgment, a triangular probability density function was assigned for emission factors. There is also
uncertainty associated with the use or destruction of CH4 generated from the process, in addition to
uncertainty associated with levels of production, net imports, consumption levels, and the percent of
total consumption that is attributed to metallurgical and other non-abrasive uses. EPA assigned an
uncertainty range of ±5 percent for the primary data inputs for consumption (i.e., crude imports, ground
and refined imports, crude exports, ground and refined exports, utilization [metallurgical applications])
to calculate overall uncertainty from SiC production, and using this suggested uncertainty provided in
Section 3.6.3.2 of the 2006 IPCC Guidelines is appropriate based on expert judgment (RTI 2023).
The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 4-46. Silicon
carbide production and consumption C02 emissions from 2023 were estimated to be between 0.17 and
Industrial Processes and Product Use 4-63
-------
0.20 MMT C02 Eq. at the 95 percent confidence level. These values indicate a range of 10 percent below
and 10 percent above the emission estimate of 0.2 MMT C02 Eq. at the 95 percent confidence level.
Silicon carbide production CH4 emissions were estimated to be between 0.012 and 0.014 MMT C02 Eq.
at the 95 percent confidence level. These values indicate a range of 10 percent below and 11 percent
above the emission estimate of 0.01 MMT C02 Eq. at the 95 percent confidence level.
Table 4-46: Approach 2 Quantitative Uncertainty Estimates for CH4 and C02 Emissions
from Silicon Carbide Production and Consumption (MMT C02 Eq. and Percent)
2023 Emission Uncertainty Range Relative to Emission
Estimate Estimate"
Source
Gas
(MMT CO2 Eq.)
(MMT CO2 Eq.)
(%)
Lower
Upper
Lower
Upper
Bound
Bound
Bound
Bound
Silicon Carbide Production
and Consumption
CO2
0.2
0.2
0.2
-10%
+ 10%
Silicon Carbide Production
cm
+
+
+
-10%
+ 11%
+ Does not exceed 0.05 MMT C02 Eq.
a Range of emission estimates predicted by Monte Carlo stochastic simulation for a 95 percent confidence interval.
QA/QC and Verification
General quality assurance/quality control (QA/QC) procedures were applied consistent with the U.S.
Inventory QA/QC plan, which is in accordance with Volume 1, Chapter 6 of 2006IPCC Guidelines as
described in the introduction of the IPPU chapter (see Annex 8 for more details).
Recalculations Discussion
During annual QC, a transcription error for the 1990 value of total C02 and CH4 emissions (MMT C02 Eq.)
from silicon carbide production and consumption was identified and corrected in Table 4-43. No
recalculations were performed due to this transcription error, and no other recalculations were
performed for the 1990 through 2022 portion of the time series.
Planned Improvements
EPA has initiated research for data on SiC consumption by end-use for consideration in updating
emissions estimates from SiC consumption and to account for emissions where they occur and has not
identified data to disaggregate emissions and allocate to specific metallurgical or other industrial
applications. This planned improvement is subject to data availability and will be implemented in the
medium- to long-term given significance of emissions.
EPA has not integrated aggregated facility-level GHGRP information to inform estimates of C02 and CH4
from SiC production and consumption. The aggregated information (e.g., activity data and emissions)
associated with silicon carbide did not meet criteria to shield underlying confidential business
information (CBI) from public disclosure. EPA plans to examine the use of GHGRP silicon carbide
emissions data for possible use in emission estimates consistent with the latest IPCC guidance on the
use of facility-level data in national inventories included in Volume 1, Chapter 2.3 of the 2019
Refinement to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories. This planned
4-64 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
improvement is ongoing and has not been incorporated into this Inventory report. This is a medium-term
planned improvement given significance of emissions from this industry.
4.11 Titanium Dioxide Production (Source
Category 2B6)
Titanium dioxide (Ti02) is manufactured using one of two processes: the chloride process and the
sulfate process. The chloride process uses petroleum coke and chlorine as raw materials and emits
process-related carbon dioxide (C02). The sulfate process does not use petroleum coke or other forms
of carbon as a raw material and does not emit C02. The reporting category (2B6) includes emissions
from production of Ti02. In accordance with the IPCC methodological guidance, emissions from fuels
consumed for energy purposes during the production of titanium dioxide are accounted for as part of
fossil fuel combustion in the industrial end-use sector reported under the Energy chapter. The chloride
process is based on the following chemical reactions and does emit C02:
2FeTi03 -I- 7Cl2 "I" 3C 2.TICI4 -I- 2.F6CI2 -I- 3CO2
2TiCl4 + 2(?2 2Ti02 "I"
The carbon in the first chemical reaction is provided by petroleum coke, which is oxidized in the
presence of the chlorine and FeTi03 (rutile ore) to form C02. Since 2004, all Ti02 produced in the United
States has been produced using the chloride process, and a special grade of "calcined" petroleum coke
is manufactured specifically for this purpose.
The principal use of Ti02 is as a white pigment in paint, lacquers, and varnishes. It is also used as a
pigment in the manufacture of paints, plastics, paper, and other products. In 2023, U.S. Ti02 production
totaled 920,000 metric tons (USGS 2024). Five plants produced Ti02 in the United States in 2023.
Emissions of C02 from titanium dioxide production in 2023 were estimated to be 1.2 MMT C02 Eq. (1,233
kt C02), which represents a decrease of 3.1 percent since 1990 (see Table 4-47 and Table 4-48).
Compared to 2022, emissions from titanium dioxide production decreased by 20% because production
decreased by 20% from 2022 to 2023. Production reduced from 2022 to 2023 due to a decrease in both
exports and imports of Ti02 pigments in 2023 as a result of reduced global and domestic demand (USGS
2024).
Table 4-47: C02 Emissions from Titanium Dioxide (MMT C02 Eq.)
Year 1990
2005
2019 2020
2021
2022
2023
Titanium Dioxide 1.2
1.8
1.3 1.3
1.5
1.5
1.2
Table 4-48: C02 Emissions from Titanium Dioxide (kt C02)
Year 1990
2005
2019 2020
2021
2022
2023
Titanium Dioxide 1,195
1,755
1,340 1,340
1,541
1,541
1,233
Industrial Processes and Product Use 4-65
-------
Methodology and Time-Series Consistency
Emissions of C02 from Ti02 production are calculated using a Tier 1 method from the 2006IPCC
Guidelines, in accordance with the IPCC methodological decision tree and available data. Annual
national Ti02 production is multiplied by chloride process-specific emission factors provided by IPCC
(IPCC 2006). The Tier 1 equation is as follows:
Equation 4-8: 2006 IPCC Guidelines Tier 1: C02 Emissions from Titanium Production
(Equation 3.12)
Etd = EFta x Qtd
where,
Etd = C02 emissions from Ti02 production, metric tons
EFtd = Emission factor (chloride process), metric ton C02/metric ton Ti02
Qtd = Quantity of Ti02 produced, metric tons
The petroleum coke portion of the total C02 process emissions from Ti02 production is adjusted for
within the Energy chapter as these fuels were consumed during non-energy related activities. Additional
information on the adjustments made within the Energy sector for Non-Energy Use of Fuels is described
in both the Methodology section of C02 from Fossil Fuel Combustion (Section 3.1 Fossil Fuel
Combustion) and Annex 2.1, Methodology for Estimating Emissions of C02 from Fossil Fuel
Combustion.
Production data and capacity data for the total amount of Ti02 produced using the chloride and sulfate
processes are based on data from the USGS.
Annual Ti02 production data for 1990 through 2018 were obtained from the U.S. Geological Survey
(USGS) Minerals Yearbook: Titanium (USGS 1994 through 2022). Production data for 2019 and 2020 were
obtained from the USGS Minerals Yearbook: Titanium, advanced data release of the 2020 tables (USGS
2023). Production data for 2021, 2022, and 2023 were obtained from the USGS Minerals Commodity
Summaries: Titanium and Titanium Dioxide (USGS 2024).37
The chloride process capacity data for 1994 through 2013 and the sulfate process capacity data for 1994
through 2004 were obtained from annual USGS Minerals Yearbook: Titanium. Starting with 2014, the
chloride process capacity data were obtained from annual USGS Minerals Commodity Summaries:
Titanium and Titanium Dioxide. Process capacity data were not available for 1990 through 1993, so data
from the 1994 USGS Minerals Yearbookwere used as proxy for these prior years. Because a sulfate
process plant closed in September 2001, the chloride process capacity data for 2001 was estimated
(Gambogi 2002). By 2002, only one sulfate process plant remained online in the United States, and this
plant closed in 2004 (USGS 2005).
As production data was not specified by process type, and the sulfate process does not produce C02,
annual production of the chloride process from 1990 through 2003 was estimated based on the ratio of
the chloride process production capacity to the total production capacity (i.e., the combined chloride
process and sulfate process production capacities). As the last remaining sulfate process plant in the
37 EPA has not integrated aggregated facility-level GHGRP information for titanium dioxide production facilities (40 CFR
Part 98 Subpart EE). The relevant aggregated information (activity data, emission factor) from these facilities did not
meet criteria to shield underlying CBI from public disclosure.
4-66 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
United States closed in 2004, 100 percent of production since 2004 used the chloride process (USGS
2005). The 2006IPCC Guidelines emission factor of 1.34 metric tons C02/metric ton Ti02 was applied to
the estimated chloride process production (IPCC 2006). It was assumed that all Ti02 produced using the
chloride process was produced using petroleum coke, although some Ti02 may have been produced
with graphite or other carbon inputs.
Table 4-49: Titanium Dioxide Production (kt)
Year
1990
2005
2019
2020
2021
2022
2023
Production
979
1,310
1,000
1,000
1,150
1,150
920
Methodological approaches were applied to the entire time series to ensure consistency in emissions
from 1990 through 2023.
Uncertainty
Each year, the USGS collects titanium industry data for titanium mineral and pigment production
operations. If Ti02 pigment plants do not respond, production from the operations is estimated based on
prior year production levels and industry trends. Variability in response rates fluctuates from 67 to 100
percent of Ti02 pigment plants over the time series. EPA currently uses an uncertainty range of ±5
percent and a normal probability density function for the primary data inputs (i.e., Ti02 production and
chloride process capacity values) to calculate overall uncertainty from Ti02 production, and using this
suggested uncertainty provided in Section 3.7.3.2 of the 2006 IPCC Guidelines is appropriate based on
expert judgment (RTI 2023). Additionally, the EPA uses an uncertainty range of ±15 percent and a
triangular probability density function for the C02 chloride process carbon consumption rate, and using
this uncertainty provided in Section 3.7.2.2 of the 2006 IPCC Guidelines is representative of operations
in the United States, based on expert judgment (RTI 2023).
Although some Ti02 may be produced using graphite or other carbon inputs, information and data
regarding these practices were not available. Titanium dioxide produced using graphite inputs, for
example, may generate differing amounts of C02 per unit of Ti02 produced as compared to that
generated using petroleum coke in production. While the most accurate method to estimate emissions
would be to base calculations on the amount of reducing agent used in each process rather than on the
amount of Ti02 produced, sufficient data were not available to do so.
As of 2004, the last remaining sulfate-process plant in the United States closed. Since annualTi02
production was not reported by USGS by the type of production process used (chloride or sulfate) prior
to 2004 and only the percentage of total production capacity by process was reported, the percent of
total Ti02 production capacity that was attributed to the chloride process was multiplied by total Ti02
production to estimate the amount of Ti02 produced using the chloride process. Finally, the emission
factor was applied uniformly to all chloride-process production, and no data were available to account
for differences in production efficiency among chloride-process plants. In calculating the amount of
petroleum coke consumed in chloride-process Ti02 production, literature data were used for petroleum
coke composition. Certain grades of petroleum coke are manufactured specifically for use in the Ti02
chloride process; however, this composition information was not available. EPA assigned an uncertainty
range of ±15 percent and a triangular probability density function for the Tier 1 C02 emission factor for
the titanium dioxide (chloride route) production process, and using this uncertainty provided in Table 3.9
Industrial Processes and Product Use 4-67
-------
of the 2006IPCC Guidelines is representative of operations in the United States based on expert
judgment (RTI 2023).
The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 4-49. Titanium
dioxide consumption C02 emissions from 2023 were estimated to be between 1.1 and 1.4 MMT C02 Eq.
at the 95 percent confidence level. This indicates a range of approximately 13 percent below and 13
percent above the emission estimate of 1.2 MMT C02 Eq.
Table 4-50: Approach 2 Quantitative Uncertainty Estimates for C02 Emissions from
Titanium Dioxide Production (MMT C02 Eq. and Percent)
2023 Emission
Uncertainty Range Relative to Emission
Estimate
Estimate3
Source
Gas
(MMT CO2 Eq.)
(MMT CO2 Eq.)
(%)
Lower
Upper
Lower
Upper
Bound
Bound
Bound
Bound
Titanium Dioxide Production
C02
1.2
1.1
1.4
-13%
+ 13%
a Range of emission estimates predicted by Monte Carlo stochastic simulation for a 95 percent confidence interval.
QA/QC and Verification
General quality assurance/quality control (QA/QC) procedures were applied consistent with the U.S.
Inventory QA/QC plan, which is in accordance with Volume 1, Chapter 6 of the 2006 IPCC Guidelines as
described in the introduction of the IPPU chapter (see Annex 8 for more details).
Recalculations Discussion
Updated USGS data on Ti02 production was available for 2021 and 2022, resulting in updated emissions
estimates for those years. Compared to the previous Inventory, emissions for 2021 increased by 5
percent (67 kt C02), and emissions for 2022 increased by 5 percent (67 kt C02).
Planned Improvements
EPA is continuing to exam the use of GHGRP titanium dioxide emissions and other data for possible use
in emission estimates consistent with the latest IPCC guidance on the use of facility-level data in
national inventories.38 This planned improvement is ongoing and has not been incorporated into this
Inventory report. This is a long-term planned improvement given the significance of these emissions.
4.12 Soda Ash Production (Source Category
2B7)
Carbon dioxide (C02) is generated as a byproduct of calcining trona ore to produce soda ash (sodium
carbonate, Na2C03) and is eventually emitted into the atmosphere. This reporting category (2B7)
38 See http://www.ipcc-nggip.iges.or.jp/public/tb/TFI Technical Bulletin 1 .pdf and Volume 1. Chapter ?.3 of the ?019
Refinement to the 2006 IPCC Guidelines for National GHG Inventories.
4-68 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
includes emissions from the production of soda ash by any of four processes, of which calcining trona
ore is the only emissive process used in the United States. In addition, C02 may also be released when
soda ash is consumed. Commercial soda ash is used as a raw material in a variety of industrial
processes and in many familiar consumer products such as glass, soap and detergents, paper, textiles,
and food. Emissions from soda ash consumption associated with glass production are reported under
Section 4.3, glass production. In addition, soda ash is used primarily to manufacture many sodium-
based inorganic chemicals, including sodium bicarbonate, sodium chromates, sodium phosphates, and
sodium silicates (USGS 2018b). Emissions from soda ash consumption not associated with glass
production are reported under Section 4.4, other process uses of carbonates. Emissions from fuels
consumed for energy purposes during the production and consumption of soda ash are accounted for
as part of fossil fuel combustion in the industrial end-use sector reported under the Energy chapter.
During the soda ash production process, trona ore is calcined in a rotary kiln and chemically
transformed into a crude soda ash that requires further processing. Calcining involves placing crushed
trona ore into a kiln to convert sodium bicarbonate into crude sodium carbonate that will later be filtered
into pure soda ash. Carbon dioxide and water are generated as byproducts of the calcination process.
Carbon dioxide emissions from the calcination of trona ore can be estimated based on the chemical
reaction shown above. The emission of C02 during trona-based production is based on the following
reaction:
2Na2C03 ¦ NaHC03 ¦ 2H20(Trona) -> 3Na2C03(Soda Ash) + 5H20 + C02
Internationally, two types of soda ash are produced: natural and synthetic. The United States produces
only natural soda ash and is second only to China in total soda ash production. Trona is the principal ore
from which natural soda ash is made.
The United States represents about one-fifth of total global soda ash output (USGS 2023a). Only two
states still produce natural soda ash: Wyoming and California. Of these two states, net emissions of
C02 from soda ash production were only calculated for Wyoming where trona ore is used.39 Soda ash
end uses in 2023 (excluding glass production) consisted of chemical production, 55 percent; other uses,
17 percent; wholesale distributors (e.g., for use in agriculture, water treatment, and grocery wholesale),
9 percent; soap and detergent manufacturing, 9 percent; flue gas desulfurization, 6 percent; water
treatment, 2 percent; and pulp and paper production, 2 percent (USGS 2024b).40
U.S. natural soda ash is competitive in world markets because it is generally considered a better-quality
raw material than synthetically produced soda ash, and most of the world's soda ash is synthetic.
Although the United States continues to be a major supplier of soda ash, China surpassed the United
States in soda ash production in 2003, becoming the world's leading producer.
39 In California, soda ash is manufactured using sodium carbonate-bearing brines instead of trona ore. To extract the
sodium carbonate, the complex brines are first treated with CO2 in carbonation towers to convert the sodium carbonate
into sodium bicarbonate, which then precipitates from the brine solution. The precipitated sodium bicarbonate is then
calcined back into sodium carbonate. Although CO2 is generated as a byproduct, the CO2 is recovered and recycled for
use in the carbonation stage and is not emitted. A facility in a third state, Colorado, produced soda ash until the plant
was idled in 2004. The lone producer of sodium bicarbonate no longer mines trona ore in the state. For a brief time,
sodium bicarbonate was produced using soda ash feedstocks mined in Wyoming and shipped to Colorado. Prior to
2004, because the trona ore was mined in Wyoming, the production numbers given by the USGS included the
feedstocks mined in Wyoming and shipped to Colorado. In this way, the sodium bicarbonate production that took place
in Colorado was accounted for in the Wyoming numbers.
40 Percentages may not add up to 100 percent due to independent rounding.
Industrial Processes and Product Use 4-69
-------
In 2023, C02 emissions from the production of soda ash from trona ore were 1.7 MMT C02 Eq. (1,723 kt
C02) (see Table 4-51 and Table 4-52). Total emissions from soda ash production in 2023 increased by
approximately 1 percent compared to emissions in 2022. Emissions have increased by approximately 20
percent from 1990 levels. Trends in emissions have remained relatively constant over the time series
with some fluctuations since 1990. In general, these fluctuations were related to the behavior of the
export market and the U.S. economy. The U.S. soda ash industry saw a decline in domestic and export
sales caused by adverse global economic conditions in 2009, followed by a steady increase in
production through 2019 before a significant decrease in 2020 due to the COVID-19 pandemic and
increase since 2020 as the economy rebounded from the height of the pandemic.
Table 4-51: C02 Emissions from Soda Ash Production (MMT C02 Eq.)
Year 1990
2005
2019 2020
2021
2022
2023
Soda Ash Production 1.4
1.7
1.8 1.5
1.7
1.7
1.7
Table 4-52: C02 Emissions from Soda Ash Production (kt C02)
Year 1990
2005
2019 2020
2021
2022
2023
Soda Ash Production 1,431
1,655
1,792 1,461
1,714
1,704
1,723
Methodology and Time-Series Consistency
Carbon dioxide emissions from soda ash production are calculated using a Tier 1 method from the 2006
IPCC Guidelines, in accordance with the IPCC methodological decision tree and available data. Based
on the reaction shown above, the IPCC default emission factor is 0.0974 metric tons C02 per metric ton
of trona ore, or one metric ton of C02 is emitted when approximately 10.27 metric tons of trona ore are
processed (IPCC 2006).
Data are not currently available for the quantity of trona used in soda ash production. Because trona ore
is used primarily for soda ash production, EPA assumes that all trona ore production was used in soda
ash production. The activity data for trona ore production (see Table 4-53) for 1990 through 2023 were
obtained from the U.S. Geological Survey (USGS) Minerals Yearbook for Soda Ash (1994 through 2015b)
and USGS Mineral Industry Surveys for Soda Ash (USGS 2016 through 2017, 2018a, 2019, 2020, 2021,
2022b, 2023b, 2024b). Soda ash production data were collected by the USGS from voluntary surveys of
the U.S. soda ash industry.41
Table 4-53: Trona Ore Used in Soda Ash Production (kt)
Year
1990
2005
2019
2020
2021
2022
2023
Trona Ore Use®
14,700
17,000
18,400
15,000
17,600
17,500
17,700
a Trona ore use is assumed to be equal to trona ore production.
Methodological approaches were applied to the entire time series to ensure consistency in emissions
estimates from 1990 through 2023.
41 EPA has assessed the feasibility of using emissions information (including activity data) from EPA's GHGRP program. At
this time, the aggregated information associated with production of soda ash did not meet criteria to shield underlying
confidential business information (CBI) from public disclosure.
4-70 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Uncertainty
Emission estimates from soda ash production have relatively low associated uncertainty levels because
reliable and accurate data sources are available for the emission factor and activity data for trona-based
soda ash production. One source of uncertainty is the purity of the trona ore used for manufacturing
soda ash. The emission factor used for this estimate assumes the ore is 100 percent pure and likely
overestimates the emissions from soda ash manufacture. The average water-soluble sodium carbonate-
bicarbonate content for ore mined in Wyoming ranges from 85.5 to 93.8 percent (USGS 1995c).
EPA is aware of one facility producing soda ash from a liquid alkaline feedstock process, based on EPA's
GHGRP. Soda ash production data was collected by the USGS from voluntary surveys. A survey request
was sent to each of the five soda ash producers, all of which responded, representing 100 percent of the
total production data (USGS 2024b). EPA assigned an uncertainty range of ±5 percent for trona
production, and using the suggested uncertainty provided in Section 3.8.2.2 of the 2006IPCC
Guidelines is appropriate based on expert judgment (RTI 2023). EPA assigned an uncertainty range of -15
percent to 0 percent range for the trona emission factor, based on expert judgment on the purity of
mined trona (USGS 1995c). Per this expert judgment, a normal probability density function was assigned
for activity data, and a triangular probability density function was assigned for the emission factor.
The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 4-54. Soda ash
production C02 emissions for 2023 were estimated to be between 1.5 and 1.8 MMT C02 Eq. at the 95
percent confidence level. This indicates a range of approximately 9 percent below and 8 percent above
the emission estimate of 1.6 MMT C02 Eq.
Table 4-54: Approach 2 Quantitative Uncertainty Estimates for C02 Emissions from
Soda Ash Production (MMT C02 Eq. and Percent)
Source
Gas
2023 Emission Estimate
(MMT CO2 Eq.)
Uncertainty Range Relative to Emission Estimate"
(MMT CO2 Eq.) (%)
Lower
Upper
Lower
Upper
Bound
Bound
Bound
Bound
Soda Ash Production
C02
1.7
1.5
1.8
-9%
+8%
a Range of emission estimates predicted by Monte Carlo stochastic simulation for a 95 percent confidence interval.
QA/QC and Verification
General quality assurance/quality control (QA/QC) procedures were applied consistent with the U.S.
Inventory QA/QC plan, which is in accordance with Volume 1, Chapter 6 of 2006 IPCC Guidelines as
described in the introduction of the IPPU chapter (see Annex 8 for more details).
Recalculations Discussion
No recalculations were performed for the 1990 through 2022 portion of the time series.
Planned Improvements
EPA continues to analyze and assess opportunities to use facility-level data from EPA's GHGRP to
improve the emission estimates for the soda ash production source category consistent with latest
Industrial Processes and Product Use 4-71
-------
IPCC guidance on the use of facility-level data in national inventories included in in Volume 1, Chapter
2.3 of the 2019 Refinement to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories.
EPA plans to assess the use of trona ore in applications other than for soda ash production and evaluate
impacts of greenhouse gas emissions from those uses.
4.13 Petrochemical Production (Source
Category 2B8)
The production of some petrochemicals results in carbon dioxide (C02) and methane (CH4) emissions.
Petrochemicals are chemicals isolated or derived from petroleum or natural gas. This reporting category
(2B8) includes C02 emissions from the production of acrylonitrile, carbon black, ethylene, ethylene
dichloride, ethylene oxide, and methanol, and CH4 emissions from the production of acrylonitrile. The
petrochemical industry uses primary fossil fuels (i.e., natural gas, coal, petroleum, etc.) for non-fuel
purposes in the production of carbon black and other petrochemicals. Per the IPCC methodological
guidance, emissions from fuels and feedstocks transferred out of the system for use in energy purposes
(e.g., indirect or direct process heat or steam production) are currently accounted for as part of fossil
fuel combustion in the industrial end-use sector reported under the Energy chapter.
Worldwide, more than 90 percent of acrylonitrile (vinyl cyanide, C3H3N) is made by way of direct
ammoxidation of propylene with ammonia (NH3) and oxygen over a catalyst. This process is referred to
as the SOHIO process, named after the Standard Oil Company of Ohio (SOHIO) (IPCC 2006). The
primary use of acrylonitrile is as the raw material for the manufacture of acrylic and modacrylic fibers.
Other major uses include the production of plastics (acrylonitrile-butadiene-styrene [ABS] and styrene-
acrylonitrile [SAN]), nitrile rubbers, nitrile barrier resins, adiponitrile, and acrylamide. All U.S.
acrylonitrile facilities use the SOHIO process (AN 2014). The SOHIO process involves a fluidized bed
reaction of chemical-grade propylene, ammonia, and oxygen over a catalyst. The process produces
acrylonitrile as its primary product, and the process yield depends on the type of catalyst used and the
process configuration. The ammoxidation process produces byproduct C02, carbon monoxide (CO),
and water from the direct oxidation of the propylene feedstock and produces other hydrocarbons from
side reactions.
Carbon black is a black powder generated by the incomplete combustion of an aromatic petroleum- or
coal-based feedstock at a high temperature. Most carbon black produced in the United States is added
to rubber to impart strength and abrasion resistance, and the tire industry is by far the largest consumer.
The other major use of carbon black is as a pigment. The predominant process used in the United States
to produce carbon black is the furnace black (or oil furnace) process. In the furnace black process,
carbon black oil (a heavy aromatic liquid) is continuously injected into the combustion zone of a natural
gas-fired furnace. Furnace heat is provided by the natural gas and a portion of the carbon black
feedstock; the remaining portion of the carbon black feedstock is pyrolyzed to carbon black. The
resultant C02 and uncombusted CH4 are released from thermal incinerators used as control devices,
process dryers, and equipment leaks. Three facilities in the United States use other types of carbon
black processes. Specifically, one facility produces carbon black by the thermal cracking of acetylene-
containing feedstocks (i.e., acetylene black process), a second facility produces carbon black by the
4-72 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
thermal cracking of other hydrocarbons (i.e., thermal black process), and a third facility produces
carbon black by the open burning of carbon black feedstock (i.e., lamp black process) (EPA 2000).
Ethylene (C2H4) is consumed in the production processes of the plastics industry including polymers
such as high, low, and linear low density polyethylene (HDPE, LDPE, LLDPE); polyvinyl chloride (PVC);
ethylene dichloride; ethylene oxide; and ethylbenzene. Virtually all ethylene is produced from steam
cracking of ethane, propane, butane, naphtha, gas oil, and other feedstocks. The representative
chemical equation for steam cracking of ethane to ethylene is shown below:
c2h6 -> c2h4 + h2
Small amounts of CH4 are also generated from the steam cracking process. In addition, C02 and CH4
emissions result from combustion units.
Ethylene dichloride (C2H4CI2) is used to produce vinyl chloride monomer, which is the precursor to
polyvinyl chloride (PVC). Ethylene dichloride was also used as a fuel additive until 1996 when leaded
gasoline was phased out. Ethylene dichloride is produced from ethylene by either direct chlorination,
oxychlorination, or a combination of the two processes (i.e., the "balanced process"); most U.S.
facilities use the balanced process. The direct chlorination and oxychlorination reactions are shown
below:
C2H4 + Cl2 -> C2H4Cl2 (direct chlorination)
C2H4 + i02 + 2HCl -> C2H4Cl2 + 2H20 (oxychlorination)
C2H4 + 302 -> 2C02 + 2H20 (direct oxidation of ethylene during oxychlorination)
In addition to the byproduct C02 produced from the direct oxidation of the ethylene feedstock, C02 and
CH4emissions are also generated from combustion units.
Ethylene oxide (C2H4O) is used in the manufacture of glycols, glycol ethers, alcohols, and amines.
Approximately 70 percent of ethylene oxide produced worldwide is used in the manufacture of glycols,
including monoethylene glycol. Ethylene oxide is produced by reacting ethylene with oxygen over a
catalyst. The oxygen may be supplied to the process through either an air (air process) or a pure oxygen
stream (oxygen process). The byproduct C02 from the direct oxidation of the ethylene feedstock is
removed from the process vent stream using a recycled carbonate solution, and the recovered C02 may
be vented to the atmosphere or recovered for further utilization in other sectors, such as food
production (IPCC 2006). The combined ethylene oxide reaction and byproduct C02 reaction is
exothermic and generates heat, which is recovered to produce steam for the process. The ethylene
oxide process also produces other liquid and off-gas byproducts (e.g., ethane) that may be burned for
energy recovery within the process. Almost all facilities, except one in Texas, use the oxygen process to
manufacture ethylene oxide (EPA 2008).
Methanol (CH3OH) is a chemical feedstock most often converted into formaldehyde, acetic acid and
olefins. It is also an alternative transportation fuel, as well as an additive used by municipal wastewater
treatment facilities in the denitrification of wastewater. Methanol is most commonly synthesized from a
synthesis gas (i.e., "syngas" - a mixture containing H2, CO, and C02) using a heterogeneous catalyst.
There are a number of process techniques that can be used to produce syngas. Worldwide, steam
reforming of natural gas is the most common method; most methanol producers in the United States
also use steam reforming of natural gas to produce syngas. Other syngas production processes in the
United States include partial oxidation of natural gas and coal gasification.
Industrial Processes and Product Use 4-73
-------
Emissions of C02 and CH4 from petrochemical production in 2023 were 30.5 MMT C02 Eq. (30,540 kt
C02) and 0.005 MMT C02 Eq. (0.19 kt CH4), respectively (see Table 4-55 and Table 4-56). Carbon dioxide
emissions from petrochemical production are driven primarily from ethylene production, while CH4
emissions are only from acrylonitrile production. Since 1990, total C02emissions from petrochemical
production increased by 52 percent, and CH4 emissions declined by 12 percent. Emissions of C02 were
6 percent higher in 2023 than in 2022, and emissions of CH4 were 13 percent higher in 2023 than in
2022. The increase in C02 emissions since 1990 is due primarily to increased ethylene and methanol
production, which have been driven by the increased natural gas production in the United States. The
increase in C02 emissions since 2022 primarily is due to an increase in ethylene production and in
emissions from ethylene production. Production and emissions from all other petrochemicals, except
carbon black, also increased by smaller amounts in 2023. Since CH4 emissions from acrylonitrile are
calculated using a Tier 1 approach based on production as the activity data, the decrease in CH4
emissions since 1990 and the increase since 2022 correspond with changes in the production levels for
acrylonitrile.
Table 4-55: C02 and CH4 Emissions from Petrochemical Production (MMT C02 Eq.)
Year
1990
2005
2019
2020
2021
2022
2023
CO2
20.1
26.9
28.5
27.9
30.7
28.8
30.5
Carbon Black
3.4
4.3
3.3
2.6
3.0
3.1
2.6
Ethylene
13.1
19.0
20.7
20.7
22.8
20.7
22.6
Ethylene Dichloride
0.3
0.5
0.5
0.5
0.4
0.4
0.5
Ethylene Oxide
1.1
1.5
1.4
1.7
1.9
1.7
1.7
Methanol
1.0
0.3
1.6
1.6
1.7
2.0
2.1
Acrylonitrile
1.2
1.3
1.0
0.9
0.9
1.0
1.1
CH4
+
+
+
+
+
+
+
Acrylonitrile
+
+
+
+
+
+
+
Total
20.1
26.9
28.5
27.9
30.7
28.8
30.5
+ Does not exceed 0.05 MMT C02 Eq.
Note: Totals may not sum due to independent rounding.
Table 4-56: C02andCH4
Emissions from Petrochemical Production (kt)
Year
1990
2005
2019
2020
2021
2022
2023
CO2
20,075
26,882
28,483
27,926
30,656
28,788
30,540
Carbon Black
3,381
4,269
3,300
2,610
3,000
3,060
2,550
Ethylene
13,126
19,024
20,700
20,700
22,800
20,700
22,600
Ethylene Dichloride
254
455
503
456
376
428
460
Ethylene Oxide
1,123
1,489
1,370
1,680
1,930
1,650
1,730
Methanol
977
319
1,620
1,630
1,700
2,000
2,130
Acrylonitrile
1,214
1,325
990
850
850
950
1,070
CH4
+
+
+
+
+
+
+
Acrylonitrile
+
+
+
+
+
+
+
+ Does not exceed 0.5 kt ChU.
Note: Totals by gas may not sum due to independent rounding.
4-74 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Methodology and Time-Series Consistency
Emissions of C02 and CH4 were calculated using the estimation methods provided by the 2006IPCC
Guidelines, in accordance with the IPCC methodological decision tree and available data, and country-
specific methods from EPA's GHGRP. The 2006 IPCC Guidelines Tier 1 method was used to estimate
C02 and CH4 emissions from production of acrylonitrile,42 and a country-specific approach similar to
the IPCC Tier 2 method was used to estimate C02 emissions from production of carbon black, ethylene
oxide, ethylene, ethylene dichloride, and methanol, as C02 emissions from petrochemical production is
a key category. The Tier 2 method for petrochemicals is a total feedstock carbon mass balance method
used to estimate total C02 emissions, but it is not applicable for estimating CH4 emissions.
As noted in the 2006 IPCC Guidelines, the Tier 2 total feedstock carbon mass balance method is based
on the assumption that all of the carbon input to the process is converted either into primary and
secondary products or into C02. Further, the guideline states that while the total carbon mass balance
method estimates total carbon emissions from the process, it does not directly provide an estimate of
the amount of the total carbon emissions emitted as C02, CH4, or non-CH4 volatile organic compounds
(NMVOCs). This method accounts for all the carbon as C02, including CH4.
A methodology refinement for emissions from methanol production was implemented in the previous
Inventory to transition from a Tier 1 method to a country-specific approach similar to a Tier 2 method,
using the process C02 emissions reported to Subpart X of the GHGRP. As part of this refinement, CH4
emissions from methanol production for every year in the time series are now included in the C02
emissions estimates to avoid double counting because the GHGRP reporting method is a mass balance
method under which all carbon input to the process is assumed to be converted either into primary and
secondary products or into C02.
Note, a subset of facilities reporting under EPA's GHGRP use Continuous Emission Monitoring Systems
(CEMS) to monitor C02 emissions from process vents and/or stacks from stationary combustion units or
use the optional combustion methodology for ethylene production facilities. These facilities are
required to also report C02, CH4 and N20 emissions from combustion of process off-gas in flares. The
C02 emissions from flares are included in aggregated C02 results. Analysis of aggregated annual reports
from those facilities shows that flared CH4 and N20 emissions are less than 300 kt C02 Eq./year. Since
data is only available from a subset of facilities and not consistently reported over time and since CH4
and N20 emissions are shown to be insignificant, they are excluded from this analysis. See the planned
improvements section below and Annex 5.
Carbon Black, Ethylene, Ethylene Dichloride, and Ethylene Oxide
2010 through 2023
Carbon dioxide emissions and national production for carbon black, ethylene, ethylene dichloride, and
ethylene oxide were aggregated directly from EPA's GHGRP dataset for 2010 through 2023 (EPA 2024).
These emissions reflect application of a country-specific approach similar to the IPCC Tier 2 method
and were used to estimate C02 emissions from the production of carbon black, ethylene, ethylene
42 EPA has not integrated aggregated facility-level GHGRP information for acrylonitrile production. The aggregated
information associated with production of these petrochemicals did not meet criteria to shield underlying CBI from
public disclosure.
Industrial Processes and Product Use 4-75
-------
dichloride, ethylene oxide. In 2023, data reported to the GHGRP included 2,550,000 metric tons of C02
emissions from carbon black production; 22,600,000 metric tons of C02from ethylene production;
460,000 metric tons of C02 from ethylene dichloride production; and 1,730,000 metric tons of C02 from
ethylene oxide production.
Since 2010, EPA's GHGRP requires all domestic producers of petrochemicals to report annual
emissions and supplemental emissions information (e.g., production data, etc.) under Subpart X to
facilitate verification of reported emissions. Most petrochemical production facilities are required to use
either a mass balance approach or CEMS to measure and report emissions for each petrochemical
process unit to estimate facility-level process C02 emissions; ethylene production facilities also have a
third option. The mass balance method is used by most facilities43 and assumes that all the carbon
input is converted into primary and secondary products or is emitted to the atmosphere as C02. To
apply the mass balance, facilities must measure the volume or mass of each gaseous and liquid
feedstock and product, mass rate of each solid feedstock and product, and carbon content of each
feedstock and product for each process unit. These data are used to calculate the difference in the
amount of carbon input and carbon output for each petrochemical process unit. The carbon difference
is converted to C02 emissions for each process unit, which are summed over all process units for their
facility. To apply the optional combustion methodology, ethylene production facilities must measure the
quantity, carbon content, and molecular weight of the fuel to a stationary combustion unit when that
fuel includes any ethylene process off-gas. These data are used to calculate the total C02 emissions
from the combustion unit. The facility must also estimate the fraction of the emissions that is
attributable to burning the ethylene process off-gas portion of the fuel. This fraction is multiplied by the
total emissions to estimate the emissions from ethylene production. The QA/QC and Verification
section below has a discussion of non-C02 emissions from ethylene production facilities.
All non-energy uses of residual fuel and some non-energy uses of "other oil" are assumed to be used in
the production of carbon black; therefore, consumption of these fuels is adjusted for within the Energy
chapter to avoid double-counting of emissions from fuel used in the carbon black production presented
here within IPPU sector. Additional information on the adjustments made within the Energy sector for
non-energy use of fuels is described in both the Methodology section of C02from Section 3.1 and Annex
2.1.
1990 through 2009
Prior to 2010, for carbon black, ethylene, ethylene dichloride, and ethylene oxide processes, an average
national C02 emission factor was calculated based on the GHGRP data and applied to production for
earlier years in the time series (i.e., 1990 through 2009) to estimate C02 emissions. For these 4 types of
petrochemical processes, C02 emission factors were derived from EPA's GHGRP data by dividing annual
C02 emissions for petrochemical type "i"with annual production for petrochemical type "i" and then
averaging the derived emission factors obtained for each calendar year 2010 through 2013 (EPA 2024).
The years 2010 through 2013 were used in the development of carbon dioxide emission factors as these
years are more representative of operations in 1990 through 2009 for these facilities. The average
emission factors for each petrochemical type were applied across all prior years because
43 A few facilities producing ethylene dichloride, ethylene, and methanol used CO2 CEMS; those CO2 emissions have been
included in the aggregated GHGRP emissions presented here.
4-76 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
petrochemical production processes in the United States have not changed significantly since 1990,
though some operational efficiencies have been implemented at facilities over the time series.
The average country-specific C02 emission factors that were calculated from the GHGRP data are as
follows:
• 2.59 metric tons C02/metric ton carbon black produced
• 0.79 metric tons C02/metric ton ethylene produced
• 0.040 metric tons C02/metric ton ethylene dichloride produced
• 0.46 metric tons C02/metric ton ethylene oxide produced
Annual production data for carbon black for 1990 through 2009 were obtained from the International
Carbon Black Association (Johnson 2003 and 2005 through 2010). Annual production data for ethylene,
ethylene dichloride, and ethylene oxide for 1990 through 2009 were obtained from the American
Chemistry Council's (ACC) Business of Chemistry (ACC 2024).
Methanol
2015 through 2023
Carbon dioxide emissions and national production for methanol were aggregated directly from EPA's
GHGRP data for 2015 through 2023 (EPA 2024). These emissions reflect application of a country-
specific approach similar to the IPCC Tier 2 method and were used to estimate C02 emissions from the
production of methanol. In 2023, data reported to the GHGRP included 2,130,000 metric tons of C02
emissions from methanol production.
As noted above, since 2010, EPA's GHGRP requires all domestic producers of petrochemicals to report
annual emissions and supplemental emissions information (e.g., production data, etc.) under Subpart X
to facilitate verification of reported emissions. Methanol production facilities are required to use either a
mass balance approach or CEMS to measure and report emissions for each methanol process unit to
estimate facility-level process C02 emissions. Most methanol production facilities use the mass
balance method. As noted above, when using the mass balance method, facilities must measure the
volume or mass of each gaseous and liquid feedstock and product, mass rate of each solid feedstock
and product, and carbon content of each feedstock and product for each process unit and sum for their
facility. For 2010 to 2014, the methanol data reported to GHGRP is considered CBI; therefore, the direct
use of the GHGRP data starts with the 2015 reported information.
1990 through 2014
In this Inventory, similar to the methodology for other petrochemicals that utilize GHGRP data, an
average national C02 emission factor for years prior to 2015 was calculated for methanol production
based on the GHGRP data and applied to production for earlier years in the time series (i.e., 1990
through 2014) to estimate C02 emissions. Methanol C02 emission factors were derived from EPA's
GHGRP data by dividing annual C02 emissions for methanol with annual production for methanol and
then averaging the derived emission factors obtained for each year 2015 through 2022. The average
country-specific C02 emission factor from the GHGRP data for these years was determined to be 0.26
metric tons C02/metric ton methanol produced. Annual methanol production data for 1990 through
2014were obtained from the ACC's Business of Chemistry (ACC 2024). The average country-specific
Industrial Processes and Product Use 4-77
-------
C02 emission factor from the GHGRP data is lower than the IPCC Tier 1 emission factor of 0.67 metric
tons C02/metric ton methanol produced value that was used in previous versions of the Inventory. The
main difference between the IPCC Tier 1 emission factor and the GHGRP emission factor is that the
IPCC emission factor includes emissions from combustion of natural gas fuel in the reformer as well as
vented C02 from the process; therefore, the use of the IPCC Tier 1 emission factor would double count
emissions from natural gas combustion in the IPPU chapter and the Energy chapter. EPA already
accounts for emissions from combustion of natural gas fuel in the reformer as part of fossil fuel
combustion in the industrial end-use sector reported under the Energy chapter.
Acrylonitrile
Carbon dioxide and methane emissions from acrylonitrile production were estimated using the Tier 1
method in the 2006 IPCC Guidelines. Acrylonitrile emissions represent about 3 percent of total
petrochemical emissions in 2023 so a Tier 1 approach is deemed acceptable, and higher Tier methods
could not be used due to data sensitivities which are described below. Annual acrylonitrile production
data were used with IPCC default Tier 1 C02 and CH4 emission factors to estimate emissions for 1990
through 2023. Emission factors used to estimate acrylonitrile production emissions are as follows:
• 0.18 kg CH4/metric ton acrylonitrile produced
• 1.00 metric tons C02/metric ton acrylonitrile produced
Annual acrylonitrile production data for 1990 through 2023 were obtained from ACC's Business of
Chemistry (ACC 2024). EPA is unable to apply the aggregated facility-level GHGRP information for
acrylonitrile production needed for a Tier 2 approach due to sensitive nature of reported data. The
aggregated information associated with production of these petrochemicals did not meet criteria to
shield underlying CBI from public disclosure.
Production of each type of petrochemical are shown in Table 4-57.
Table 4-57: Production of Selected Petrochemicals (kt)
Chemical
1990
2005
2019
2020
2021
2022
2023
Carbon Black
1,307
1,651
1,210
990
1,140
1,170
1,010
Ethylene
16,542
23,975
32,400
33,500
34,700
35,400
39,400
Ethylene Dichloride
6,283
11,260
12,600
11,900
11,500
12,100
11,500
Ethylene Oxide
2,429
3,220
3,800
4,680
4,860
5,310
5,430
Methanol
3,750
1,225
6,460
6,580
7,110
8,030
8,640
Acrylonitrile
1,214
1,325
990
850
850
950
1,070
As noted earlier in the introduction section of the Petrochemical Production section, the allocation and
reporting of emissions from both fuels and feedstocks transferred out of the system for use in energy
purposes to the Energy chapter differs slightly from the 2006 IPCC Guidelines. According to the 2006
IPCC Guidelines, emissions from fuel combustion from petrochemical production should be allocated
to this source category within the IPPU chapter. Due to national circumstances, ElAdata on primary fuel
for feedstock use within the energy balance are presented by commodity only, with no resolution on
data by industry sector (i.e., petrochemical production). In addition, under EPA's GHGRP, reporting
facilities began reporting in 2014 on annual feedstock quantities for mass balance and CEMS
methodologies (79 FR 63794), as well as the annual average carbon content of each feedstock (and
molecular weight for gaseous feedstocks) for the mass balance methodology beginning in reporting year
4-78 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
2017 (81 FR 89 2 60).44 The United States is currently unable to report non-energy fuel use from
petrochemical production under the IPPU chapter due to CBI issues. Therefore, consistent with 2006
IPCC Guidelines, fuel consumption data reported by EIA are adjusted to account for these overlaps to
avoid double-counting. More information on the non-energy use of fossil fuel feedstocks for
petrochemical production can be found in Annex 2.3.
Methodological approaches were applied to the entire time series to ensure consistency in emissions
from 1990 through 2023. The methodology for ethylene production, ethylene dichloride production, and
ethylene oxide production spliced activity data from two different sources: ACC for 1990 through 2009
and GHGRP for 2010 through 2023. The methodology for methanol production spliced activity data from
two different sources: ACC for 1990 through 2014 and GHGRP for 2015 through 2023. The methodology
for carbon black production spliced activity data from two different sources: ICBA for 1990 through 2009
and GHGRP for 2010 through 2023. Consistent with the 2006 IPCC Guidelines, the overlap technique
was applied to compare the three data sets for years where there was overlap. For ethylene production
and carbon black production, the data sets were determined to be consistent, and adjustments were
not needed. For ethylene dichloride production, ethylene oxide production, and methanol production,
the data sets were determined to be inconsistent. The GHGRP data includes production of ethylene
dichloride and ethylene oxide as intermediates, while it is unclear if the ACC data does. Methanol
production data from GHGRP are significantly higher than the ACC data for every year since 2015; the
reason for the difference is not clear. Therefore, no adjustments were made to the ethylene dichloride,
ethylene oxide, and methanol activity data for 1990 through 2009 because the 2006 IPCC Guidelines
indicate that it is not good practice to use the overlap technique when the data sets are inconsistent.
Uncertainty
The C02 and CH4 emission factors used for acrylonitrile production are based on a limited number of
studies. Using plant-specific factors instead of default or average factors could increase the accuracy of
the emission estimates; however, such data were not available for the current Inventory report. For
acrylonitrile, EPA assigned an uncertainty range of ±60 percent for the C02 emission factor, ±10 percent
for the CH4 emission factor, and a normal probability density function for both, and using the suggested
uncertainty provided in Table 3.27 of the 2006 IPCC Guidelines is appropriate based on expert judgment,
(RTI 2023). The results of the quantitative uncertainty analysis for the C02 emissions from carbon black
production, ethylene, ethylene dichloride, ethylene oxide, and methanol are based on reported GHGRP
data. Refer to the Methodology section for more details on how these emissions were calculated and
reported to EPA's GHGRP. EPA assigned an uncertainty range of ±5 percent and a normal probability
density function for C02 emissions from carbon black, ethylene, ethylene dichloride, and ethylene oxide
production, and using the suggested uncertainty provided in Table 3.27 of the 2006 IPCC Guidelines is
appropriate based on expert judgment (RTI 2023). There is some uncertainty in the applicability of the
average emission factors for each petrochemical type across all prior years. While petrochemical
production processes in the United States have not changed significantly since 1990, some operational
efficiencies have been implemented at facilities over the time series.
The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 4-58.
Petrochemical production C02 emissions from 2023 were estimated to be between 29.2 and 31.9 MMT
C02 Eq. at the 95 percent confidence level. This indicates a range of approximately 4 percent below to 4
44 See https://www.epa.gov/ghgreporting/historical-rulemakings.
Industrial Processes and Product Use 4-79
-------
percent above the emission estimate of 30.5 MMT C02 Eq. Petrochemical production CH4 emissions
from 2023 were estimated to be between 0.0 and 0.01 MMT C02 Eq. at the 95 percent confidence level.
This indicates a range of approximately 14 percent below to 14 percent above the emission estimate of
0.005 MMT C02 Eq.
Table 4-58: Approach 2 Quantitative Uncertainty Estimates for CH4 Emissions from
Petrochemical Production and C02 Emissions from Petrochemical Production (MMT
C02 Eq. and Percent)
2023 Emission
Uncertainty Range Relative to Emission Estimate"
Source
Gas
Estimate
(MMT CO2 Eq.)
(MMT CO2 Eq.)
(%)
Lower
Upper
Lower
Upper
Bound
Bound
Bound
Bound
Petrochemical
Production
C02
30.5
29.2
31.9
-4%
+4%
Petrochemical
Production
cm
+
0.0
0.01
-14%
+ 14%
+ Does not exceed 0.05 MMT C02 Eq.
a Range of emission estimates predicted by Monte Carlo stochastic simulation for a 95 percent confidence interval.
QA/QC and Verification
For petrochemical production, QA/QC activities were conducted consistent with the U.S. Inventory
QA/QC plan, as described in the QA/QC and Verification Procedures section of the IPPU chapter and
Annex 8. Source-specific quality control measures for this category included the QA/QC requirements
and verification procedures of EPA's GHGRP. More details on the greenhouse gas calculation,
monitoring and QA/QC methods applicable to petrochemical facilities can be found under Subpart X
(Petrochemical Production) of the regulation (40 CFR Part 98).45 EPA verifies annual facility-level GHGRP
reports through a multi-step process (e.g., combination of electronic checks and manual reviews) to
identify potential errors and ensure that data submitted to EPA are accurate, complete, and consistent
(EPA 2015).46 Based on the results of the verification process, EPA follows up with facilities to resolve
mistakes that may have occurred. The post-submittals checks are consistent with a number of general
and category-specific QC procedures, including range checks, statistical checks, algorithm checks, and
year-to-year checks of reported data and emissions. EPA also conducts QA checks of GHGRP reported
production data by petrochemical type against external datasets.
For ethylene, ethylene dichloride and ethylene oxide, it is possible to compare C02 emissions
calculated using the GHGRP data to the C02 emissions that would have been calculated using the Tier 1
approach if GHGRP data were not available. For ethylene, the GHGRP emissions were within ±8 percent
of the emissions calculated using the Tier 1 approach prior to 2018; for 2018 through 2023, the GHGRP
emissions were between 73 percent and 85 percent of what would be calculated using the Tier 1
approach. For ethylene dichloride, the GHGRP emissions are typically higher than the Tier 1 emissions
by up to 25 percent, but in 2010 and 2021, GHGRP emissions were slightly lower than the Tier 1
emissions. For ethylene oxide, GHGRP emissions typically vary from the Tier 1 emissions by up to ±20
percent, but in 2021 through 2023, the GHGRP emissions were significantly higher than the Tier 1
45 See http://www.ecfr.gov/cgi-bin/text-idx7tpN/ecfrbrowse/Title40/40cfr98 main 0?.tpl.
46 See https://www.epa.gov/sites/prodijction/files/9015-07/documents/ghgrp verification factsheet.pdf.
4-80 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
emissions. This was likely due to GHGRP data capturing the production of ethylene oxide at new
facilities as an intermediate in the onsite production of ethylene glycol.
For methanol, GHGRP production data was consistently higher than ACC production data in all years
between 2015 and 2023. Even though the GHGRP production was higher than the ACC production, the
GHGRP C02 emissions estimated are significantly lower than the emissions calculated using the Tier 1
approach in all years between 2015 and 2023. Additionally, there is a trend towards increasing
differences over these years starting with an 873 kt C02 difference in 2015 and increasing to a 2,900 kt
C02 difference in 2022 and 2,800 kt C02 difference in 2023. GHGRP emissions were between 43 percent
and 61 percent of the Tier 1 emissions in 2015 and 2018, respectively. As discussed in the Methodology
and Time-Series Consistency section above, EPA has determined that using the IPCC Tier 1 emissions
factor to calculate methanol emissions results in double counting of natural gas combustion emissions
in both this chapter and in the Energy chapter; therefore, use of the GHGRP derived emissions is
deemed appropriate. For the years 1990 through 2014, the use of the GHGRP derived emission factor
also results in lower emissions than those calculated using the IPCC Tier 1 emission factor. While this
avoids the double counting of emissions with the Energy chapter, as described below in the Planned
Improvements section, EPA intends to examine the emissions from methanol facilities that report to the
GHGRP and may have been operating prior to 2010 to assess whether a more specific process-only
emission factor can be developed from the GHGRP data for use in estimating C02 emissions from
methanol production in 1990 through 2014.
EPA's GHGRP mandates that all petrochemical production facilities report their annual emissions of
C02, CH4, and N20 from each of their petrochemical production processes. Source-specific quality
control measures for the Petrochemical Production category included the QA/QC requirements and
verification procedures of EPA's GHGRP. The QA/QC requirements differ depending on the calculation
methodology used.
As part of a planned improvement effort, EPA has assessed the potential of using GHGRP data to
estimate CH4 emissions from ethylene production. As discussed in the Methodology section above, C02
emissions from ethylene production in this chapter are based on data reported under the GHGRP, and
these emissions are calculated using a Tier 2 approach that assumes all of the carbon in the fuel (i.e.,
ethylene process off-gas) is converted to C02. Ethylene production facilities also calculate and report
CH4 emissions under the GHGRP when they use the optional combustion methodology. The facilities
calculate CH4 emissions from each combustion unit that burns off-gas from an ethylene production
process unit using a Tier 1 approach based on the total quantity of fuel burned, a default or measured
higher heating value, and a default emission factor. Because multiple other types of fuel in addition to
the ethylene process unit off-gas may be burned in these combustion units, the facilities also report an
estimate of the fraction of emissions that is due to burning the ethylene process off-gas component of
the total fuel. Multiplying the total emissions by the estimated fraction provides an estimate of the CH4
emissions from the ethylene production process unit. These ethylene production facilities also
calculate CH4 emissions from flares that burn process vent emissions from ethylene processes. The
C02 emissions are calculated using either a Tier 2 approach based on measured gas volumes and
measured carbon content or higher heating value, or a Tier 1 approach based on the measured gas flow
and a default emission factor; the CH4 emissions are calculated based on a Tier 1 approach using the
C02 emissions and default emission factors. Nearly all ethylene production facilities use the optional
combustion methodology under the GHGRP. The CH4 emissions from ethylene production under the
GHGRP have not been included in this chapter because this approach double counts carbon (i.e., all of
Industrial Processes and Product Use 4-81
-------
the carbon in the CH4 emissions is also included in the C02 emissions from the ethylene process units).
EPA continues to assess the GHGRP data for ways to better disaggregate the data and incorporate it into
the Inventory.
These facilities are also required to report emissions of N20 from combustion of ethylene process off-
gas in both stationary combustion units and flares. Facilities using CEMS (consistent with a Tier 3
approach) are also required to report emissions of CH4 and N20 from combustion of petrochemical
process-off gases in flares. Preliminary analysis of the aggregated reported CH4 and N20 emissions
from facilities using CEMS and N20 emissions from facilities using the optional combustion
methodology suggests that these annual emissions are less than 0.4 percent of total petrochemical
emissions, which is not significant enough to prioritize for inclusion in the report at this time. Pending
resources and significance, EPA may include these N20 emissions in future reports to enhance
completeness. Future QC efforts to validate the use of Tier 1 default emission factors and report on the
comparison of Tier 1 emission estimates and GHGRP data are described below in the Planned
Improvements section.
Recalculations Discussion
No recalculations were performed for the 1990 through 2022 portion of the time series.
Planned Improvements
Improvements include completing category-specific QC of activity data and emission factors, along
with further assessment of CH4 and N20 emissions to enhance completeness in reporting of emissions
from U.S. petrochemical production, pending resources, significance and time-series consistency
considerations. For example, EPA is planning additional assessment of fuel combustion emissions data
reported by methanol production facilities for ways to estimate process-based emissions in the
Inventory separately from combustion emissions for 1990 through 2014. If the GHGRP data can be
categorized by type of methanol process design, it may be possible to use GHGRP data for single
reformer process units to develop a ratio of process-to-total emissions to adjust the IPCC emission
factor. Potential difficulties with this analysis are that some of the methanol producing facilities also
produce other chemicals and the combustion unit names may not clearly identify the process unit to
which they apply, and some combustion unit data may be aggregated for multiple combustion units. The
EPA is also planning additional assessment of ways to use CH4 data from the GHGRP in the Inventory.
One possible approach EPA is assessing would be to adjust the C02 emissions from the GHGRP
downward by subtracting the carbon that is also included in the reported CH4 emissions, per the
discussion in the Petrochemical Production QA/QC and Verification section, above. As of this current
report, timing and resources have not allowed EPA to complete these analyses of activity data,
emissions, and emission factors but they remain priority improvements within the IPPU chapter.
Pending resources, a secondary potential improvement for this source category would focus on
continuing to analyze the fuel and feedstock data from EPA's GHGRP to better disaggregate energy-
related emissions and allocate them more accurately between the Energy and IPPU sectors of the
Inventory. EPA will continue to look for ways to incorporate this data into future Inventories that will
allow for easier data integration between the non-energy uses of fuels category and the petrochemicals
category presented in this chapter. This planned improvement is still under development and has not
been completed to report on progress in this current Inventory.
4-82 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
4.14 HCFC-22 Production (Source Category
2B9a)
This reporting category (2B9a) includes by-product emissions of HCFC-23 (trifluoromethane or CHF3)
from production of HCFC-22 (chlorodifluoromethane). HFC-23 is generated as a byproduct during the
manufacture of HCFC-22, which is primarily employed in refrigeration and air conditioning systems and
as a chemical feedstock for manufacturing synthetic polymers. Between 1990 and 2000, U.S.
production of HCFC-22 increased significantly as HCFC-22 replaced chlorofluorocarbons (CFCs) in
many applications. Between 2000 and 2007, U.S. production fluctuated but generally remained above
1990 levels. In 2008 and 2009, U.S. production declined markedly and has remained near 2009 levels
since. Because HCFC-22 depletes stratospheric ozone, its production for non-feedstock uses was
phased out in 2020 under the U.S. Clean Air Act. Feedstock production, however, is permitted to
continue indefinitely. Per the IPCC methodological guidance, emissions from energy use are currently
accounted for as part of fossil fuel combustion in the industrial end-use sector reported under the
Energy chapter.
HCFC-22 is produced by the reaction of chloroform (CHCl3) and hydrogen fluoride (HF) in the presence
of a catalyst, SbCl5. The reaction of the catalyst and HF produces SbClxFy, (where x + y = 5), which reacts
with chlorinated hydrocarbons to replace chlorine atoms with fluorine. The HF and chloroform are
introduced by submerged piping into a continuous-flow reactor that contains the catalyst in a
hydrocarbon mixture of chloroform and partially fluorinated intermediates. The vapors leaving the
reactor contain HCFC-21 (CHCl2F), HCFC-22 (CHCIF2), HFC-23 (CHF3), HCl, chloroform, and HF. The
under-fluorinated intermediates (HCFC-21) and chloroform are then condensed and returned to the
reactor, along with residual catalyst, to undergo further fluorination. The final vapors leaving the
condenser are primarily HCFC-22, HFC-23, HCl and residual HF. The HCl is recovered as a useful
byproduct, and the HF is removed. Once separated from HCFC-22, the HFC-23 may be released to the
atmosphere, recaptured for use in a limited number of applications, or destroyed.
Two facilities produced HCFC-22 in the United States in 2023. Emissions of HFC-23 from this activity in
2023 were estimated to be 0.39 MMT C02 Eq. (0.03 kt) (see Table 4-59). This quantity represents a 79
percent decrease from 2022 emissions and a 99 percent decrease from 1990 emissions. The decrease
from 1990 emissions was caused primarily by changes in the HFC-23 emission rate (kg HFC-23
emitted/kg HCFC-22 produced). The decrease from 2022 emissions was caused by a large decrease in
the HFC-23 emission rate at one plant and a decrease in the total quantity of HCFC-22 produced. The
long-term decrease in the emission rate is primarily attributable to six factors: (a) five plants that did not
capture and destroy the HFC-23 generated have ceased production of HCFC-22 since 1990; (b) one
plant that captures and destroys the HFC-23 generated began to produce HCFC-22; (c) one plant
implemented and documented a process change that reduced the amount of HFC-23 generated; (d) the
same plant began recovering HFC-23, primarily for destruction and secondarily for sale; (e) another
plant began destroying HFC-23; and (f) the same plant, whose emission rate was higher than that of the
other two plants, ceased production of HCFC-22 in 2013.
Emissions from HCFC-22 production are reported under fluorochemical production (category 2B9) in
this Inventory, which also includes the production of fluorochemicals other than HCFC-22 described
further in section 4.15 of this chapter.
Industrial Processes and Product Use 4-83
-------
Table 4-59: HFC-23 Emissions from HCFC-22 Production (MMT C02 Eq.)
Year 1990
2005
2019 2020 2021
2022
2023
HCFC-22 Production 38.6
16.8
3.1 1.8 2.2
1.8
0.4
Table 4-60: HFC-23 Emissions from HCFC-22 Production (kt HFC-23)
Year 1990
2005
2019 2020 2021
2022
2023
HCFC-22 Production 3
1
+ + +
+
+
+ Does not exceed 0.5 kt.
Methodology and Time-Series Consistency
To estimate HFC-23 emissions for five of the eight HCFC-22 plants that have operated in the United
States since 1990, methods comparable to the Tier 3 methods in the 2006IPCC Guidelines (IPCC 2006)
were used throughout the time series. Emissions for 2010 through 2023 were obtained through reports
submitted by U.S. HCFC-22 production facilities to EPA's Greenhouse Gas Reporting Program (GHGRP).
EPA's GHGRP mandates that all HCFC-22 production facilities report their annual emissions of HFC-23
from HCFC-22 production processes and HFC-23 destruction processes. Previously, data were
obtained by EPA through collaboration with an industry association that received voluntarily reported
HCFC-22 production and HFC-23 emissions annuallyfrom all U.S. HCFC-22 producers from 1990
through 2009. These emissions were aggregated and reported to EPA on an annual basis.
For the other three plants, the last of which closed in 1993, methods comparable to the Tier 1 method in
the 2006 IPCC Guidelines were used. Emissions from these three plants have been calculated using the
recommended emission factor for unoptimized plants operating before 1995 (0.04 kg HCFC-23/kg
HCFC-22 produced).
The five plants that have operated since 1994 measure (or, for the plants that have since closed,
measured) concentrations of HFC-23 as well as mass flow rates of process streams to estimate their
generation of HFC-23. Plants using thermal oxidation to abate their HFC-23 emissions monitor the
performance of their oxidizers to verify that the HFC-23 is almost completely destroyed. One plant that
releases a small fraction of its byproduct HFC-23 periodically measures HFC-23 concentrations at
process vents using gas chromatography. This information is combined with information on quantities of
products (e.g., HCFC-22) to estimate HFC-23 emissions.
To estimate 1990 through 2009 emissions, reports from an industry association were used that
aggregated HCFC-22 production and HFC-23 emissions from all U.S. HCFC-22 producers and reported
them to EPA (ARAP 1997, 1999, 2000, 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008, 2009, and 2010).
To estimate 2010 through 2023 emissions, facility-level data (including both HCFC-22 production and
HFC-23 emissions) reported through EPA's GHGRP were analyzed. In 1997 and 2008, comprehensive
reviews of plant-level estimates of HFC-23 emissions and HCFC-22 production were performed (RTI
1997; RTI 2008). The 1997 and 2008 reviews enabled U.S. totals to be reviewed, updated, and where
necessary, corrected. The reviews also allowed plant-level uncertainty analyses (Monte-Carlo
simulations) to be performed for 1990, 1995, 2000, 2005, and 2006. Estimates of annual U.S. HCFC-22
production are presented in Table 4-61.
4-84 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Table 4-61: HCFC-22 Production (kt)
Year
1990
2005
2012
2019
2020
2021
2022
2023
Production
139
156
96
C
C
C
C
C
C(CBI)
Note: HCFC-22 production in 2013 through 2023 is considered confidential business information (CBI) as there were only two
producers of HCFC-22 in those years.
Uncertainty
The uncertainty analysis presented in this section was based on a plant-level Monte Carlo stochastic
simulation for 2006. The Monte Carlo analysis used estimates of the uncertainties in the individual
variables in each plant's estimating procedure. This analysis was based on the generation of 10,000
random samples of model inputs from the probability density functions for each input. A normal
probability density function was assumed for all measurements and biases except the equipment leak
estimates for one plant; a log-normal probability density function was used for this plant's equipment
leak estimates. The simulation for 2006 yielded a 95-percent confidence interval for U.S. emissions of
6.8 percent below to 9.6 percent above the reported total.
The relative errors yielded by the Monte Carlo stochastic simulation for 2006 were applied to the U.S.
emission estimate for 2023. The resulting estimates of absolute uncertainty are likely to be reasonably
accurate because (1) the methods used by the two remaining plants to estimate their emissions are not
believed to have changed significantly since 2006, and (2) although the distribution of emissions among
the plants has changed between 2006 and 2023 (because one plant has closed), the plant that currently
accounts for most emissions had a relative uncertainty in its 2006 (as well as 2005) emissions estimate
that was similar to the relative uncertainty for total U.S. emissions. Thus, the closure of one plant is not
likely to have a large impact on the uncertainty of the national emission estimate.
The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 4-62. HFC-23
emissions from HCFC-22 production were estimated to be between 0.36 and 0.42 MMT C02 Eq. at the
95 percent confidence level. This indicates a range of approximately 7 percent below and 10 percent
above the emission estimate of 0.39 MMT C02 Eq.
Table 4-62: Approach 2 Quantitative Uncertainty Estimates for HFC-23 Emissions from
HCFC-22 Production (MMT C02 Eq. and Percent)
Uncertainty Range Relative to Emission Estimate"
2023 Emission
(MMT CO2 Eq.)
(%)
Estimate
Lower
Upper
Lower
Upper
Source
Gas
(MMT CO2 Eq.)
Bound
Bound
Bound
Bound
HCFC-22 Production
HFC-23
0.39
0.36
0.42
-7%
+ 10%
a Range of emissions reflects a 95 percent confidence interval.
QA/QC and Verification
General quality assurance/quality control (QA/QC) procedures were applied consistent with the U.S.
Inventory QA/QC plan, which is in accordance with Volume 1, Chapter 6 of 2006IPCC Guidelines as
described in the introduction of the IPPU chapter (see Annex 8 for more details). Under the GHGRP, EPA
verifies annual facility-level reports through a multi-step process (e.g., including a combination of pre-
Industrial Processes and Product Use 4-85
-------
and post-submittal electronic checks and manual reviews by staff) to identify potential errors and
ensure that data submitted to EPA are accurate, complete, and consistent (EPA 20 1 5).47 Based on the
results of the verification process, EPA follows up with facilities to resolve mistakes that may have
occurred. The post-submittals checks are consistent with a number of general and category-specific QC
procedures, including: range checks, statistical checks, algorithm checks, and year-to-year checks of
reported data and emissions.
The GHGRP also requires source-specific quality control measures for the HCFC-22 Production
category. Under EPA's GHGRP, HCFC-22 producers are required to (1) measure concentrations of HFC-
23 and HCFC-22 in the product stream at least weekly using equipment and methods (e.g., gas
chromatography) with an accuracy and precision of 5 percent or better at the concentrations of the
process samples, (2) measure mass flows of HFC-23 and HCFC-22 at least weekly using measurement
devices (e.g., flowmeters) with an accuracy and precision of 1 percent of full scale or better, (3) calibrate
mass measurement devices at the frequency recommended by the manufacturer using traceable
standards and suitable methods published by a consensus standards organization, (4) calibrate gas
chromatographs at least monthly through analysis of certified standards, and (5) document these
calibrations.
Recalculations Discussion
No recalculations were performed for the 1990 to 2022 portion of the time series.
Planned Improvements
At this time, there are no specific planned improvements for estimating HFC-23 emissions from HCFC-
22 production.
4.15 Production of Fluorochemicals Other
Than HCFC-22 (Source Category 2B9b)
In this reporting category, fluorochemical production (2B9b), facilities in the United States produced or
transformed approximately 200 fluorinated gases other than HCFC-22 in 2023, including saturated and
unsaturated hydrofluorocarbons (HFCs), saturated and unsaturated perfluorocarbons (PFCs), sulfur
hexafluoride (SF6), nitrogen trifluoride (NF3), hydrofluoroethers (HFEs), perfluoroalkylamines, and
dozens of others. Emissions from fluorochemical production may include emissions of the intentionally
manufactured chemical as well as reactant and by-product emissions. The compounds emitted depend
upon the production or transformation process, but may include, e.g., HFCs, PFCs, SF6, nitrous oxide
(N20), NF3, and many others. Potential sources of fluorinated GHG emissions at fluorochemical
production facilities include process vents, equipment leaks, and evacuating returned containers48
Production-related emissions of fluorinated GHGs occur from both process vents and equipment leaks.
47 EPA (2015). Greenhouse Gas Reporting Program Report Verification. Available online at:
https://www.epa.gov/sites/production/files/9015-07/documents/ghgrp verification factsheet.pdf.
48 The totals presented below also include emissions from destruction of previously produced fluorinated GHGs that are
shipped to production facilities for destruction, e.g., because they are found to be irretrievably contaminated.
4-86 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Process vent emissions occur from manufacturing equipment such as reactors, distillation columns,
and packaging equipment. Equipment leak emissions, or fugitive emissions, occur from valves, flanges,
pump seals, compressor seals, pressure relief valves, connectors, open-ended lines, and sampling
connections. In addition, users of fluorinated GHGs may return empty containers (e.g., cylinders) to the
production facility for reuse; prior to reuse, the residual fluorinated GHGs (often termed "heels") may be
evacuated from the container and are a potential emission source. In many cases, these "heels" are
recovered or exhausted to a treatment device for destruction. In other cases, however, they are released
into the atmosphere.49
Emissions of all HFCs, PFCs, NF3, and SF6 from production of fluorochemicals other than
hydrochlorofluorocarbon (HCFC)-22 are presented in Table 4-63 below for the years 1990, 2005, and the
period 2019 to 2023. Per the IPCC methodological guidance, emissions from energy use are currently
accounted for as part of fossil fuel combustion in the industrial end-use sector reported under the
Energy chapter.
The fluorinated GHG emissions reported under the Greenhouse Gas Reporting Program (GHGRP)
include emissions of HFCs, PFCs, SF6, NF3, and numerous "other" fluorinated GHGs, such as
octafluorotetrahydrofuran (C4F80), trifluoromethyl sulfur pentafluoride (SF5CF3), and
hexafluoropropylene oxide. Because they are not included among the seven reportable gases or gas
groups, the "other" fluorinated GHGs are not included in Inventory totals. However, their emissions are
presented below because they often have high GWPs and large GWP-weighted emissions.
Total emissions of HFCs, PFCs, SF6, and NF3 from fluorochemical production are estimated to have
increased from 32 MMT C02 Eq. in 1990 to a peak of 45 MMT C02 Eq. in 1999, declined to 3.9 MMT C02
Eq. in 201650, and fluctuated between 3.9 and 6.2 MMTC02 Eq. thereafter, reaching4.3 MMTC02 Eq. in
2023. These trends reflect estimated changes in fluorinated gas production and increasing use of
control devices. Prior to 2000, only 2 facilities are known to have operated control devices to destroy
fluorinated GHG emissions. After 2000, additional production facilities began to install and use control
devices to destroy fluorinated GHG emissions,51 and fluorinated GHG emissions declined sharplyfrom
45 MMT C02 Eq. in 1999 to 13 MMT C02 Eq. in 2005. Emissions continued to fall more slowly through
2016, reflecting the installation of controls at an additional 4 facilities in 2011, 2012, 2015, and 2016.
Total fluorinated GHG emissions fluctuated from 2017 to 2022, and total fluorinated GHG emissions
declined in 2023 as some high-emitting facilities reduced both production and emission rates.
Emissions from the production of fluorochemicals other than HCFC-22 are reported under
fluorochemical production (category 2B9) in conjunction with emissions from HCFC-22 production
described in Section 4.14 of this chapter.
HFC Emissions
Estimated emissions of saturated HFCs increased from 8.7 MMT C02 Eq. in 1990 to a peak of 14 MMT
C02 Eq. in 1999, declining with some fluctuation to 1.3 MMT C02 Eq. in 2023. Emissions in 1990 were
49 IPCC (2019) 2019 Refinement to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories, Calvo Buendia, E.,
Tanabe, K., Kranjc, A., Baasansuren, J., Fukuda, M., Ngarize S., Osako, A., Pyrozhenko, Y., Shermanau, P. and Federici, S.
(eds). Published: IPCC, Switzerland.
50 Emissions in MMT CO2 Eq. were similar in 2017, but the 2017 emissions in MT were considerably higher (4,500 MT) due
to anomalously high emissions of one low-GWP, unsaturated HFC at one facility.
51 One facility is assumed to have installed controls in 2000, another installed controls in 2003, and three facilities are
assumed to have installed controls in 2005.
Industrial Processes and Product Use 4-87
-------
primarily from facilities producing compounds other than saturated HFCs. The subsequent trends in
emissions were driven by the growth in production of saturated HFCs and the implementation of
controls. Production of saturated HFCs is estimated to have increased from around 0.3 MMT C02 Eq. in
1990 to over 300 MMT C02 Eq. by 2010 as HFCs replaced ozone-depleting substances. This increase in
HFC production drove HFC emissions to their 1999 peak. However, estimated emissions declined
significantly from 1999 to 2005 due to the assumed addition of controls in 2000 and subsequent years.
Estimated emissions of HFCs resumed their increase from 2005 to 2010, reaching 7.3 MMT C02 Eq., but
again declined sharply in 2011 to 4.6 MMT C02 Eq. based on addition of controls. Since 2012, HFC
emissions have continued to trend downward with some fluctuations, hitting a minimum of 1.3 MMT
C02 Eq. in 2023.
PFC Emissions
Emissions of PFCs increased gradually from 18 MMT C02 Eq. in 1990 to 24 MMT C02 Eq. in 1999 but
dropped to 4.1 MMT C02 Eq. by 2005, reflecting the addition of controls at high-emitting facilities and
apparent changes to the mix of products produced at another facility.52 Overall PFC emissions from
2005 to 2023 have remained relatively steady, oscillating around 2.5 MMT C02 Eq. The upward trend
between 1990 and 1999 was largely driven by the three facilities that reported their historical emissions
to the EPA. In the absence of historical emissions data for other facilities, the quantities of fluorinated
GHGs produced or transformed at other facilities emitting PFCs are estimated to have remained
generally steady between 1990 and 2009 and therefore do not contribute to the emissions trend before
2010. For most of the fluorinated GHGs produced at these facilities, there was no available industry
information to inform activity estimates or trends for 1990 to 2009. Therefore, as discussed in the
Methodology section below, 2010 production values from EPA's GHGRP were assumed to have held
constant for these compounds from 1990 to 2010.
SFe Emissions
Emissions of SF6 are estimated to have risen gradually from 5.8 MMT C02 Eq. in 1990 to a peak of 7.5
MMT C02 Eq. in 1995, to have declined slowly to 7.0 MMT C02 Eq. in 2000, and then to have declined
more rapidly to a minimum of 0.0002 MMT C02 Eq. in 2017, after which emissions rose and fluctuated
between 0.0057 MMT C02 Eq. (in 2020) and 0.0027 MMT C02 Eq. (through 2023). The rapid emissions
decline after 2000 was driven first by the imposition of controls at one facility and then by the cessation
of production in 2010 at a major U.S. SF6-producing facility.
52 In a summary of 1990 through 2010 emissions submitted to EPA (described more below), 3M, which owns several
facilities that historically emitted PFCs, noted that the mix of products produced at its various facilities had changed
over time, leading to changes in the magnitude and contents of emissions. This change in magnitude and contents was
particularly pronounced at 3M's Decatur facility (referred to elsewhere in this document as "3M Company"), where
emissions declined from 15.8 MMT CO2 Eq. in 2000 to 0.53 MMT CO2 Eq. in 2002, and where the contents of emissions
changed from HFCs, PFCs, SFe and other fluorinated GHGs in 2000 to PFCs and other fluorinated GHGs in 2003.
(Emissions in 2002 were not differentiated by group). Emissions were also reduced afterthe installation of a control
device at the Cordova facility. 3M noted that Initial start-up of the thermal oxidizer occurred in 2003, but that it took time
to optimize the operation of the thermal oxidizer and treatment of the various gas streams, leading to a decrease in
emissions over several years.
4-88 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
NF3 Emissions
Since 1990, estimated emissions of NF3 have fluctuated between 0.14 MMT C02 Eq. and 0.72 MMT C02
Eq., with peaks occurring in 2002 (0.50 MMTC02 Eq.), 2010 (0.70 MMT C02 Eq.), and 2020 (0.72 MMT
C02 Eq.), and lows occurring in 1990 (0.14 MMTC02 Eq.), 2003 (0.33 MMTC02 Eq.), and 2018 (0.11 MMT
C02 Eq.). NF3 may be emitted both from the production of NF3 and from the production of other
fluorochemicals. The dominant source since 2010 has been production of NF3. Trends after 2010 were
driven by changes both in NF3 production and in the emission rate (kg NF3 emitted/kg NF3 produced) for
NF3 production, with both contributing to increased emissions since 2018. For 1990 through 2009, the
NF3 that is emitted from the production of NF3 is assumed to be influenced by the trajectory of NF3
production, which is generally assumed to follow production trends in the semiconductor industry
except where NF3 facility capacity limits production further. Semiconductor production increased from
1995 to 2007 but is estimated to have declined in 2008 and 2009. As described in the Methodology
section under "Estimated Emissions for 3M facilities," the NF3 that is emitted from production of other
fluorochemicals is assumed to have been emitted as a constant fraction of the "other" fluorinated
GHGs whose 1990 through 2010 emissions were reported by 3M facilities. This fraction was estimated
based on the fraction of "other" fluorinated GHG emissions accounted for by NF3 between 2011 and
2015 and is highly uncertain. Nevertheless, because the highest-emitting 3M facilities reported
decreasing emissions of all other fluorinated GHG groups between 2000 and 2005 (due to the
installation of a control device at one facility and apparent production changes at another), NF3
emissions also appear likely to have decreased during this period.
Other Fluorinated GHG Emissions
Other fluorinated GHGs, i.e., those not included in the reportable gases or gas groups, are also emitted
in significant quantities from fluorinated gas production and transformation processes. Estimated
emissions of these other fluorinated GHGs are provided in Table 4-65 for the years 1990, 2005, and the
period 2019 to 2023. The other fluorinated GHGs with the highest estimated emissions in 2023 are
presented separately, and the remaining other fluorinated GHGs are aggregated.
Total emissions of other fluorinated GHGs increased from 4.9 MMT C02 Eq. in 1990 to a peak of 10.4
MMT C02 in 2000, declining rapidly to 0.87 MMT C02 Eq. in 2009 and then declining more slowly to 0.13
MMT C02 Eq. in 2020 through 2022. Emissions in 2023 were 0.14 MMTC02 Eq.. Between 1990 and 2009,
estimated emissions of other fluorinated GHGs were primarily driven by the emissions reported by 3M
facilities, which showed significant declines between 2000 and 2005, reflecting apparent production
changes at one facility and the installation of a control device at another. The decline in emissions from
2019 to 2020 was due to a decrease in the emission rate at one facility.
Table 4-63: Emissions of HFCs, PFCs, SF6, and NF3 from Production of
Fluorochemicals Other Than HCFC-22 (MMT C02 Eq.)
Gas
1990
2005
2019
2020
2021
2022
2023
HFC-23
6.8
1.7
1.1
0.9
0.7
1.0
0.7
HFC-143a
0.2
0.8
0.6
0.3
0.3
0.3
0.2
HFC-134a
+
0.4
0.3
0.2
0.2
0.3
0.2
HFC-125
0.1
1.9
0.4
0.4
0.4
0.3
0.2
HFC-32
+
0.1
0.1
0.1
0.1
0.1
+
Industrial Processes and Product Use 4-89
-------
Gas
1990
2005
2019
2020
2021
2022
2023
HFC-227ea
1.5
0.1
0.1
0.1
0.1
0.1
+
Other HFCs
0.1
0.3
0.1
0.1
0.1
0.5
+
Perfluorocyclobutane
11
0.7
1.4
1.1
1.2
1.2
1.3
PFC-14 (Perfluoromethane)
3.0
1.4
0.9
0.9
0.9
1.0
1.0
Other PFCs
3.5
2.0
0.7
0.4
0.4
0.6
0.5
Nitrogen trifluoride
0.1
0.6
0.6
0.7
0.5
0.5
0.3
Sulfur hexafluoride
5.8
3.3
+
+
+
+
+
Total
32.4
13.3
6.2
5.2
4.8
5.8
4.3
+ Does not exceed 0.05 MMT C02 Eq.
Table 4-64: Emissions of HFCs, PFCs, SF6, and NF3 from Production of
Fluorochemicals Other Than HCFC-22 (Metric Tons)
Gas
1990
2005
2019
2020
2021
2022
2023
HFC-23
550
140
90
72
56
77
55
HFC-143a
32
170
120
67
53
60
41
HFC-134a
37
330
220
180
180
190
120
HFC-125
35
600
130
120
110
110
49
HFC-32
22
100
110
93
100
99
56
HFC-227ea
460
42
25
33
26
23
5.2
Other HFCsa
120,000
810
340
460
360
540
200
Perfluorocyclobutane
1,200
70
150
120
130
120
130
PFC-14 (Perfluoromethane)
450
220
130
140
140
160
140
Other PFCs
370
210
79
47
47
62
64
Nitrogen trifluoride
8.7
36
35
45
31
31
19
Sulfur hexafluoride
250
140
+
+
+
+
+
Total
120,000
2,900
1,400
1,400
1,200
1,500
880
aThe metric ton total for HFCs is highly uncertain because, as described further below in the Methodology section, it is ultimately
based on assumptions regarding the chemical identity of emissions that were reported after 2011 only in metric tons of C02
Eq. by fluorinated GHG group. The metric ton total is very sensitive to the GWP used to convert the C02 Eq. emissions to metric
tons, and the GWPs of the unsaturated compounds span a factor of 6000.
Note: Totals may not sum due to independent rounding.
Table 4-65: Emissions of Other Fluorinated GHGs from Production of Fluorochemicals
Other Than HCFC-22 (MMT C02 Eq.)
Gas
1990
2005
2019
2020
2021
2022
2023
1,1,1,2,2,3,3-Heptafluoro-3-(1,2,2,2-
tetrafluoroethoxy)-propane
+
+
+
+
+
+
+
Hexafluoropropylene oxide
0.4
0.4
0.3
+
+
+
+
Octafluorotetrahydrofuran
1.0
1.9
0.1
+
+
+
+
Trifluoromethyl sulfur pentasulfide
pentafluoride
0.5
0.9
0.1
+
+
+
+
HFE-449sl, (HFE-7100) Isomer blend
+
+
+
+
+
+
+
Others
3.1
0.5
0.1
0.1
+
+
+
Total Other Fluorinated GHGs
4.9
3.7
0.6
0.1
0.1
0.1
0.1
+ Does not exceed 0.05 MMT C02 Eq.
Note: Totals may not sum due to independent rounding.
4-90 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Table 4-66: Emissions of Other Fluorinated GHGs from Production of Fluorochemicals
Other Than HCFC-22 (Metric Tons)
Gas
1990
2005
2019
2020
2021
2022
2023
1,1,1,2,2,3,3-Heptafluoro-3-(1,2,2,2-
tetrafluoroethoxy)-propane
6.1
3.9
5.8
3.2
5.9
3.3
6.8
Hexafluoropropylene oxide
35
35
31
1.7
1.8
1.8
1.8
Octafluorotetrahydrofuran
69
140
3.9
0.92
0.64
1.5
1.2
Trifluoromethyl sulfur pentasulfide
27
54
3.7
0.86
0.83
1.1
0.50
HFE-449sl, (HFE-7100) Isomer blend
2.3
14
35
21
23
28
24
Others
1,500
110
54
18
18
13
12
Total Other Fluorinated GHGs
1,600
360
130
46
50
49
46
Note: Totals may not sum due to independent rounding.
Table 4-67: Production and Transformation of Fluorinated GHGs (kt)a
Set of Facilities
1990
2005
2019
2020
2021
2022
2023
Facilities reporting their F-GHG emissions,
production, and transformation to GHGRP
after 2010b
86
271
371
352
348
370
341
Facilities reporting only their F-GHG
production and transformation to GHGRP
after 2010
3.3
3.3
9.7
8.2
7.5
11
9.1
Total Production and Transformation
89
274
381
360
356
381
350
a Totals are presented across species to protect confidential business information.
b Includes 1 facility that reported production, but not emissions, of SF6 through 2010.
Note: Tables may not sum due to independent rounding
Methodology
The 2006IPCC Guidelines as elaborated by the 2079 Refinement include Tier 1, Tier 2, and Tier 3
methods for estimating fluorinated GHG emissions from production of fluorinated compounds. The Tier
1 method calculates emissions by multiplying a default emission factor by total production. Specific
default emission factors exist for production of SF6 and NF3; a more general default emission factor
covers production of all other fluorinated GHGs. (The more general default emission factor was
developed based on data from U.S. facilities collected under the GHGRP between 2011 and 2016.) The
Tier 2 method calculates emissions using a mass-balance approach. The Tier 3 method is based on the
collection of plant-specific data on the types and quantities of fluorinated GHGs emitted from process
vents, leaks, container venting, and other sources, considering any abatement technology. The Tier 3
method is often implemented by developing and applying facility-specific emission factors indexed to
production.
Based on available data on emissions and activity, EPA used a form of the IPCC Tier 3 method to
estimate fluorinated GHG emissions from most U.S. production of fluorinated compounds. Emissions
from U.S. production for which there are fewer data are based on the Tier 1 method.
As discussed further in Annex 3.9, much of the data used to develop the estimates presented here come
from the GHGRP. The data were collected under two sections of the GHGRP regulation—Subpart L,
Fluorinated Gas Production; and Subpart OO, Suppliers of Industrial Greenhouse Gases. Under Subpart
L, certain fluorinated gas production facilities must report their emissions from a range of processes and
Industrial Processes and Product Use 4-91
-------
sources. Data collected under Subpart L include emissions data for calendar years 2011 through 2023.
Under Subpart OO, fluorinated GHG suppliers (including fluorinated GHG producers) must report the
quantities of each fluorinated GHG that they produce, transform, destroy, import, or export. Data
collected under Subpart OO include production and transformation data for calendar years 2010
through 2023. Facilities' individual production and transformation data are not shown here because they
are considered confidential business information under the GHGRP.
1990-2010 Emissions Estimates
For 14 of the 17 fluorinated gas production facilities that have reported their emissions under the
GHGRP, 1990 through 2010 emissions are estimated using (1) facility- and chemical-specific emission
factors based on the emissions data discussed under "2011-2023 Emissions" below, (2) reported or
estimated production and transformation of fluorinated GHGs at each facility in each year, i.e., activity
data, and (3) reported and estimated levels of emissions control at each facility in each year. (For the
other 3 fluorinated gas production facilities that have reported their emissions under the GHGRP, 1990
through 2010 emissions were estimated using data submitted by the company, as explained further
below.)
Facility- and Chemical-Specific Emission Factors Reflecting Emissions Controls
Facility- and chemical-specific emission factors were developed based on the 2011 to 2015 emissions
reported under the GHGRP and the 2011 to 2015 production and transformation of fluorinated GHGs
reported under the GHGRP. (Production and transformation of CFCs and HCFCs are not reported under
the GHGRP.) For each emitted fluorinated GHG at each facility, emissions of the fluorinated GHG were
summed over the five-year period. This sum was then divided by the sum of the quantities of all
fluorinated GHGs produced or transformed at the facility over the five-year period.53
Facility- and Chemical-Specific Emission Factors Reflecting No Emissions Controls
The 2011 to 2015 emissions reported under the GHGRP reflect emissions controls to the extent those
are implemented at each facility. Because facilities have not always controlled their fluorinated GHG
emissions since 1990, uncontrolled emission factors were developed for each facility to apply to years
when the facility's emissions were not believed to be controlled. To estimate uncontrolled emissions,
GHGRP data were first used to assess the 2011 to 2015 levels of control for each production or
transformation process at each facility as described in Annex 3.9. Then, information from the GHGRP
and other sources was used to determine whether and when emissions from facilities were likely to
have been controlled from 1990 to 2010. For the estimated status of emissions controls at each facility
reporting under Subpart L, and, where relevant, the starting year for those controls, see Table A-108 in
Annex 3.9.
Activity Data
The activity data for production and transformation of fluorinated compounds for 1990 to 2010 are
based on production and transformation data reported to EPA by certain facilities for certain years, on
53 Permit data for two facilities indicated that they began controlling emissions at some point between 2011 and 2015.
However, the actual emissions reported by these facilities did not change substantially after the date when the permit
indicated that controls were imposed. For this reason, the reported 2011 to 2015 emissions and emission factors are
believed to be representative of emissions for these facilities before 2011.
4-92 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
production capacity data, and on fluorinated GHG production and consumption trends estimated for
the various fluorinated GHG-consuming industries.
Production and Production Capacity Data
Production data are available from reporting to the U.S. GHGRP under Subpart OO, Suppliers of
Industrial Greenhouse Gases, and from an industry survey conducted by U.S. EPA in 2008 and 2009.
Production and transformation data were reported under Subpart OO for 2010 and later years. The
responses to the industry survey included production data for certain fluorinated gases at certain
facilities for the years 2004, 2005, and 2006. 2004 to 2006 production data are available for 15
fluorinated compounds. Year 2006 production at an SF6-producing facility was estimated based on
production capacity data as described in Annex 3.9 (Rand 2007). Production of certain compounds at
one other facility was estimated based on 2003 production capacity estimates from SRI 2004.
Estimated Production
Estimated production for facilities and fluorinated GHGs for which production or production capacity
data were available for some years before 2010.
For facilities and fluorinated GHGs for which production or production capacity data were available for
2006 or 2003, production between 2006 or 2003 (as applicable) and 2010 (or 2011) was estimated by
interpolating between the 2006 production or 2003 production capacity value and the 2010 (or 2011)
production value reported under Subpart OO. To account for production occurring in some years but not
others, production for 2009 was estimated to be the average production for 2010 to 2015.
For the years before the earliest year with production or production capacity data (e.g., years 1990 to
2002 or 2003), production was estimated based on growth or consumption trends for the major
industries using each fluorinated GHG as described in Annex 3.9.
Estimated Production for Facilities and Fluorinated GHGs for which Production Data
before 2010 were Not Available
In the absence of production data for the period 1990 to 2009, the production data reported to the
GHGRP under Subpart OOwere extrapolated backward based on the industry trends discussed above.
For compounds for which industry trend data were unavailable, production was assumed to have
remained constant over the time series.
In both cases, 2009 production was estimated by conducting a trend analysis on the Subpart OO
production data for years 2010 to 2015. In instances where there did not appear to be a trend, the
average of the production values for years 2010 to 2015 was used as the estimated production for year
2009. In instances where there was a trend, the year 2010 (or 2011) production value was used as the
estimated production for year 2009.
If the industry trend information discussed above was applicable to a fluorinated compound, it was
assumed that production varied with the industry trend from 1990 to 2009. If no industry trend
information was available, it was assumed that production from 1990 to 2008 remained constant at the
2009 value.
For facilities and fluorinated compounds where information was available on annual production
capacity, the estimated activity data were reviewed and compared to the known production capacity.
Industrial Processes and Product Use 4-93
-------
For instances where the estimated activity data exceeded known production capacity for a certain year,
the production estimate was set equal to the capacity value. In addition, where information was
available on the starting year for production of a fluorinated GHG at a facility, production was only
estimated beginning in the process startup year through 2009.
Estimated Emissions for 3M Facilities
3M provided 1990,1995, 2000, and 2002 through 2010 emissions data for three facilities: 3M Cordova,
3M Company, and 3M Cottage Grove Center - Site.54 Therefore, speciated 1990-2010 emissions at these
facilities were estimated using a different methodology than that described above.55
3M emissions data were provided by facility and by fluorinated GHG group in metric tons of C02 Eq.,
weighted by 100-year GWPs from various IPCC Assessment Reports. As detailed in Annex 3.9, EPA
disaggregated the data provided by 3M to present emissions estimates by compound for 1990,1995,
2000, and subsequent years. EPA assumed that emissions of each fluorinated GHG group before 2011
consisted of the same fluorinated GHGs, in the same proportions, as from 2011 through 2015. EPA then
used linear interpolation to estimate emissions for 1991 to 1994, 1996 to 1999, and 2001 for each
compound for these three facilities.
Estimated Emissions for Facilities that Produce Fluorinated GHGs but Do Not Report
Under Subpart L
There is a subset of facilities that report production and transformation of fluorinated gases under
Subpart OO and that also have emission levels less than the threshold value for reporting under Subpart
L (i.e., uncontrolled emissions below the 25,000-MT C02 Eq. threshold). For these facilities, EPA
developed emission estimates based on aggregated production estimates and the Tier 1 default
emission factor in the 2019 Refinement. Because the specific fluorinated GHGs emitted are not known,
the emissions were assumed to consist of the fluorinated GHGs shown in Table 3.28b of chapter 3.10.2
ofVolume 3 IPPU (IPCC 2019), in the proportions shown in that table. Emissions are assumed to have
been flat at the 2010 value in the years before 2010.
Estimated Emissions for SFe Production Facility
For an SF6 production facility that ceased production in 2010, the year before emissions from fluorinated
gas production were required to be reported under the GHGRP, SF6 emissions were estimated using
historical production capacity, the global growth rate of SF6 sales reported in RAND 2007, and the Tier 1
default emission factor for production of SF6 in the 2019 Refinement. For this plant, a 1982 SF6
production capacity of 1,200 short tons (Perkins 1982) was multiplied by the ratio between the RAND
survey SF6 sales totals for 2006 and 1982,1.52 (RAND 2007), resulting in estimated production of 1,652
metric tons in 2006. This production was assumed to have declined linearly to zero in 2011.
54 For 1990, 1995, and 2000, 3M provided emissions data for a Pilot Development Center in addition to the other three
facilities. Emissions by group from the Pilot Development Center were added to and are represented by the emissions by
group for 3M Cottage Grove Center-Site.
55 3M's methods for estimating its emissions are described in detail in "3M Global EHS Laboratory Response to EPA Data
Request on Fluorochemical Emissions," February 2024 (3M, 2024). In brief, 3M estimated emissions from its processes
using emission factors that were developed using methods similar to those used for developing emission factors under
the GHGRP. As under the GHGRP, emission factors were multiplied by different types of activity data (e.g., production) to
estimate emissions for each facility and year. In 2003 and lateryears, 3M also accounted for emission reductions
attributable to operation of the thermal oxidizer at the Cordova plant.
4-94 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
2011-2023 Emissions Estimates
For the 17 fluorinated gas production facilities that have reported their emissions under the GHGRP,
2011 to 2023 emissions are estimated using the fluorinated GHG emissions reported under Subpart L of
the GHGRP.
As discussed above, most emissions reported under Subpart L are reported by chemical, but some
emissions are reported only by fluorinated GHG group in MT C02 Eq. Between 2011 and 2023, the share
of total C02 Eq. emissions reported only by fluorinated GHG group has ranged between 1 and 2 percent.
In this analysis, to ensure that all emissions are reported by species, emissions that are reported only by
fluorinated GHG group are assumed to consist of the fluorinated GHGs in that group that are reported by
chemical at the facility in that year. When no fluorinated GHGs in the group are reported by chemical by
that facility in that year, the emissions are assumed to consist of fluorinated GHGs in that group
reported in other years at that facility. If no fluorinated GHGs in that group were ever reported by
chemical by the facility, the emissions are assumed to consist of fluorinated GHGs in that group
reported across the industry for that year. Because 3M facilities emitted many more individual
compounds than the rest of the industry, fluorinated GHG groups at non-3M facilities were assumed to
consist of fluorinated GHGs in groups as reported at other non-3M facilities. In each of these scenarios,
fractions of gases emitted in MT C02 Eq from each fluorinated GHG group were established and applied
to the total MT C02 Eq. emitted from a fluorinated GHG group to calculate emissions in MT C02 Eq of
each individual fluorinated GHG. As discussed further in the Uncertainty section, this is likely to result in
incorrect speciation of some emissions, but the impact of this incorrect speciation is expected to be
small.
Estimated Emissions for Facilities that Produce Fluorinated GHGs but Do Not Report
Under Subpart L
As discussed above, for facilities that produce fluorinated GHGs but that do not report their emissions
under Subpart L, EPA developed emission estimates based on aggregated production estimates and the
Tier 1 default emission factor in the 2019 Refinement. Because the specific fluorinated GHGs emitted
are not known, the emissions were assumed to consist of the fluorinated GHGs shown in Table 3.28b of
chapter 3.10.2 of Volume 3 IPPU (IPCC 2019), in the proportions shown in that table.
Uncertainty
The estimates shown here are subject to a number of uncertainties. These uncertainties are generally
greater for years before 2011, when reporting of fluorinated GHG emissions from fluorinated gas
production began under the GHGRP, than for 2011 and following years. However, the emissions
estimated from 2011 to 2023 are also subject to various uncertainties. Important sources of uncertainty
in the 2010 through 2023 estimates include uncertainties regarding the identity of processes that emit
particular fluorinated GHGs, process vent emission factors, equipment leak estimates, the quantities of
residual gas vented from containers, and emissions from facilities that produce fluorinated gases but do
not report their emissions to the GHGRP. Important sources of uncertainty in the 1990 through 2010
estimates include many of the uncertainties that affect the 2010 through 2023 estimates as well as
uncertainties regarding changes in the set of gases produced and emitted over time, the quantities of
gases produced before 2010, and the magnitudes and trends of the facility-specific emission factors,
Industrial Processes and Product Use 4-95
-------
which vary based on the compounds produced and transformed and the level of control at the facility.
See Annex 3.9 for a more detailed discussion of the uncertainties in the estimates.
The uncertainties in process vent emission factors (±35% for emission factors based on emissions
testing; ±50% for emission factors based on engineering calculations), equipment leak estimates
(±90%), the quantities of residual gas vented from containers (±30%), and emissions from facilities that
produce fluorinated gases but do not report their emissions to the GHGRP (±98%) were convolved using
error propagation to arrive at an overall uncertainty estimate for 2023. The results of the Approach 1
quantitative uncertainty analysis are summarized in Table 4-68. Emissions of HFCs, PFCs, SF6, and NF3
from production of fluorochemicals other than HCFC-22 were estimated to fall between 3.40 and 5.15
MMT C02 Eq. at the 95 percent confidence level. This indicates a range of approximately 20 percent
below and 20 percent above the emission estimate of 4.27 MMT C02 Eq.
Table 4-68: Approach 1 Quantitative Uncertainty Estimates for HFC, PFC, SF6, and NF3
from Production of Fluorochemicals other than HCFC-22 (MMT C02 Eq. and Percent)
Source
Gas
2023 Emission
Estimate
(MMT CO2 Eq.)
Uncertainty Range Relative to Emission Estimate
(MMT CO2 Eq.) (%)
Lower
Bound"
Upper
Bound"
Lower
Bound
Upper
Bound
Production of
Fluorochemicals
other than HCFC-22
HFCs, PFCs,
SFe, and NF3
4.27
3.40
5.15
-20%
+20%
"Absolute lower and upper bounds were calculated using the corresponding lower and upper bounds in percentages.
QA/QC and Verification
General quality assurance/quality control (QA/QC) procedures were applied consistent with the U.S.
Inventory QA/QC plan, which is in accordance with Volume 1, Chapter 6 of 2006IPCC Guidelines as
described in the introduction of the IPPU chapter (see Annex 8 for more details). Under the GHGRP, EPA
verifies annual facility-level reports through a multi-step process (e.g., including a combination of pre-
and post-submittal electronic checks and manual reviews by staff) to identify potential errors and
ensure that data submitted to EPA are accurate, complete, and consistent (EPA 20 1 5).56 Based on the
results of the verification process, EPA follows up with facilities to resolve mistakes that may have
occurred. The post-submittals checks are consistent with a number of general and category-specific QC
procedures, including: range checks, statistical checks, algorithm checks, and year-to-year checks of
reported data and emissions.
The GHGRP also requires source-specific quality control measures for the Fluorinated Gas Production
category. Under the GHGRP, fluorinated gas producers are required to (1) develop and periodically
update process vent-specific emission factors using either measurements or engineering calculations,
depending on the nature of the process (continuous vs. batch) and the magnitude of emissions from the
vent, (2) take more measurements of vent emissions where variability is high, (3) use methods for
sampling, measuring volumetric flow rates, non-fluorinated-GHG gas analysis, and measuring stack gas
moisture that have been validated using a scientifically sound validation protocol, (4) use a quality-
assured analytical measurement technology capable of detecting the analyte of interest at the
56 EPA (2015). Greenhouse Gas Reporting Program Report Verification. Available online at:
https://www.epa.gov/sites/production/files/9015-07/documents/ghgrp verification factsheet.pdf.
4-96 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
concentration of interest and use a sampling and analytical procedure validated with the analyte of
interest at the concentration of interest, (5) periodically test the performance of destruction devices
used to control emissions, (6) account for any malfunctions in the process or destruction device, (7)
account for emissions from equipment leaks, (8) measure the quantities of residual gas that are vented
from returned containers (or develop an emission factor based on at least 30 measurements per gas
and container size and type), (9) calibrate mass measurement devices at the frequency recommended
by the manufacturer using traceable standards and suitable methods published by a consensus
standards organization, (10) calibrate analytical equipment used to determine the concentration of
fluorinated GHGs, and (11) document all measurements and calibrations.
The 1990, 1995, 2000, and 2002 through 2010 emissions data reported by 3M for three facilities was
compared to the 1990 through 2010 emissions previously calculated for those facilities using the same
calculation method used for other facilities that have reported their emissions under the GHGRP since
2011. The overall trajectory of the 3M-reported emissions, as well as the minima and maxima of those
emissions, were similar to those previously calculated, but the increases and decreases in the 3M-
reported emissions were more gradual. 3M explained that the gradual changes were due to changes in
the compounds and quantities produced and to the gradual deployment and optimization of the
destruction device at the 3M Cordova facility.
Recalculations
Recalculations were performed on the fluorinated GHG emissions that are reported only by fluorinated
GHG group from production and transformation processes over the full time series. The recalculations
corrected (1) an error that led to the inadvertent exclusion of certain fluorinated GHG groups from the
totals, (2) double-counting of production of one compound at one facility for years before 2010, (3)
under-counting of production of a few other compounds for years before 2010.
These updates resulted in an average annual increase of 0.07 MMT C02 Eq. (0.4 percent) for
fluorochemical production across the time series compared to the previous Inventory.
Planned Improvements
EPA is planning to refine its estimates of emissions from non-reporting facilities after confirming with the
facilities that their actual per-facility uncontrolled emissions fall below 25,000 MT C02 Eq. EPA is also
planning to refine its estimates of emissions for other facilities between 1990 and 2009, e.g., by
comparing these against emissions inferred from atmospheric measurements. Moreover, EPA is
continuing to seek datasets that can be used to improve and/or QA/QC emissions estimates,
particularly for the years 1990 to 2009. These datasets may include, for example, real-time facility-
specific estimates or additional global "top-down," atmosphere-based emissions estimates that could
be used to establish an upper limit on emissions of certain compounds.
Industrial Processes and Product Use 4-97
-------
4.16 Non-EOR Carbon Dioxide Utilization
(Source Category 2H2 and 2H3)
Carbon dioxide (C02) is used for a variety of commercial applications, including food processing,
chemical production, carbonated beverage production, and refrigeration, and is also used in petroleum
production for enhanced oil recovery (EOR). C02 used for EOR is injected underground to enable
additional petroleum to be produced. For the purposes of this analysis, C02 used in food and beverage
(category 2H2) as well as other non-EOR applications (category 2H3) is assumed to be emitted to the
atmosphere. Reporting category 2H3 includes emissions that do not fall within any other source
category, which includes emissions from non-EOR C02 utilization. A further discussion of C02 used in
EOR is described in the Energy chapter in Section 3.9 and is not included in this section.
Carbon dioxide is produced from naturally-occurring C02 reservoirs, as a byproduct from the energy and
industrial production processes (e.g., ammonia production, fossil fuel combustion, ethanol
production), and as a byproduct from the production of crude oil and natural gas, which contain
naturally occurring C02 as a component.
Several ethanol plants capture biogenic C02 as a source of C02 sequestration. This biogenic C02,
absent capture, would not be included in the Inventory as an emission source.57 Where this C02 is
captured by the ethanol plant before it can be released to the atmosphere and then sequestered, it is a
C02 emission reduction. This approach is consistent with the IPCC Guidance, which states: "Once
captured, there is no differentiated treatment between biogenic carbon and fossil carbon. Emissions
and storage of both biogenic and fossil carbon will be estimated and reported." The biogenic C02
captured is likely from biomass fermentation and not necessarily a combustion source, therefore, the
C02 captured for sequestration is subtracted from the food and beverage source category (2H2) that
includes ethanol facilities. See Section 3.9 for more detail on including C02 sequestration in the
Inventory.
Regarding the treatment of biogenic C02 in the Inventory, it should be noted that the Inventory does not
quantify lifecycle emissions of individual products. For example, a lifecycle analysis of ethanol
production with CCS would account for positive emissions associated with any land use change
(including direct and indirect land use change as appropriate) from feedstock production. It would also
account for emissions from energy use at the facility and other upstream and downstream emissions
including the subtraction of captured C02that was permanently sequestered.
The Inventory accounts for emissions and sinks as part of their specific source category in which they
occur. In line with IPCC methodological guidance net carbon fluxes from changes in biogenic carbon
reservoirs in croplands are accounted for in the estimates for Land Use, Land-Use Change, and Forestry
(LULUCF) in the Inventory. Any feedstock production emissions would also be captured under the
agricultural sector and energy use emissions at the facility are captured under the fossil fuel
combustion emissions from the industrial sector in the Inventory. To avoid double counting biogenic C02
emissions are not included as part of energy emissions but are reported as memo or informational items
in the Inventory for tracking purposes. The net carbon flux accounting in the LULUCF sector accounts for
57 Net carbon fluxes from changes in biogenic carbon reservoirs are accounted for in the estimates for Land Use, Land-Use
Change, and Forestry sector of the Inventory.
4-98 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
any C02 emissions associated with harvested carbon. Therefore, if biogenic C02 is captured and
sequestered it would need to be netted out of the source category where it is captured and as noted
previously it may give negative emissions at that source (e.g., ethanol facility) since any positive
emissions are being accounted for elsewhere under LULUCF.
In 2023, the amount of C02 produced and captured for commercial applications and subsequently
emitted to the atmosphere was 3.1 MMT C02 Eq. (3,050 kt). The total C02 captured from ethanol
production for sequestration in 2023 was 0.9 MMT C02 Eq. (903 kt). The total net emissions (excluding
sequestration) from C02 utilization in non-EOR applications was 2.1 MMT C02 Eq. (2,150 kt) in 2023 (see
Table 4-69 and Table 4-67).
Table 4-69: Net C02 Emissions from Non-EOR C02 Utilization (MMT C02 Eq.)
Year
1990
2005
2019
2020
2021
2022
2023
Net CO2 from Food and
Beverage
IE
IE
1.5
1.8
1.9
1.8
1.1
CO2 Emitted from Food
and Beverage
IE
IE
2.1
2.4
2.4
2.4
2.0
CO2 Sequestered from
Food and Beverage
0.0
0.0
(0.5)
(0.5)
(0.4)
(0.6)
(0.9)
CO2 Emitted from Other
Non-EOR Applications
IE
IE
0.9
1.0
0.9
1.0
1.0
Total CO2 Emitted
1.5
1.4
2.4
2.8
2.9
2.8
2.1
IE (Included Elsewhere), meaning included in totals.
Notes: Parentheses indicate negative values. Totals may not sum due to independent rounding.
Table 4-70: NetC02
Emissions from Non-EOR C02
Utilization (kt C02)
Year
1990
2005
2019
2020
2021
2022
2023
Net CO2 from Food and
Beverage
IE
IE
1,540
1,840
1,940
1,799
1,125
CO2 Emitted from
Food and Beverage
IE
IE
2,060
2,362
2,384
2,402
2,028
CO2 Sequestered from
Food and Beverage
0
0
(520)
(522)
(444)
(603)
(903)
CO2 Emitted from Other
Non-EOR Applications
IE
IE
875
1,001
949
1,013
1,024
Total CO2 Emitted
1,472
1,375
2,415
2,842
2,889
2,812
2,150
IE (Included Elsewhere), meaning included in totals.
Notes: Parentheses indicate negative values. Totals may not sum due to independent rounding.
Methodology and Time-Series Consistency
As per IPCC guidance, C02 capture that is used in emissive uses should not be subtracted out of the
Inventory and is assumed to reach the atmosphere on a relatively short time-frame. Also, C02 produced
from natural domes is an anthropogenic activity (i.e., it would not have been emitted otherwise).
Therefore, C02 from natural domes that is used for emissive uses should be counted as an emission in
the Inventory. However, captured C02 from industrial sources are not currently being netted out with the
Industrial Processes and Product Use 4-99
-------
exception of natural gas processing and petroleum refining.58 Therefore, C02 used in emissive uses from
natural gas processing and petroleum refining are the only industrial source C02 capture that need to be
counted in the Inventory under Non-EOR C02 utilization.
Carbon dioxide emission estimates for 1990 through 2023 use a country-specific method and were
based on the quantity of C02 extracted, captured and transferred for industrial applications (i.e., non-
EOR end-uses). Some of the C02 produced by these facilities is used for EOR, and some is used in other
commercial applications (e.g., chemical manufacturing, food and beverage). The IPCC does not have
specific methodological guidelines for C02 utilization, but the country-specific methodology used is
consistent with a Tier 3 approach since it relies on facility-specific information.
2010 through 2023
For 2010 through 2023, data from EPA's GHGRP (Subpart PP) were aggregated from facility-level reports
to develop a national-level estimate for use in the Inventory (EPA 2024). Facilities report C02 extracted or
produced from natural reservoirs and industrial sites, and C02 captured from energy and industrial
processes and transferred to various end-use applications to EPA's GHGRP. This analysis includes
reported C02 transferred to food and beverage end-uses and other end-uses of C02 from Subpart PP.
Other uses include cleaning and solvent use, industrial and municipal water/wastewater treatment, and
metal fabrication. Additionally, a small amount of C02 is used as a refrigerant; use and emissions from
this application are reported under Section 4.25 Substitution of Ozone Depleting Substances (Source
Category 2F).
Reporters subject to EPA's GHGRP Subpart PP are also required to report the quantity of C02 that is
imported and/or exported. Currently, these data are not publicly available through the GHGRP due to
data confidentiality reasons and hence are excluded from this analysis.
The other end uses of C02 are included for the first time in this Inventory, incorporating feedback from
recent expert review periods.
The updated methodology includes all of the C02 that is extracted from natural domes and transferred
to food and beverage use and other uses of C02 as well as a portion of the C02 that is captured from
natural gas processing and petroleum refining industrial sources and transferred to food and beverage
use and other uses of C02. The portion corresponding to the two categories can not be derived directly
from the Subpart PP data for those facilities since the facility level data is considered CBI. Therefore, the
amount of C02 capture from natural gas processing and petroleum refining industrial sources is
estimated based on the assumption that the total amount of the industrial sector C02 that is captured
and transferred are distributed equally across the eleven industrial sector categories assumed to
capture C02 (i.e. 2/11 or 18.2% of the C02 is from natural gas processing and petroleum refining). This is
effectively assuming that each sector that captured and supplied C02 each supplied an equal amount.
The different sectors and total amount of C02 captured is shown in Section 3.9.
Data on C02 capture from ethanol facilities for 2017 through 2023 were obtained from GHGRP. The
approach to account for C02 capture and sequestration in the Inventory in a consistent and
comprehensive manner is to:
58 Capture of CO2 for urea production and for CO2 export are also being netted out, but emissions from those sources are
presented elsewhere in the Inventory, see sections 4.5 Ammonia Production and 3.1 Fossil Fuel Combustion for more
detail.
4-100 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
• Allocate sequestered C02 to the source directly if known based on data from the GHGRP
Subpart RR.
• If unknown or if multiple sources are listed in Subpart RR, allocate sequestrated C02 across
sources based on GHGRP Subpart PP data.
While some Subpart RR facilities vary the source of C02 by year, sequestered C02 can be directly
allocated to an Inventory source category. For some Subpart RR facilities, the sequestered C02 needs to
be allocated across natural domes and other sources. This is done based on GHGRP Subpart PP data on
the total amount of C02 captured that is supplied to EOR since that is felt to best represent C02
supplied for sequestration. See Section 3.9 for more detail on including C02 sequestration in the
Inventory.
Facilities subject to Subpart PP of EPA's GHGRP are required to measure C02 extracted or produced.
More details on the calculation and monitoring methods applicable to extraction and production
facilities can be found under Subpart PP: Suppliers of Carbon Dioxide of the regulation, Part 98.59 The
number of facilities that reported data to EPA's GHGRP Subpart PP (Suppliers of Carbon Dioxide) for
2010 through 2023 is much higher (ranging from 44 to 53) than the number of facilities included in the
Inventory for the 1990 to 2009 time period prior to the availability of GHGRP data (4 facilities). The
difference is largely due to the fact the 1990 to 2009 data includes only C02 transferred to end-use
applications from naturally occurring C02 reservoirs and excludes industrial sites.
1990 through 2009
For 1990 through 2009, data from EPA's GHGRP are not available. For this time period, C02 production
data from four naturally-occurring C02 reservoirs were used to estimate annual C02 emissions. These
facilities were Jackson Dome in Mississippi, Bravo and West Bravo Domes in New Mexico, and
McCallum Dome in Colorado. The facilities in Mississippi and New Mexico produced C02 for use in both
EOR and in other commercial applications (e.g., chemical manufacturing, food production). The fourth
facility in Colorado (McCallum Dome) produced C02 for commercial applications only (New Mexico
Bureau of Geology and Mineral Resources 2006).
Carbon dioxide production data and the percentage of production that was used for non-EOR
applications for the Jackson Dome, Mississippi facility were obtained from Advanced Resources
International (ARI 2006, 2007) for 1990 to 2000, and from the Annual Reports of Denbury Resources
(Denbury Resources 2002 through 2010) for 2001 to 2009 (see Table 4-71). Denbury Resources reported
the average C02 production in units of MMCF C02 per day for 2001 through 2009 and reported the
percentage of the total average annual production that was used for EOR. Production from 1990 to 1999
was set equal to 2000 production, due to lack of publicly available production data for 1990 through
1999. Carbon dioxide production data for the Bravo Dome and West Bravo Dome were obtained from
ARI for 1990 through 2009 (ARI 1990 to 2010). Data for the West Bravo Dome facility were only available
for 2009. The percentage of total production that was used for non-EOR applications for the Bravo Dome
and West Bravo Dome facilities for 1990 through 2009 were obtained from New Mexico Bureau of
Geology and Mineral Resources (Broadhead 2003; New Mexico Bureau of Geology and Mineral
Resources 2006). Production data for the McCallum Dome (Jackson County), Colorado facility were
obtained from the Colorado Oil and Gas Conservation Commission (COGCC) for 1999 through 2009
59 See http://www.ecfr.gov/cgi-bin/text-idx7tpN/ecfrbrowse/Title40/40cfr98 main 0?.tpl.
Industrial Processes and Product Use 4-101
-------
(COGCC 2014). Production data for 1990 to 1998 and percentage of production used for EOR were
assumed to be the same as for 1999, due to lack of publicly available data.
Table 4-71: C02 Production (kt C02) and the Percent Used for Non-EOR Applications
Year
Jackson Dome,
MS
CO2 Production
(kt) (% Non-
EOR)
Bravo Dome,
NM
CO2
Production
(kt) (% Non-
EOR)
West Bravo
Dome, NM
CO2 Production
(kt) (% Non-EOR)
McCallum
Dome, CO
CO2 Production
(kt) (% Non-EOR)
Total CO2
Production
from Extraction
and Capture
Facilities (kt)
%
Non-
EOR3
1990
1,344(100%)
63(1%)
+
65(100%)
NE
NE
2005
1,254 (27%)
58(1%)
+
63(100%)
NE
NE
2019
IE
IE
IE
IE
61,300b
5%
2020
IE
IE
IE
IE
44,800b
8%
2021
IE
IE
IE
IE
44,000b
8%
2022
IE
IE
IE
IE
46,700b
7%
2023
IE
IE
IE
IE
42,900b
7%
+ Does not exceed 0.5 percent.
NE (Not Estimated)
IE (Included Elsewhere), meaning included in totals.
a Includes food and beverage applications and other end uses.
bFor 2010 through 2023, the publicly available GHGRP data were aggregated at the national level based on GHGRP CBI criteria.
The Dome-specific C02 production values are accounted for (i.e., included elsewhere) in the Total C02 Production from
Extraction and Capture Facilities values starting in 2010 and are notable to be disaggregated.
Methodological approaches were applied to the entire time series to ensure consistency in emissions
from 1990 through 2023. The methodology for non-EOR C02 utilization spliced activity data from two
different sources: Industry data for 1990 through 2009 and GHGRP data starting in 2010. Consistent
with the 2006IPCC Guidelines, the overlap technique was applied to compare the two data sets for
years where there was overlap (IPCC 2006). The data sets were determined to be inconsistent; the
GHGRP data include C02 from industrial sources while the industry data do not. No adjustments were
made to the activity data for 1990 through 2009 because prior to 2010, GHGRP data was not available to
net out industrial source C02 capture from natural gas processing and petroleum refining, so those
emissions are accounted for in the Inventory, therefore adjustments were not needed in the 1990-2009
timeframe.
Uncertainty
There is uncertainty associated with the data reported through EPA's GHGRP Specifically, there is
uncertainty associated with the amount of C02 utilized for food and beverage applications, given the
GHGRP does have provisions that Subpart PP reporters are not required to report to the GHGRP if their
emissions fall below certain thresholds, in addition to the exclusion of the amount of C02 transferred to
all other end-use categories. This latter category might include C02 quantities that are being used for
non-EOR industrial applications such as firefighting. Second, uncertainty is associated with the
exclusion of imports/exports data for C02 suppliers. Currently these data are not publicly available
through EPA's GHGRP and hence are excluded from this analysis. EPA verifies annual facility-level
reports through a multi-step process (e.g., combination of electronic checks and manual reviews by
staff) to identify potential errors and ensure that data submitted to EPA are accurate, complete, and
consistent. Based on the results of the verification process, EPA follows up with facilities to resolve
4-102 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
mistakes that may have occurred.60 Given the lack of specific uncertainty ranges available on the data
used, EPA assigned an uncertainty range of ±5 percent and a normal probability density function for C02
utilized for food and beverage applications. The uncertainty range is derived from the default range for
solvent use in Section 5.5 of Chapter 3 of the 2006IPCC Guidelines. These values are representative of
C02 used in food and beverage based on expert judgment (RTI 2023).
The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 4-72. Non-EOR
C02utilization emissions for 2023 were estimated to be between 2.9 and 3.2 MMT C02 Eq. at the 95
percent confidence level. This indicates a range of approximately 5 percent below to 5 percent above
the emission estimate of 3.1 MMT C02 Eq.
Table 4-72: Approach 2 Quantitative Uncertainty Estimates for C02 Emissions from Non-
EOR C02 Utilization (MMT C02 Eq. and Percent)
Source
Gas
2023 Emission Estimate
(MMT CO2 Eq.)
Uncertainty Range Relative to Emission
Estimate3
(MMT CO2 Eq.) (%)
Lower
Upper Lower Upper
Bound
Bound Bound Bound
Non-EOR C02
Utilization
CO2
3.1
2.9
3.2 -5% +5%
a Range of emission estimates predicted by Monte Carlo stochastic simulation for a 95 percent confidence interval.
QA/QC and Verification
General quality assurance/quality control (QA/QC) procedures were applied consistent with the U.S.
Inventory QA/QC plan, which is in accordance with Volume 1, Chapter 6 of 2006 IPCC Guidelines as
described in the introduction of the IPPU chapter (see Annex 8 for more details). More details on the
greenhouse gas calculation, monitoring and QA/QC methods applicable to non-EOR C02 utilization can
be found under Subpart PP (Suppliers of Carbon Dioxide) of the regulation (40 CFR Part 98).61 EPA
verifies annual facility-level GHGRP reports through a multi-step process (e.g., combination of
electronic checks and manual reviews) to identify potential errors and ensure that data submitted to
EPA are accurate, complete, and consistent (EPA 20 1 5).62 Based on the results of the verification
process, EPA follows up with facilities to resolve mistakes that may have occurred. The post-submittals
checks are consistent with a number of general and category-specific QC procedures, including range
checks, statistical checks, algorithm checks, and year-to-year checks of reported data and emissions.
Recalculations Discussion
For the current Inventory, refinements to the methodology were implemented, to incorporate more
complete activity data from GHGRP Subpart PP for 2010 through 2023. These refinements are described
under the Methodology and Time-Series Consistency section. The revised values for 2010 through 2022
resulted in decreased emissions estimates for 2011-2013 and 2016-2022 and increased emissions
estimates for 2010, 2014, and 2015. Across the 2010 to 2022 time series, based on this methodology
refinement emissions decreased by an average of 15 percent compared to the previous Inventory.
Annual emission changes during the time series ranged from a 40 percent decrease in 2019 (1,935 kt
60 See https://www.epa.gov/sites/prodijction/files/9015-07/documents/ghgrp verification factsheet.pdf.
61 See http://www.ecfr.gov/cgi-bin/text-idx7tpN/ecfrbrowse/Title40/40cfr98 main 0?.tpl.
62 See https://www.epa.gov/sites/prodijction/files/7015-07/documents/ghgrp verification factsheet.pdf.
Industrial Processes and Product Use 4-103
-------
C02) to a 35 percent increase in 2014 (1,562 kt C02). In addition to the methodology refinement,
captured C02 for 2017-2023 from ethanol facilities is being included in the inventory for the first time.
These combined updates resulted in an average annual decrease of 0.4 MMT C02 Eq. (8.1 percent) for
non-EOR C02 utilization across the full 1990-2022 time series compared to the previous Inventory.
Planned Improvements
EPA will continue to evaluate the potential to include additional GHGRP data on other emissive end-
uses to improve the accuracy and completeness of estimates for this source category. Particular
attention will be made to ensuring time-series consistency of the emissions estimates presented in
future Inventory reports, consistent with IPCC guidelines. This is required as the facility-level reporting
data from EPA's GHGRP, with the program's initial requirements for reporting of emissions in calendar
year 2010, are not available for all inventory years (i.e., 1990 through 2009) as required for this Inventory.
In implementing improvements and integration of data from EPA's GHGRP, EPA will rely on the latest
guidance from the IPCC on the use of facility-level data in national inventories.63
These improvements are still in process and will be incorporated into future Inventory reports. These are
near-to medium-term improvements.
4.17 Phosphoric Acid Production (Source
Category 2B10)
Phosphoric acid (H3P04) is a basic raw material used in the production of phosphate-based fertilizers.
Phosphoric acid production from natural phosphate rock is a source of carbon dioxide (C02) emissions,
due to the chemical reaction of the inorganic carbon (calcium carbonate) component of the phosphate
rock. These emissions are included under reporting category (2B10) because they reflect a country-
specific source that does not fall within any other existing source category. Emissions from fuels
consumed for energy purposes during the production of phosphoric acid are accounted for as part of
fossil fuel combustion in the industrial end-use sector reported under the Energy chapter.
The phosphoric acid production process involves chemical reaction of the calcium phosphate
(Ca3(P04)2) component of the phosphate rock with sulfuric acid (H2S04) and recirculated phosphoric
acid (H3PO4) (EFMA 2000). Phosphate rock also contains naturally occurring limestone (CaC03), ranging
from 0.2 to 4.5 percent (as C02). The generation of C02from limestone in the phosphate rock is from the
associated limestone-sulfuric acid reaction, as shown below:
CctCO3 + H2SO4 + H2O CaS04 • 2.H2O -I- CO2
Phosphate rock mined in Florida and North Carolina, accounts for more than 75 percent of total
domestic output, with lesser production in Idaho and Utah (USGS 2024). It is used primarily as a raw
material for wet-process phosphoric acid production. The composition of natural phosphate rock varies,
depending on the location where it is mined. Natural phosphate rock mined in the United States
generally contains inorganic carbon in the form of calcium carbonate (limestone) and may also contain
63 See http://www.ipcc-nggip.iges.or.jp/public/tb/TFI Technical Bulletin 1 .pdf and Volume 1. Chapter ?.3 of the ?019
Refinement to the 2006 IPCC Guidelines for National GHG Inventories.
4-104 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
organic carbon. For example, phosphate rock mined from Florida contains 3.1 percent limestone (as
C02) (EFMA 2000). Total U.S. phosphate rock production in 2023 was an estimated 21 million metric
tons (USGS 2024). Between 1990 and 2023, domestic phosphate rock production decreased by
approximately 58 percent. Total imports of phosphate rock to the United States in 2023 were 2.4 million
metric tons (USGS 2024). In 2023, most of the imported phosphate rock (98 percent) came from Peru,
with 2 percent from Morocco (USGS 2024). All phosphate rock mining companies in the United States
are vertically integrated with fertilizer plants that produce phosphoric acid located near the mines.
Total C02 emissions from phosphoric acid production were 0.9 MMT C02 Eq. (850 kt C02) in 2023 (see
Table 4-73 and Table 4-74). Domestic consumption of phosphate rock in 2023 was estimated to have
increased 6.1 percent relative to 2022 levels.
Table 4-73: C02 Emissions from Phosphoric Acid Production (MMT C02 Eq.)
Year
1990
2005
2019
2020
2021
2022
2023
Phosphoric Acid Production
1.5
1.3
0.9
0.9
0.9
0.8
0.9
Table 4-74: C02 Emissions from
Phosphoric Acid Production (kt
CM
O
O
Year
1990
2005
2019
2020
2021
2022
2023
Phosphoric Acid Production
1,529
1,342
909
901
874
804
850
Methodology and Time-Series Consistency
The United States uses a country-specific methodology consistent with and comparable to an IPCC Tier
1 approach to calculate emissions from production of phosphoric acid from phosphate rock based on
the stoichiometry of the process reaction shown above. The 2006 IPCC Guidelines do not provide a
method for estimating process emissions (C02) from phosphoric acid production. Carbon dioxide
emissions from production of phosphoric acid from phosphate rock are estimated by multiplying the
average amount of inorganic carbon (expressed as C02) contained in the natural phosphate rock as
calcium carbonate by the amount of phosphate rock that is used annually to produce phosphoric acid,
accountingfor domestic production and net imports for consumption. The estimation methodology is
as follows:
Equation 4-9: C02 Emissions from Phosphoric Acid Production
Epa Cpr X Qpr
where,
Epa = C02 emissions from phosphoric acid production, metric tons
Cpr = Average amount of carbon (expressed as C02) in natural phosphate rock, metric ton C02/
metric ton phosphate rock
Qpr = Quantity of phosphate rock used to produce phosphoric acid
The C02 emissions calculation methodology assumes that all of the inorganic carbon (calcium
carbonate) content of the phosphate rock reacts to produce C02 in the phosphoric acid production
process and is emitted with the stack gas. The methodology also assumes that none of the organic
carbon content of the phosphate rock is converted to C02 and that all of the organic carbon content
precipitates out of solution or remains in the phosphoric acid product (RTI 2024).
Industrial Processes and Product Use 4-105
-------
From 1993 to 2004, the U.S. Geological Survey (USGS) Mineral Yearbook: Phosphate Rock disaggregated
phosphate rock mined annually in Florida and North Carolina from phosphate rock mined annually in
Idaho and Utah, and reported the annual amounts of phosphate rock exported and imported for
consumption (see Table 4-75). For the years 1990 through 1992, and 2005 through 2023, only nationally
aggregated mining data was reported by USGS. For the years 1990, 1991, and 1992, the breakdown of
phosphate rock mined in Florida and North Carolina and the amount mined in Idaho and Utah are
approximated using data reported by USGS for the average share of U.S. production in those states from
1993 to 2004. For the years 2005 through 2023, the same approximation method is used, but the share
of U.S. production was assumed to be consistent with the ratio of production capacity in those states,
which were obtained from the USGS commodity specialist for phosphate rock (USGS 2012; USGS
2021 b). For 1990 through 2023, data on U.S. domestic consumption of phosphate rock, consisting of
domestic reported sales and use of phosphate rock, exports of phosphate rock (primarily from Florida
and North Carolina), and imports of phosphate rock for consumption, were obtained from USGS
Minerals Yearbook: Phosphate Rock (USGS 1994 through 2015b) and from USGS Minerals Commodity
Summaries: Phosphate Rock (USGS 2016 through 2023). From 2004 through 2023, the USGS reported
no exports of phosphate rock from U.S. producers (USGS 2024).
The carbonate content of phosphate rock varies depending upon where the material is mined.
Composition data for domestically mined and imported phosphate rock were provided by the Florida
Institute of Phosphate Research, now known as the Florida Industrial and Phosphate Research Institute
(FIPR 2003a). Phosphate rock mined in Florida contains approximately 1 percent inorganic C, and
phosphate rock imported from Morocco contains approximately 1.46 percent inorganic C. Calcined
phosphate rock mined in North Carolina and Idaho contains approximately 0.41 percent and 0.27
percent inorganic C, respectively (see Table 4-76). Phosphate rock from Utah is assumed to have similar
characteristics as of phosphate rock mined in Idaho. Similar to the phosphate rock mined in Morocco,
phosphate rock mined in Peru contains approximately 5 percent C02 (Golder Associates and M3
Engineering 2016).
Carbonate content data for phosphate rock mined in Florida are used to calculate the C02 emissions
from consumption of phosphate rock mined in Florida and North Carolina (more than 75 percent of
domestic production), and carbonate content data for phosphate rock mined in Morocco and Peru are
used to calculate C02 emissions from consumption of imported phosphate rock. The C02 emissions
calculation assumes that all of the domestic production of phosphate rock is used in uncalcined form.
As of 2006, the USGS noted that one phosphate rock producer in Idaho produces calcined phosphate
rock; however, no production data were available for this single producer (USGS 2006). The USGS
confirmed that no significant quantity of domestic production of phosphate rock is in the calcined form
(USGS 2012).
Table 4-75: Phosphate Rock Domestic Consumption, Exports, and Imports (kt)
Location/Year
1990
2005
2019
2020
2021
2022
2023
U.S. Domestic Consumption®
49,800
35,200
23,400
22,600
21,900
19,800
21,000
FL and NC
42,494
28,160
18,250
17,630
17,080
15,444
16,380
ID and UT
7,306
7,040
5,150
4,970
4,820
4,356
4,620
Exports—FLand NC
6,240
0
0
0
0
0
0
Imports
451
2,630
2,140
2,520
2,460
2,500
2,600
Total U.S. Consumption
44,011
37,830
25,540
25,120
24,360
22,300
23,600
4-106 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Notes: Regional production data for 2021 through 2023 are estimates (USGS 2022-2024a). Totals may not sum due to
independent rounding.
Table 4-76: Chemical Composition of Phosphate Rock (Percent by Weight)
Composition
Central
Florida
North
Florida
North
Carolina
(calcined)
Idaho
(calcined)
Morocco
Peru
Total Carbon (as C)
1.60
1.76
0.76
0.60
1.56
NA
Inorganic Carbon (as C)
1.00
0.93
0.41
0.27
1.46
NA
Organic Carbon (as C)
0.60
0.83
0.35
0.00
0.10
NA
Inorganic Carbon (as CO2)
3.67
3.43
1.50
1.00
5.00
5.00
NA (Not Available)
Sources: FIPR (2003a), Golder Associates and M3 Engineering (2016)
Methodological approaches were applied to the entire time series to ensure consistency in emissions
estimates from 1990 through 2023.
Uncertainty
Phosphate rock production data used in the emission calculations were developed by the USGS through
monthly and semiannual voluntary surveys of the active phosphate rock mines during 2021. Prior to
2006, USGS provided the data disaggregated regionally; however, beginning in 2006, only total U.S.
phosphate rock production was reported. Regional production for 2021 was estimated based on
regional production data from 2017 to 2020 and multiplied by regionally-specific emission factors. While
total U.S. phosphate rock production data are not considered to be a significant source of uncertainty
because all the domestic phosphate rock producers report their annual production to the USGS, there is
uncertainty associated with the degree to which the estimated 2021 regional production data represents
actual production in those regions. Data for exports of phosphate rock used in the emission calculations
are reported to the USGS by phosphate rock producers and are not considered to be a significant source
of uncertainty. Data for imports for consumption are based on international trade data collected by the
U.S. Census Bureau. These U.S. government economic data are not considered to be a significant
source of uncertainty. Based on expert judgement of the USGS, EPA assigned an uncertainty range of ±5
percent to the percentage of phosphate rock produced from Florida and North Carolina, and ±5 percent
to phosphoric acid production and imports (USGS 2012). Per this expert judgment, a normal probability
density function was assigned for all activity data.
An additional source of uncertainty in the calculation of C02 emissions from phosphoric acid
production is the carbonate composition of phosphate rock, as the composition of phosphate rock
varies depending upon where the material is mined and may also vary over time. The Inventory relies on
one study (FIPR 2003a) of chemical composition of the phosphate rock; limited data are available
beyond this study. Another source of uncertainty is the disposition of the organic carbon content of the
phosphate rock. A representative of FIPR indicated that in the phosphoric acid production process, the
organic carbon content of the mined phosphate rock generally remains in the phosphoric acid product,
which is what produces the color of the phosphoric acid product (FIPR 2003b). Organic carbon is
therefore not included in the calculation of C02 emissions from phosphoric acid production.
A third source of uncertainty is the assumption that all domestically-produced phosphate rock is used in
phosphoric acid production and used without first being calcined. Calcination of the phosphate rock
Industrial Processes and Product Use 4-107
-------
would result in conversion of some of the organic carbon in the phosphate rock into C02; however,
according to air permit information available to the public, at least one facility has calcining units
permitted for operation (NCDENR 2013).
Finally, USGS indicated that between 2021 and 2023, less than 5 percent of domestically-produced
phosphate rock was used to manufacture elemental phosphorus and other phosphorus-based
chemicals, rather than phosphoric acid (USGS 2022 through 2024). According to USGS, there is only one
domestic producer of elemental phosphorus, in Idaho, and no data were available concerning the
annual production of this single producer. Elemental phosphorus is produced by reducing phosphate
rock with coal coke, and it is therefore assumed that 100 percent of the carbonate content of the
phosphate rock will be converted to C02 in the elemental phosphorus production process. The
calculation for C02 emissions assumes that phosphate rock consumption, for purposes other than
phosphoric acid production, results in C02 emissions from 100 percent of the inorganic carbon content
in phosphate rock, but none from the organic carbon content.
The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 4-77. 2023
phosphoric acid production C02 emissions were estimated to be between 0.7 and 1.1 MMT C02 Eq. at the
95 percent confidence level. This indicates a range of approximately 18 percent below and 20 percent
above the emission estimate of 0.9 MMT C02 Eq.
Table 4-77: Approach 2 Quantitative Uncertainty Estimates for C02 Emissions from
Phosphoric Acid Production (MMT C02 Eq. and Percent)
2023 Emission
Uncertainty Range Relative to Emission
Gas
Estimate
Estimate3
(MMT CO2 Eq.)
(MMT CO2 Eq.)
(%)
Lower Upper
Lower Upper
Source
Bound Bound
Bound Bound
Phosphoric Acid Production
C02
0.9
0.7 1.1
-18% +20%
a Range of emission estimates predicted by Monte Carlo stochastic simulation for a 95 percent confidence interval.
QA/QC and Verification
For more information on the general QA/QC process applied to this source category, consistent with the
U.S. Inventory QA/QC plan, which is in accordance with Volume 1, Chapter 6 of 2006IPCC Guidelines
as described in the introduction of the IPPU chapter (see Annex 8 for more details).
Recalculations Discussion
Recalculations were performed for 2022 to reflect updated USGS data on the total U.S. production of
phosphate rock. This update resulted in a decrease of 36 kt C02 in 2022 (4 percent).
Planned Improvements
EPA continues to incrementally advance, evaluation of potential improvements to the Inventory
estimates for this source category, which include direct integration of EPA's GHGRP data for 2010
through 2023 along with assessing applicability of reported GHGRP data to update the inorganic carbon
content of phosphate rock for prior years to ensure time-series consistency. Specifically, EPA would
need to assess that averaged inorganic carbon content data (by region or other approaches) meets
4-108 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
GHGRP confidential business information (CBI) screening criteria. EPA would then need to assess the
applicability of GHGRP data for the averaged inorganic carbon content (by region or other approaches)
from 2010 through 2023, along with other information to inform estimates in prior years in the time
series (1990 through 2009) based on the sources of phosphate rock used in production of phosphoric
acid over time. In implementing improvements and integration of data from EPA's GHGRP, EPA will rely
upon the latest guidance from the IPCC on the use of facility-level data in national inventories.64 These
are long-term planned improvements and have not been implemented into the current Inventory.
4.18 Iron and Steel Production (Source
Category 2C1) and Metallurgical Coke
Production
Iron and steel production is a multi-step process that generates process-related emissions of carbon
dioxide (C02) and methane (CH4) as raw materials are refined into iron and then transformed into raw
steel. This reporting category (2C1) includes emissions from the production of iron and steel. Per the
IPCC methodological guidance, emissions from conventional fuels (e.g., natural gas, fuel oil) consumed
for energy purposes during the production of iron and steel are accounted for as part of fossil fuel
combustion in the industrial end-use sector reported under the Energy chapter.
Iron and steel production includes seven distinct production processes: metallurgical coke production,
sinter production, direct reduced iron (DRI) production, pellet production, pig iron65 production, electric
arc furnace (EAF) steel production, and basic oxygen furnace (BOF) steel production. The number of
production processes at a particular plant is dependent upon the specific plant configuration. Most
process C02 generated from the iron and steel industry is a result of the production of crude iron.
In addition to the production processes mentioned above, C02 is also generated at iron and steel mills
through the consumption of process byproducts (e.g., blast furnace gas, coke oven gas) used for various
purposes including heating, annealing, and electricity generation. Process byproducts sold off-site for
use as synthetic natural gas are also accounted for in these calculations. In general, C02 emissions are
generated in these production processes through the reduction and consumption of various carbon-
containing inputs (e.g., ore, scrap, flux, coke byproducts). Fugitive CH4 emissions can also be generated
from these processes, as well as from sinter, direct iron, and pellet production.
In 2023, twelve integrated iron and steel steelmaking facilities utilized BOFs to refine and produce steel
from iron, and raw steel was produced at 105 facilities across the United States. In 2023 approximately
29 percent of steel production was attributed to BOFs and 71 percent to EAFs (USGS 2024a). The trend
in the United States for integrated facilities has been a shift towards fewer BOFs and more EAFs. EAFs
64 See http://www.ipcc-nggip.iges.or.jp/public/tb/TFI Technical Bulletin 1.pdf and the 2019 Refinement, Volume 1,
Chapter 2, Section 2.3, Use of Facility Data in Inventories at https://www.ipcc-
nggip.iges.or.ip/public/2019rf/pdf/1 Volume1/19R V1 Ch02 DataCollection.pdf.
65 Pig iron is the common industry term to describe what should technically be called crude iron. Pig iron is a subset of
crude iron that has lost popularity over time as industry trends have shifted. Throughout this report, pig iron will be used
interchangeably with crude iron, but it should be noted that in other data sets or reports pig iron and crude iron may not
be used interchangeably and may provide different values.
Industrial Processes and Product Use 4-109
-------
use scrap steel as their main input and use significantly less energy than BOFs. There are also 14
cokemaking facilities, of which 3 facilities are co-located with integrated iron and steel facilities (ACCCI
2021). In 2023, four states accounted for approximately 46 percent of total raw steel production:
Indiana, Ohio, Pennsylvania, and Texas (USGS 2024a).
Total annual production of raw steel in the United States was fairly constant between 2000 and 2008 and
ranged from a low of 99,320,000 tons to a high of 109,880,000 tons (2001 and 2004, respectively). Due to
the decrease in demand caused by the global economic downturn (particularly from the automotive
industry), raw steel production in the United States sharply decreased to 65,459,000 tons in 2009. Raw
steel production was fairly constant from 2011 through 2014, and after a dip in production from 2014 to
2015, raw steel production steadily increased. Raw steel production dipped again in 2020 due to the
COVID-19 pandemic and returned to pre-pandemic levels in 2021. Production declined by
approximately 6 percent from the prior year in 2022 (AISI 2023) and remained approximately at that level
in 2023 (USGS 2024a). This decline may be attributable to projections for decreased global end-use
consumption due to multiple factors includingthe conflict in Ukraine, rising energy costs and interest
rates, and global inflation (USGS 2024a). The United States was the fourth largest producer of raw steel
in the world, behind China, India, and Japan, accounting for approximately 4.2 percent of world
production in 2023 (USGS 2024a).
The majority of C02 emissions from the iron and steel production process come from the use of
metallurgical coke in the production of pig iron and from the consumption of other process byproducts,
with lesser amounts emitted from the use of carbon-containing flux and from the removal of carbon
from pig iron used to produce steel.
According to the 2006IPCC Guidelines, the production of metallurgical coke from coking coal is
considered to be an energy use of fossil fuel, and the use of coke in iron and steel production is
considered to be an industrial process source. The 2006 IPCC Guidelines suggest that emissions from
the production of metallurgical coke should be reported separately in the Energy sector, while emissions
from coke consumption in iron and steel production should be reported in the Industrial Processes and
Product Use sector. The approaches and emission estimates for both metallurgical coke production and
iron and steel production, however, are presented here because much of the relevant activity data is
used to estimate emissions from both metallurgical coke production and iron and steel production. For
example, some byproducts (e.g., coke oven gas) of the metallurgical coke production process are
consumed during iron and steel production, and some byproducts of the iron and steel production
process (e.g., blast furnace gas) are consumed during metallurgical coke production. Emissions
associated with the consumption of these byproducts are attributed at the point of consumption.
Emissions associated with the use of conventional fuels (e.g., natural gas, fuel oil) for electricity
generation, heating and annealing, or other miscellaneous purposes downstream of the iron and
steelmaking furnaces are reported in the Energy chapter. As further discussed in the Planned
Improvements section, EPA is considering methodological refinements to account for estimates of
emissions from the production of metallurgical coke in the Energy sector as well as better identifying the
coke production inputs and outputs including at merchant coke plants.
Metallurgical Coke Production
Emissions of C02 from metallurgical coke production in 2023 were 3.0 MMT C02 Eq. (2,986 kt C02) (see
Table 4-78 and Table 4-79). Emissions increased by 1 percent from 2022 to 2023 and have decreased by
4-110 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
47 percent since 1990. Coke production in 2023 was about 1 percent lower than in 2022 and 59 percent
below 1990 (EIA 2024).
Significant activity data for 2020 through 2023 were not available in time for publication of this report
due to industry consolidation that impacts the publication of data without revealing confidential
business information. Activity data for these years were estimated using 2019 values adjusted based on
GHGRP emissions data, as described in the Methodology and Time-Series Consistency section below.
Table 4-78: C02 Emissions from Metallurgical Coke Production (MMT C02 Eq.)
Gas 1990
2005
2019 2020 2021 2022 2023
CO2 5.6
3.9
O
CO*
O
CO*
CNJ
CO*
CO
CN
O
CO*
Table 4-79: C02 Emissions from Metallurgical Coke Production (kt C02)
Gas 1990
2005
2019 2020 2021 2022 2023
CO2 5,608
3,921
3,006 2,325 3,224 2,954 2,986
Iron and Steel Production
Emissions of C02 and CH4from iron and steel production in 2023 were 43.3 MMT C02 Eq. (43,254 kt) and
0.0080 MMT C02 Eq. (0.3 kt CH4), respectively (see Table 4-80 through Table 4-83). Emissions from iron
and steel production increased by 2.5 percent from 2022 to 2023 and have decreased by 56 percent
since 1990, due to restructuring of the industry, technological improvements, and increased scrap steel
utilization. Carbon dioxide emission estimates include emissions from the consumption of
carbonaceous materials in the blast furnace, EAF, and BOF, as well as blast furnace gas and coke oven
gas consumption for other activities at the steel mill.
Significant activity data for 2020 through 2023 were not available in time for publication of this report
due to industry consolidation that impacts the publication of data without revealing confidential
business information. Activity data for these years were estimated using 2019 values adjusted based on
GHGRP emissions data, as described in the Methodology and Time-Series Consistency section below.
In 2023, domestic production of pig iron increased by 6 percent from 2022 levels. Overall, domestic pig
iron production has declined since the 1990s; pig iron production in 2023 was 56 percent lower than in
2000 and 58 percent below 1990. Carbon dioxide emissions from iron production have decreased by 73
percent (33.4 MMT C02 Eq.) since 1990. Carbon dioxide emissions from steel production have
decreased by 2 percent (0.2 MMT C02 Eq.) since 1990, while overall C02 emissions from iron and steel
production have declined by 56 percent (55.9 MMT C02 Eq.) from 1990 to 2023. The magnitude of
reductions in carbon dioxide emissions from steel production may be underestimated due to data
availability and time series consistency for process inputs in steel production that are further discussed
in the Methodology and Time-Series Consistency section.
Table 4-80: C02 Emissions from Iron and Steel Production (MMT C02 Eq.)
Source/Activity Data
1990
2005
2019
2020
2021
2022
2023
Sinter Production
2.4
1.7
0.9
0.7
0.8
0.8
0.8
Iron Production
45.7
17.7
11.3
10.0
12.2
12.3
12.4
Pellet Production
1.8
1.5
0.9
0.8
0.8
0.8
0.8
Steel Production
8.0
9.4
7.6
7.0
8.0
7.5
7.8
Industrial Processes and Product Use 4-111
-------
Source/Activity Data
1990
2005
2019
2020
2021
2022
2023
Other Activities®
41.2
35.9
23.2
19.8
22.1
20.8
21.5
Total
99.1
66.2
43.8
38.3
44.0
42.2
43.3
a Includes emissions from blast furnace gas and coke oven gas combustion for activities at the steel mill other than consumption
in blast furnace, EAFs, or BOFs.
Note: Totals may not sum due to independent rounding.
Table 4-81: C02 Emissions from Iron and Steel Production (kt C02)
Source/Activity Data
1990
2005
2019
2020
2021
2022
2023
Sinter Production
2,448
1,663
876
749
836
787
812
Iron Production
45,707
17,663
11,315
10,023
12,244
12,301
12,353
Pellet Production
1,817
1,503
878
751
838
789
814
Steel Production
7,964
9,395
7,602
7,006
7,956
7,511
7,797
Other Activities®
41,194
35,934
23,158
19,820
22,119
20,814
21,478
Total
99,130
66,158
43,829
38,350
43,994
42,202
43,254
a Includes emissions from blast furnace gas and coke oven gas combustion for activities at the steel mill other than consumption
in blast furnace, EAFs, or BOFs.
Note: Totals may not sum due to independent rounding.
Table 4-82: CH4 Emissions from Iron and Steel Production (MMT C02 Eq.)
Sou rce/Activity Data 1990
2005
2019 2020 2021
2022
2023
Sinter Production +
+
+ + +
+
+
+ Does not exceed 0.05 MMT C02 Eq.
Table 4-83: CH4 Emissions from Iron and Steel Production (kt CH4)
Sou rce/Activity Data 1990
2005
2019 2020 2021
2022
2023
Sinter Production 0.9
1
+ + +
+
+
+ Does not exceed 0.5 kt.
Methodology and Time-Series Consistency
Emission estimates for metallurgical coke, EAF steel production, and BOF steel production presented in
this chapter utilize a country-specific approach based on Tier 2 methodologies provided by the 2006
IPCC Guidelines, in accordance with the IPCC methodological decision tree and available data. These
Tier 2 methodologies call for a mass balance accounting of the carbonaceous inputs and outputs during
the iron and steel production process and the metallurgical coke production process. Estimates for pig
iron production apply Tier 2 methods consistent with the 2006 IPCC Guidelines, in accordance with the
IPCC methodological decision tree and available data. Tier 1 methods are used for certain iron and steel
production processes (i.e., sinter production, pellet production and DRI production) for which available
data are insufficient to apply a Tier 2 method (e.g., country-specific carbon contents of inputs and
outputs are not known). The majority of emissions are captured with higher tier methods, as sinter
production, pellet production, and DRI production only account for roughly 16 percent of total iron and
steel production emissions.
The Tier 2 methodology equation is as follows:
4-112 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Equation 4-10: C02 Emissions from Coke, Pig Iron, EAF Steel, and BOF Steel
Production, based on 2006IPCC Guidelines Tier 2 Methodologies
where,
Ec02
a
b
Qa
Ca
Qb
Cb
44/12
Eco2 ~
X Ca) ~ X Q)
44
12
Emissions from coke, pig iron, EAF steel, or BOF steel production, metric tons
Input material a
Output material to
Quantity of input material a, metric tons
Carbon content of input material a, metric tons C/metric ton material
Quantity of output material b, metric tons
Carbon content of output material b, metric tons C/metric ton material
Stoichiometric ratio of C02 to C
The Tier 1 methodology equations are as follows:
Equation 4-11:2006 IPCC Guidelines Tier 1: Emissions from Sinter, Direct Reduced
Iron, and Pellet Production (Equations 4.6,4.7, and 4.8)
es,v Qsx
Ed,C02 = Qd X EFd C02
Ep,C02 = Qp X EFp,co2
where,
Es,p —
Emissions from sinter production process for pollutant p (C02or CH4), metric ton
Qs
Quantity of sinter produced, metric tons
EFs,p
Emission factor for pollutant p (C02or CH4), metric ton p/metric ton sinter
Ed,C02 =
Emissions from DRI production process for C02, metric ton
Qd
Quantity of DRI produced, metric tons
EFd,C02 =
Emission factor for C02, metric ton C02/metric ton DRI
Ep,C02 =
Emissions from pellet production process for C02, metric ton
QP
Quantity of pellets produced, metric tons
EFp,C02 =
Emission factor for C02, metric ton C02/metric ton pellets produced
A significant number of activity data that serve as inputs to emissions calculations were unavailable for
2020 through 2023 at the time of publication and were estimated using 2019 values. To estimate annual
emissions for these years, EPA used process emissions data from EPA's Greenhouse Gas Reporting
Program (GHGRP) Subpart Q for the iron and steel sector to adjust the estimated values for 2020
through 2023. GHGRP process emissions data decreased by approximately 14 percent from 2019 to
2020, increased by approximately 12 percent from 2020 to 2021, decreased by approximately 6 percent
from 2021 to 2022, and increased by approximately 3 percent from 2022 to 2023 (EPA 2024). These
percentage changes were applied to 2019 activity data values to produce estimates for 2020 through
2023.
Industrial Processes and Product Use 4-113
-------
Metallurgical Coke Production
Coking coal is used to manufacture metallurgical coke that is used primarily as a reducing agent in the
production of iron and steel but is also used in the production of other metals including zinc and lead
(see Zinc Production and Lead Production sections of this chapter). Emissions associated with
producing metallurgical coke from coking coal are estimated and reported separately from emissions
that result from the iron and steel production process. To estimate emissions from metallurgical coke
production, a Tier 2 method provided by the 2006IPCC Guidelines was utilized. The amount of carbon
contained in materials produced during the metallurgical coke production process (i.e., coke, coke
breeze and coke oven gas) is deducted from the amount of carbon contained in materials consumed
during the metallurgical coke production process (i.e., natural gas, blast furnace gas, and coking coal).
For calculations, activity data for these inputs, including natural gas, blast furnace gas, and coking coke
consumed for metallurgical coke production, are in units consistent with the carbon content values.
Light oil, which is produced during the metallurgical coke production process, is excluded from the
deductions due to data limitations. The amount of carbon contained in these materials is calculated by
multiplying the material-specific carbon content by the amount of material consumed or produced (see
Table 4-84). The amount of coal tar produced was approximated using a production factor of 0.03 tons
of coal tar per ton of coking coal consumed. The amount of coke breeze produced was approximated
using a production factor of 0.075 tons of coke breeze per ton of coking coal consumed (Steiner 2008;
DOE 2000). Data on the consumption of carbonaceous materials (other than coking coal) as well as
coke oven gas production were available for integrated steel mills only (i.e., steel mills with co-located
coke plants); therefore, carbonaceous material (other than coking coal) consumption and coke oven gas
production were excluded from emission estimates for merchant coke plants. Carbon contained in coke
oven gas used for coke oven underfiring was not included in the deductions to avoid double-counting.
Table 4-84: Material Carbon Contents for Metallurgical Coke Production
Material
kgC/kg
CoalTar®
0.62
Coke®
0.83
Coke Breeze®
0.83
Coking Coalb
0.75
Material
kg C/GJ
Coke Oven Gas0
12.1
Blast Furnace Gas0
70.8
a Source: IPCC (2006), Vol. 3 Chapter 4, Table 4.3
"Source: EIA(2017b)
0 Source: IPCC (2006), Vol. 2 Chapter 1, Table 1.3
Although the 2006 IPCC Guidelines provide a Tier 1 CH4 emission factor for metallurgical coke
production (i.e., 0.1 g CH4 per metric ton of coke production), it is not appropriate to use because C02
emissions were estimated using the Tier 2 mass balance methodology. The mass balance methodology
makes a basic assumption that all carbon that enters the metallurgical coke production process either
exits the process as part of a carbon-containing output or as C02 emissions. This is consistent with a
preliminary assessment of aggregated facility-level greenhouse gas CH4 emissions reported by coke
production facilities under EPA's GHGRP. The assessment indicates that CH4 emissions from coke
production are insignificant and below 500 kt or 0.05 percent of total national emissions. Pending
4-114 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
resources and significance, EPA continues to assess the possibility of including these emissions in
future Inventories to enhance completeness but has not incorporated these emissions into this report.
Data relating to the mass of coking coal consumed at metallurgical coke plants and the mass of
metallurgical coke produced at coke plants were taken from the Energy Information Administration (EIA)
Quarterly Coal Report: October through December (EIA 1998 through 2019) and EIA Quarterly Coal
Report: January through March (EIA 2021 through 2024) (see Table 4-85). Data on the volume of natural
gas consumption, blast furnace gas consumption, and coke oven gas production for metallurgical coke
production at integrated steel mills were obtained from the American Iron and Steel Institute (AISI)
Annual Statistical Report (AISI 2004 through 2023) and through personal communications with AISI
(Steiner 2008) (see Table 4-86). These data from the AISI Annual Statistical Report were withheld for
2020 through 2023, so the 2019 values were used as estimated data for the missing 2020 through 2023
values and adjusted using GHGRP emissions data, as described earlier in this Methodology and Time-
Series Consistency section.
The factor for the quantity of coal tar produced per ton of coking coal consumed was provided by AISI
(Steiner 2008). The factor for the quantity of coke breeze produced per ton of coking coal consumed was
obtained through Table 2-1 of the report Energy and Environmental Profile of the U.S. Iron and Steel
Industry (DOE 2000). Data on natural gas consumption and coke oven gas production at merchant coke
plants were not available and were excluded from the emission estimate. Carbon contents for
metallurgical coke, coal tar, coke oven gas, and blast furnace gas were provided by the 2006IPCC
Guidelines. The carbon content for coke breeze was assumed to equal the carbon content of coke.
Carbon contents for coking coal was from EIA.
Table 4-85: Production and Consumption Data for the Calculation of C02 Emissions
from Metallurgical Coke Production (Thousand Metric Tons)
Source/Activity Data
1990
2005
2019
2020
2021
2022
2023
Metallurgical Coke Production
Coking Coal Consumption at Coke Plants
35,269
21,259
16,261
13,076
15,957
14,523
14,378
Coke Production at Coke Plants
25,054
15,167
11,676
9,392
11,381
10,337
10,193
Coke Breeze Production
2,645
1,594
1,220
981
1,197
1,089
1,078
Coal Tar Production
1,058
638
488
392
479
436
431
Table 4-86: Production and Consumption Data for the Calculation of C02
from Metallurgical Coke Production (Million ft3)
Emissions
Source/Activity Data
1990
2005
2019
2020
2021
2022
2023
Metallurgical Coke Production
Coke Oven Gas Production
250,767
114,213
77,692
66,492
74,206
69,829
72,054
Natural Gas Consumption
599
2,996
2,189
1,873
2,091
1,967
2,030
Blast Furnace Gas Consumption
24,602
4,460
3,914
3,350
3,738
3,518
3,630
Iron and Steel Production
To estimate emissions from pig iron production in the blast furnace, the amount of carbon contained in
the produced pig iron and blast furnace gas were deducted from the amount of carbon contained in
inputs (i.e., metallurgical coke, sinter, natural ore, pellets, natural gas, fuel oil, coke oven gas, carbonate
Industrial Processes and Product Use 4-115
-------
fluxes or slagging materials, and direct coal injection). For calculations, activity data for these inputs,
including coke consumed for pig iron production, are in units consistent with the carbon content values.
The carbon contained in the pig iron, blast furnace gas, and blast furnace inputs was estimated by
multiplying the material-specific carbon content by each material type (see Table 4-87). In the absence
of a default carbon content value from the 2006IPCC Guidelines for pellet, sinter, or natural ore
consumed for pig iron production, a country-specific approach based on Tier 2 methodology is used.
Pellet, sinter, and natural ore used as an input for pig iron production is assumed to have the same
carbon content as direct reduced iron (2 percent), based on expert judgment (RTI 2024). Carbon in blast
furnace gas used to pre-heat the blast furnace air is combusted to form C02 during this process. Carbon
contained in blast furnace gas used as a blast furnace input was not included in the deductions to avoid
double-counting.
Emissions from steel production in EAFs were estimated by deducting the carbon contained in the steel
produced from the carbon contained in the EAF anode, charge carbon, and scrap steel added to the EAF.
Small amounts of carbon from DRI and pig iron to the EAFs were also included in the EAF calculation.
For BOFs, estimates of carbon contained in BOF steel were deducted from carbon contained in inputs
such as natural gas, coke oven gas, fluxes (i.e., limestone and dolomite), and pig iron. In each case, the
carbon was calculated by multiplying material-specific carbon contents by each material type (see
Table 4-87). For EAFs, the amount of EAF anode consumed was approximated by multiplying total EAF
steel production by the amount of EAF anode consumed per metric ton of steel produced (0.002 metric
tons EAF anode per metric ton steel produced [Steiner 2008]). The amount of carbon-containing flux
(i.e., limestone and dolomite) used in EAF and BOF steel production was deducted from the "Other
Process Uses of Carbonates" source category (Source Category 2A4) to avoid double-counting.
Carbon dioxide emissions from the consumption of blast furnace gas and coke oven gas for other
activities occurring at the steel mill were estimated by multiplying the amount of these materials
consumed for these purposes by the material-specific carbon content (see Table 4-87).
Table 4-87: Material Carbon Contents for Iron and Steel Production
Material kg C/kg
Coke 0.83
Direct Reduced Iron 0.02
Dolomite 0.13
EAF Carbon Electrodes 0.82
EAF Charge Carbon 0.83
Limestone 0.12
Pig Iron 0.04
Steel 0.01
Coke Oven Gas 12.1
Blast Furnace Gas 70.8
Source: IPCC (2006), Table 4.3. Coke Oven Gas and Blast Furnace Gas, Table 1.3.
Carbon dioxide emissions associated with sinter production, direct reduced iron production, pellet
production, pig iron production, steel production, and other steel mill activities were summed to
calculate the total C02 emissions from iron and steel production (see Table 4-80 and Table 4-81).
The sinter production process results in fugitive emissions of CH4, which are emitted via leaks in the
production equipment, rather than through the emission stacks or vents of the production plants. The
4-116 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
fugitive emissions were calculated by applying Tier 1 emission factors taken from the 2006IPCC
Guidelines for sinter production (see Table 4-88). Although the 2006 IPCC Guidelines also provide a Tier
1 methodology for CH4 emissions from pig iron production, it is not appropriate to use because C02
emissions for pig iron production are estimated using the Tier 2 mass balance methodology. The mass
balance methodology makes a basic assumption that all carbon that enters the pig iron production
process either exits the process as part of a carbon-containing output or as C02 emissions; the
estimation of CH4 emissions is precluded. Annual analysis of facility-level emissions reported during
iron production further supports this assumption and indicates that CH4 emissions are below 500 kt C02
Eq. and well below 0.05 percent of total national emissions. The production of direct reduced iron could
also result in emissions of CH4 through the consumption of fossil fuels (e.g., natural gas, etc.); however,
these emission estimates are excluded due to data limitations. Pending further analysis and resources,
EPA may include these emissions in future reports to enhance completeness. EPA is still assessing the
possibility of including these emissions in future reports and have not included this data in the current
report.
Table 4-88: CH4 Emission Factors for Sinter and Pig Iron Production
Material Produced
Factor
Unit
Sinter
0.07
kg ChU/metric ton
Source: IPCC (2006), Table 4.2.
Emissions of C02 from sinter production, direct reduced iron production, and pellet production were
estimated by multiplying total national sinter production, total national direct reduced iron production,
and total national pellet production by Tier 1 C02 emission factors (see Table 4-89). Because estimates
of sinter production, direct reduced iron production, and pellet production were not available,
production was assumed to equal consumption.
Table 4-89: C02 Emission Factors for Sinter Production, Direct Reduced Iron
Production, and Pellet Production
Material Produced
Metric Ton C02/Metric Ton
Sinter
0.2
Direct Reduced Iron
0.7
Pellet Production
0.03
Source: IPCC (2006), Table 4.1.
The consumption of coking coal, natural gas, distillate fuel, and coal used in iron and steel production
are adjusted for within the Energy chapter to avoid double-counting of emissions reported within the
IPPU chapter as these fuels were consumed during non-energy related activities. More information on
this methodology and examples of adjustments made between the IPPU and Energy chapters are
described in Annex 2.1, Methodology for Estimating Emissions of C02 from Fossil Fuel Combustion.
Sinter consumption and pellet consumption data for 1990 through 2020 were obtained from AISI's
Annual Statistical Report (AISI 2004 through 2022) and through personal communications with AISI
(Steiner 2008) (see Table 4-90). These data from the AISI Annual Statistical Report were withheld for
2020 through 2023, so the 2019 values were used as estimated data for the missing 2020 through 2023
values and adjusted using GHGRP emissions data, as described earlier in this Methodology and Time-
Series Consistency section.
Industrial Processes and Product Use 4-117
-------
In general, DRI consumption data were obtained from the U.S. Geological Survey (USGS) Minerals
Yearbook- Iron and Steel Scrap (USGS 1991 through 2023; USGS 2024b) and personal communication
with the USGS Iron and Steel Commodity Specialist (Tuck 2024). In 2024 the USGS implemented revised
data collection and estimation methodology to more accurately reflect total steel industry consumption
of DRI, ferrous scrap, and pig iron. These improvements resulted in significant increases in estimated
consumption data for 2019 through 2023. Revised data for prior years was not available at the time of
preparation of this chapter. Data for DRI consumed in EAFs were not available for the years 1990 and
1991. EAF DRI consumption in 1990 and 1991 was calculated by multiplying the total DRI consumption
for all furnaces by the EAF share of total DRI consumption in 1992. Data for DRI consumed in BOFs were
not available for the years 1990 through 1993. BOF DRI consumption in 1990 through 1993 was
calculated by multiplying the total DRI consumption for all furnaces (excluding EAFs and cupola) by the
BOF share of total DRI consumption (excluding EAFs and cupola) in 1994.
The Tier 1 C02 emission factors for sinter production, direct reduced iron production and pellet
production were obtained through the 2006IPCC Guidelines (IPCC 2006). Time-series data for pig iron
production, coke, natural gas, fuel oil, sinter, and pellets consumed in the blast furnace; pig iron
production; and blast furnace gas produced at the iron and steel mill and used in the metallurgical coke
ovens and other steel mill activities were obtained from AISI's Annual Statistical Report (AISI 2004
through 2021) and through personal communications with AISI (Steiner 2008) (see Table 4-90 and Table
4-91). Data including blast furnace gas, coke oven gas, natural gas, limestone, sinter, and natural ore
consumption for blast furnaces, coke production, and steelmaking furnaces (EAFs and BOFs) from the
AISI Annual Statistical Report were withheld for 2020 through 2023, so the 2019 values were used as
estimated data for the missing 2020 through 2023 values and adjusted using GHGRP emissions data, as
described earlier in this Methodology and Time-Series Consistency section. Similarly, the percent of
total steel production for EAF and BOF steelmaking processes were withheld for 2021 through 2023, so
the 2020 values were used as estimated data for the missing values and adjusted usingGHGRP
emissions data, as described earlier in this Methodology and Time-Series Consistency section.
Data for EAF steel production, carbon-containing flux, EAF charge carbon, and natural gas consumption
were obtained from AISI's Annual Statistical Report (AISI 2004 through 2022) and through personal
communications with AISI (AISI 2006 through 2016, Steiner 2008). The factor for the quantity of EAF
anode consumed per ton of EAF steel produced was provided by AISI (Steiner 2008). Data for BOF steel
production, carbon-containing flux, natural gas, natural ore, pellet, sinter consumption as well as BOF
steel production were obtained from AISI's Annual Statistical Report (AISI 2004 through 2023) and
through personal communications with AISI (Steiner 2008). Data for EAF consumption of natural gas and
BOF consumption of coke oven gas, limestone, and natural ore from the AISI Annual Statistical Report
were not available for 2021 through 2023, so 2020 values were used as estimated data for the missing
values and adjusted using GHGRP emissions data, as described earlier in this Methodology and Time-
Series Consistency section. Data for EAF and BOF scrap steel, pig iron, and DRI consumption were
obtained from the USGS Minerals Yearbook- Iron and Steel Scrap (USGS 1991 through 2023; USGS
2024b) and personal communication with the USGS Iron and Steel Commodity Specialist (Tuck 2024).
Data on coke oven gas and blast furnace gas consumed at the iron and steel mill (other than in the EAF,
BOF, or blast furnace) were obtained from AISI's Annual Statistical Report (AISI 2004 through 2021) and
through personal communications with AISI (Steiner 2008). These data were not available for 2021
through 2023, so 2020 values were used as estimated data for the missing values and adjusted using
GHGRP emissions data, as described earlier in this Methodology and Time-Series Consistency section.
Some data from the AISI Annual Statistical Report on natural gas consumption were withheld for 2020
4-118 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
through 2023, so the 2019 values were used as estimated data for the missing values and adjusted using
GHGRP emissions data, as described earlier in this Methodology and Time-Series Consistency section.
Data on blast furnace gas and coke oven gas sold for use as synthetic natural gas were obtained from
ElA's Natural Gas Annual 2019 (EIA 2020). Carbon contents for direct reduced iron, EAF carbon
electrodes, EAF charge carbon, limestone, dolomite, pig iron, and steel were provided by the 2006IPCC
Guidelines. The carbon contents for natural gas, fuel oil, and direct injection coal were obtained from
EIA (EIA 2017b) and EPA (EPA 2010). Heat contents for fuel oil and direct injection coal were obtained
from EIA (EIA 1992, 2011); natural gas heat content was obtained from Table 37 of AISI's Annual
Statistical Report (AISI 2004 through 2021). Heat contents for coke oven gas and blast furnace gas were
provided in Table 37 of AISI's Annual Statistical Report (AISI 2004 through 2021) and confirmed by AISI
staff (Carroll 2016).
Table 4-90: Production and Consumption Data for the Calculation of C02 and CH4
Emissions from Iron and Steel Production (Thousand Metric Tons)
Source/Activity Data
1990
2005
2019
2020
2021
2022
2023
Sinter Production
12,239
8,315
4,378
3,747
4,182
3,935
4,060
Direct Reduced Iron Production
517
1,303
C
C
C
C
C
Pellet Production
60,563
50,096
29,262
25,044
27,949
26,300
27,139
Pig Iron Production
Coke Consumption
24,946
13,832
7,291
6,240
6,964
6,553
6,762
Pig Iron Production
49,669
37,222
22,302
18,320
22,246
19,791
21,000
Direct Injection Coal Consumption
1,485
2,573
2,465
2,110
2,354
2,216
2,286
EAF Steel Production
EAF Anode and Charge Carbon Consumption
67
1,127
1,137
1,118
1,130
1,123
1,126
Scrap Steel Consumption
42,691
46,600
C
C
C
C
C
Flux Consumption
319
695
998
998
998
998
1,030
EAF Steel Production
33,511
52,194
61,172
51,349
57,307
53,926
55,645
BOF Steel Production
Pig Iron Consumption
47,307
34,400
C
C
C
C
C
Scrap Steel Consumption
14,713
11,400
C
C
C
C
C
Flux Consumption
576
582
363
311
347
326
337
BOF Steel Production
43,973
42,705
26,591
21,384
23,865
22,457
23,172
C (Confidential)
Table 4-91: Production and Consumption Data for the Calculation of C02 Emissions
from Iron and Steel Production (Million ft3 unless otherwise specified)
Source/Activity Data
1990
2005
2019
2020
2021
2022
2023
Pig Iron Production
Natural Gas Consumption
56,273
59,844
37,934
32,465
36,232
37,387
38,578
Fuel Oil Consumption (thousand gallons)
163,397
16,170
2,321
1,986
2,217
2,086
2,153
Coke Oven Gas Consumption
22,033
16,557
12,926
11,063
12,346
11,618
11,988
Blast Furnace Gas Production
1,439,380
1,299,980
836,033
715,509
798,522
751,418
775,364
EAF Steel Production
Natural Gas Consumption
15,905
19,985
9,115
7,801
8,706
8,192
8,454
Industrial Processes and Product Use 4-119
-------
Source/Activity Data
1990
2005
2019
2020
2021
2022
2023
BOF Steel Production
Coke Oven Gas Consumption
3,851
524
389
333
372
350
361
Other Activities
Coke Oven Gas Consumption
224,883
97,132
64,377
55,096
61,489
57,861
59,705
Blast Furnace Gas Consumption
1,414,778
1,295,520
832,119
712,159
794,783
747,900
771,734
Methodological approaches were applied to the entire time series to ensure consistency in emissions
from 1990 through 2023.
Uncertainty
The estimates of C02 emissions from metallurgical coke production are based on assessing
uncertainties in material production and consumption data and average carbon contents. Uncertainty is
associated with the total U.S. coking coal consumption, total U.S. coke production, and materials
consumed during this process. Data for coking coal consumption and metallurgical coke production are
from different data sources (EIA) than data for other carbonaceous materials consumed at coke plants
(AISI), which does not include data for merchant coke plants. There is uncertainty associated with the
fact that coal tar and coke breeze production were estimated based on coke production because coal
tar and coke breeze production data were not available. Since merchant coke plant data is not included
in the estimate of other carbonaceous materials consumed at coke plants, the mass balance equation
for C02 from metallurgical coke production cannot be reasonably completed; therefore, for the purpose
of this analysis, uncertainty parameters are applied to primary data inputs to the calculation (i.e., coking
coal consumption and metallurgical coke production) only.
The estimates of C02 emissions from iron and steel production are based on material production and
consumption data and average carbon contents. There is uncertainty associated with the assumption
that pellet production, direct reduced iron and sinter consumption are equal to production. There is
uncertainty with the representativeness of the associated IPCC default emission factors. There is
uncertainty associated with the assumption that all coal used for purposes other than coking coal is for
direct injection coal. There is also uncertainty associated with the carbon contents for pellets, sinter,
and natural ore, which are assumed to equal the carbon contents of direct reduced iron, when
consumed in the blast furnace. There is uncertainty associated with the consumption of natural ore
under current industry practices. For EAF steel production, there is uncertainty associated with the
amount of EAF anode and charge carbon consumed due to inconsistent data throughout the time
series. Also for EAF steel production, there is uncertainty associated with the assumption that 100
percent of the natural gas attributed to "steelmaking furnaces" by AISI is process-related and nothing is
combusted for energy purposes. Uncertainty is also associated with the use of process gases such as
blast furnace gas and coke oven gas. Data are not available to differentiate between the use of these
gases for processes at the steel mill versus for energy generation (i.e., electricity and steam generation);
therefore, all consumption is attributed to iron and steel production. These data and carbon contents
produce a relatively accurate estimate of C02 emissions; however, there are uncertainties associated
with each.
For calculating the emissions estimates from iron and steel and metallurgical coke production, EPA
utilizes a number of data points taken from the AISI Annual Statistical Report (ASR). This report serves as
a benchmark for information on steel companies in United States, regardless if they are a member of
4-120 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
AISI, which represents integrated producers (i.e., blast furnace and EAF). During the compilation of the
1990 through 2016 Inventory report EPA initiated conversation with AISI to better understand and update
the qualitative and quantitative uncertainty metrics associated with AISI data elements. AISI estimates
their data collection response rate to range from 75 to 90 percent, with certain sectors of the iron and
steel industry not being covered by the ASR; therefore, there is some inherent uncertainty in the values
provided in the AISI ASR, including material production and consumption data. There is also some
uncertainty to which materials produced are exported to Canada. As indicated in the introduction to this
section, the trend for integrated facilities has moved to more use of EAFs and fewer BOFs. This trend
may not be completely captured in the current data which also increases uncertainty. EPA assigned an
uncertainty range of ±10 percent for the primary data inputs (i.e., consumption and production values
for each production process, heat and carbon content values), a normal probability density function for
consumption and production values for each production process, and a triangular probability density
function for heat and carbon content values to calculate overall uncertainty from iron and steel
production, and using this suggested uncertainty provided in Table 4.4 of the 2006IPCC Guidelines is
appropriate based on expert judgment (RTI 2023). During EPA's discussion with AISI, AISI noted that an
uncertainty range of ±5 percent would be a more appropriate approximation to reflect their coverage of
integrated steel producers in the United States. EPA will continue to assess the best range of uncertainty
for these values. EPA assigned an uncertainty range of ±25 percent and a triangular probability density
function for the Tier 1 C02 emission factors for the sinter, direct reduced iron, and pellet production
processes, and using this suggested uncertainty provided in Table 4.4 of the 2006 IPCC Guidelines is
appropriate based on expert judgment (RTI 2023).
The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 4-92 for
metallurgical coke production and iron and steel production. Total C02 emissions from metallurgical
coke production and iron and steel production for 2023 were estimated to be between 39.0 and 53.5
MMT C02 Eq. at the 95 percent confidence level. This indicates a range of approximately 16 percent
below and 16 percent above the emission estimate of 46.2 MMT C02 Eq. Total CH4 emissions from
metallurgical coke production and iron and steel production for 2023 were estimated to be between
0.006 and 0.01 MMT C02 Eq. at the 95 percent confidence level. This indicates a range of approximately
20 percent below and 21 percent above the emission estimate of 0.008 MMT C02 Eq.
Table 4-92: Approach 2 Quantitative Uncertainty Estimates for C02 and CH4 Emissions
from Iron and Steel Production and Metallurgical Coke Production (MMT C02 Eq. and
Percent)
Source
Gas
2023 Emission Estimate
Uncertainty Range Relative to Emission
Estimate3
(MMT CO2 Eq.)
(MMT CO2 Eq.)
(%)
Lower
Upper
Lower
Upper
Bound
Bound
Bound
Bound
Metallurgical Coke & Iron
and Steel Production
C02
46.2
39.0
53.5
-16%
+ 16%
Metallurgical Coke & Iron
and Steel Production
cm
+
+
+
-20%
+21%
+ Does not exceed 0.05 MMT C02 Eq.
a Range of emission estimates predicted by Monte Carlo stochastic simulation for a 95 percent confidence interval.
Industrial Processes and Product Use 4-121
-------
QA/QC and Verification
General quality assurance/quality control (QA/QC) procedures were applied consistent with the U.S.
Inventory QA/QC plan, which is in accordance with Volume 1, Chapter 6 of 2006IPCC Guidelines as
described in the introduction of the IPPU chapter (see Annex 8 for more details). As part of a multiyear
improvement effort, EPA is reviewing the iron and steel methodology and available data, conducting
additional category specific QC checks and will report on findings when that review is complete (i.e.,
projected to be complete at the earliest for the 2025 report). More information is provided under
Planned Improvements below.
Recalculations Discussion
Recalculations were performed for the emissions estimates for 2019 through 2022 based upon updated
USGS values for DRI, pig iron, and scrap steel consumption for both BOF and EAF steel production. As a
result of improvements to USGS data collection and estimation methodology, estimated consumption
values were increased significantly from those previously presented (Tuck 2024). Additionally, revisions
to GHGRP data for 2020 through 2022 resulted in minor changes to activity data that were adjusted
using GHGRP data, as described in the Methodology and Time-Series Consistency section. The changes
to estimated C02 emissions compared to the previous Inventory are summarized in Table 4-93.
Estimated emissions from production processes not included in the table (i.e., sinter production, pellet
production, and other activities) were not impacted by these recalculations.
These updates resulted in an average annual increase for iron and steel production and metallurgical
coke production of 0.5 MMT C02 Eq. (1.2 percent) in C02 emissions and no change in CH4 emissions
across the time series compared to the previous Inventory.
Table 4-93: Changes from Previous Inventory in C02 Emissions from Iron and Steel
Production (kt C02, % change)
Source/Activity Data
2019
2020
2021
2022
Iron Production
Steel Production
1,954 (+21%)
1,790 (+31%)
1,606 (+19%)
1,349 (+24%)
3,209 (+36%)
2,140 (+37%)
3,631 (+42%)
856 (+13%)
Total
3,743 (+9.3%)
2,954 (+8.3%)
5,349 (+14%)
4,487 (+12%)
Planned Improvements
Significant activity data for 2020 through 2023 were not available for this report and were estimated
using 2019 values and adjusted using GHGRP emissions data. EPA will continue to explore sources of
2020 through 2023 data and other estimation approaches. EPA will evaluate and analyze data reported
under EPA's GHGRP to improve the emission estimates for Iron and Steel Production process categories.
Particular attention will be made to ensure time-series consistency of the emissions estimates
presented in future Inventory reports, consistent with IPCC guidelines. This is required as the facility-
level reporting data from EPA's GHGRP, with the program's initial requirements for reporting of emissions
in calendar year 2010, are not available for all inventory years (i.e., 1990 through 2009) as required for
this Inventory. In implementing improvements and integration of data from EPA's GHGRP, EPA will rely on
4-122 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
the latest guidance from the IPCC on the use of facility-level data in national inventories.66 This is a near
to medium-term improvement, and per preliminary work, EPA estimates that the earliest this
improvement could be incorporated is the next (i.e., 2026) Inventory.
In conjunction with considering use of the GHGRP data to improve estimates, additional improvements
include updated accounting for emission estimates for the production of metallurgical coke including
better accounting for merchant coke plants. Additional efforts will be made to improve the reporting and
transparency in accounting for fuels between the IPPU and Energy chapters. EPA estimates that the
earliest this improvement could be incorporated is the next (i.e., 2026) Inventory.
4.19 Ferroalloy Production (Source Category
2C2)
Ferroalloys are composites of iron (Fe) and other elements such as silicon (Si), manganese (Mn), and
chromium (Cr). This reporting category (2C2) includes emissions of carbon dioxide (C02) and methane
(CH4) from the production of several ferroalloys. Per the IPCC methodological guidance, emissions from
fuels consumed for energy purposes during the production of ferroalloys are accounted for as part of
fossil fuel combustion in the industrial end-use sector reported under the Energy chapter. Emissions
from the production of two types of ferrosilicon (25 to 55 percent and 56 to 95 percent silicon), silicon
metal (96 to 99 percent silicon), and miscellaneous alloys (32 to 65 percent silicon) have been
calculated.
Emissions from the production of ferrochromium and ferromanganese are not included because of the
small number of manufacturers of these materials in the United States. Government information
disclosure rules prevent the publication of production data for these production facilities. Additionally,
production of ferrochromium in the United States ceased in 2009 (USGS 2013).
Similar to emissions from the production of iron and steel, C02 is emitted when metallurgical coke is
oxidized during a high-temperature reaction with iron and the selected alloying element. Due to the
strong reducing environment, CO is initially produced and eventually oxidized to C02. A representative
reaction equation for the production of 50 percent ferrosilicon (FeSi) is given below:
Fe203 + 2Si02 + 7C —> 2FeSi + 7C0
While most of the carbon contained in the process materials is released to the atmosphere as C02, a
percentage is also released as CH4 and other volatiles. The amount of CH4 that is released is dependent
on furnace efficiency, operation technique, and control technology.
Ferroalloys are used to alter the material properties of the steel. Ferroalloys are produced in conjunction
with the iron and steel industry, often at co-located facilities, and production trends closely follow that
of the iron and steel industry. As of 2021, 11 facilities in the United States produce ferroalloys (USGS
2024b).
66 See http://www.ipcc-nggip.iges.or.jp/public/tb/TFI Technical Bulletin 1.pdf and the 2019 Refinement, Volume 1,
Chapter 2, Section 2.3, Use of Facility Data in Inventories at https://www.ipcc-
nggip.iges.or.jp/public/2019rf/pdf/1 Volume1/19R V1 Ch02 DataCollection.pdf.
Industrial Processes and Product Use 4-123
-------
Emissions of C02 from ferroalloy production in 2023 were 1.2 MMT C02 Eq. (1,245 kt C02) (see Table
4-94 and Table 4-95), which is a 6 percent reduction since 2022 and a 42 percent reduction since 1990
Emissions of CH4 from ferroalloy production in 2023 were 0.01 MMT C02 Eq. (0.3 kt CH4), which is a 6
percent decrease since 2022 and a 49 percent decrease since 1990. Variability in emissions over the
past five years is attributable to one facility shutting down in 2020 (USGS 2021) and reopening in 2021,
owing to increased demand for ferrosilicon products and improved domestic pricing (USGS 2022).
Table 4-94: C02 and CH4 Emissions from Ferroalloy Production (MMT C02 Eq.)
Gas 1990
2005
2019
2020
2021
2022
2023
CO2 2.2
1.4
1.6
1.4
1.4
1.3
1.2
cm +
+
+
+
+
+
+
Total 2.2
1.4
1.6
1.4
1.4
1.3
1.3
+ Does not exceed 0.05 MMT C02 Eq.
Note: Totals may not sum due to independent rounding.
Table 4-95: C02andCH4
Emissions from Ferroalloy Production (kt)
Gas 1990
2005
2019
2020
2021
2022
2023
CO2 2,152
1,392
1,598
1,377
1,426
1,327
1,245
CH4 1
+
+
+
+
+
+
+ Does not exceed 0.05 MMT C02 Eq.
Methodology and Time-Series Consistency
Emissions of C02 and CH4 from ferroalloy production are calculated67 using a Tier 1 method from the
2006IPCC Guidelines, in accordance with the IPCC methodological decision tree and available data.
Annual ferroalloy production is multiplied by material-specific emission factors provided by IPCC (IPCC
2006). The Tier 1 equations for C02 and CH4 emissions are as follows:
Equation 4-12:2006 IPCC Guidelines Tier 1: C02 Emissions for Ferroalloy Production
(Equation 4.15)
Eco2 = x EF^
i
where,
Ec02 = C02 emissions, metric tons
MPi = Production of ferroalloy type/', metric tons
EFi = Generic emission factor for ferroalloy type /', metric tons C02/metric ton specific
ferroalloy product
67 EPA has not integrated aggregated facility-level GHGRP information to inform these estimates. The aggregated
information (e.g., activity data and emissions) associated with production of ferroalloys did not meet criteria to shield
underlying confidential business information (CBI) from public disclosure.
4-124 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Equation 4-13:2006IPCC Guidelines Tier 1: CH4 Emissions for Ferroalloy Production
(Equation 4.18)
Default emission factors were used because country-specific emission factors are not currently
available. The following emission factors were used to develop annual C02 and CH4 estimates:
Ferrosilicon, 25 to 55 percent Si and Miscellaneous Alloys, 32 to 65 percent Si: 2.5 metric tons
C02/metric ton of alloy produced; 1.0 kg CH4/metric ton of alloy produced.
Ferrosilicon, 56 to 95 percent Si: 4.0 metric tons C02/metric ton alloy produced; 1.0 kg CH4/metric ton of
alloy produced.
Silicon Metal: 5.0 metric tons C02/metric ton metal produced; 1.2 kg CH4/metric ton metal produced.
It was assumed that 100 percent of the ferroalloy production was produced using petroleum coke in an
electric arc furnace process (IPCC 2006), although some ferroalloys may have been produced with
coking coal, wood, other biomass, or graphite carbon inputs. The amount of petroleum coke consumed
in ferroalloy production was calculated assuming that the petroleum coke used is 90 percent carbon (C)
and 10 percent inert material (Onder and Bagdoyan 1993).
The use of petroleum coke for ferroalloy production is adjusted for within the Energy chapter as this fuel
was consumed during non-energy related activities. Additional information on the adjustments made
within the Energy sector for non-energy use of fuels is described in both the Methodology section of C02
from Fossil Fuel Combustion (3.1 Fossil Fuel Combustion [Source Category 1 A]) and Annex 2.1,
Methodology for Estimating Emissions of C02 from Fossil Fuel Combustion.
Ferroalloy production data for 1990 through 2022 (see Table 4-96) were obtained from the U.S.
Geological Survey (USGS) through the Minerals Yearbook: Silicon (USGS 1996 through 2023) and the
Minerals Industry Survey: Silicon (USGS 2024a). The following data were available from the USGS
publications for the time series:
• Ferrosilicon, 25 to 55 percent Si: Annual production data were available from 1990 through
2010.
• Ferrosilicon, 56 to 95 percent Si: Annual production data were available from 1990 through
2010.
• Silicon Metal: Annual production data were available from 1990 through 2005. Production data
for 2005 were used as estimates for 2006 through 2010 because data for these years were not
available due to government information disclosure rules.
• Miscellaneous Alloys, 32 to 65 percent Si: Annual production data were available from 1990
through 1998. Starting 1999, USGS reported miscellaneous alloys and ferrosilicon containing 25
to 55 percent silicon as a single category.
where,
CH4 emissions, kg
Production of ferroalloy type/', metric tons
Generic emission factor for ferroalloy type /', kg CH4/metric ton specific ferroalloy
product
Industrial Processes and Product Use 4-125
-------
Because production data for 2023 was withheld to avoid disclosing proprietary information (USGS
2024a), production data for 2022 was used as proxy for 2023 data. The EPA then used process
emissions data (metric tons) from the EPA's Greenhouse Gas Reporting Program (GHGRP) Subpart K for
ferroalloys to adjust the 2022 production values. For reference, the annual GHGRP emissions from
ferroalloys were 6.2 percent less in 2023 than in 2022 (EPA 2024).
Starting with the 2011 publication, USGS ceased publication of production quantity by ferroalloy
product and began reporting all the ferroalloy production data as a single category (i.e., Total Silicon
Materials Production). This is due to the small number of ferroalloy manufacturers in the United States
and government information disclosure rules. Ferroalloy product shares developed from the 2010
production data (i.e., ferroalloy product production divided by total ferroalloy production) were used
with the total silicon materials production quantity to estimate the production quantity by ferroalloy
product type for 2011 through 2023.
Table 4-96: Production of Ferroalloys (Metric Tons)
Year
1990
2005
2019
2020
2021
2022
2023
Ferrosilicon 25%-55%
321,385
123,000
147,034
126,681
131,280
122,119
114,581
Ferrosilicon 56%-95%
109,566
86,100
129,736
111,778
115,835
107,752
101,101
Silicon Metal
145,744
148,000
142,229
122,541
126,989
118,128
110,837
Misc. Alloys 32-65%
72,442
NA
NA
NA
NA
NA
NA
Methodological approaches were applied to the entire time series to ensure consistency in emissions
from 1990 through 2023.
Uncertainty
Annual ferroalloy production was reported by the USGS in three broad categories until the 2010
publication: ferroalloys containing 25 to 55 percent silicon (including miscellaneous alloys), ferroalloys
containing 56 to 95 percent silicon, and silicon metal (through 2005 only, 2005 value used as an
estimate for 2006 through 2010). Starting with the 2011 Minerals Yearbook: Silicon, USGS started
reporting all the ferroalloy production under a single category: total silicon materials production. The
total silicon materials quantity was allocated across the three categories, based on the 2010 production
shares for the three categories. Refer to the Methodology section for further details. Additionally,
production data for silvery pig iron (alloys containing less than 25 percent silicon) are not reported by the
USGS to avoid disclosing proprietary company data. Emissions from this production category, therefore,
were not estimated.
Some ferroalloys may be produced using wood or other biomass as a primary or secondary carbon
source (carbonaceous reductants); however, information and data regarding these practices were not
available. Emissions from ferroalloys produced with wood or other biomass would not be counted under
this source because wood-based carbon is of biogenic origin.68 Even though emissions from ferroalloys
produced with coking coal or graphite inputs would be counted in national trends, they may be
generated with varying amounts of C02 per unit of ferroalloy produced. The most accurate method for
these estimates would be to base calculations on the amount of reducing agent used in the process,
Emissions and sinks of biogenic carbon are accounted for in the Land Use, Land-Use Change, and Forestry chapter.
4-126 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
rather than the amount of ferroalloys produced. These data, however, were not available, and are also
often considered confidential business information.
Emissions of CH4 from ferroalloy production will vary depending on furnace specifics, such as type,
operation technique, and control technology. Higher heating temperatures and techniques such as
sprinkle charging would reduce CH4 emissions; however, specific furnace information was not available
or included in the CH4 emission estimates.
EPA assigned a uncertainty range of ±25 percent for the primary emission factors (i.e., ferrosilicon 25-
55% Si, ferrosilicon 56-95% Si, and silicon metal), and an uncertainty range of ±5 percent for the 2010
production values for ferrosilicon 25-55% Si, ferrosilicon 56-95% Si, and silicon metal production and
the 2021 total silicon materials production value used to calculate emissions from overall ferroalloy
production. Using these suggested uncertainties provided in in Table 4.9 of Section 4.3.3.2 of the 2006
IPCC Guidelines is appropriate based on expert judgment (RTI 2023). Per this expert judgment, a normal
probability density function was assumed for all activity data, and a triangular probability density
function was assumed for emission factors.
The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 4-97. Ferroalloy
production C02 emissions from 2023 were estimated to be between 1.1 and 1.4 MMT C02 Eq. at the 95
percent confidence level. This indicates a range of approximately 13 percent below and 13 percent
above the emission estimate of 1.2 MMT C02 Eq. Ferroalloy production CH4 emissions were estimated
to be between 0.009 and 0.011 at the 95 percent confidence level. This indicates a range of
approximately 12 percent below and 13 percent above the emission estimate of 0.01 MMT C02 Eq.
Table 4-97: Approach 2 Quantitative Uncertainty Estimates for C02 Emissions from
Ferroalloy Production (MMT C02 Eq. and Percent)
2023 Emission
Uncertainty Range Relative to Emission
Source
Gas
Estimate
Estimate3
(MMT CO2 Eq.)
(MMT CO2 Eq.)
(%)
Lower
Upper
Lower
Upper
Bound
Bound
Bound
Bound
Ferroalloy Production
C02
1.2
1.1
1.4
-13%
+ 13%
Ferroalloy Production
cm
+
+
+
-12%
+ 13%
+ Does not exceed 0.05 MMT C02 Eq.
a Range of emission estimates predicted by Monte Carlo stochastic simulation for a 95 percent confidence interval.
QA/QC and Verification
General quality assurance/quality control (QA/QC) procedures were applied consistent with the U.S.
Inventory QA/QC plan, which is in accordance with Volume 1, Chapter 6 of 2006 IPCC Guidelines as
described in the introduction of the IPPU chapter (see Annex 8 for more details).
Recalculations Discussion
Recalculations were implemented for 2021 based on revised USGS production data. The 2021
production data, which had been previously released, were withheld in the Minerals Yearbook: Silicon
for 2022 to avoid disclosing priority data (USGS 2023c). To estimate revised production data for 2021,
the EPA used process emissions data from the GHGRP Subpart K for ferroalloys to adjust the 2020
production values. GHGRP process emissions data increased by 3.6 percent from 2020 to 2021 (EPA
Industrial Processes and Product Use 4-127
-------
2024). Compared to the previous Inventory, emissions decreased by 9 percent (141 kt C02) for 2021, as
a result of the revised production values. USGS production data will be reimplemented when it becomes
available.
Planned Improvements
Pending available resources and prioritization of improvements for more significant sources, EPA will
continue to evaluate and analyze data reported under EPA's GHGRP that would be useful to improve the
emission estimates and category-specific QC procedures for the Ferroalloy Production source category.
Given the small number of facilities and reporting thresholds, particular attention will be made to ensure
completeness and time-series consistency of the emissions estimates presented in future Inventory
reports. This is required as the facility-level reporting data from EPA's GHGRP, with the program's initial
requirements for reporting of emissions in calendar year 2010, are not available for all inventory years
(i.e., 1990 through 2009) as required for this Inventory. In implementing improvements and integration of
data from EPA's GHGRP, the latest guidance from the IPCC on the use of facility-level data in national
inventories will be relied upon.69 This is a long-term planned improvement, and EPA is still assessing the
possibility of incorporating this improvement into the Inventory. This improvement has not been
included in the current Inventory report.
4.20 Aluminum Production (Source Category
2C3)
Aluminum is a lightweight, malleable, and corrosion-resistant metal that is used in many manufactured
products, including aircraft, automobiles, bicycles, and kitchen utensils. As of recent reporting, the
United States was the eleventhl 870 largest producer of primary aluminum with an estimated aluminum
production of 750 thousand metric tons, with approximately 1.1 percent of the world total production
(USGS 2024). The United States was also a major importer of primary aluminum. This reporting category
(2C3) includes emissions from the production of primary aluminum—in addition to consuming large
quantities of electricity—results in process-related emissions of carbon dioxide (C02) and two
perfluorocarbons (PFCs): perfluoromethane (CF4) and perfluoroethane (C2F6).
Carbon dioxide is emitted during the aluminum smelting process when alumina (aluminum oxide, Al203)
is reduced to aluminum using the Hall-Heroult reduction process. The reduction of the alumina occurs
through electrolysis in a molten bath of natural or synthetic cryolite (Na3AlF6). The reduction cells
contain a carbon (C) lining that serves as the cathode. Carbon is also contained in the anode, which can
be a carbon mass of paste, coke briquettes, or prebaked carbon blocks from petroleum coke. During
reduction, most of this carbon is oxidized and released to the atmosphere as C02.
69 See http://www.ipcc-nggip.iges.or.jp/public/tb/TFI Technical Bulletin 1.pdf and the 2019 Refinement, Volume 1,
Chapter 2, Section 2.3, Use of Facility Data in Inventories at https://www.ipcc-
nggip.iges.or.ip/public/2019rf/pdf/1 Volume1/19R V1 Ch02 DataCollection.pdf.
70 Based on the U.S. USGS (2024) Aluminum factsheet, assuming all countries grouped under the "other countries"
categories all have lower production than the U.S. Available at: https://pubs.usgs.gov/periodicals/mcs2024/mcs2024-
aluminum.pdf
4-128 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Process emissions of C02 from aluminum production were estimated to be 1.2 MMT C02 Eq. (1,237 kt)
in 2023 (see Table 4-98 and Table 4-99). The carbon anodes consumed during aluminum production
consist of petroleum coke and, to a minor extent, coal tar pitch. The petroleum coke portion of the total
C02 process emissions from aluminum production is considered to be a non-energy use of petroleum
coke and is accounted for here and not under the C02 from fossil fuel combustion source category of
the Energy sector. Similarly, the coal tar pitch portion of these C02 process emissions is accounted for
here.
Table 4-98: C02 Emissions from Aluminum Production (MMT C02 Eq.)
Year 1990
2005
2019 2020
2021
2022
2023
Aluminum Production 6.8
4.1
1.9 1.7
1.5
1.4
1.2
Table 4-99: C02 Emissions from Aluminum Production (kt C02)
Year 1990
2005
2019 2020
2021
2022
2023
Aluminum Production 6,831
4,142
1,880 1,748
1,541
1,446
1,237
In addition to C02 emissions, the aluminum production industry is also a source of PFC emissions.
During the smelting process, when the alumina ore content of the electrolytic bath falls below critical
levels required for electrolysis, rapid voltage increases occur, which are termed High Voltage Anode
Effects (HVAEs). HVAEs cause carbon from the anode and fluorine from the dissociated molten cryolite
bath to combine, thereby producing fugitive emissions of CF4 and C2F6. In general, the magnitude of
emissions for a given smelter and level of production depends on the frequency and duration of these
anode effects. As the frequency and duration of the anode effects increase, emissions increase.
Another type of anode effect, Low Voltage Anode Effects (LVAEs), became a concern in the early 2010s
as the aluminum industry increasingly began to use cell technologies with higher amperage and
additional anodes (IPCC 2019). LVAEs emit CF4 and are included in PFC emission totals from 2006
forward.
Since 1990, emissions of CF4 and C2F6 have both declined by 97 and 99 percent respectively, to
0.42 MMT C02 Eq. of CF4 (0.1 kt) and 0.04 MMT C02 Eq. of C2F6 (0.004 kt) in 2023, respectively, as shown
in Table 4-100 and Table 4-101. This decline is due both to reductions in domestic aluminum production
and to actions taken by aluminum smelting companies to reduce the frequency and duration of anode
effects. These actions include technology and operational changes such as employee training, use of
computer monitoring, and changes in alumina feeding techniques. Since 1990, aluminum production
has declined by 81 percent, while the combined CF4 and C2F6 emission rate (per metric ton of aluminum
produced) has been reduced by 87 percent. PFC emissions decreased by approximately 39 percent
between 2022 and 2023.
Table 4-100: PFC Emissions from Aluminum Production (MMT C02 Eq.)
Gas
1990
2005
2019
2020
2021
2022
2023
LL
O
16.1
2.6
1.1
1.2
0.8
0.7
0.4
C2F6
3.2
0.5
0.3
0.2
0.1
0.1
+
Total
19.3
3.1
1.4
1.4
0.9
0.8
0.5
+ Does not exceed 0.05 MMT C02 Eq.
Note: Totals may not sum due to independent rounding.
Industrial Processes and Product Use 4-129
-------
Table 4-101: PFC Emissions from Aluminum Production (kt)
Gas
1990
2005
2019
2020
2021
2022
2023
LL
O
2
+
+
+
+
+
+
C2F6
+
+
+
+
+
+
+
+ Does not exceed 0.5 kt.
In 2023, U.S. primary aluminum production totaled approximately 0.75 million metric tons, a 13 percent
decrease from 2022 production levels (USGS 2024). In 2023, three companies managed production at
five operational primary aluminum smelters across five states. Two smelters operated at full capacity
during 2023. The other three smelters operated at reduced capacity. A sixth smelter in Kentucky has
been temporarily shutdown since 2022 (USGS 2024). Domestic smelters were operating at about 55
percent of capacity of 1.36 million tons per year at year end 2023 (USGS 2024).
Methodology and Time-Series Consistency
Process C02 and PFC (i.e., CF4 and C2F6) emission estimates from primary aluminum production for
2010 through 2023 are available from EPA's GHGRP Subpart F (Aluminum Production) (EPA 2024). Under
EPA's GHGRP, facilities began reporting primary aluminum production process emissions (for 2010) in
2011; as a result, GHGRP data (for 2010 through 2023) are available to be incorporated into the
Inventory. EPA's GHGRP mandates that all facilities that contain an aluminum production process must
report: CF4 and C2F6 emissions from anode effects in all prebake and Soderberg electrolysis cells, C02
emissions from anode consumption during electrolysis in all prebake and Soderberg cells, and all C02
emissions from onsite anode baking. To estimate the process emissions, EPA's GHGRP uses the
process-specific equations detailed in Subpart F (aluminum production).71 These equations are based
on the Tier 2/Tier 3 IPCC (2006) methods for primary aluminum production, and Tier 1 methods when
estimating missing data elements. It should be noted that the same methods (i.e., 2006 IPCC
Guidelines) were used for estimating the emissions prior to the availability of the reported GHGRP data
in the Inventory. Prior to 2010, aluminum production data were provided through EPA's Voluntary
Aluminum Industrial Partnership (VAIP).
As previously noted, the use of petroleum coke for aluminum production is adjusted for within the
Energy chapter to avoid double counting emissions as this fuel was consumed during non-energy
related activities. Additional information on the adjustments made within the Energy sector for non-
energy use of fuels is described in both the Methodology section of C02 from Fossil Fuel Combustion
(3.2 Carbon Emitted from Non-Energy Uses of Fossil Fuels [Source Category 1 A]) and Annex 2.3,
Methodology for Estimating Carbon Emitted from Non-Energy Uses of Fossil Fuels.
Process CO2 Emissions from Anode Consumption and Anode Baking
Carbon dioxide emission estimates for the years prior to the introduction of EPA's GHGRP in 2010 were
estimated using 2006 IPCC Guidelines methods, but individual facility reported data were combined
with process-specific emissions modeling. These estimates were based on information previously
gathered from EPA's Voluntary Aluminum Industrial Partnership (VAIP) program, U.S. Geological Survey
(USGS) Mineral Commodity reviews, and The Aluminum Association (USAA) statistics, among other
71 Code of Federal Regulations, Title 40: Protection of Environment, Part 98: Mandatory Greenhouse Gas Reporting,
Subpart F—Aluminum Production. See https://www.ecfr.gov/cgi-bin/text-
idx?SID=?4a41781dfe4?18h339e914de03e87?7&mc=true&node=pt40.?3.98&rgn=div5#sp40.?3.98.f.
4-130 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
sources. Since pre- and post-GHGRP estimates use the same methodology, emission estimates are
comparable across the time series.
Most of the C02 emissions released during aluminum production occur during the electrolysis reaction
of the carbon anode, as described by the following reaction:
2A1203 + 3C -> 4A1 + 3C02
For prebake smelter technologies, C02 is also emitted during the anode baking process. These
emissions can account for approximately 10 percent of total process C02 emissions from prebake
smelters.
Depending on the availability of smelter-specific data, the C02 emitted from electrolysis at each smelter
was estimated from: (1) the smelter's annual anode consumption, (2) the smelter's annual aluminum
production and rate of anode consumption (per ton of aluminum produced) for previous and/or
following years, or (3) the smelter's annual aluminum production and IPCC default C02 emission
factors. The first approach tracks the consumption and carbon content of the anode, assuming that all
carbon in the anode is converted to C02. Sulfur, ash, and other impurities in the anode are subtracted
from the anode consumption to arrive at a carbon consumption figure. This approach corresponds to
either the IPCC Tier 2 or Tier 3 method, depending on whether smelter-specific data on anode impurities
are used. The second approach interpolates smelter-specific anode consumption rates to estimate
emissions duringyears for which anode consumption data are not available. This approach avoids
substantial errors and discontinuities that could be introduced by reverting to Tier 1 methods for those
years. The last approach corresponds to the IPCC Tier 1 method (IPCC 2006) and is used in the absence
of present or historic anode consumption data.
The equations used to estimate C02 emissions in the Tier 2 and 3 methods vary depending on smelter
type (IPCC 2006). For Prebake cells, the process formula accounts for various parameters, including net
anode consumption, and the sulfur, ash, and impurity content of the baked anode. For anode baking
emissions, the formula accounts for packing coke consumption, the sulfur and ash content of the
packing coke, as well as the pitch content and weight of baked anodes produced. For Soderberg cells,
the process formula accounts for the weight of paste consumed per metric ton of aluminum produced,
and pitch properties, including sulfur, hydrogen, and ash content.
Through the VAIP, anode consumption (and some anode impurity) data have been reported for 1990,
2000, 2003, 2004, 2005, 2006, 2007, 2008, and 2009. Where available, smelter-specific process data
reported under the VAIP were used; however, if the data were incomplete or unavailable, information
was supplemented using industry average values recommended by IPCC (2006). Smelter-specific C02
process data were provided by 18 of the 23 operating smelters in 1990 and 2000, by 14 out of 16
operating smelters in 2003 and 2004,14 out of 15 operating smelters in 2005, 13 out of 14 operating
smelters in 2006, 5 out of 14 operating smelters in 2007 and 2008, and 3 out of 13 operating smelters in
2009. For years where C02 emissions data or C02 process data were not reported by these companies,
estimates were developed through linear interpolation, and/or assuming representative (e.g., previously
reported or industry default) values.
In the absence of any previous historical smelter-specific process data (i.e., 1 out of 13 smelters in 2009;
1 out of 14 smelters in 2006, 2007, and 2008; 1 out of 15 smelters in 2005; and 5 out of 23 smelters
between 1990 and 2003), C02 emission estimates were estimated using Tier 1 Soderberg and/or
Prebake emission factors (metric ton of C02 per metric ton of aluminum produced) from IPCC (2006).
Industrial Processes and Product Use 4-131
-------
Process PFC Emissions from Anode Effects
High Voltage Anode Effects
Smelter-specific PFC emissions from aluminum production for 2010 through 2023 were reported to EPA
under its GHGRP. To estimate their PFC emissions from HVAEs and report them under EPA's GHGRP,
smelters use an approach identical to the Tier 3 approach in the 2006IPCC Guidelines (IPCC 2006).
Specifically, they use a smelter-specific slope coefficient as well as smelter-specific operating data to
estimate an emission factor using the following equation:
PFC = S XAE
AE = F XD
where,
PFC
CF4 or C2F6, kg/MT aluminum
S
Slope coefficient, PFC/AE
AE
Anode effect, minutes/cell-day
F
Anode effect frequency per cell-day
D
Anode effect duration, minutes
They then multiply this emission factor by aluminum production to estimate PFC emissions from HVAEs.
All U.S. aluminum smelters are required to report their emissions under EPA's GHGRP.
Perfluorocarbon emissions for the years prior to 2010 were estimated using the same equation, but the
slope-factor used for some smelters was technology-specific rather than smelter-specific, making the
method a Tier 2 rather than a Tier 3 approach for those smelters. Emissions and background data were
reported to EPA under the VAI P. For 1990 through 2009, smelter-specific slope coefficients were
available and were used for smelters representing between 30 and 94 percent of U.S. primary aluminum
production. The percentage changed from year to year as some smelters closed or changed hands and
as the production at remaining smelters fluctuated. For smelters that did not report smelter-specific
slope coefficients, IPCC technology-specific slope coefficients were applied (IPCC 2006). The slope
coefficients were combined with smelter-specific anode effect data collected by aluminum companies
and reported under the VAIP to estimate emission factors over time. For 1990 through 2009, smelter-
specific anode effect data were available for smelters representing between 80 and 100 percent of U.S.
primary aluminum production. Where smelter-specific anode effect data were not available,
representative values (e.g., previously reported or industry averages) were used.
For all smelters, emission factors were multiplied by annual production to estimate annual emissions at
the smelter level. For 1990 through 2009, smelter-specific production data were available for smelters
representing between 30 and 100 percent of U.S. primary aluminum production. (For the years after
2000, this percentage was near the high end of the range.) Production at non-reporting smelters was
estimated by calculating the difference between the production reported under VAIP and the total U.S.
production supplied by USGS, and then allocating this difference to non-reporting smelters in
proportion to their production capacity. Emissions were then aggregated across smelters to estimate
national emissions (see Table 4-105).
4-132 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Table 4-102: Summary of HVAE Emissions (MMT C02 Eq.)
Year
1990
2005
2019
2020
2021
2022
2023
HVAE Emissions
19.3
3.1
1.4
1.4
0.9
0.7
0.4
Low Voltage Anode Effects
LVAE emissions of CF4 were estimated for 2006 through 2023 (see Table 4-106) based on the Tier 1
(technology-specific, production-based) method in the 207 9 Refinement to the 2006IPCC Guidelines
for National Greenhouse Gas Inventories (IPCC 2019). Prior to 2006, LVAE emissions are believed to
have been negligible.72 The Tier 1 method is used in the LVAE emissions calculations from aluminum
production in the absence of smelter-specific data available to quantify the LVAE-specific process
emissions. National aluminum production estimates (allocated to smelters as described below) and the
technology used in individual smelters were the best available data to perform the emissions
calculations, as smelter-specific production data is not publicly available.
The following equation was used to estimate LVAE PFC emissions:
Equation 4-14: CF4 Emissions Resulting from Low Voltage Anode Effects
LVAEEcp4 = LVAE EFCF4 X MP
where,
LVAE ECf4 = LVAE emissions of CF4 from aluminum production, kg CF4
LVAE EFCf4 = LVAE emission factor for CF4 (default by cell technology type)
MP = Metal production by cell technology type, tons Al.
In the LVAE emissions calculations, the Metal Production (MP) factor is calculated differently for the
years 2006 through 2009 than for 2010 and beyond. For years prior to GHGRP reporting (2006 through
2009), the MP factor is calculated by dividing the annual production reported by USGS with the total U.S.
capacity reported for this specific year, based on the USGS yearbook and applying this national
utilization factor to each facility's production capacity to obtain an estimated facility production value.
For GHGRP reporting years (2010+), the methodology to calculate the MP value was changed to allocate
the total annual production reported by USAA, based on the distribution of C02 emissions amongst the
operating smelters in a specific year. The latter improves the accuracy of the LVAE emissions estimates
over assuming capacity utilization is the same at all smelters. The main drawback of using this
methodology to calculate the MP factor is that, in some instances, it led to production estimates that
are slightly larger (less than six percent) than the production capacity reported that year. In practice, this
is most likely explained by the differences in process efficiencies at each facility and to a lesser extent,
differences in measurements and methods used by each facility to obtain their C02 estimates and the
degree of uncertainty in the USGS annual production reporting.
Once LVAE emissions were estimated, they were then combined with HVAE emissions estimates to
calculate total PFC emissions from aluminum production.
72 The 2019 Refinement states, "Since 2006, the global aluminum industry has undergone changes in technology and
operating conditions that make LVAE emissions much more prevalent12; these changes have occurred not only through
uptake of newer technologies (e.g., PFPBl to PFPBm) but also during upgrades within the same technology in order to
maximize productivity and reduce energy use" (IPCC 2019). Footnote #12 uses the example of PFPBl, which is prevalent
in the United States, as an older technology that has been upgraded.
Industrial Processes and Product Use 4-133
-------
Table 4-103: Summary of LVAE Emissions (MMT C02 Eq.)
Year
2006
2019
2020
2021
2022
2023
LVAE Emissions
0.1
0.1
0.1
0.1
0.1
+
+ Does not exceed 0.05 MMT C02 Eq.
Production Data
Between 1990 and 2009, production data were provided under the VAIP by 21 of the 23 U.S. smelters
that operated during at least part of that period. For the non-reporting smelters, production was
estimated based on the difference between reporting smelters and national aluminum production levels
as reported to USGS, with allocation to specific smelters based on reported production capacities
(USGS 1990 through 2009).
National primary aluminum production data for 2010 through 2023 were compiled using USGS Mineral
Industry Surveys, and the USGS Mineral Commodity Summaries (see Table 4-107).
Table 4-104: Production of Primary Aluminum (kt)
Year
1990
2005
2019
2020
2021
2022
2023
Production (kt)
4,048
2,481
1,093
1,012
889
861
750
Methodological approaches were applied to the entire time-series to ensure time-series consistency
from 1990 through 2023.
Uncertainty
Uncertainty was estimated for the C02, CF4, and C2F6 emission values reported by each individual
facility to EPA's GHGRP, taking into consideration the uncertainties associated with aluminum
production, anode effect minutes, and slope factors. The uncertainty bounds used for these parameters
were established based on information collected under the VAIP and held constant through 2023.
Uncertainty surrounding the reported C02, CF4, and C2F6 emission values were determined to have a
normal distribution with uncertainty ranges of approximately 6 percent below to 6 percent above, 16
percent below to 16 percent above, and 20 percent below to 20 percent above their 2023 emission
estimates, respectively.
For LVAE, since emission values were not reported through EPA's GHGRP but estimated instead through
a Tier 1 methodology, the uncertainty analysis examined uncertainty associated with primary capacity
data as well as technology-specific emission factors. Uncertainty for each facility's primary capacity,
reported in the USGS Yearbook, was estimated to have a Pert Beta distribution with an uncertainty range
of 7 percent below to 7 percent above the capacity estimates based on the uncertainty of reported
capacity data, the number of years since the facility reported new capacity data, and uncertainty in
capacity utilization. Uncertainty was applied to LVAE emission factors according to technology using the
uncertainty ranges provided in the 2019 Refinement to the 2006IPCC Guidelines. An uncertainty range
for Horizontal Stud Soderberg (HSS) technology was not provided in the 2019 Refinement to the 2006
IPCC Guidelines due to insufficient data, so a normal distribution and uncertainty range of ±99 percent
was applied for that technology based on expert judgment. A Monte Carlo analysis was applied to
estimate the overall uncertainty of the C02, CF4, and C2F6 emission estimates for the U.S. aluminum
industry as a whole, and the results are provided below.
4-134 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
The results of this Approach 2 quantitative uncertainty analysis are summarized in Table 4-105.
Aluminum production-related C02 emissions were estimated to be between 1.21 and 1.28 MMT C02 Eq.
at the 95 percent confidence level. This indicates a range of approximately 3 percent below to 3 percent
above the emission estimate of 1.243 MMT C02 Eq. Also, production-related CF4 emissions were
estimated to be between 0.38 and 0.46 MMT C02 Eq. at the 95 percent confidence level. This indicates a
range of approximately 9 percent below to 11 percent above the emission estimate of 0.415 MMT C02
Eq. Aluminum production-related C2F6 emissions were estimated to be between 0.04 and 0.05 MMT C02
Eq. at the 95 percent confidence level. This indicates a range of approximately 9 percent below to 9
percent above the emission estimate of 0.043 MMT C02 Eq. Finally, Aluminum production-related
aggregated PFCs emissions were estimated to be between 0.42 and 0.50 MMT C02 Eq. at the 95 percent
confidence level. This indicates a range of approximately 8 percent below to 10 percent above the
emission estimate of 0.459 MMT C02 Eq.
Table 4-105: Approach 2 Quantitative Uncertainty Estimates for C02 and PFC
Emissions from Aluminum Production (MMT C02 Eq. and Percent)
2023 Emission
Source
Gas
Estimate
Uncertainty Range Relative to Emission Estimate"
(MMT CO2 Eq.)
(MMT CO2 Eq.)
(%)
Lower
Upper
Lower
Upper
Bound
Bound
Bound
Bound
Aluminum
Production
C02
1.243
1.21
1.28
-3%
+3%
Aluminum
Production
LL
O
0.415
0.38
0.46
-9%
+ 11%
Aluminum
Production
C2F6
0.043
0.04
0.05
-9%
+9%
Aluminum
Production
PFCs
0.459
0.42
0.50
-8%
+ 10%
a Range of emission estimates predicted by Monte Carlo stochastic simulation for a 95 percent confidence interval.
QA/QC and Verification
General quality assurance/quality control (QA/QC) procedures were applied consistent with the U.S.
Inventory QA/QC plan, which is in accordance with Volume 1, Chapter 6 of 2006IPCC Guidelines as
described in the introduction of the IPPU chapter (see Annex 8 for more details). For the GHGRP data,
EPA verifies annual facility-level reports through a multi-step process (e.g., including a combination of
pre-and post-submittal electronic checks and manual reviews by staff) to identify potential errors and
ensure that data submitted to EPA are accurate, complete, and consistent (EPA 20 1 5).73 Based on the
results of the verification process, EPA follows up with facilities to resolve mistakes that may have
occurred. The post-submittals checks are consistent with a number of general and category-specific QC
procedures, including: range checks, statistical checks, algorithm checks, and year-to-year checks of
reported data and emissions.
73 GHGRP Report Verification Factsheet. See https://www.epa.gov/sites/production/files/7015-
07/rtocuments/ghgrp verification factsheet.pdf.
Industrial Processes and Product Use 4-135
-------
Recalculations Discussion
Duplicate emission data were identified and removed from GHGRP Subpart F for Century Aluminum
facility of South Carolina Inc. for the years 2010 to 2015. Additionally, the total aluminum production for
2022 was updated from 860,000 metric tons to 861,000 metric tons based on data from the latest
available reporting (USGS 2024).
These updates resulted in an average annual increase of less than 0.5 MMT C02 Eq. (less than 0.05
percent) in PFC emissions and no change in C02 across the time series compared to the previous
Inventory.
Planned Improvements
EPA is assessing planned improvements for future reports, but at this time has no specific planned
improvements for estimating C02and PFC emissions from aluminum production.
4.21 Magnesium Production (Source
Category 2C4)
The magnesium metal production and casting industry uses sulfur hexafluoride (SF6) as a cover gas to
prevent the rapid oxidation of molten magnesium in the presence of air. This reporting category (2C4)
includes emissions from magnesium metal production and processing. Sulfur hexafluoride has been
used in this application around the world for more than 30 years. A dilute gaseous mixture of SF6 with dry
air and/or carbon dioxide (C02) is blown over molten magnesium metal to induce and stabilize the
formation of a protective crust. A small portion of the SF6 reacts with the magnesium to form a thin
molecular film of mostly magnesium oxide and magnesium fluoride. The amount of SF6 reacting in
magnesium production and processing is considered to be negligible and thus all SF6 used is assumed
to be emitted into the atmosphere. Alternative cover gases, such as AM-cover™ (containing HFC-134a),
Novec™ 612 (FK-5-1-12) and dilute sulfur dioxide (S02) systems can and are being used by some facilities
in the United States. However, many facilities in the United States are still using traditional SF6 cover gas
systems. Carbon dioxide is also released during primary magnesium production if carbonate based raw
materials, such as dolomite, are used. During the processing of these raw materials to produce
magnesium, calcination occurs which results in a release of C02 emissions.
The magnesium industry emitted 1.1 MMT C02 Eq. (0.05 kt) of SF6, 0.01 MMT C02 Eq. (0.01 kt) of HFC-
134a, and 0.002 MMT C02 Eq. (2.3 kt) of C02 in 2023. This represents a decrease of approximately 1
percent from total 2022 emissions (see Table 4-106 and Table 4-107) and an increase in SF6 emissions by
less than 1 percent. In 2023, total HFC-134a emissions decreased from 0.029 MMT C02 Eq. to 0.008
MMT C02 Eq., or a 71 percent decrease as compared to 2022 emissions. FK 5-1-12 emissions in 2023
were consistent with 2022. The emissions of the carrier gas, C02, decreased from 2.94 kt in 2022 to 2.34
kt in 2023, or 20 percent.
4-136 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Table 4-106: SF6, HFC-134a, FK 5-1-12 and C02 Emissions from Magnesium Production
(MMT C02 Eq.)
Year
1990
2005
2019
2020
2021
2022
2023
SFe
5.6
3.0
0.9
0.9
1.2
1.1
1.1
HFC-134a
0.0
0.0
0.1
0.1
+
+
+
C02
0.1
+
+
+
+
+
+
FK 5-l-12a
0.0
0.0
+
+
+
+
+
Total
5.7
3.0
1.0
0.9
1.2
1.1
1.1
+ Does not exceed 0.05 MMT C02 Eq.
a Emissions of FK 5-1-12 are not included in totals.
Note: Totals may not sum due to independent rounding.
Table 4-107: SF6, HFC-134a, FK 5-1-12 and C02 Emissions from Magnesium Production
(kt)
Year
1990
2005
2019
2020
2021
2022
2023
SFe
+
+
+
+
+
+
+
HFC-134a
0
0
+
+
+
+
+
CO2
129
4
2
3
3
3
2
FK 5-l-12a
0
0
+
+
+
+
+
+ Does not exceed 0.5 kt
a Emissions of FK 5-1-12 are not included in totals.
Methodology and Time-Series Consistency
Emission estimates for the magnesium industry incorporate information provided by industry
participants in EPA's SF6 Emission Reduction Partnership for the Magnesium Industry as well as
emissions data reported through Subpart T (Magnesium Production and Processing) of EPA's GHGRP.
The Partnership started in 1999 and, in 2010, participating companies represented 100 percent of U.S.
primary and secondary production and 16 percent of the casting sector production (i.e., die, sand,
permanent mold, wrought, and anode casting). SF6 emissions for 1999 through 2010 from primary
production, secondary production (i.e., recycling), and die casting were generally reported by
Partnership participants. Partners reported their SF6 consumption, which is assumed to be equivalent to
emissions. Along with SF6, some Partners reported their HFC-134a and FK 5-1-12 consumed, which is
also assumed to be equalto emissions. The last reportingyear under the Partnership was 2010.
Emissions data for 2011 through 2023 are obtained through EPA's GHGRP Under the program, owners or
operators of facilities that have a magnesium production or casting process must report emissions from
use of cover or carrier gases, which include SF6, HFC-134a, FK 5-1-12 and C02. Consequently, cover and
carrier gas emissions from magnesium production and processing were estimated for three time
periods, depending on the source of the emissions data: 1990 through 1998 (pre-EPA Partnership), 1999
through 2010 (EPA Partnership), and 2011 through 2023 (EPA GHGRP). The methodologies described
below also make use of magnesium production data published by the U.S. Geological Survey (USGS) as
available.
Industrial Processes and Product Use 4-137
-------
1990 through 1998
To estimate emissions for 1990 through 1998, industry SF6 emission factors were multiplied by the
corresponding metal production and consumption (casting) statistics from USGS. For this period, it was
assumed that there was no use of HFC-134a or FK 5-1-12 cover gases, and hence emissions were not
estimated for these alternatives.
Sulfur hexafluoride emission factors from 1990 through 1998 were based on a number of sources and
assumptions. Emission factors for primary production were available from U.S. primary producers for
1994 and 1995. The primary production emission factors were 1.2 kg SF6 per metric ton for 1990 through
1993, and 1.1 kg SF6 per metric ton for 1994 through 1997. The emission factor for secondary production
from 1990 through 1998 was assumed to be constant at the 1999 average Partner value. An emission
factor for die casting of 4.1 kg SF6 per metric ton, which was available for the mid-1990s from an
international survey (Gjestland and Magers 1996), was used for years 1990 through 1996. For 1996
through 1998, the emission factor for die casting was assumed to decline linearly to the level estimated
based on Partner reports in 1999. This assumption is consistent with the trend in SF6 sales to the
magnesium sector that was reported in the RAND survey of major SF6 manufacturers, which showed a
decline of 70 percent from 1996 to 1999 (RAND 2002). Sand casting emission factors for 1990 through
2001 were assumed to be the same as the 2002 emission factor for all but one facility, which used an
emission factor derived from 2011 GHGRP data and held constant to back cast emissions for 1990-
1998. The emission factors for the other processes (i.e., permanent mold, wrought, and anode casting),
about which less is known, were assumed to remain constant at levels defined in Table 4-107. The
emission factors for the other processes (i.e., permanent mold, wrought, and anode casting) were based
on discussions with industry representatives.
The quantities of C02 carrier gas used for each production type have been estimated using the 1999
estimated C02 emissions data and the annual calculated rate of change of SF6 use in the 1990 through
1999 time period. For each year and production type, the rate of change of SF6 use between the current
year and the subsequent year was first estimated. This rate of change was then applied to the C02
emissions of the subsequent year to determine the C02 emission of the current year.
Carbon dioxide emissions from the calcination of dolomite in the primary production of magnesium
were calculated based on the 2006IPCC Guidelines Tier 2 method by multiplying the estimated primary
production of magnesium by an emissions factor of 3.62 kilogram of C02 per kilogram of magnesium
produced.74 For 1990 through 1998, production was estimated to be equal to the production capacity of
the facility.
1999 through 2010
The 1999 through 2010 emissions from primary and secondary production were based on information
provided by EPA's industry Partners. In some instances, there were years of missing Partner data,
including SF6 consumption and metal processed. For these situations, emissions were estimated
through interpolation where possible, or by holding company-reported emissions (as well as production)
constant from the previous year. For alternative cover gases, including HFC-134a and FK 5-1-12, mainly
reported data was relied upon. That is, unless a Partner reported using an alternative cover gas, it was
74 See https://www.ipcc-nggip.iges.or.jp/public/?006gl/pdf/3 Volume3/V3 4 Ch4 Metal lndustrv.pdf.
4-138 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
not assumed it was used. Emissions of alternate gases were also estimated through linear interpolation
where possible.
The die casting emission estimates for 1999 through 2010 were also based on information supplied by
industry Partners. When a Partner was determined to be no longer in production, its metal production
and usage rates were set to zero. Missing data on emissions or metal input was either interpolated or
held constant at the last available reported value. In 1999 through 2010, Partners were assumed to
account for all die casting tracked by USGS. For 1999, die casters who were not Partners were assumed
to be similar to Partners who cast small parts. Due to process requirements, these casters consume
larger quantities of SF6 per metric ton of processed magnesium than casters that process large parts.
Consequently, emission estimates from this group of die casters were developed using an average
emission factor of 5.2 kg SF6 per metric ton of magnesium. This emission factor was developed using
magnesium production and SF6 usage data for the year 1999. In 2008, the derived emission factor for die
casting began to increase after many years of largely decreasing emission factors. As determined
through an analysis of activity data reported from the USGS, this increase is due to a temporary
decrease in production at many facilities between 2008 and 2010, which reflects the change in
production that occurred during the recession.
The emissions from other casting operations were estimated by multiplying emission factors (kg SF6 per
metric ton of metal produced or processed) by the amount of metal produced or consumed from USGS,
with the exception of some years for which Partner sand casting emissions data are available. The
emission factors for sand casting activities were acquired through the data reported by the Partnership
for 2002 to 2006. For 1999 through 2001, the sandcasting emission factor was held constant at the 2002
Partner-reported level. For 2007 through 2010, the sandcasting Partner did not report and the reported
emission factor from 2005 was applied to the Partner and to a non GHGRP sand casters. Activity data for
2005 was obtained from USGS (USGS 2005b). One non partner sand casting facility reported to GHGRP
in 2011 and had an emission factor derived for 2011, this factor was used to back cast emissions for this
facility from 1999 to 2010.
The emission factors for primary production, secondary production and sand casting for the 1999 to
2010 are not published to protect company-specific production information. However, the emission
factor for primary production has not risen above the average 1995 Partner value of 1.1 kg SF6 per metric
ton. The emission factors for the other industry sectors (i.e., permanent mold, wrought, and anode
casting) were based on discussions with industry representatives. The emission factors for casting
activities are provided below in Table 4-108.
The emissions of HFC-134a and FK-5-1-12 were included in the estimates for only instances where
Partners reported that information to the Partnership. Emissions of these alternative cover gases were
not estimated for instances where emissions were not reported.
Carbon dioxide carrier gas emissions were estimated using the emission factors developed based on
GHGRP-reported carrier gas and cover gas data, by production type. It was assumed that the use of
carrier gas, by production type, is proportional to the use of cover gases. Therefore, an emission factor,
in kg C02 per kg cover gas and weighted by the cover gases used, was developed for each of the
production types. GHGRP data, on which these emissions factors are based, was available for primary,
secondary, die casting and sand casting. The emission factors were applied to the quantity of all cover
gases used (SF6, HFC-134a, and FK-5-1-12) by production type in this time period for producers that
reported C02 emissions from 2011-2022 through the GHGP. Carrier gas emissions for the 1999 through
Industrial Processes and Product Use 4-139
-------
2010 time period were only estimated for those Partner companies that reported using C02 as a carrier
gas through the GHGRP. Using this approach helped ensure time-series consistency. Emissions of
carrier gases for permanent mold, wrought, and anode processes were estimated using the ratio of total
C02 emissions to total cover gas emissions for primary, secondary, die and sand in a given year and the
total SF6 emissions from each permanent mold, wrought, and anodes processes respectively in that
same year. C02 emissions from the calcination of dolomite were estimated using the same approach as
described above. At the end of 2001, the sole magnesium production plant operating in the United
States that produced magnesium metal using a dolomitic process that resulted in the release of C02
emissions ceased its operations (USGS 1995b through 2024).
Table 4-108: SF6 Emission Factors (kg SF6 per metric ton of magnesium)
Year
Die Casting"
Permanent Mold Wrought Anodes
1999
1.75b
2 1 1
2000
0.72
2 1 1
2001
0.72
2 1 1
2002
0.71
2 1 1
2003
0.81
2 1 1
2004
0.79
2 1 1
2005
0.77
2 1 1
2006
0.88
2 1 1
2007
0.65
2 1 1
2008
0.97
2 1 1
2009
0.55
2 1 1
2010
0.64
2 1 1
a Weighted average includes all die casters, Partners and non-Partners. For the majority of the time series (2000 through 2010),
Partners made up 100 percent of die casters in the United States.
b Weighted average that includes an estimated emission factor of 5.2 kgSFe per metric ton of magnesium for die casters that do
not participate in the Partnership.
2011 through 2023
For 2011 through 2023, for the primary and secondary producers, GHGRP-reported cover and carrier
gases emissions data were used. For sand and die casting, some emissions data was obtained through
EPA's GHGRP. Additionally, in 2018 a new GHGRP reporter began reporting permanent mold emissions.
The balance of the emissions for this industry segment was estimated based on previous Partner
reporting (i.e., for Partners that did not report emissions through EPA's GHGRP) or were estimated by
multiplying emission factors by the amount of metal produced or consumed. Partners who did not
report through EPA's GHGRP were assumed to have continued to emit SF6 at the last reported level,
which was from 2010 in most cases, unless publicly available sources indicated that these facilities
have closed or otherwise eliminated SF6emissions from magnesium production (ARB 2015). Many
Partners that did report through the GHGRP showed increases in SF6 emissions driven by increased
production related to a continued economic recovery after the 2008 recession. One Partner in particular
reported an anonymously large increase in SF6 emissions from 2010 to 2011, further driving increases in
emissions between the two time periods of inventory estimates. All Partners were assumed to have
continued to consume magnesium at the last reported level. Where the total metal consumption
estimated for the Partners fell below the U.S. total reported by USGS, the difference was multiplied by
the emission factors discussed in the section above, i.e., non-partner emission factors. For the other
4-140 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
types of production and processing (i.e., permanent mold, wrought, and anode casting), emissions were
estimated by multiplying the industry emission factors with the metal production or consumption
statistics obtained from USGS (USGS 1995b-2024). USGS data for 2023 were not yet available at the
time of the analysis, so the 2022 values were held constant through 2023 as an estimate.
Emissions of carrier gases for permanent mold, wrought, and anode processes were estimated using an
approach consistent with the 1999 through 2010 time series.
Methodological approaches were applied to the entire time series to ensure time-series consistency
from 1990 through 2023. 2006IPCC Guidelines methodologies were used throughout the time series,
mainly either a Tier 2 or Tier 3 approach depending on available data.
Uncertainty
Uncertainty surrounding the total estimated emissions in 2023 is attributed to the uncertainties around
SF6, HFC-134a, and C02 emission estimates. To estimate the uncertainty surrounding the estimated
2022 SF6 emissions from magnesium production and processing, the uncertainties associated with
three variables were estimated: (1) emissions reported by magnesium producers and processors for
2023 through EPA's GHGRP, (2) emissions estimated for magnesium producers and processors that
reported via the Partnership in prior years but did not report 2023 emissions through EPA's GHGRP, and
(3) emissions estimated for magnesium producers and processors that did not participate in the
Partnership or report through EPA's GHGRP. An uncertainty of 5 percent was assigned to the emissions
(usage) data reported by each GHGRP reporter for all the cover and carrier gases (per the 2006 IPCC
Guidelines). If facilities did not report emissions data during the current reporting year through EPA's
GHGRP, SF6 emissions data were held constant at the most recent available value reported through the
Partnership. The uncertainty associated with these values was estimated to be 30 percent for each year
of extrapolation (per the 2006 IPCC Guidelines). The uncertainty of the total inventory estimate
remained relatively constant between 2022 and 2023.
Alternate cover gas and carrier gases data was set equal to zero if the facilities did not report via the
GHGRP. For those industry processes that are not represented in the Partnership, such as permanent
mold and wrought casting, SF6 emissions were estimated using production and consumption statistics
reported by USGS and estimated process-specific emission factors (see Table 4-108). The uncertainties
associated with the emission factors and USGS-reported statistics were assumed to be 75 percent and
25 percent, respectively. Emissions associated with die casting and sand casting activities utilized
emission factors based on Partner reported data with an uncertainty of 75 percent. In general, where
precise quantitative information was not available on the uncertainty of a parameter, a conservative
(upper-bound) value was used.
Additional uncertainties exist in these estimates that are not addressed in this methodology, such as the
basic assumption that SF6 neither reacts nor decomposes during use. The melt surface reactions and
high temperatures associated with molten magnesium could potentially cause some gas degradation.
Previous measurement studies have identified SF6 cover gas degradation in die casting applications on
the order of 20 percent (Bartos et al. 2007). Sulfur hexafluoride may also be used as a cover gas for the
casting of molten aluminum with high magnesium content; however, the extent to which this technique
is used in the United States is unknown.
The results of this Approach 2 quantitative uncertainty analysis are summarized in Table 4-109. Total
emissions associated with magnesium production and processing were estimated to be between 1.04
Industrial Processes and Product Use 4-141
-------
and 1.22 MMT C02 Eq. at the 95 percent confidence level. This indicates a range of approximately 8.2
percent below to 8.1 percent above the 2022 emission estimate of 1.13 MMT C02 Eq. The uncertainty
estimates for 2023 are slightly higher to the uncertainty reported for 2021 in the previous Inventory. This
increase in uncertainty is attributed to the increased number of facilities with interpolated emissions
and the increasing number of years for facilities with emissions held constant.
Table 4-109: Approach 2 Quantitative Uncertainty Estimates for SF6, HFC-134a and C02
Emissions from Magnesium Production (MMT C02 Eq. and Percent)
2023 Emission
Uncertainty Range Relative to Emission
Source
Gas
Estimate
Estimate3
(MMT CO2 Eq.)
(MMT CO2 Eq.)
(%)
Lower
Upper
Lower
Upper
Bound
Bound
Bound
Bound
Magnesium
SFe, HFC-
1.13
1.04
1.22
-8.2%
+8.1%
Production
134a, C02
a Range of emission estimates predicted by Monte Carlo stochastic simulation for a 95 percent confidence interval.
QA/QC and Verification
General quality assurance/quality control (QA/QC) procedures were applied consistent with the U.S.
Inventory QA/QC plan, which is in accordance with Volume 1, Chapter 6 of 2006IPCC Guidelines as
described in the introduction of the IPPU chapter (see Annex 8 for more details). For the GHGRP data,
EPA verifies annual facility-level reports through a multi-step process (e.g., including a combination of
pre-and post-submittal electronic checks and manual reviews by staff) to identify potential errors and
ensure that data submitted to EPA are accurate, complete, and consistent (EPA 2015).75 Based on the
results of the verification process, EPA follows up with facilities to resolve mistakes that may have
occurred. The post-submittals checks are consistent with a number of general and category-specific QC
procedures, including: range checks, statistical checks, algorithm checks, and year-to-year checks of
reported data and emissions.
Recalculations Discussion
One die casting facility updated GHGRP reported emissions of SF6 from 2022, leading to an increase in
SF6 emissions.
Sand Casting Emissions for 2021 and 2022 were updated based on 2021 and 2022 specific data
available in the 2022 data tables release from USGS's Mineral Yearbook. 2021 and 2022 data were
previously held constant at 2021 levels due to USGS Mineral Yearbook data only going through 2021. The
updated production of sand cast magnesium was larger than what was estimated for 2021 and smaller
than what was estimated in 2022 in the previous Inventory cycle leading to an increase in SF6 emissions
in 2021 and a decrease in SF6 emissions in 2022.
Review of facility responses indicate that changes over time in the emission factors for this industry
have occurred as facilities switch to using systems with cover gases other than SF6 (e.g. S02) and also
during time periods where back-up SF6-based systems are used due to the failure of the primary (non-
75 GHGRP Report Verification Factsheet. See https://www.epa.gov/sites/production/files/7015-
07/rtocuments/ghgrp verification factsheet.pdf.
4-142 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
SF6) system have occurred, leading to the periodic spike in SF6 usage rates. These updates resulted in an
average annual increase of less than 0.5 MMT C02 Eq. (less than 0.05 percent) in emissions across the
time series compared to the previous Inventory.
Planned Improvements
Cover gas research conducted over the last decade has found that SF6 used for magnesium melt
protection can have degradation rates on the order of 20 percent in die casting applications (Bartos et al.
2007). Current emission estimates assume (per the 2006IPCC Guidelines) that all SF6 utilized is emitted
to the atmosphere. Additional research may lead to a revision of the 2006 IPCC Guidelines to reflect this
phenomenon and until such time, developments in this sector will be monitored for possible application
to the Inventory methodology.
Additional emissions are generated as byproducts from the use of alternate cover gases, which are not
currently accounted for. Research on this topic is developing, and as reliable emission factors become
available, these emissions will be incorporated into the Inventory
4.22 Lead Production (Source Category 2C5)
In 2023, lead was produced in the United States using only secondary production processes. Until 2014,
lead production in the United States involved both primary and secondary processes—both of which
emit carbon dioxide (C02) (Sjardin 2003). This reporting category (2C5) includes emissions from the
production of lead. Per the IPCC methodological guidance, emissions from fuels consumed for energy
purposes during the production of lead are accounted for as part of fossil fuel combustion in the
industrial end-use sector reported under the Energy chapter.
Primary production of lead through the direct smelting of lead concentrate produces C02 emissions as
the lead concentrates are reduced in a furnace using metallurgical coke (Sjardin 2003). Primary lead
production, in the form of direct smelting, previously occurred at a single smelter in Missouri. This
primary lead smelter was closed at the end of 2013, and a small amount of residual lead was processed
during demolition of the facility in 2014 (USGS 2015). Beginning in 2015, primary lead production no
longer occurred in the United States.
Similar to primary lead production, C02 emissions from secondary lead production result when a
reducing agent, usually metallurgical coke, is added to the smelter to aid in the reduction process.
Carbon dioxide emissions from secondary production also occur through the treatment of secondary
raw materials (Sjardin 2003). Secondary production primarily involves the recycling of lead acid
batteries and post-consumer scrap at secondary smelters. Secondary lead production in the United
States has fluctuated over the past 20 years, reaching a high of 1,180,000 metric tons in 2007. In 2023,
secondary lead production accounted for 100 percent of total U.S. lead production. The lead-acid
battery industry accounted for about 85 percent of the reported U.S. lead consumption in 2023 (USGS
2024a).
In 2023, secondary lead production in the United States decreased by approximately 1 percent
compared to 2022 (USGS 2024a). Secondary lead production in 2023 is 8 percent higher than in 1990
(USGS 1994-2023 and 2024a). The United States has become more reliant on imported refined lead,
owing to the closure of the last primary lead smelter in 2013. Exports of spent starting-lighting-ignition
Industrial Processes and Product Use 4-143
-------
(SLI) batteries decreased between 2014 and 2017, and subsequently recovered beginning in 2018.
Exports were 38 percent higher in the first 9 months of 2023 compared to the same time period in 2014
(USGS 1994 through 2023 and USGS 2024a). In the first 9 months of 2023, 31 million spent SLI lead-acid
batteries were exported, 26 percent more than that in the same time period in 2022 (USGS 2024a).
Emissions of C02 from lead production in 2023 were 0.5 MMT C02 Eq. (450 kt), which is a 1 percent
decrease compared to 2022 and a 13 percent decrease compared to 1990 (see Table 4-110 and Table
4-111) (USGS 1994-2023; USGS 2024a; USGS 2024b).
The United States and Mexico were tied as the third largest mine producers of lead in the world, behind
China and Australia, and the United States accounted for approximately 6 percent of world production
in 2023 (USGS 2024a).
Table 4-110: C02 Emissions from Lead Production (MMT C02 Eq.)
Year 1990
2005
2019 2020
2021
2022
2023
Lead Production 0.5
0.6
0.5 0.5
0.5
0.5
0.5
Table 4-111: C02 Emissions from Lead Production (kt C02)
Year 1990
2005
2019 2020
2021
2022
2023
Lead Production 516
553
518 491
473
455
450
Methodology and Time-Series Consistency
Carbon dioxide emissions from lead production76 are calculated based on Sjardin's work (Sjardin 2003)
for lead production emissions and use Tier 1 methods from the 2006IPCC Guidelines, in accordance
with the IPCC methodological decision tree and available data. The Tier 1 equation is as follows:
Equation 4-15:2006 IPCC Guidelines Tier 1: C02 Emissions From Lead Production
(Equation 4.32)
C02 Emissions = (DS x EFDS) + (S x EFS)
where,
DS = Lead produced by direct smelting, metric ton
S = Lead produced from secondary materials
EFds = Emission factor for direct smelting, metric tons C02/metric ton lead product
EFS = Emission factor for secondary materials, metric tons C02/metric ton lead product
For primary lead production using direct smelting, Sjardin (2003) and the 2006 IPCC Guidelines provide
an emission factor of 0.25 metric tons C02/metric ton lead. For secondary lead production, Sjardin
(2003) and the 2006 IPCC Guidelines provide an emission factor of 0.25 metric tons C02/metric ton lead
for direct smelting, as well as an emission factor of 0.2 metric tons C02/metric ton lead produced for the
treatment of secondary raw materials (i.e., pretreatment of lead acid batteries). Since the secondary
production of lead involves both the use of the direct smelting process and the treatment of secondary
raw materials, Sjardin recommends an additive emission factor to be used in conjunction with the
76 EPA has not integrated aggregated facility-level Greenhouse Gas Reporting Program (GHGRP) information to inform
these estimates. The aggregated information (e.g., activity data and emissions) associated with Lead Production did not
meet criteria to shield underlying confidential business information (CBI) from public disclosure.
4-144 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
secondary lead production quantity. The direct smelting factor (0.25) and the sum of the direct smelting
and pretreatment emission factors (0.45) are multiplied by total U.S. primary and secondary lead
production, respectively, to estimate C02 emissions.
The production and use of coking coal for lead production is adjusted for within the Energy chapter as
this fuel was consumed during non-energy related activities. Additional information on the adjustments
made within the Energy sector for Non-Energy Use of Fuels is described in both the Methodology section
of C02 from Fossil Fuel Combustion (Section 3.1 Fossil Fuel Combustion (Source Category 1 A)) and
Annex 2.1, Methodology for Estimating Emissions of C02 from Fossil Fuel Combustion.
The 1990 through 2023 activity data for primary and secondary lead production (see Table 4-112) were
obtained from the U.S. Geological Survey (USGS 1994-2023 and 2024a).
Table 4-112: Lead Production (Metric Tons)
Year
1990
2005
2019
2020
2021
2022
2023
Primary
404,000
143,000
0
0
0
0
0
Secondary
922,000
1,150,000
1,150,000
1,090,000
1,050,000
1,010,000
1,000,000
Methodological approaches discussed below were applied to applicable years to ensure time-series
consistency in emissions from 1990 through 2023.
Uncertainty
Uncertainty associated with lead production relates to the emission factors and activity data used. The
direct smelting emission factor used in primary production is taken from Sjardin (2003) who averaged
the values provided by three other studies (Dutrizac et al. 2000; Morris et al. 1983; Ullman 1997). For
secondary production, Sjardin (2003) added a C02 emission factor associated with battery treatment.
The applicability of these emission factors to plants in the United States is uncertain. EPA assigned an
uncertainty range of ±20 percent for these emission factors, and using this suggested uncertainty
provided in Table 4.23 of the 2006IPCC Guidelines for a Tier 1 emission factor by process type is
appropriate based on expert judgment (RTI 2023). Per this expert judgment, a triangular probability
density function was assumed for emission factors.
There is also a smaller level of uncertainty associated with the accuracy of primary and secondary
production data provided by the USGS which is collected via voluntary surveys; the uncertainty of the
activity data is a function of the reliability of reported plant-level production data and the completeness
of the survey response. EPA currently uses an uncertainty range of ±10 percent for primary and
secondary lead production, and using this suggested uncertainty provided in Table 4.23 of the 2006
IPCC Guidelines for Tier 1 national production data is appropriate based on expert judgment (RTI 2023).
Per this expert judgment, a normal probability density function was assumed for all activity data.
The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 4-113. Lead
production C02 emissions in 2023 were estimated to be between 0.4 and 0.5 MMT C02 Eq. at the 95
percent confidence level. This indicates a range of approximately 15 percent below and 15 percent
above the emission estimate of 0.5 MMT C02 Eq.
Industrial Processes and Product Use 4-145
-------
Table 4-113: Approach 2 Quantitative Uncertainty Estimates for C02 Emissions from
Lead Production (MMT C02 Eq. and Percent)
Source
Gas
2023 Emission Estimate
(MMT CO2 Eq.)
Uncertainty Range Relative to Emission Estimate"
(MMT CO2 Eq.) (%)
Lower Upper
Bound Bound
Lower Upper
Bound Bound
Lead Production
C02
0.5
0.4 0.5
-15% +15%
a Range of emission estimates predicted by Monte Carlo stochastic simulation for a 95 percent confidence interval.
QA/QC and Verification
General quality assurance/quality control (QA/QC) procedures were applied consistent with the U.S.
Inventory QA/QC plan, which is in accordance with Volume 1, Chapter 6 of 2006IPCC Guidelines as
described in the introduction of the IPPU chapter (see Annex 8 for more details).
Initial review of activity data show that EPA's GHGRP Subpart R lead production data and resulting
emissions are fairly consistent with those reported by USGS. EPA is still reviewing available GHGRP
data, reviewing QC analysis to understand differences in data reporting (i.e., threshold implications),
and assessing the possibility of including this planned improvement in future Inventory reports (see
Planned Improvements section below). Currently, GHGRP data are used for QA purposes only.
Recalculations Discussion
Recalculations were implemented for 2019 through 2022 based on revised USGS data for secondary
lead production. Compared to the previous Inventory, emissions decreased by 3 percent (14 kt C02) for
2019 and increased by 9 percent (41 kt C02) for 2020, by 8 percent (34 kt C02) for 2021, and by 6 percent
(27 kt C02) for 2022 (USGS 2024b).
Planned Improvements
Pending resources and prioritization of improvements for more significant sources, EPA will continue to
evaluate and analyze data reported under EPA's GHGRP that would be useful to improve the emission
estimates and category-specific QC for the Lead Production source category, in particular considering
completeness of reported lead production given the reporting threshold. Particular attention will be
made to ensuring time-series consistency of the emissions estimates presented in future Inventory
reports. This is required as the facility-level reporting data from EPA's GHGRP, with the program's initial
requirements for reporting of emissions in calendar year 2010, are not available for all inventory years
(i.e., 1990 through 2009) as required for this Inventory. In implementing improvements and integration of
data from EPA's GHGRP, the latest guidance from the IPCC on the use of facility-level data in national
inventories will be relied upon.77
77 See http://www.ipcc-nggip.iges.or.jp/public/tb/TFI Technical Bulletin 1.pdf.
4-146 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
4.23 Zinc Production (Source Category 2C6)
Zinc production in the United States consists of both primary and secondary processes. Of the primary
and secondary processes currently used in the United States, only the electrothermic and Waelz kiln
secondary processes result in non-energy carbon dioxide (C02) emissions (Viklund-White 2000). This
reporting category (2C6) includes emissions from the production of zinc. Per the IPCC methodological
guidance, emissions from fuels consumed for energy purposes during the production of zinc are
accounted for as part of fossil fuel combustion in the industrial end-use sector reported under the
Energy chapter.
The majority of zinc produced in the United States is used for galvanizing. Galvanizing is a process where
zinc coating is applied to steel in order to prevent corrosion. Zinc is used extensively for galvanizing
operations in the automotive and construction industry. Zinc is also used in the production of zinc alloys
and brass and bronze alloys (e.g., brass mills, copper foundries, and copper ingot manufacturing). Zinc
compounds and dust are also used, to a lesser extent, by the agriculture, chemicals, paint, and rubber
industries.
Production of zinc can be conducted with a range of pyrometallurgical (e.g., electrothermic furnace,
Waelz kiln, flame reactor, batch retorts, Pinto process, and PIZO process) and hydrometallurgical (e.g.,
hydrometallurgical recovery, solvent recovery, solvent extraction-electrowinning, and electrolytic)
processes. Hydrometallurgical production processes are assumed to be non-emissive since no carbon
is used in these processes (Sjardin 2003). Primary production in the United States is conducted through
the non-emissive electrolytic process, while secondary techniques include the electrothermic and
Waelz kiln processes, as well as a range of other processes. Worldwide primary zinc production also
employs a pyrometallurgical process using an Imperial Smelting Furnace; however, this process is not
used in the United States (Sjardin 2003).
In the electrothermic process, roasted zinc concentrate and secondary zinc products enter a sinter feed
where they are burned to remove impurities before entering an electric retort furnace. Metallurgical coke
is added to the electric retort furnace as a carbon-containing reductant. This concentration step, using
metallurgical coke and high temperatures, reduces the zinc oxides and produces vaporized zinc, which
is then captured in a vacuum condenser. This reduction process also generates non-energy C02
emissions.
ZnO + C -> Zn(gas) + C02 (Reaction 1)
ZnO + CO -> Zn(gas) + C02 (Reaction 2)
In the Waelz kiln process, electric arc furnace (EAF) dust, which is captured during the recycling of
galvanized steel, enters a kiln along with a reducing agent (typically carbon-containing metallurgical
coke). When kiln temperatures reach approximately 1,100 to 1,200 degrees Celsius, zinc fumes are
produced, which are combusted with air entering the kiln. This combustion forms zinc oxide, which is
collected in a baghouse or electrostatic precipitator, and is then leached to remove chloride and
fluoride. The use of carbon-containing metallurgical coke in a high-temperature fuming process results
in non-energy C02 emissions. Through this process, approximately 0.33 metric tons of zinc is produced
for every metric ton of EAF dust treated (Viklund-White 2000).
In the flame reactor process, a waste feed stream, which can include EAF dust, is processed in a high-
temperature environment (greater than 2,000 °C) created by the combustion of natural gas or coal and
Industrial Processes and Product Use 4-147
-------
oxygen-enriched air. Volatile metals, including zinc, are forced into the gas phase and drawn into a
combustion chamber, where air is introduced and oxidation occurs. The metal oxide product is then
collected in a dust collection system (EPA 1992).
In 2023, the only companies in the United States that used emissive technology to produce secondary
zinc products were Befesa Holding US Inc (Befesa) and Steel Dust Recycling (SDR). The secondary zinc
facilities operated by Befesa were acquired from American Zinc Recycling (AZR) (formerly "Horsehead
Corporation") in 2021. PIZO Operating Company, LLC (PIZO) operated a secondary zinc production
facility that processed EAF dust in Blytheville, AR from 2009 to 2012.
For Befesa, EAF dust is recycled in Waelz kilns at their Calumet, IL; Palmerton, PA; Rockwood, TN; and
Barnwell, SC facilities. The former AZR facility in Beaumont, TX processed EAF dust via flame reactor
from 1993 through 2009 (AZR 2021, Horsehead 2014). These Waelz kiln and flame reactor facilities
produce intermediate zinc products (crude zinc oxide or calcine). Prior to 2014, most of output from
these facilities were transported to their Monaca, PA facility where the products were smelted into
refined zinc using electrothermic technology. In April 2014, the Monaca smelter was permanently
closed and replaced by a new facility in Mooresboro, NC in 2014.
The Mooresboro facility uses a hydrometallurgical process (i.e., solvent extraction with electrowinning
technology) to produce zinc products, which is assumed to be non-emissive as described above.
Production at the Mooresboro facility was idled in April 2016 and re-started in March 2020 (Recycling
Today 2020). Direct consumption of coal, coke, and natural gas were replaced with electricity
consumption (Horsehead 2012b). The Mooresboro facility uses leaching and solvent extraction (SX)
technology combined with electrowinning, melting, and casting technology. In this process, Waelz
Oxide (WOX) is first washed in water to remove soluble elements such as chlorine, potassium, and
sodium, and then is leached in a sulfuric acid solution to dissolve the contained zinc creating a pregnant
liquor solution (PLS). The PLS is then processed in a solvent extraction step in which zinc is selectively
extracted from the PLS using an organic solvent creating a purified zinc-loaded electrolyte solution. The
loaded electrolyte solution is then fed into the electrowinning process in which electrical energy is
applied across a series of anodes and cathodes submerged in the electrolyte solution causing the zinc
to deposit on the surfaces of the cathodes. As the zinc metal builds up on these surfaces, the cathodes
are periodically harvested in order to strip the zinc from their surfaces (Horsehead 2015).
SDR recycles EAF dust into intermediate zinc products using Waelz kilns and sells the intermediate
products to companies who smelt it into refined products.
Emissions of C02 from zinc production in 2023 were estimated to be 0.9 MMT C02 Eq. (920 kt C02) (see
Table 4-114). All 2023 C02 emissions resulted from secondary zinc production processes. Emissions
from zinc production in the United States have increased overall since 1990 due to a gradual shift from
non-emissive primary production to emissive secondary production. In 2023, emissions were estimated
to be 46 percent higher than they were in 1990. Emissions decreased 9 percent from 2021 levels.
Table 4-114: C02 Emissions from Zinc Production (MMT C02 Eq.)
Year
1990
2005
2019
2020
2021
2022
2023
Zinc Production
0.6
1.0
1.0
1.0
1.0
0.9
0.9
4-148 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Table 4-115: C02 Emissions from Zinc Production (kt C02)
Year
1990
2005
2019
2020
2021
2022
2023
Zinc Production
632
1,030
1,026
977
1,007
947
920
U.S. zinc mine production decreased slightly in 2023 compared to 2022, due to production being
suspended at two zinc-producing mines during the second half of the year (USGS 2024). In 2023, United
States primary and secondary refined zinc production were estimated to total 220,000 metric tons
(USGS 2024, USGS 2022) (see Table 4-116), remaining at approximately the same production level as in
2022. Secondary zinc production has increased significantly since the reopening of the idled
Mooresboro facility in March 2020 (USGS 2021; AZP 2021).
Table 4-116: Zinc Production (Metric Tons)
Year
1990
2005
2019
2020
2021
2022
2023
Primary
262,704
191,120
99,900
110,000
110,000
110,000
110,000
Secondary
95,708
156,000
15,100
70,000
110,000
110,000
110,000
Total
358,412
347,120
115,000
180,000
220,000
220,000
220,000
Note: Totals may not sum due to independent rounding.
Methodology and Time-Series Consistency
Emissions of C02 emissions from zinc production78 using the electrothermic primary production and
Waelz kiln secondary production processes are calculated using a Tier 1 method from the 2006IPCC
Guidelines, in accordance with the IPCC methodological decision tree and available data (IPCC 2006).
The Tier 1 equation used to estimate emissions from zinc production is as follows:
Equation 4-16: 2006 IPCC Guidelines Tier 1: C02 Emissions from Zinc Production
(Equation 4.33)
Ec02 Zfl X £^de/aii(t
where,
ECo2 = C02 emissions from zinc production, metric tons
Zn = Quantity of zinc produced, metric tons
EFdefault = Default emission factor, metric tons C02/metric ton zinc produced
The Tier 1 emission factors provided by IPCC for Waelz kiln-based secondary production were derived
from metallurgical coke consumption factors and other data presented in Vikland-White (2000). These
coke consumption factors as well as other inputs used to develop the Waelz kiln emission factors are
shown below. IPCC does not provide an emission factor for electrothermic processes due to limited
information; therefore, the Waelz kiln-specific emission factors were also applied to zinc produced from
electrothermic processes. Starting in 2014, refined zinc produced in the United States used
hydrometallurgical processes and is assumed to be non-emissive.
78 EPA has not integrated aggregated facility-level Greenhouse Gas Reporting Program (GHGRP) information to inform
these estimates. The aggregated information (e.g., activity data and emissions) associated with Zinc Production did not
meet criteria to shield underlying confidential business information (CBI) from public disclosure.
Industrial Processes and Product Use 4-149
-------
For Waelz kiln-based production, IPCC recommends the use of emission factors based on EAF dust
consumption, if possible, rather than the amount of zinc produced since the amount of reduction
materials used is more directly dependent on the amount of EAF dust consumed. Since only a portion of
emissive zinc production facilities consume EAF dust, the emission factor based on zinc production is
applied to the non-EAF dust consumingfacilities, while the emission factor based on EAF dust
consumption is applied to EAF dust consumingfacilities.
The Waelz kiln emission factor based on the amount of zinc produced was developed based on the
amount of metallurgical coke consumed for non-energy purposes per ton of zinc produced (i.e., 1.19
metric tons coke/metric ton zinc produced) (Viklund-White 2000), and the following equation:
Equation 4-17: Waelz Kiln C02 Emission Factor for Zinc Produced
1.19 metric tons coke 0.85 metric tons C 3.67 metric tons C02
f /'i/l'aslz Kiln ^ ^
metric tons zinc metric tons coke metric tons C
3.70 metric tons C02
metric tons zinc
Refined zinc production levels for AZR's Monaca, PA facility (utilizing electrothermic technology) were
available from the company for years 2005 through 2013 (Horsehead 2008, 2011, 2012, 2013, and
2014). The Monaca facility was permanently shut down in April 2014 and replaced by AZR's new facility
in Mooresboro, NC. The new facility uses hydrometallurgical process to produce refined zinc products.
Hydrometallurgical production processes are assumed to be non-emissive since no carbon is used in
these processes (Sjardin 2003).
Metallurgical coke consumption for non-EAF dust consuming facilities for 1990 through 2004 were
extrapolated using the percentage change in annual refined zinc production at secondary smelters in
the United States, as provided by the U.S. Geological Survey (USGS) Minerals Yearbook: Zinc (USGS
1994 through 2006). Metallurgical coke consumption for 2005 through 2013 were based on the
secondary zinc production values obtained from the Horsehead Corporation Annual Report Form 10-K:
2005 through 2008 from the 2008 10-K (Horsehead Corp 2009); 2009 and 2010 from the 2010 10-K
(Horsehead Corp. 2011); and 2011 through 2013 from the associated 10-K (Horsehead Corp. 2012a,
2013, 2014). Metallurgical coke consumption levels for 2014 and later were zero due to the closure of
the AZR (formerly "Horsehead Corporation") electrothermic furnace facility in Monaca, PA. The
secondary zinc produced values for each year were then multiplied by the 3.70 metric tons C02/metric
ton zinc produced emission factor to develop C02 emission estimates for the AZR electrothermic
furnace facility.
The Waelz kiln emission factor based on the amount of EAF dust consumed was developed based on
the amount of metallurgical coke consumed per ton of EAF dust consumed (i.e., 0.4 metric tons
coke/metric ton EAF dust consumed) (Viklund-White 2000), and the following equation:
Equation 4-18: Waelz Kiln C02 Emission Factor for EAF Dust Consumed
0.4 metric tons coke 0.85 metric tons C 3.67 metric tons C02
EP _ x x -
eaf Dust metric tons EAF Dust metric tons coke metric tons C
1.24 metric tons C02
metric tons EAF Dust
4-150 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Metallurgical coke consumption for EAF dust consuming facilities for 1990 through 2023 were
calculated based on the values of EAF dust consumed. The total amount of EAF dust consumed by the
Waelz kilns currently operated by Befesa was available from AZR (formerly "Horsehead Corporation") in
financial reports for years 2006 through 2015 (Horsehead 2007, 2008, 2010a, 2011, 2012a, 2013, 2014,
2015, and 2016), from correspondence with AZR for 2016 through 2019 (AZR 2020), and from
correspondence with Befesa for 2020 through 2024 (Befesa 2022, 2023, 2024). The EAF dust
consumption values for each year were then multiplied by the 1.24 metric tons C02/metric ton EAF dust
consumed emission factor to develop C02 emission estimates for Befesa's Waelz kiln facilities.
The amount of EAF dust consumed by SDR and their total production capacity were obtained from
SDR's facility in Alabama for the years 2011 through 2022 (SDR 2012, 2014, 2015, 2017, 2018, 2021,
2022, 2023, 2024). The SDR facility has been operational since 2008, underwent expansion in 2011 to
include a second unit (operational since early- to mid-2012), and expanded its capacity again in 2017
(SDR 2018). Annual consumption data for SDR was not publicly available for the years 2008, 2009, and
2010. These data were estimated using data for AZR's Waelz kilns for 2008 through 2010 (Horsehead
2007, 2008, 2010a, 2010b, 2011). Annual capacity utilization ratios were calculated using AZR's annual
consumption and total capacity for the years 2008 through 2010. AZR's annual capacity utilization ratios
were multiplied with SDR's total capacity to estimate SDR's consumption for each of the years, 2008
through 2010 (SDR 2013). The 1.24 metric tons C02/metric ton EAF dust consumed emission factor was
then applied to SDR's estimated EAF dust consumption to develop C02 emission estimates for those
Waelz kiln facilities.
PIZO's facility in Arkansas was operational from 2009 to 2012 (PIZO 2021). The amount of EAF dust
consumed by PIZO's facility for 2009 through 2012 was not publicly available. EAF dust consumption for
PIZO's facility for 2009 and 2010 were estimated by calculating annual capacity utilization of AZR's
Waelz kilns and multiplying this utilization ratio by PIZO's total capacity (PIZO 2012). EAF dust
consumption for PIZO's facility for 2011 through 2012 were estimated by applying the average annual
capacity utilization rates for AZR and SDR (Grupo PROMAX) to PIZO's annual capacity (Horsehead 2012;
SDR 2012; PIZO 2012). The 1.24 metric tons C02/metric ton EAF dust consumed emission factor was
then applied to PIZO's estimated EAF dust consumption to develop C02 emission estimates for those
Waelz kiln facilities.
The production and use of coking coal for zinc production is adjusted for within the Energy chapter as
this fuel was consumed during non-energy related activities. Additional information on the adjustments
made within the Energy sector for non-energy use of fuels is described in both the Methodology section
of 002 from Fossil Fuel Combustion (3.1 Fossil Fuel Combustion (Source Category 1 A)) and Annex 2.1,
Methodology for Estimating Emissions of C02 from Fossil Fuel Combustion.
Beginning with the 2017 USGS Minerals Commodity Summary: Zinc, United States primary and
secondary refined zinc production were reported as one value, total refined zinc production. Prior to this
publication, primary and secondary refined zinc production statistics were reported separately. For
years 2016 through 2023, only one facility produced primary zinc. Primary zinc produced from this
facility was subtracted from the USGS 2016 to 2023 total zinc production statistic to estimate secondary
zinc production for these years.
Methodological approaches were applied to the entire time series to ensure consistency in emissions
from 1990 through 2023.
Industrial Processes and Product Use 4-151
-------
Uncertainty
There is uncertainty associated with the amount of EAF dust consumed in the United States to produce
secondary zinc using emission-intensive Waelz kilns. The estimate for the total amount of EAF dust
consumed in Waelz kilns is based on combining the totals for (1) the EAF dust consumption value
obtained for the kilns currently operated by Befesa (and formerly operated byAZR or Horsehead
Corporation) and (2) an EAF dust consumption value obtained from the Waelz kiln facility operated by
SDR. For the 1990 through 2015 estimates, EAF dust consumption values for the kilns currently
operated by Befesa were obtained from annual financial reports to the Securities and Exchange
Commission (SEC) byAZR. In 2016, AZR reorganized as a private company and ceased providing annual
reports to the SEC (Recycling Today 2017). EAF dust consumption values for subsequent years from the
Befesa kilns and SDR have been obtained from personal communication with facility representatives.
Since actual EAF dust consumption information is not available for PIZO's facility (2009 through 2010)
and SDR's facility (2008 through 2010), the amount is estimated by multiplying the EAF dust recycling
capacity of the facility (available from the company's website) by the capacity utilization factor for AZR
(which was available from Horsehead Corporation financial reports).The EAF dust consumption for
PIZO's facility for 2011 through 2012 was estimated by multiplying the average capacity utilization factor
developed from AZR and SDR's annual capacity utilization rates by PIZO's EAF dust recycling capacity.
Therefore, there is uncertainty associated with the assumption used to estimate PIZO's annual EAF dust
consumption values for 2009 through 2012 and SDR's annual EAF dust consumption values for 2008
through 2010. EPA uses an uncertainty range of ±5 percent for these EAF dust consumption data inputs,
based upon expert elicitation from the USGS commodity specialist. Per this expert judgment, a normal
probability density function was assigned for EAF dust consumption data inputs.
There is also uncertainty associated with the emission factors used to estimate C02 emissions from
secondary zinc production processes. The Waelz kiln emission factors are based on materials balances
for metallurgical coke and EAF dust consumed as provided by Viklund-White (2000). Therefore, the
accuracy of these emission factors depends upon the accuracy of these materials balances. Data
limitations prevented the development of emission factors for the electrothermic process. Therefore,
emission factors for the Waelz kiln process were applied to both electrothermic and Waelz kiln
production processes. Consistent with the ranges in Table 4.25 of the 2006IPCC Guidelines, EPA
assigned an uncertainty range of ±20 percent for the Tier 1 Waelz kiln emission factors, which are
provided by Viklund-White in the form of metric tons of coke per metric ton of EAF dust consumed and
metric tons of coke per metric ton of zinc produced. In order to convert coke consumption rates to C02
emission rates, values for the heat and carbon content of coke were obtained from Table 4.2 - Tier 2 of
the 2006 IPCC Guidelines. An uncertainty range of ±10 percent was assigned to these coke data
elements, and using the suggested uncertainty provided in Table 4.25, Tier 2 - National Reducing Agent
& Process Materials Data of the 2006 IPCC Guidelines is appropriate based on expert judgment (RTI
2023). Per this expert judgment, a triangular probability density function was assigned for emission
factors and the heat and carbon content of coke.
The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 4-117. Zinc
production C02 emissions from 2023 were estimated to be between 0.7 and 1.1 MMT C02 Eq. at the 95
percent confidence level. This indicates a range of approximately 19 percent below and 21 percent
above the emission estimate of 0.9 MMT C02 Eq.
4-152 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Table 4-117: Approach 2 Quantitative Uncertainty Estimates for C02 Emissions from
Zinc Production (MMT C02 Eq. and Percent)
Source Gas
2023 Emission Estimate
Uncertainty Range Relative to Emission Estimate3
(MMTCOz Eq.)
(MMTCOz Eq.)
(%)
Lower Upper
Lower Upper
Bound Bound
Bound Bound
Zinc Production C02
0.9
0.7 1.1
-19% +21%
a Range of emission estimates predicted by Monte Carlo stochastic simulation for a 95 percent confidence interval.
QA/QC and Verification
General quality assurance/quality control (QA/QC) procedures were applied consistent with the U.S.
Inventory QA/QC plan, which is in accordance with Volume 1, Chapter 6 of 2006 IPCC Guidelines as
described in the introduction of the IPPU chapter (see Annex 8 for more details).
Recalculations Discussion
No recalculations were implemented for the 1990 to 2022 portion of the time series.
Planned Improvements
Pending resources and prioritization of improvements for more significant sources, EPA will continue to
evaluate and analyze data reported under EPA's GHGRP that would be useful to improve the emission
estimates and category-specific QC for zinc production, in particular considering completeness of
reported zinc production given the reporting threshold. Given the small number of facilities in the United
States, particular attention will be made to risks for disclosing CBI and ensuring time-series consistency
of the emissions estimates presented in future Inventory reports. This is required as the facility-level
reporting data from EPA's GHGRP, with the program's initial requirements for reporting of emissions in
calendar year 2010, are not available for all inventory years (i.e., 1990 through 2009) as required for this
Inventory. In implementing improvements and integration of data from EPA's GHGRP, the latest guidance
from the IPCC on the use of facility-level data in national inventories will be relied upon.79 This is a long-
term planned improvement, and EPA is still assessing the possibility of including this improvement in
future Inventory reports.
4.24 Electronics Industry (Source Category
2E)
The electronics industry uses multiple greenhouse gases in its manufacturing processes. In
semiconductor manufacturing, these include long-lived fluorinated greenhouse gases used for plasma
etching and chamber cleaning (Source Category 2E1), fluorinated heat transfer fluids used for
temperature control and other applications (Source Category 2E4), and nitrous oxide (N20) used to
79 See http://www.ipcc-nggip.iges.or.jp/public/tb/TFI Technical Bulletin 1 .pdf and the P019 Refinement. Volume 1,
Chapter 2, Section 2.3, Use of Facility Data in Inventories at https://www.ipcc-
nggip.iges.or.jp/public/2019rf/pdf/1 Volume1/19R V1 Ch02 DataCollection.pdf.
Industrial Processes and Product Use 4-153
-------
produce thin films through chemical vapor deposition and in other applications (reported under Source
Category 2H3). Similar to semiconductor manufacturing, the manufacturing of micro-electro-
mechanical systems (MEMS) devices (reported under Source Category 2E5 Other) and photovoltaic (PV)
cells (Source Category 2E3) requires the use of multiple long-lived fluorinated greenhouse gases for
various processes.
The gases most commonly employed in the electronics industry are trifluoromethane
(hydrofluorocarbon (HFC)-23 or CHF3), perfluoromethane (CF4), perfluoroethane (C2F6), nitrogen
trifluoride (NF3), and sulfur hexafluoride (SF6), although other fluorinated compounds such as
perfluoropropane (C3F8) and perfluorocyclobutane (c-C4F8) are also used. The exact combination of
compounds is specific to the process employed.
In addition to emission estimates for these seven commonly used fluorinated gases, this Inventory
contains emissions estimates for N20 and other HFCs and unsaturated, low-GWP PFCs including C5F8,
C4F6, HFC-32, HFC-41, and HFC-134a. These additional HFCs and PFCs are emitted from etching and
chamber cleaning processes in much smaller amounts, accounting for 0.02 percent of emissions (in
C02 Eq.) from these processes.
For semiconductors, a single 300 mm silicon wafer that yields between 400 to 600 semiconductor
products (devices or chips) may require more than 100 distinct fluorinated-gas-using process steps,
principally to deposit and pattern dielectric films. Plasma etching (or patterning) of dielectric films, such
as silicon dioxide and silicon nitride, is performed to provide pathways for conducting material to
connect individual circuit components in each device. The patterning process uses plasma-generated
fluorine atoms, which chemically react with exposed dielectric film to selectively remove the desired
portions of the film. The material removed as well as undissociated fluorinated gases flow into waste
streams and, unless emission abatement systems are employed, into the atmosphere. Plasma
enhanced chemical vapor deposition (PECVD) chambers, used for depositing dielectric films, are
cleaned periodically using fluorinated and other gases. During the cleaning cycle the gas is converted to
fluorine atoms in plasma, which etches away residual material from chamber walls, electrodes, and
chamber hardware. Undissociated fluorinated gases and other products pass from the chamber to
waste streams and, unless abatement systems are employed, into the atmosphere.
In addition to emissions of unreacted gases, some fluorinated compounds can also be transformed in
the plasma processes into different fluorinated compounds which are then exhausted, unless abated,
into the atmosphere. For example, when C2F6 is used in cleaning or etching, CF4 is typically generated
and emitted as a process byproduct. In some cases, emissions of the byproduct gas can rival or even
exceed emissions of the input gas, as is the case for NF3 used in remote plasma chamber cleaning,
which often generates CF4 as a byproduct.
Besides dielectric film etching and PECVD chamber cleaning, much smaller quantities of fluorinated
gases are used to etch polysilicon films and refractory metal films like tungsten.
Nitrous oxide is used in manufacturing semiconductor devices to produce thin films by CVD and
nitridation processes as well as for N-doping of compound semiconductors and reaction chamber
conditioning (Doering 2000).
Liquid perfluorinated compounds are also used as heat transfer fluids (F-HTFs) for temperature control,
device testing, cleaning substrate surfaces and other parts, and soldering in certain types of
semiconductor manufacturing production processes. Leakage and evaporation of these fluids during
use is a source of fluorinated gas emissions (EPA 2006). Unweighted F-HTF emissions consist primarily
4-154 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
of perfluorinated amines, hydrofluoroethers, perfluoropolyethers (specifically, PFPMIEs), and
perfluoroalkylmorpholines. Three percent or less consist of HFCs, PFCs, and SF6 (where PFCs are
defined as compounds including only carbon and fluorine). With the exceptions of the hydrofluoroethers
and most of the HFCs, all of these compounds are very long-lived in the atmosphere and have global
warming potentials (GWPs) near 10,000.80
MEMS and photovoltaic cell manufacturing require thin film deposition and etching of material with a
thickness of one micron or more, so the process is less intricate and complex than semiconductor
manufacturing. The manufacturing process is different than semiconductors, but generally employs
similar techniques. Like semiconductors, MEMS and photovoltaic cell manufacturers use fluorinated
compounds for etching, cleaning reactor chambers, and temperature control. CF4, SF6, and the Bosch
process (which consists of alternating steps of SF6 and C4F8) are used to manufacture MEMS (EPA 2010).
Photovoltaic cell manufacturing predominately uses CF4, to etch crystalline silicon wafers, and C2F6 or
NF3 during chamber cleaning after deposition of SiNx films (IPCC 2006), although other F-GHGs may be
used. Similar to semiconductor manufacturing, both MEMS and photovoltaic cell manufacturing use
N20 in depositing films and other manufacturing processes. MEMS and photovoltaic manufacturing may
also employ HTFs for cooling process equipment (EPA 2010).
Emissions from all fluorinated greenhouse gases (including F-HTFs) and N20 for semiconductors, MEMS
and photovoltaic cells manufacturing are presented in Table 4-118 below for the years 1990, 2005, and
the period 2018 to 2023. The rapid growth of the electronics industry and the increasing complexity
(growing number of layers and functions)81 of electronic products led to an increase in emissions of 152
percent between 1990 and 1999, when emissions peaked at 8.4 MMT C02 Eq. Emissions began to
decline after 1999, reaching a low point in 2009 before rebounding to 2006 emission levels and more or
less plateauing at the current level, although emissions did decrease in both 2022 and 2023 relative to
the previous year. Together, industrial growth, increasing chip complexity, adoption of emissions
reduction technologies (including but not limited to abatement technologies) and shifts in gas usages
resulted in a net increase in emissions of approximately 27 percent in the electronics manufacturing
industry between 1990 and 2023. Total emissions from semiconductor manufacturing in 2023 were
lower than 2022 emissions, decreasing by 12.8 percent, primarily due to a large decrease in CF4, C2F6,
C4F8and NF3emissions. This decrease in emissions is consistent with data from the U.S. Census 2023
Quarterly Survey of Plant Capacity Utilization that shows semiconductor utilized capacity decreased in
2023 compared to 2022.
For U.S. semiconductor manufacturing in 2023, total C02-equivalent emissions of all fluorinated
greenhouse gases and N20 from deposition, etching, and chamber cleaning processes were estimated
to be 4.1 MMT C02 Eq. This is a decrease in emissions from 1999 of 51 percent, and an increase in
emissions from 1990 of 43 percent. These trends are driven by the above-stated reasons.
The GWP of PFPMIE, a perfluoropolyether used as an F-HTF, is included in the IPCC Fourth Assessment Report with a
value of 10,300. The GWPs of the perfluorinated amines and perfluoroalkylmorpholines that are used as F-HTFs have
not been evaluated in the peer-reviewed literature. However, evaluations by the manufacturer indicate that their GWPs
are near 10,000 (78 FR 20632), which is expected given that these compounds are both saturated and fully fluorinated.
EPA assigns a default GWP of 10,000 to compounds that are both saturated and fully fluorinated and that do not have
chemical-specific GWPs in either the Fourth or the Fifth Assessment Reports.
Complexity is a term denoting the circuit required to connect the active circuit elements (transistors) on a chip.
Increasing miniaturization, for the same chip size, leads to increasing transistor density, which, in turn, requires more
complex interconnections between those transistors. This increasing complexity is manifested by increasing the levels
(i.e., layers) of wiring, with each wiring layer requiring fluorinated gas usage for its manufacture.
Industrial Processes and Product Use 4-155
-------
Photovoltaic cell and MEMS manufacturing emissions of all fluorinated greenhouse gases are in Table
4-118. While EPA has developed a simple methodology to estimate emissions from non-reporters and to
back-cast emissions from these sources for the entire time series, there is very high uncertainty
associated with these emission estimates.
The emissions reported by facilities manufacturing MEMS included emissions of C2F6, C3F8, c-C4F8, CF4,
HFC-23, NF3, N20 and SF6,82 and were equivalent to only 0.102 percent to 0.255 percent of the total
reported emissions from electronics manufacturing in 2011 to 2023. F-GHG emissions, the primary type
of emissions for MEMS, ranged from 0.0003 to 0.012 MMT C02 Eq. from 1991 to 2023. Based upon
information in the World Fab Forecast (WFF), it appears that some GHGRP reporters that manufacture
both semiconductors and MEMS are reporting their emissions as onlyfrom semiconductor
manufacturing (GHGRP reporters must choose a single classification per fab). Emissions from non-
reporters have not been estimated.
Total C02-equivalent emissions from manufacturing of photovoltaic cells were estimated to range from
0.0003 MMT C02 Eq. to 0.0330 MMT C02 Eq. between 1998 to 2023 and were equivalent to between
0.003 percent to 0.76 percent of the total reported emissions from electronics manufacturing. F-GHG
emissions, the primary type of emissions for photovoltaic cells, ranged from 0.0003 to 0.0318 MMT C02
Eq. from 1998 to 2023. Emissions from manufacturing of photovoltaic cells were estimated using an
emission factor developed from reported data from a single manufacturer between 2015 and 2016. This
emission factor was then applied to production capacity estimates from non-reporting facilities.
Reported emissions from photovoltaic cell manufacturing consisted of CF4, C2F6, c-C4F8, CHF3, NF3, and
n2o.83
Emissions of F-HTFs, grouped by HFCs, PFCs or SF6 are presented in Table 4-118. Emissions of F-HTFs
that are not HFCs, PFCs or SF6 are not included in inventory totals and are included for informational
purposes only. Since reporting of F-HTF emissions began under EPA's GHGRP in 2011, total F-HTF
emissions (reported and estimated non-reported) have fluctuated between 0.4 MMT C02 Eq. and 0.8
MMT C02 Eq., with an overall declining trend between 2011 to 2023. An analysis of the data reported to
EPA's GHGRP indicates that F-HTF emissions account for anywhere between 8 percent and 14 percent
of total annual emissions (F-GHG, N20 and F-HTFs) from semiconductor manufacturing.30F It is
important to note that EPA recalculated HTF emissions for years 1990 to 2023 to align with updated
GWPs from EPA's April 2024 rule to amend specific provisions in the GHGRP Provisions.84 Overall, the
impact of these recalculations led to an average annual decrease of 0.078 MMT C02 Eq. (13.1 percent)
from 2001-2022, compared to last year's inventory (there are no HTF emissions before 2001). Table
4-120 shows F-HTF emissions in tons by compound group based on reporting to EPA's GHGRP and the
interpolated share of F-HTF emissions to F-GHG emissions for select years prior to reporting.85
82 Gases not reported by MEMS manufacturers to the GHGRP are currently listed as "NE" in the tables. Since no facilities
report using these gases, emissions of these gases are not estimated for this sub-sector. However, there is insufficient
data to definitively conclude that they are not used by non-reporting facilities.
83 Gases not reported by PV manufacturers to the GHGRP are currently listed as "NE" in the tables. Since no facilities
report using these gases, emissions of these gases are not estimated for this sub-sector. However, there is insufficient
data to definitively conclude that they are not used by non-reporting facilities.
84 Revisions and Confidentiality Determinations for Data Elements Under the Greenhouse Gas Reporting Rule. See
https://www.govinfo.gov/content/pkg/FR-9094-04-95/pdf/9094-07413.pdf
85 Many fluorinated heat transfer fluids consist of perfluoropolymethylisopropyl ethers (PFPMIEs) of different molecular
weights and boiling points that are distilled from a mixture. "BP 200 °C" (and similar terms below) indicate the boiling
4-156 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Table 4-118: PFC, HFC, SF6, NF3, and N20 Emissions from Electronics Industry (MMT
C02 Eq.)
Year
1990
2005
2019
2020
2021
2022
2023
LL
O
0.8
1.0
1.5
1.5
1.6
1.7
1.5
C2Fe
1.8
1.8
0.9
0.8
0.9
0.9
0.7
c3f8
+
0.1
0.1
0.1
0.1
0.1
0.1
C4F8
0.0
0.1
0.1
0.1
0.1
0.1
+
HFC-23
0.2
0.2
0.3
0.3
0.4
0.3
0.3
SFe
0.5
0.8
0.8
0.8
0.9
0.8
0.7
NF3
+
0.4
0.5
0.6
0.6
0.6
0.5
C4F6
+
+
+
+
+
+
+
CsFe
+
+
+
+
+
+
+
CH2F2
+
+
+
+
+
+
+
CH3F
+
+
+
+
+
+
+
CH2FCF3
+
+
+
+
+
0.0
+
Total Semiconductors
3.3
4.3
4.2
4.2
4.5
4.4
3.9
LL
O
0.0
+
+
+
+
+
+
C2F6
0.0
+
+
+
+
+
+
C3F8
0.0
+
0.0
0.0
0.0
0.0
0.0
C4F8
0.0
+
+
+
+
+
+
HFC-23
0.0
+
+
+
+
+
+
SFe
0.0
+
+
+
+
+
+
NF3
0.0
0.0
+
+
+
+
+
Total MEMS
0.0
+
+
+
+
+
+
LL
O
0.0
+
+
+
+
+
+
C2F6
0.0
+
+
+
+
+
+
C4F8
0.0
+
+
+
+
+
+
HFC-23
0.0
+
+
+
+
+
+
SFe
0.0
0.0
0.0
0.0
0.0
0.0
0.0
NF3
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Total PV
0.0
+
+
+
+
+
+
N2O (Semiconductors)
+
0.1
0.2
0.3
0.3
0.3
0.3
N20(MEMS)
0.0
+
+
+
+
+
+
N2O (PV)
0.0
+
+
+
+
+
+
Total N2O
+
0.1
0.2
0.3
0.3
0.3
0.3
HFC, PFC and SFe F-
HTFs
0.0
+
0.1
0.1
0.1
0.1
0.1
Total Electronics
Industry
3.3
4.5
4.5
4.5
4.9
4.8
4.2
+ Does not exceed 0.05 MMT C02 Eq.
Note: Totals by gas may not sum due to independent rounding.
point of the fluid in degrees Celsius. For more information, see https://www.regulations.gov/document?D=FPA-HO-
OAR-?009-09?7-Q?76.
Industrial Processes and Product Use 4-157
-------
Table 4-119: PFC, HFC, SF6, NF3, and N20 Emissions from Semiconductor Manufacture
(Metric Tons)
Year
1990
2005
2019
2020
2021
2022
2023
LL
O
114.8
145.3
224.1
227.9
238.5
250.9
222.8
C2F6
160.0
163.4
85.0
76.1
78.9
82.3
66.3
C3F8
0.4
7.3
10.7
9.6
11.2
13.6
11.9
C4F8
0.0
10.9
5.7
5.8
6.3
5.9
4.8
HFC-23
14.6
14.1
25.7
26.6
30.4
26.2
22.2
SFe
21.7
33.4
33.3
32.4
38.4
31.9
30.1
NF3
2.8
26.2
33.5
36.2
39.3
38.4
31.6
C4F6
0.7
0.9
0.9
0.8
1.0
0.8
0.9
CsFe
0.5
0.6
0.4
0.4
0.4
0.4
0.4
CH2F2
0.6
0.8
1.0
1.0
1.0
1.1
0.9
CH3F
1.4
1.8
2.5
2.8
2.9
2.4
2.1
CH2FCF3
+
+
+
+
+
0.0
+
N2o
135.9
463.3
816.0
1,020.8
1,083.0
1,097.3
1,039.6
+ Does not exceed 0.05 MT.
Table 4-120: F-HTF Emissions from Electronics Manufacture by Compound Group (kt
C02 Eq.)
Year
1990
2005
2019
2020
2021
2022
2023
HFCs
0
1
3
1
2
2
4
PFCs
0
37
67
54
63
53
58
SFe
0
6
6
13
9
4
3
HFEs
0
4
2
7
3
17
3
PFPMIEs
0
105
168
146
144
146
137
Perfluoalkylromorpholines
0
60
53
56
50
18
9
Perfluorotrialkylamines
0
154
275
300
275
164
186
Total F-HTFs
0
367
574
577
547
404
401
Notes: Emissions of F-HTFs that are not HFCs, PFCs or SFe are not included in inventory totals and are included
for informational purposes only. Emissions presented for informational purposes include HFEs, PFPMIEs,
perfluoroalkylmorpholines, and perfluorotrialkylamines. Totals may not sum due to independent rounding.
Methodology and Time-Series Consistency
Emissions are based on data reported through Subpart I, Electronics Manufacture, of EPA's GHGRP,
semiconductor manufacturing Partner-reported emissions data received through EPA's PFC86
Reduction/Climate Partnership, EPA's PFC Emissions Vintage Model (PEVM)—a model that estimates
industry emissions from etching and chamber cleaning processes in the absence of emission control
In the context of the EPA Partnership and PEVM, PFC refers to perfluorocompounds, not perfluorocarbons.
4-158 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
strategies (Burton and Beizaie 2001 )87—and estimates of industry activity (i.e., total manufactured layer
area and manufacturing capacity). The availability and applicability of reported emissions data from the
EPA Partnership and EPA's GHGRP and activity data differ across the 1990 through 2023 time series.
Consequently, fluorinated greenhouse gas (F-GHG) emissions from etching and chamber cleaning
processes for semiconductors were estimated using seven distinct methods, one each for the periods
1990 through 1994, 1995 through 1999, 2000 through 2006, 2007 through 2010, 2011 and 2012, 2013
and 2014, and 2015 through 2023. Nitrous oxide emissions were estimated using five distinct methods,
one each for the period 1990 through 1994, 1995 through 2010, 2011 and 2012, 2013 and 2014, and
2015 through 2023. The methodology discussion below for these time periods focuses on
semiconductor emissions from etching, chamber cleaning, and uses of N20. Other emissions for MEMS,
photovoltaic cells, and HTFs were estimated using the approaches described immediately below.
MEMS
GHGRP-reported emissions (F-GHG and N20) from the manufacturing of MEMS are available for the
years 2011 to 2023. Emissions from manufacturing of MEMS for years prior to 2011 were calculated by
linearly interpolating emissions between 1990 (at zero MMT C02 Eq.) and 2011, the first year where
emissions from manufacturing of MEMS was reported to the GHGRP. Based upon information in the
World Fab Forecast (WFF), it appears that some GHGRP reporters that manufacture both
semiconductors and MEMS are reporting their emissions as only from semiconductor manufacturing;
however, emissions from MEMS manufacturing are likely being included in semiconductor totals.
Emissions were not estimated for non-reporters.
Photovoltaic Cells
GHGRP-reported emissions (F-GHG and N20) from the manufacturing of photovoltaic cells are available
for 2011, 2012, 2015, and 2016 from two manufacturers. EPA estimates the emissions from
manufacturing of PVs from non-reporting facilities by multiplying the estimated capacity of non-
reporters by a calculated F-GHG emission factor and N20 emission factor based on GHGRP reported
emissions from the manufacturer (in MMT C02 Eq. per megawatt) that reported emissions in 2015 and
2016. This manufacture's emissions are expected to be more representative of emissions from the
sector, as their emissions were consistent with consuming only CF4for etching processes and are a
large-scale manufacturer, representing 28 percent of the U.S. production capacity in 2016. The second
photovoltaic manufacturer only produced a small fraction of U.S. production (<4 percent). They also
reported the use of NF3in remote plasma cleaning processes, which does not have an emission factor in
Part 98 for PV manufacturing, requiring them to report emissions equal to consumption. The total F-
GHG emissions from non-reporters are then disaggregated into individual gases using the gas
distribution from the 2015 to 2016 manufacturer. Manufacturing capacities in megawatts were drawn
from DisplaySearch, a 2015 Congressional Research Service Report on U.S. Solar Photovoltaic
Manufacturing, and self-reported capacity by GHGRP reporters. EPA estimated that during the 2015 to
2016 period, 28 percent of manufacturing capacity in the United States was represented through
reported GHGRP emissions. Capacities are estimated for the full time series by linearly scaling the total
U.S. capacity between zero in 1997 to the total capacity reported of crystalline silicon (c-Si) PV
87 A Partner refers to a participant in the U.S. EPA PFC Reduction/Climate Partnership for the Semiconductor Industry.
Through a Memorandum of Understanding (MoU) with the EPA, Partners voluntarily reported their PFC emissions to the
EPA by way of a third party, which aggregated the emissions through 2010.
Industrial Processes and Product Use 4-159
-------
manufacturing in 2000 in DisplaySearch and then linearly scaling between the total capacity of c-Si PV
manufacturing in DisplaySearch in 2009 to the total capacity of c-Si PV manufacturing reported in the
Congressional Research Service report in 2012. Capacities were held constant for non-reporters for
2012 to 2019. In 2020, non-reporter capacity declined due to the closure of several PV manufacturing
plants. This capacity was held constant for 2021 to 2023. Average emissions per MW from the GHGRP
reporter in 2015 and 2016 were then applied to the total capacity prior to 2015. Emissions for 2014 from
the GHGRP reporter that reported in 2015 and 2016 were scaled to the number of months open in 2014.
For 1998 through 2023, emissions per MW (capacity) from the GHGRP reporter were applied to the non-
reporters. For 2017 through 2023, there are no reported PV emissions. Therefore, emissions were
estimated using the EPA-derived emission factor and estimated manufacturing capacity from non-
reporters only.
HTFs
Facility emissions of F-HTFs from semiconductor manufacturing are reported to EPA under its GHGRP
and are available for the years 2011 through 2023. EPA estimates the emissions of F-HTFs from non-
reporting semiconductor facilities by calculating the ratio of GHGRP-reported fluorinated HTF emissions
to GHGRP reported F-GHG emissions from etching and chamber cleaning processes, and then
multiplying this ratio by the F-GHG emissions from etching and chamber cleaning processes estimated
for non-reporting facilities. Fluorinated HTF use in semiconductor manufacturing is assumed to have
begun in the early 2000s and to have gradually displaced other HTFs (e.g., de-ionized water and glycol)
in semiconductor manufacturing (EPA 2006). For time-series consistency, EPA interpolated the share of
F-HTF emissions to F-GHG emissions between 2000 (at 0 percent) and 2011 (at 17 percent) and applied
these shares to the unadjusted F-GHG emissions during those years to estimate the emissions.
EPA recalculated HTF emissions for years 1990 to 2023 to align with updated GWPs from EPA's April
2024 rule to amend specific provisions in the GHGRP Provisions.88
Semiconductors
1990 through 1994
From 1990 through 1994, Partnership data were unavailable, and emissions were modeled using PEVM
(Burton and Beizaie 2001 ).89 The 1990 to 1994 emissions are assumed to be uncontrolled, since
reduction strategies such as chemical substitution and abatement were yet to be developed.
PEVM is based on the recognition that fluorinated greenhouse gas emissions from semiconductor
manufacturing vary with: (1) the number of layers that comprise different kinds of semiconductor
devices, including both silicon wafer and metal interconnect layers, and (2) silicon consumption (i.e.,
the area of semiconductors produced) for each kind of device. The product of these two quantities, Total
Manufactured Layer Area (TMLA), constitutes the activity data for semiconductor manufacturing. PEVM
Revisions and Confidentiality Determinations for Data Elements Under the Greenhouse Gas Reporting Rule. See
https://www.govinfo.gov/content/pkg/FR-2024-04-25/pdf/2024-Q7413.pdf
Various versions of the PEVM exist to reflect changing industrial practices. From 1990 to 1994 emissions estimates are
from PEVM v1.0, completed in September 1998. The emission factor used to estimate 1990 to 1994 emissions is an
average of the 1995 and 1996 emissions factors, which were derived from Partner reported data for those years.
4-160 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
also incorporates an emission factor that expresses emissions per unit of manufactured layer-area.
Emissions are estimated by multiplying TMLA by this emission factor.
PEVM incorporates information on the two attributes of semiconductor devices that affect the number
of layers: (1) linewidth technology (the smallest manufactured feature size),90 and (2) product type
(discrete, memory or logic).91 For each linewidth technology, a weighted average number of layers is
estimated using VLSI product-specific worldwide silicon demand data in conjunction with complexity
factors (i.e., the number of layers per Integrated Circuit (IC) specific to product type (Burton and Beizaie
2001; ITRS 2007). PEVM derives historical consumption of silicon (i.e., square inches) by linewidth
technology from published data on annual wafer starts and average wafer size (VLSI Research, Inc.
2012).
The emission factor in PEVM is the average of four historical emission factors, each derived by dividing
the total annual emissions reported by the Partners for each of the four years between 1996 and 1999 by
the total TM LA estimated for the Partners in each of those years. Over this period, the emission factors
varied relatively little (i.e., the relative standard deviation for the average was 5 percent). Since Partners
are believed not to have applied significant emission reduction measures before 2000, the resulting
average emission factor reflects uncontrolled emissions and hence may be use here to estimate 1990
through 1994 emissions. The emission factor is used to estimate U.S. uncontrolled emissions using
publicly available data on world (including U.S.) silicon consumption.
As it was assumed for this time period that there was no consequential adoption of fluorinated-gas-
reducing measures, a fixed distribution of fluorinated-gas use was assumed to apply to the entire U.S.
industry to estimate gas-specific emissions. This distribution was based upon the average fluorinated-
gas purchases made by semiconductor manufacturers duringthis period and the application of IPCC
default emission factors for each gas (Burton and Beizaie 2001).
PEVM only addressed the seven main F-GHGs (CF4, C2F6, C3F8, c-C4F8, HFC-23, SF6, and NF3) used in
semiconductor manufacturing. Through reporting under Subpart I of EPA's GHGRP, data on other F-
GHGs (C4F6, C5F8, HFC-32, HFC-41, HFC-134a) used in semiconductor manufacturing became available
and EPA was therefore able to extrapolate this data across the entire 1990 to 2023 timeseries. To
estimate emissions for these "other F-GHGs", emissions data from Subpart I between 2014 to 2016
were used to estimate the average share or percentage contribution of these gases as compared to total
F-GHG emissions. Subpart I emission factors were updated for 2014 by EPA as a result of a larger set of
emission factor data becoming available, so reported data from 2011 through 2013 was not utilized for
the average. To estimate non-reporter emissions from 2011 -2023, the average emissions data from
Subpart I of 2011 to 2023 was used.
90 By decreasing features of Integrated Circuit components, more components can be manufactured per device, which
increases its functionality. However, as those individual components shrink it requires more layers to interconnect them
to achieve the functionality. For example, a microprocessor manufactured with 65 nm feature sizes might contain as
many as 1 billion transistors and require as many as 11 layers of component interconnects to achieve functionality,
while a device manufactured with 130 nm feature size might contain a few hundred million transistors and require 8
layers of component interconnects (ITRS 2007).
91 Memory devices manufactured with the same feature sizes as microprocessors (a logic device) require approximately
one-half the number of interconnect layers, whereas discrete devices require only a silicon base layer and no
interconnect layers (ITRS 2007). Since discrete devices did not start using PFCs appreciably until 2004, they are only
accounted for in the PEVM emissions estimates from 2004 onwards.
Industrial Processes and Product Use 4-161
-------
To estimate N20 emissions, it was assumed the proportion of N20 emissions estimated for 1995
(discussed below) remained constant for the period of 1990 through 1994.
1995 through 1999
For 1995 through 1999, total U.S. emissions were extrapolated from the total annual emissions reported
by the Partners (1995 through 1999). Partner-reported emissions are considered more representative
(e.g., in terms of capacity utilization in a given year) than PEVM-estimated emissions and are used to
generate total U.S. emissions when applicable. The emissions reported by the Partners were divided by
the ratio of the total capacity of the plants operated by the Partners and the total capacity of all of the
semiconductor plants in the United States; this ratio represents the share of capacity attributable to the
Partnership. This method assumes that Partners and non-Partners have identical capacity utilizations
and distributions of manufacturing technologies. Plant capacity data is contained in the World Fab
Forecast (WFF) database and its predecessors, which is updated quarterly. Gas-specific emissions were
estimated using the same method as for 1990 through 1994.
For this time period emissions of other F-GHGs (C4F6, C5F8, HFC-32, HFC-41, HFC-134a) were estimated
using the method described above for 1990 to 1994.
For this time period, the N20 emissions were estimated using an emission factor that was applied to the
annual, total U.S. TMLA manufactured. The emission factor was developed using a regression-through-
the-origin (RTO) model: GHGRP reported N20 emissions were regressed against the corresponding
TMLA of facilities that reported no use of abatement systems. Details on EPA's GHGRP reported
emissions and development of emission factor using the RTO model are presented in the 2011 through
2012 section. The total U.S. TMLA for 1995 through 1999 was estimated using PEVM.
2000 through 2006
Emissions for the years 2000 through 2006—the period during which Partners began the consequential
application of fluorinated greenhouse gas-reduction measures—were estimated using a combination of
Partner-reported emissions and adjusted PEVM modeled emissions. The emissions reported by Partners
for each year were accepted as the quantity emitted from the share of the industry represented by those
Partners. Remaining emissions, those from non-Partners, were estimated using PEVM, with one change.
To ensure time-series consistency and to reflect the increasing use of remote clean technology (which
increases the efficiency of the production process while lowering emissions of fluorinated greenhouse
gases), the average non-Partner emission factor (PEVM emission factor) was assumed to begin declining
gradually during this period. Specifically, the non-Partner emission factor for each year was determined
by linear interpolation, using the end points of 1999 (the original PEVM emission factor) and 2011 (a new
emission factor determined for the non-Partner population based on GHGRP-reported data, described
below).
The portion of the U.S. total emissions attributed to non-Partners is obtained by multiplying PEVM's total
U.S. emissions figure by the non-Partner share of U.S. total silicon capacity for each year as described
above.92 Gas-specific emissions from non-Partners were estimated using linear interpolation between
the gas-specific emissions distributions of 1999 (assumed to be the same as that of the total U.S.
Industry in 1994) and 2011 (calculated from a subset of non-Partners that reported through the GHGRP
92 This approach assumes that the distribution of linewidth technologies is the same between Partners and non-Partners.
As discussed in the description of the method used to estimate 2007 emissions, this is not always the case.
4-162 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
as a result of emitting more than 25,000 MT C02 Eq. peryear). Annual updates to PEVM reflect published
figures for actual silicon consumption from VLSI Research, Inc., revisions and additions to the world
population of semiconductor manufacturing plants, and changes in IC fabrication practices within the
semiconductor industry (see ITRS 2008 and Semiconductor Equipment and Materials Industry 2011 ).93,
94,95 For this time period emissions of other F-GHGs (C4F6, C5F8, HFC-32, HFC-41, HFC-134a) were
estimated using the method described above for 1990 to 1994.
Nitrous oxide emissions were estimated using the same methodology as the 1995 through 1999
methodology.
2007 through 2010
For the years 2007 through 2010, emissions were also estimated using a combination of Partner
reported emissions and adjusted PEVM modeled emissions to provide estimates for non-Partners;
however, two improvements were made to the estimation method employed for the previous years in the
time series. First, the 2007 through 2010 emission estimates account for the fact that Partners and non-
Partners employ different distributions of manufacturing technologies, with the Partners using
manufacturing technologies with greater transistor densities and therefore greater numbers of layers.96
Second, the scope of the 2007 through 2010 estimates was expanded relative to the estimates for the
years 2000 through 2006 to include emissions from research and development (R&D) fabs. This
additional enhancement was feasible through the use of more detailed data published in the WFF. PEVM
databases were updated annually as described above. The published world average capacity utilization
93 Special attention was given to the manufacturing capacity of plants that use wafers with 300 mm diameters because the
actual capacity of these plants is ramped up to design capacity, typically over a 2 to 3 year period. To prevent overstating
estimates of partner-capacity shares from plants using 300 mm wafers, design capacities contained in WFF were
replaced with estimates of actual installed capacities for 2004 published by Citigroup Smith Barney (2005). Without this
correction, the partner share of capacity would be overstated, by approximately 5 percent. For perspective,
approximately 95 percent of all new capacity additions in 2004 used 300 mm wafers, and by year-end those plants, on
average, could operate at approximately 70 percent of the design capacity. For 2005, actual installed capacities were
estimated using an entry in the World Fab Watch database (April 2006 Edition) called "wafers/month, 8-inch equivalent,"
which denoted the actual installed capacity instead of the fully-ramped capacity. For 2006, actual installed capacities of
new fabs were estimated using an average monthly ramp rate of 1100 wafer starts per month (wspm) derived from
various sources such as semiconductor fabtech, industry analysts, and articles in the trade press. The monthly ramp
rate was applied from the first-quarter of silicon volume (FQSV) to determine the average design capacity over the 2006
period.
94 In 2006, the industry trend in co-ownership of manufacturing facilities continued. Several manufacturers, who are
Partners, now operate fabs with other manufacturers, who in some cases are also Partners and in other cases are not
Partners. Special attention was given to this occurrence when estimating the Partner and non-Partner shares of U.S.
manufacturing capacity.
95 Two versions of PEVM are used to model non-Partner emissions during this period. For the years 2000 to 2003 PEVM
v3.2.0506.0507 was used to estimate non-Partner emissions. During this time, discrete devices did not use PFCs during
manufacturing and therefore only memory and logic devices were modeled in the PEVM v3.2.0506.0507. From 2004
onwards, discrete device fabrication started to use PFCs, hence PEVM v4.0.0701.0701, the first version of PEVM to
account for PFC emissions from discrete devices, was used to estimate non-Partner emissions for this time period.
96 EPA considered applying this change to years before 2007 but found that it would be difficult due to the large amount of
data (i.e., technology-specific global and non-Partner TMLA) that would have to be examined and manipulated for each
year. This effort did not appear to be justified given the relatively small impact of the improvement on the total estimate
for 2007 and the fact that the impact of the improvement would likely be lower for earlier years because the estimated
share of emissions accounted for by non-Partners is growing as Partners continue to implement emission-reduction
efforts.
Industrial Processes and Product Use 4-163
-------
for 2007 through 2010 was used for production fabs, while for R&D fabs a 20 percent figure was
assumed (SIA2009).
In addition, publicly available utilization data was used to account for differences in fab utilization for
manufacturers of discrete and IC products for 2010 emissions for non-Partners. The Semiconductor
Capacity Utilization (SICAS) Reports from SIA provides the global semiconductor industry capacity and
utilization, differentiated by discrete and IC products (SIA 2009 through 2011). PEVM estimates were
adjusted using technology-weighted capacity shares that reflect the relative influence of different
utilization. Gas-specific emissions for non-Partners were estimated using the same method as for 2000
through 2006.
For this time period emissions of other F-GHGs (C5F8, CH2F2, CH3F, CH2FCF3, C2H2F4) were estimated
using the method described above for 1990 to 1994. Nitrous oxide emissions were estimated using the
same methodology as the 1995 through 1999 methodology.
2011 through 2012
The fifth method for estimating emissions from semiconductor manufacturing covers the period 2011
through 2012. This methodology differs from previous years because the EPA's Partnership with the
semiconductor industry ended (in 2010) and reporting under EPA's GHGRP began. Manufacturers whose
estimated uncontrolled emissions equal or exceed 25,000 MT C02 Eq. per year (based on default F-
GHG-specific emission factors and total capacity in terms of substrate area) are required to report their
emissions to EPA. This population of reporters to EPA's GHGRP included both historical Partners of EPA's
PFC Reduction/Climate Partnership as well as non-Partners some of which use gallium arsenide (GaAs)
technology in addition to Si technology.97 Emissions from the population of manufacturers that were
below the reporting threshold were also estimated for this time period using EPA-developed emission
factors and estimates of facility-specific production obtained from WFF. Inventory totals reflect the
emissions from both reporting and non-reporting populations.
Under EPA's GHGRP, semiconductor manufacturing facilities report emissions of F-GHGs (for all types
of F-GHGs) used in etch and clean processes as well as emissions of fluorinated heat transfer fluids.
(Fluorinated heat transfer fluids are used to control process temperatures, thermal test devices, and
clean substrate surfaces, among other applications.) They also report N20 emissions from CVD and
other processes. The F-GHGs and N2Owere aggregated, by gas, across all semiconductor
manufacturing GHGRP reporters to calculate gas-specific emissions for the GHGRP-reporting segment
of the U.S. industry. At this time, emissions that result from heat transfer fluid use that are HFC, PFC and
SF6 are included in the total emission estimates from semiconductor manufacturing, and these GHGRP-
reported emissions have been compiled and presented in Table 4-118. F-HTF emissions resulting from
other types of gases (e.g., HFEs) are not presented in semiconductor manufacturing totals in Table 4-118
and Table 4-119 but are shown in Table 4-120 for informational purposes.
Changes to the default emission factors and default destruction or removal efficiencies (DREs) used for
GHGRP reporting affected the emissions trend between 2013 and 2014. These changes did not reflect
actual emission rate changes but data improvements. Therefore, for the current Inventory, EPA adjusted
the time series of GHGRP-reported data for 2011 through 2013 to ensure time-series consistency using
a series of calculations that took into account the characteristics of a facility (e.g., wafer size and
97 GaAs and Si technologies refer to the wafer on which devices are manufactured, which use the same PFCs but in
different ways.
4-164 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
abatement use). To adjust emissions for facilities that did not report abatement in 2011 through 2013,
EPA simply applied the revised emission factors to each facility's estimated gas consumption by gas,
process type and wafer size. In 2014, EPA also started collecting information on fab-wide DREs and the
gases abated by process type, which were used in calculations for adjusting emissions from facilities
that abated F-GHGs in 2011 through 2013.
• To adjust emissions for facilities that abated emissions in 2011 through 2013, EPA first
calculated the quantity of gas abated in 2014 using reported F-GHG emissions, the revised
default DREs (or the estimated site-specific DRE,98 if a site-specific DRE was indicated), and the
fab-wide DREs reported in 2014." To adjust emissions for facilities that abated emissions in
2011 through 2013, EPA first estimated the percentage of gas passing through abatement
systems for remote plasma clean in 2014 using the ratio of emissions reported for CF4 and NF3.
• EPA then estimated the quantity of NF3 abated for remote plasma clean in 2014 using the ratio of
emissions reported for CF4 (which is not abated) and NF3. This abated quantity was then
subtracted from the total abated quantity calculated as described in the bullet above.
• To account for the resulting remaining abated quantity, EPA assumed that the percentage of gas
passing through abatement systems was the same across all remaining gas and process type
combinations where abatement was reported for 2014.
• The percentage of gas abated was then assumed to be the same in 2011 through 2013 (if the
facility claimed abatement that year) as in 2014 for each gas abated in 2014.
The revised emission factors and DREs were then applied to the estimated gas consumption for each
facility by gas, process type and wafer size.100
For the segment of the semiconductor industry that is below EPA's GHGRP reporting threshold, and for
R&D facilities, which are not covered by EPA's GHGRP, emission estimates are based on EPA-developed
emission factors for the F-GHGs and N20 and estimates of manufacturing activity. The new emission
factors (in units of mass of C02 Eq./TMLA [million square inches (MSI)]) are based on the emissions
reported under EPA's GHGRP by facilities without abatement and on the TMLA estimates for these
facilities based on the WFF (SEMI 2012, 2013).101 In a refinement of the method used to estimate
emissions for the non-Partner population for prior years, different emission factors were developed for
different subpopulations of fabs, disaggregated by wafer size (200 mm and 300 mm). For each of these
groups, a subpopulation-specific emission factor was obtained using a regression-through-the-origin
(RTO) model: facility-reported aggregate emissions of seven F-GHGs (CF4, C2F6, C3F8, c-C4F8, CHF3, SF6
98 EPA generally assumed site-specific DREs were as follows: CF4, Etch (90 percent); all other gases, Etch (98 percent);
NF3, Clean (95 percent); CF4, Clean (80 percent), and all other gases, Clean (80 percent). There were a few exceptions
where a higher DRE was assumed to ensure the calculations operated correctly when there was 100 percent
abatement.
99 If abatement information was not available for 2014 or the reported incorrectly in 2014, data from 2015 or 2016 was
substituted.
100 Since facilities did not report by fab before 2014, fab-wide DREs were averaged if a facility had more than one fab. For
facilities that reported more than one wafer size per facility, the percentages of a facility's emissions per wafer size were
estimated in 2014 and applied to earlier years, if possible. If the percentage of emissions per wafer size were unknown, a
50/50 split was used.
101 EPA does not have information on fab-wide DREs for this time period, so it is not possible to estimate uncontrolled
emissions from fabs that reported point-of-use abatement. These fabs were therefore excluded from the regression
analysis. (They are still included in the national totals.)
Industrial Processes and Product Use 4-165
-------
and NF3)102were regressed against the corresponding TMLA to estimate an aggregate F-GHG emissions
factor (C02 Eq./MSI TMLA), and facility-reported N20 emissions were regressed against the
correspondingTMLA to estimate a N20 emissions factor (C02 Eq./MSI TMLA). For each subpopulation,
the slope of the RTO model is the emission factor for that subpopulation. Information on the use of
point-of-use abatement by non-reporting fabs was not available; thus, EPA conservatively assumed that
non-reporting facilities did not use point-of-use abatement.
For 2011 and 2012, estimates of TMLA relied on the capacity utilization of the fabs published by the U.S.
Census Bureau's Historical Data Quarterly Survey of Plant Capacity Utilization (USCB 2011, 2012).
Similar to the assumption for 2007 through 2010, facilities with only R&D activities were assumed to
utilize only 20 percent of their manufacturing capacity. All other facilities in the United States are
assumed to utilize the average percent of the manufacturing capacity without distinguishing whether
fabs produce discrete products or logic products.
Non-reporting fabs were then broken out into subpopulations by wafer size (200 mm and 300 mm), using
information available through the WFF. The appropriate emission factor was applied to the total TMLA of
each subpopulation of non-reporting facilities to estimate the C02-equivalent emissions of that
subpopulation.
Gas-specific, C02-equivalent emissions for each subpopulation of non-reporting facilities were
estimated using the corresponding reported distribution of gas-specific, C02-equivalent emissions from
which the aggregate emission factors, based on GHGRP-reported data, were developed. Estimated in
this manner, the non-reporting population accounted for 4.9 and 5.0 percent of U.S. emissions in 2011
and 2012, respectively. The GHGRP-reported emissions and the calculated non-reporting population
emissions are summed to estimate the total emissions from semiconductor manufacturing.
2013 and 2014
For 2013 and 2014, as for 2011 and 2012, F-GHG and N20 emissions data received through EPA's
GHGRP were aggregated, by gas, across all semiconductor-manufacturing GHGRP reporters to
calculate gas-specific emissions for the GHGRP-reporting segment of the U.S. industry. However, for
these years WFF data was not available. Therefore, an updated methodology that does not depend on
the WFF derived activity data was used to estimate emissions for the segment of the industry that are
not covered by EPA's GHGRP. For the facilities that did not report to the GHGRP (i.e., which are below
EPA's GHGRP reporting threshold or are R&D facilities), emissions were estimated based on the
proportion of total U.S. emissions attributed to non-reporters for 2011 and 2012. EPA used a simple
averaging method by first estimating this proportion for both F-GHGs and N20 for 2011, 2012, and 2015
and 2016, resulting in one set of proportions for F-GHGs and one set for N20, and then applied the
average of each set to the 2013 and 2014 GHGRP reported emissions to estimate the non-reporters'
emissions. Fluorinated gas-specific, C02-equivalent emissions for non-reporters were estimated using
the corresponding reported distribution of gas-specific, C02-equivalent emissions reported through
EPA's GHGRP for 2013 and 2014.
GHGRP-reported emissions in 2013 were adjusted to capture changes to the default emission factors
and default destruction or removal efficiencies used for GHGRP reporting, affecting the emissions trend
102 Only seven gases were aggregated because inclusion of F-GHGs that are not reported in the Inventory results in
overestimation of emission factor that is applied to the various non-reporting subpopulations.
4-166 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
between 2013 and 2014. EPA used the same method to make these adjustments as described above for
2011 and 2012 GHGRP data.
2015 through 2023
Similar to the methods described above for 2011 and 2012, and 2013 and 2014, EPA relied upon
emissions data reported directly through the GHGRP. For 2015 through 2023, EPA took an approach
similar to the one used for 2011 and 2012 to estimate emissions for the segment of the semiconductor
industry that is below EPA's GHGRP reporting threshold, and for R&D facilities, which are not covered by
EPA's GHGRP. However, in a change from previous years, EPA was able to develop new annual emission
factors for 2015 through 2023 using TMLA from WFF and a more comprehensive set of emissions, i.e.,
fabs with as well as without abatement control, as new information about the use of abatement in
GHGRP fabs and fab-wide were available. Fab-wide DREs represent total fab C02 Eq.-weighted
controlled F-GHG and N20 emissions (emissions after the use of abatement) divided by total fab C02
Eq.-weighted uncontrolled F-GHG and N20 emissions (emission prior to the use of abatement).
Using information about reported emissions and the use of abatement and fab-wide DREs, EPA was
able to calculate uncontrolled emissions (each total F-GHG and N20) for every GHGRP reporting fab.
Using this, coupled with TMLA estimated using methods described above (see 2011 through 2012), EPA
derived emission factors by year, gas type (F-GHG or N20), and wafer size (200 mm and less or 300 mm)
by dividing the total annual emissions reported by GHGRP reporters by the total TMLA estimated for
those reporters. These emission factors were multiplied by estimates of non-reporter TMLA to arrive at
estimates of total F-GHG and N20 emissions for non-reporters for each year. For each wafer size, the
total F-GHG emissions were disaggregated into individual gases using the shares of total emissions
represented by those gases in the emissions reported to the GHGRP by unabated fabs producing that
wafer size.
Data Sources
GHGRP reporters, which consist of former EPA Partners and non-Partners, estimated their emissions
using a default emission factor method established by EPA. Like the Tier 2c Method in the 2019
Refinement to the 2006IPCC Guidelines, this method uses different emission and byproduct generation
factors for different F-GHGs and process types and uses factors for different wafer sizes (i.e., 300mm vs.
150 and 200mm) and CVD clean subtypes (in situ thermal, in situ plasma, and remote plasma). Starting
with 2014 reported emissions, EPA's GHGRP required semiconductor manufacturers to apply updated
emission factors to estimate their F-GHG emissions. For the years 2011 through 2013 reported
emissions, semiconductor manufacturers used older emission factors to estimate their F-GHG
emissions (Federal Register / Vol. 75, No. 230 /December 1, 2010, 74829). Subpart I emission factors
were updated for 2014 by EPA as a result of a larger set of emission factor data becoming available as
part of the Subpart I petition process, which took place from 2011 through 2013. In addition to
semiconductor manufacturing, GHGRP also includes reported emissions from MEMS and PV producers.
Historically, semiconductor industry partners estimated and reported their emissions using a range of
methods and uneven documentation. It is assumed that most Partners used a method at least as
accurate as the IPCC's Tier 2a Methodology, recommended in the 2006 IPCC Guidelines. Partners are
estimated to have accounted for between 56 and 79 percent of F-GHG emissions from U.S.
semiconductor manufacturing between 1995 and 2010, with the percentage declining in recent years as
Partners increasingly implemented abatement measures.
Industrial Processes and Product Use 4-167
-------
Estimates of operating plant capacities and characteristics for Partners and non-Partners were derived
from the Semiconductor Equipment and Materials Industry (SEMI) WFF (formerly World Fab Watch)
database (1996 through 2012, 2013, 2016, 2018, 2021, and 2023) (e.g., Semiconductor Materials and
Equipment Industry 2021). Actual worldwide capacity utilizations for 2008 through 2010 were obtained
from Semiconductor International Capacity Statistics (SICAS) (SIA 2009 through 2011). Estimates of the
number of layers for each linewidth was obtained from International Technology Roadmap for
Semiconductors: 2013 Edition (Burton and Beizaie 2001; ITRS 2007; ITRS 2008; ITRS 2011; ITRS 2013).
PEVM utilized the WFF, SICAS, and ITRS, as well as historical silicon consumption estimates published
by VLSI. Actual quarterly U.S. capacity utilizations for 2011, 2012, 2014 to 2023 were obtained from the
U.S. Census Bureau's Historical Data Quarterly Survey of Plant Capacity Utilization (USCB 2011, 2012,
2015, 2016, 2017, 2018, 2019, 2020, 2021, 2022, 2023).
Estimates of PV manufacturing capacity, which are used to calculate emissions from non-reporting
facilities, are based on data from two sources. A historical market analysis from DisplaySearch provided
estimates of U.S. manufacturing capacity from 2000 to 2009 (DisplaySearch 2010). Domestic PV cell
production for 2012 was obtained from a Congressional Research Service report titled U.S. Solar
Photovoltaic Manufacturing: Industry Trends, Global Competition, Federal Support (Platzer 2015).
Uncertainty
A quantitative uncertainty analysis of this source category was performed using the IPCC-
recommended Approach 2 uncertainty estimation methodology, the Monte Carlo stochastic simulation
technique. The Monte Carlo stochastic simulation was performed on the total emissions estimate from
the electronics industry, represented in equation form as:
Equation 4-19: Total Emissions from Electronics Industry
Total Emissions (Er)
= Semiconductors F-GHG and N20 Emissions (ESemi)
+ MEMS F-GHG and N20 Emissions (Emems)
+ PV F-GHG and N20 Emissions (EPF)
+ HFC, PFC and SF6 F-HTFs Emissions (EHTF)
The uncertainty in the total emissions for the electronics industry, presented in Table 4-121 below,
results from the convolution of four distributions of emissions, namely from semiconductors
manufacturing, MEMS manufacturing, PV manufacturing and emissions of heat transfer fluids. The
approaches for estimating uncertainty in each of the sources are described below:
Semiconductors Manufacture Emission Uncertainty
The Monte Carlo stochastic simulation was performed on the emissions estimate from semiconductor
manufacturing, represented in equation form as:
4-168 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Equation 4-20: Total Emissions from Semiconductor Manufacturing
Semiconductors F-GHG and N20 Emissions (ESemi)
= GHGRP Reported F-GHG Emissions (ER,F-GHG, Semi)
+ Non-Reporters' Estimated F-GHG Emissions (ENRF.GHG Semi)
+ GHGRP Reported N20 Emissions (ERjN 0jSe77li)
+ Non-Reporters' Estimated N20 Emissions (ENRjN 0jSe77li)
The uncertainty in ESemi results from the convolution of four distributions of emissions, ER,F-GHG,semi
ER,N2o,semi ENr,F-GHG,semi and ENR,N2o,semi- The approaches for estimating each distribution and combining
them to arrive at the reported 95 percent confidence interval (CI) for ESemi are described in the remainder
of this section.
The uncertainty estimate of ER,F-GHG,semi, or GHGRP-reported F-GHG emissions, is developed based on
gas-specific uncertainty estimates of emissions for two industry segments, one processing 200 mm or
less wafers and one processing 300 mm wafers. Uncertainties in emissions for each gas and industry
segment are based on an uncertainty analysis conducted during the assessment of emission estimation
methods for the Subpart I rulemaking in 2012 (see Technical Support for Modifications to the Fluorinated
Greenhouse Gas Emission Estimation Method Option for Semiconductor Facilities under Subpart I,
docket EPA-HQ-OAR-2011-0028).103 This assessment relied on facility-specific gas information by gas
and wafer size, and incorporated uncertainty associated with both emission factors and gas
consumption quantities. The 2012 analysis did not consider the use of abatement.
For the industry segment that manufactured 200 mm wafers, estimates of uncertainty at a 95 percent CI
ranged from ±29 percent for C3F8 to ±10 percent for CF4. For the corresponding 300 mm industry
segment, estimates of uncertainty at the 95 percent CI ranged from ±36 percent for C4F8 to ±16 percent
for CF4. For gases for which uncertainty was not analyzed in the 2012 assessment (e.g., CH2F2), EPA
applied the 95 percent CI range equivalent to the range for the gas and industry segment with the
highest uncertainty from the 2012 assessment. These gas and wafer-specific uncertainty estimates
were developed to represent uncertainty at a facility-level, but they are applied to the total emissions
across all the facilities that did not abate emissions as reported under EPA's GHGRP at a national level.
Hence, it is noted that the uncertainty estimates used may be overestimating the uncertainties at a
national level.
For those facilities reporting abatement of emissions under EPA's GHGRP, estimates of uncertainties for
the no abatement industry segments are modified to reflect the use of full abatement (abatement of all
gases from all cleaning and etching equipment) and partial abatement. These assumptions used to
develop uncertainties for the partial and full abatement facilities are identical for 200 mm and 300 mm
wafer processing facilities. For all facilities reporting gas abatement, a triangular distribution of
103 On November 13, 2013, EPA published a final rule revising Subpart I (Electronics Manufacturing) of the GHGRP (78 FR
68162). The revised rule includes updated default emission factors and updated default destruction and removal
efficiencies that are slightly different from those that semiconductor manufacturers were required to use to report their
2012 emissions. The uncertainty analyses that were performed during the development of the revised rule focused on
these updated defaults but are expected to be reasonably representative of the uncertainties associated with the older
defaults, particularly for estimates at the country level. (They may somewhat underestimate the uncertainties
associated with the older defaults at the facility level.) For simplicity, the 2012 estimates are assumed to be unbiased
although in some cases, the updated (and therefore more representative) defaults are higher or lower than the older
defaults. Multiple models and sensitivity scenarios were run for the Subpart I analysis. The uncertainty analysis
presented here made use of the Input gas and wafer size model (Model 1) under the following conditions: Year = 2010, f
= 20, n = SIA3.
Industrial Processes and Product Use 4-169
-------
destruction or removal efficiency is assumed for each gas. The triangular distributions range from an
asymmetric and highly uncertain distribution of zero percent minimum to 90 percent maximum with 70
percent most likely value for CF4 to a symmetric and less uncertain distribution of 85 percent minimum
to 95 percent maximum with 90 percent most likely value for C4F8, NF3, and SF6. For facilities reporting
partial abatement, the distribution of fraction of the gas fed through the abatement device, for each gas,
is assumed to be triangularly distributed as well. It is assumed that no more than 50 percent of the
gases are abated (i.e., the maximum value) and that 50 percent is the most likely value, and the
minimum is zero percent. Consideration of abatement then resulted in four additional industry
segments, two 200-mm wafer-processing segments (one fully and one partially abating each gas) and
two 300-mm wafer-processing segment (one fully and the other partially abating each gas). Gas-specific
emission uncertainties were estimated by convolving the distributions of unabated emissions with the
appropriate distribution of abatement efficiency for fully and partially abated facilities using a Monte
Carlo simulation.
The uncertainty in ER,F-GHG,semi is obtained by allocating the estimates of uncertainties to the total
GHGRP-reported emissions from each of the six industry segments, and then running a Monte Carlo
simulation which results in the 95 percent CI for emissions from GHGRP-reporting facilities (ER,F-GHG,semi).
The uncertainty in ER,N2o,semi is obtained by assuming that the uncertainty in the emissions reported by
each of the GHGRP reporting facilities results from the uncertainty in quantity of N20 consumed and the
N20 emission factor (or utilization). Similar to analyses completed for Subpart I (see Technical Support
for Modifications to the Fluorinated Greenhouse Gas Emission Estimation Method Option for
Semiconductor Facilities under Subpart I, docket EPA-HQ-OAR-2011-0028), the uncertainty of N20
consumed was assumed to be 20 percent. Consumption of N20 for GHGRP reporting facilities was
estimated by back-calculating from emissions reported and assuming no abatement. The quantity of
N20 utilized (the complement of the emission factor) was assumed to have a triangular distribution with
a minimum value of zero percent, mode of 20 percent and maximum value of 84 percent. The minimum
was selected based on physical limitations, the mode was set equivalent to the Subpart I default N20
utilization rate for chemical vapor deposition, and the maximum was set equal to the maximum
utilization rate found in ISMI Analysis of Nitrous Oxide Survey Data (ISMI 2009). The inputs were used to
simulate emissions for each of the GHGRP reporting, N20-emitting facilities. The uncertainty for the
total reported N20 emissions was then estimated by combining the uncertainties of each facilities'
reported emissions using Monte Carlo simulation.
The estimate of uncertainty in ENR,F-GHG,semi and ENR,N2o,semi entailed developing estimates of uncertainties
for the emissions factors and the corresponding estimates of TMLA.
The uncertainty in TMLA depends on the uncertainty of two variables—an estimate of the uncertainty in
the average annual capacity utilization for each level of production of fabs (e.g., full scale or R&D
production) and a corresponding estimate of the uncertainty in the number of layers manufactured. For
both variables, the distributions of capacity utilizations and number of manufactured layers are
assumed triangular for all categories of non-reporting fabs. The most probable utilization is assumed to
be 82 percent, with the highest and lowest utilization assumed to be 89 percent, and 70 percent,
respectively. For the triangular distributions that govern the number of possible layers manufactured, it
is assumed the most probable value is one layer less than reported in the ITRS; the smallest number
varied by technology generation between one and two layers less than given in the ITRS and largest
number of layers corresponded to the figure given in the ITRS.
4-170 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
The uncertainty bounds for the average capacity utilization and the number of layers manufactured are
used as inputs in a separate Monte Carlo simulation to estimate the uncertainty around the TMLA of
both individual facilities as well as the total non-reportingTMLA of each sub-population.
The uncertainty around the emission factors for non-reporting facilities is the total combined
uncertainties of individual gases and the TMLA of each reporting facility in that category. The combined
uncertainty of emissions of individual gases from non-reporters is equal to the uncertainty of total
emissions for non-reporting facilities.
The uncertainty around the emission factors for non-reporting facilities is the total combined
uncertainties of individual gases (MT units) and the TMLA of each reporting facility in that category. The
combined uncertainty of emissions of individual gases from non-reporters is equal to the uncertainty of
total emissions for non-reporting facilities. For each wafer size for reporting facilities, emissions of
individual gases were regressed on TMLA (with an intercept forced to zero) for 10,000 emission and
10,000 TMLA values in a Monte Carlo simulation, which results in 10,000 total regression coefficients
(emission factors). The 2.5th and the 97.5th percentile of these emission factors are determined, and the
bounds are assigned as the percent difference from the estimated emission factor.
The next step in estimating the uncertainty in emissions of reporting and non-reporting facilities in
semiconductor manufacture is convolving the distribution of reported emissions, emission factors, and
TMLA using Monte Carlo simulation. For this Monte Carlo simulation, the distributions of the reported F-
GHG gas- and wafer size-specific emissions are assumed to be normally distributed, and the
uncertainty bounds are assigned at 1.96 standard deviations around the estimated mean. The were
some instances, though, where departures from normality were observed for variables, including for the
distributions of the gas- and wafer size-specific N20 emissions, TMLA, and non-reporter emission
factors, both for F-GHGs and N20. As a result, the distributions for these parameters were assumed to
follow a PERT beta distribution.
MEMS Manufacture Emission Uncertainty
The Monte Carlo stochastic simulation was performed on the emissions estimate from MEMS
manufacturing, represented in equation form as:
Equation 4-21: Total Emissions from MEMS Manufacturing
MEMS F-GHG and N20 Emissions (JLmems) — GHGRP Reported F-GHG Emissions (Erj-ghg^ms) + GHGRP
Reported N2O Emissions (Er,nzo, mems)
MEMS F-GHG and N20 Emissions (EMEMS)
— GHGRP Reported F-GHG Emissions (ER F_GHG MBMS)
+ GHGRP Reported N20 Emissions (ERN20 mbms)
Emissions from MEMS manufacturing are only quantified for GHGRP reporters. MEMS manufacturers
that report to the GHGRP all report the use of 200 mm wafers. Some MEMS manufacturers report using
abatement equipment. Therefore, the estimates of uncertainty at the 95 percent CI for each gas emitted
by MEMS manufacturers are set equal to the gas-specific uncertainties for manufacture of 200mm
semiconductor wafers with partial abatement. The same assumption is applied for uncertainty levels for
GHGRP reported MEMS N20 emissions (Er,N2o,mems).
Industrial Processes and Product Use 4-171
-------
PV Manufacture Emission Uncertainty
The Monte Carlo stochastic simulation was performed on the emissions estimate from PV
manufacturing, represented in equation form as:
Equation 4-22: Total Emissions from PV Manufacturing
PV F-GHG and N20 Emissions (E«/) = Non-Reporters' Estimated F-GHG Emissions (ENr,f-ghg,/>i/) + Non-
Reporters' Estimated N2O Emissions (Enr,n20,/>i/)
PV F-GHG and N20 Emissions (EPF)
= Non-Reporters' Estimated F-GHG Emissions (Enrf.ghgpf)
+ Non-Reporters' Estimated N20 Emissions (ENRN2q pv)
Emissions from PV manufacturing are only estimated for non-GHGRP reporters in 2023. There were no
reported emissions from PV manufacturing in GHGRP in 2023. The "Non-Reporters' Estimated F-GHG
Emissions" term in Equation 4-22 was estimated using an emission factor developed using emissions
from reported data in 2015 and 2016 and total non-reporters' capacity. Due to a lack of information and
data and because they represent similar physical and chemical processes, the uncertainty at the 95
percent CI level for non-reporter PV capacity is assumed to be the same as the uncertainty in non-
reporter TMLA for semiconductor manufacturing. Similarly, the uncertainty for the PV manufacture
emission factors are assumed to be the same as the uncertainties in emission factors used for non-
reporters in semiconductor manufacture.
Heat Transfer Fluids Emission Uncertainty
There is a lack of data related to the uncertainty of emission estimates of heat transfer fluids used for
electronics manufacture. Therefore, per the 2006IPCC Guidelines (IPCC 2006, Volume 3, Chapter 6),
uncertainty bounds of 20 percent were applied to estimate uncertainty associated with the various
types of heat transfer fluids, including PFCs, HFC, and SF6, at the national level.
The results of the Approach 2 quantitative uncertainty analysis for electronics manufacturing are
summarized in Table 4-121. These results were obtained by convolving—using Monte Carlo simulation—
the distributions of emissions for each reporting and non-reporting facility that manufactures
semiconductors, MEMS, or PVs and use heat transfer fluids. The emissions estimate for total U.S. F-
GHG, N20, and HTF emissions from electronics manufacturing were estimated to be between 3.95 and
4.48 MMT C02 Eq. at a 95 percent CI level. This range represents 6 percent below to 6 percent above the
2023 emission estimate of 4.21 MMT C02 Eq. for all emissions from electronics manufacture. This range
and the associated percentages apply to the estimate of total emissions rather than those of individual
gases. Uncertainties associated with individual gases will be somewhat higher than the aggregate but
were not explicitly modeled.
4-172 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Table 4-121: Approach 2 Quantitative Uncertainty Estimates for HFC, PFC, SF6, NF3 and
N20 Emissions from Electronics Manufacture (MMT C02 Eq. and Percent)
2023 Emission Estimate
Uncertainty Range Relative to Emission Estimate8
Source
Gas
(MMT C02 Eq.)
(MMT C02 Eq.)
(%)
Lower
Upper
Lower Upper
Bound"
Bound"
Bound Bound
Electronics
Industry
HFC, PFC, SFs,
NF3, and N2O
4.21
3.95
4.48
-6% +6%
a Range of emission estimates predicted by Monte Carlo stochastic simulation for a 95 percent confidence interval.
b Absolute lower and upper bounds were calculated using the corresponding lower and upper bounds in percentages.
QA/QC and Verification
For its GHGRP data, EPA verifies annual facility-level reports through a multi-step process (e.g.,
including a combination of pre-and post-submittal electronic checks and manual reviews by staff) to
identify potential errors and ensure that data submitted to EPA are accurate, complete, and consistent
(EPA 2015).104 Based on the results of the verification process, EPA follows up with facilities to resolve
mistakes that may have occurred. The post-submittals checks are consistent with a number of general
and category-specific QC procedures including range checks, statistical checks, algorithm checks, and
year-to-year checks of reported data and emissions.
For more information on the general QA/QC process applied to this source category, consistent with
Volume 1, Chapter 6 of the 2006IPCC Guidelines, see the QA/QC and Verification Procedures section in
the introduction of the IPPU chapter and Annex 8 for more details.
Recalculations Discussion
Any resubmitted emissions data reported to EPA's GHGRP from all prior years were updated in this
Inventory. Additionally, EPA made the following changes:
• To estimate non-reporter F-GHG and N20 emissions, EPA relies on data reported through
Subpart I and the World Fab Forecast. This process requires EPA to map facilities that report
through Subpart I and which are also represented in the World Fab Forecast. For this Inventory
update, EPA identified and made corrections to a few instances of this mapping based on new
information and additional reviews of the data. This had minor effects on non-reporter emission
estimates for all gases for historical inventory years 2013 to 2017.
• EPA recalculated HTF emissions for years 1990 to 2023 to align with updated GWPs from EPA's
April 2024 rule to amend specific provisions in the GHGRP Provisions.105 Overall, the impact of
these recalculations led to an average annual decrease of 0.078 MMT C02 Eq. (13.1 percent)
from 2001-2022, compared to last year's Inventory (there are no HTF emissions before 2001).
• EPA refined the non-reporting population for 2015 to 2023 by conducting an analysis into the
criteria being used to determine which fabs should be included and excluded from this
population. This included incorporating non-reporters that use complementary metal-oxide-
104 GHGRP Report Verification Factsheet. See https://www.epa.gov/sites/production/files/2015-
07/documents/ghgrp verification factsheet.pdf.
105 Revisions and Confidentiality Determinations for Data Elements Under the Greenhouse Gas Reporting Rule. See
https://www.govinfo.gov/content/pkg/FR-9094-04-95/pdf/9094-07413.pdf
Industrial Processes and Product Use 4-173
-------
semiconductor (CMOS) technology into the inclusion criteria and reviewing and updating
product code classifications (Discrete, Mix, Logic, Memory) for WFF fabs for year 2023 only
EPA recategorized N20 emissions from two PV manufacturing facilities in 2015 and 2016 that were
miscategorized as semiconductor manufacturing facilities in previous inventory years. This affected the
N20 emissions from semiconductor manufacturing emissions marginally for the years 2013 to 2016
(2013 and 2014 emissions estimates are linked to 2015 N20 emissions in their methodology). These
updates resulted in an average annual increase of less than 0.05 MMT C02 Eq. (0.7 percent) across the
time series compared to the previous Inventory.
Planned Improvements
The Inventory methodology uses data reported through the EPA Partnership (for earlier years) and EPA's
GHGRP (for later years) to extrapolate the emissions of the non-reporting population. While these
techniques are well developed, the accuracy of the emissions estimates for the non-reporting
population could be further increased through EPA's further investigation of and improvement upon the
accuracy of estimated activity in the form of TMLA.
The Inventory uses utilization from two different sources for various time periods-SEMI to develop PEVM
and to estimate non-Partner emissions for the period 1995 to 2010 and U.S. Census Bureau for 2011
through 2023. SEMI reported global capacity utilization for manufacturers through 2011. U.S. Census
Bureau capacity utilization include U.S. semiconductor manufacturers as well as assemblers. Further
analysis on the impacts of using a new and different source of utilization data could prove to be useful in
better understanding of industry trends and impacts of utilization data sources on historical emission
estimates.
Estimates of semiconductor non-reporter and non-Partner emissions are based on EPA-developed
emission factors for the time periods pre-2010, 2011 through 2012, and 2015 through 2023. Based on
the data available for these time periods, the methods used to develop emission factors for non-
reporters and non-Partners are slightly inconsistent for semiconductors (e.g., how data representing
emissions and TMLA from the manufacture of various wafer sizes are aggregated or disaggregated for
purposes of calculating emission factors). Further analyses to support potentially adjusting the
methods for developing these emission factors could be done to better ensure consistency across the
time series.
The methodology for estimating semiconductor emissions from non-reporters uses data from the
International Technology Roadmap for Semiconductors (ITRS) on the number of layers associated with
various technology node sizes. The ITRS has now been replaced by the International Roadmap for
Devices and Systems (IRDS), which has published updated data on the number of layers used in each
device type and node size (in nanometers). Incorporating this updated dataset will improve the accuracy
of emissions estimates from non-reporting semiconductor fabs.
To improve the accuracy of the WFF to GHGRP facility mapping process, the criteria used to determine
which fabs should be included and excluded from the non-reporter emissions estimates could be
reviewed and updated, as appropriate.
Additionally, the Inventory assigns product code classifications for fabs included in WFF data that are
used to calculate TMLA. These product codes were updated for newly added fabs in the 2023 WFF data
4-174 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
and will be reviewed and updated for historical years 2014 to 2022 to further improve the accuracy of
TMLA.
4.25 Substitution of Ozone Depleting
Substances (Source Category 2F)
This reporting category (2F) includes emissions from the substitution of ozone-depleting substances
(ODS). Hydrofluorocarbons (HFCs), perfluorocarbons (PFCs), and carbon dioxide (C02) are used as
alternatives to several classes of ODS that are being phased under the Montreal Protocol and the Clean
Air Act. Ozone-depleting substances—chlorofluorocarbons (CFCs), halons, carbon tetrachloride,
methyl chloroform, and hydrochlorofluorocarbons (HCFCs)—are used in a variety of industrial
applications including refrigeration and air conditioning equipment, solvent cleaning, foam production,
sterilization, fire extinguishing, and aerosols. Although HFCs and PFCs are not harmful to the
stratospheric ozone layer, they are potent greenhouse gases. In 2020 Congress directed EPA to address
HFCs by phasing down production and consumption (i.e., production plus import minus export),
maximizing reclamation and minimizing releases from equipment, and facilitating the transition to next-
generation technologies through sector-based restrictions. Emission estimates for HFCs, PFCs, and C02
used as substitutes for ODSs are provided in Table 4-122 and Table 4-123.106
Table 4-122: Emissions of HFCs, PFCs, and C02 from ODS Substitutes (MMT C02 Eq.)
Gas
1990
2005
2019
2020
2021
2022
2023
HFC-23
0.0
+
+
+
+
+
+
HFC-32
0.0
0.3
6.9
7.7
9.4
10.5
11.4
HFC-125
+
9.4
55.9
60.5
68.9
74.4
78.9
HFC-134a
+
72.9
55.4
54.1
50.0
48.3
47.2
HFC-143a
+
12.1
34.7
34.7
34.6
34.2
33.6
HFC-236fa
0.0
1.0
0.7
0.7
0.6
0.6
0.5
LL
O
0.0
+
+
+
+
+
+
co2
+
+
+
+
+
+
+
Other Saturated HFCsa
0.3
6.9
16.1
15.9
16.3
16.9
17.3
Other PFCs and HFOsb
+
0.1
+
+
+
+
+
Total
0.3
102.7
169.7
173.7
179.9
184.9
189.0
+ Does not exceed 0.05 MMT C02 Eq.
a Other Saturated HFCs represents an unspecified mix of saturated HFCs, which includes HFC-152a, HFC-227ea, HFC-245fa,
HFC-365mfc, and HFC-43-10mee.
b Other PFCs and HFOs represents an unspecified mix of PFCs and HFOs, which includes HCFO-1233zd(E), HFO-1234yf, HFO-
1234ze(E), HFO-1336mzz(Z), C4F10, and PFC/PFPEs, the latter being a proxy for a diverse collection of PFCs and
perfluoropolyethers (PFPEs) employed for solvent applications. For estimating purposes, the GWP value used for PFC/PFPEs
was based upon n-CeFu.
Note: Totals may not sum due to independent rounding.
106 Emissions of ODS are not included here consistent with reporting guidelines for national inventories noted in Box 4-1.
See Annex 6.2 for more details on emissions of ODS. Emissions from CO2 used in the food and beverage industry are
separately reported in Chapter 4.16 Non-EOR Carbon Dioxide Utilization but does not include CO2 in ODS substitute
use sectors as a refrigerant, foam blowing agent, or fire extinguishing agent.
Industrial Processes and Product Use 4-175
-------
Table 4-123: Emissions of HFCs, PFCs, and C02 from ODS Substitution (Metric Tons)
Gas
1990
2005
2019
2020
2021
2022
2023
HFC-23
0
1
2
2
2
3
3
HFC-32
0
397
10,142
11,437
13,923
15,523
16,854
HFC-125
+
2,952
17,631
19,088
21,724
23,454
24,889
HFC-134a
+
56,054
42,590
41,608
38,448
37,144
36,274
HFC-143a
+
2,514
7,231
7,220
7,202
7,115
6,999
HFC-236fa
0
127
91
84
78
72
68
LL
O
0
3
5
4
4
4
3
co2
14
1,325
3,304
3,517
3,736
3,972
4,223
Other Saturated HFCsa
M
M
M
M
M
M
M
Other PFCs and HFOsb
M
M
M
M
M
M
M
+ Does not exceed 0.5 MT.
M (Mixture of Gases).
a Other Saturated HFCs represents an unspecified mix of saturated HFCs, which includes HFC-152a, HFC-227ea, HFC-245fa,
HFC-365mfc, and HFC-43-10mee.
b Other PFCs and HFOs represents an unspecified mix of PFCs and HFOs, which includes HCFO-1233zd(E), HFO-1234yf, HFO-
1234ze(E), HFO-1336mzz(Z), C4F10, and PFC/PFPEs, the latter being a proxy for a diverse collection of PFCs and
perfluoropolyethers (PFPEs) employed for solvent applications. For estimating purposes, the GWP value used for PFC/PFPEs
was based upon n-CeFu.
In 1990 and 1991, the only significant emissions of HFCs and PFCs as substitutes to ODSs were
relatively small amounts of HFC-152a—used as an aerosol propellant and also a component of the
refrigerant blend R-500 used in chillers. Beginning in 1992, HFC-134a was used in growing amounts as a
refrigerant in motor vehicle air-conditioners and in refrigerant blends such as R-404A.107 In 1993, the
use of HFCs in foam production began, and in 1994 ODS substitutes for halons entered widespread use
in the United States as halon production was phased out. In 1995, these compounds also found
applications as solvents. Non-fluorinated ODS substitutes, such as C02, have been used in place of
ODS in certain foam production and fire extinguishing uses since the 1990s.
The use and subsequent emissions of HFCs, PFCs, and C02 as ODS substitutes has been increasing
from small amounts in 1990 to 189.0 MMT C02 Eq. emitted in 2023. This increase was in large part the
result of efforts to phase out CFCs, HCFCs, and other ODSs in the United States. Use and emissions of
HFCs are expected to start decreasing in the next few years and continue downward as production and
consumption of HFCs are phased down to 15 percent of their baseline levels by 2036 through an
allowance allocation and trading program established by EPA. Improvements in recovery practices and
the use of alternative gases and technologies will also contribute to a reduction in HFC use and
emissions.
Table 4-124 presents emissions of HFCs, PFCs, and C02 as ODS substitutes by end-use sector for 1990
through 2023. The refrigeration and air-conditioning sector is further broken down by sub-sector. The
end-use sectors that contributed the most toward emissions of HFCs, PFCs, and C02 as ODS
substitutes in 2023 include refrigeration and air-conditioning (154.7 MMT C02 Eq., or approximately 82
percent), aerosols (17.4 MMT C02 Eq., or approximately 9 percent), and foams (12.1 MMT C02 Eq., or
approximately 6 percent). Within the refrigeration and air-conditioning end-use sector residential unitary
AC, part of the Residential Stationary Air-conditioning subsector shown below, was the highest emitting
107 R-404A contains HFC-125, HFC-143a, and HFC-134a.
4-176 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
end-use (43.9 MMT C02 Eq.), followed by large retail food, which is part of the Commercial Refrigeration
subsector. Each of the end-use sectors is described in more detail below.
Table 4-124: Emissions of HFCs, PFCs, and C02 from ODS Substitutes by Sector (MMT
C02 Eq.)
Sector
1990
2005
2019
2020
2021
2022
2023
Refrigeration/Air Conditioning
+
86.2
134.1
138.1
146.7
151.3
154.7
Commercial Refrigeration
+
14.9
40.2
40.6
41.0
41.4
41.8
Domestic Refrigeration
+
0.2
1.2
1.2
1.1
1.0
0.9
Industrial Process Refrigeration
+
5.0
22.6
23.7
24.6
25.3
26.0
Transport Refrigeration
+
1.6
7.4
7.9
8.4
8.8
9.0
Mobile Air Conditioning
+
61.5
26.6
24.6
22.9
20.8
18.8
Residential Stationary Air Conditioning
+
1.2
29.4
33.2
41.5
46.4
50.3
Commercial Stationary Air Conditioning
+
1.7
6.6
6.9
7.3
7.6
7.9
Aerosols
0.2
10.2
17.0
17.3
17.7
17.0
17.4
Foams
+
3.5
14.1
13.7
10.8
11.7
12.1
Solvents
+
1.6
2.0
2.0
2.1
2.1
2.2
Fire Protection
+
1.2
2.5
2.5
2.6
2.6
2.7
Total
0.3
102.7
169.7
173.7
179.9
184.9
189.0
+ Does not exceed 0.05 MMT C02 Eq.
Note: Totals may not sum due to independent rounding.
Refrigeration/Air Conditioning
The refrigeration and air-conditioning sector includes a wide variety of equipment types that have
historically used CFCs or HCFCs. End-uses within this sector include motor vehicle air-conditioning,
retail food refrigeration, refrigerated transport (e.g., ship holds, truck trailers, railway freight cars),
household refrigeration, residential and small commercial air-conditioning and heat pumps, chillers
(large comfort cooling), cold storage facilities, and industrial process refrigeration (e.g., systems used in
food processing, chemical, petrochemical, pharmaceutical, oil and gas, metallurgical, and other
industries). As the ODS phaseout has taken effect, most equipment has been retrofitted or replaced to
use HFC-based substitutes. Common HFCs in use today in refrigeration/air-conditioning equipment are
HFC-134a, R-41 OA,108 R-404A, R-407A,109 and R-507A.110 Lower-GWP options such as hydrofluoroolefin
(HFO)-1234yf in motorvehicle air-conditioning, R-717 (ammonia) in cold storage and industrial
applications, and R-744 (carbon dioxide) and HFC/HFO blends in retail food refrigeration, are also being
used. Manufacturers of residential and commercial air conditioning have announced their plans to use
HFC-32 and R-454B111 and some equipment using those refrigerants are available today, and at least
one manufacturer has announced the availability of chillers operating on HFC-32 as of 2023 (Carrier,
2023) and other low-GWP refrigerants such as R-513A112 and HFO-1234ze(E) are also being used
(Johnson Controls, 2022). These refrigerants are emitted to the atmosphere during equipment operation
108 R-41 OA contains HFC-32 and HFC-125.
109 R-407A contains HFC-32, HFC-125, and HFC-134a.
110 R-507A, also called R-507, contains HFC-125 and HFC-143a.
111 R-454B contains HFC-32 and HFO-1234yf.
112 R-513A contains HFO-1234yf and HFC-134a.
Industrial Processes and Product Use 4-177
-------
(as a result of component failure, leaks, and purges), as well as at manufacturing (if charged at the
factory), installation, servicing, and disposal events.
Aerosols
Aerosol propellants are used in metered dose inhalers (MDIs) and a variety of personal care products
and technical or specialty products (e.g., duster sprays and safety horns). Pharmaceutical companies
that produce MDIs—a type of inhaled therapy used to treat asthma and chronic obstructive pulmonary
disease—have replaced the use of CFCs with HFC-propellant alternatives. The earliest ozone-friendly
MDIs were produced with HFC-134a, but the industry is using HFC-227ea as well. Conversely, since the
use of CFC propellants in other types of aerosols was banned in the Unites States in 1978, most non-
medical consumer aerosol products have not transitioned to HFCs, but to "not-in-kind" technologies,
such as solid or roll-on deodorants and finger-pump sprays. The transition away from ODSs in specialty
aerosol products has also led to the introduction of non-fluorocarbon alternatives (e.g., hydrocarbon
propellants) in certain applications, in addition to HFC-134a or HFC-152a. Other low-GWP options such
as HFO-1234ze(E) are being used as well. These propellants are released into the atmosphere as the
aerosol products are used.
Foams
Chlorofluorocarbons and HCFCs have traditionally been used as foam blowing agents to produce
polyurethane (PU), polystyrene, polyolefin, and phenolic foams, which are used in a wide variety of
products and applications. Flexible PU foams as well as other types of foam, such as polystyrene sheet,
polyolefin, and phenolic foam, have transitioned almost completely away from fluorocompounds into
alternatives such as C02 and hydrocarbons. The majority of rigid PU foams have transitioned to HFCs—
primarily HFC-134a and HFC-245fa. Today, these HFCs are used to produce PU appliance, PU
commercial refrigeration, PU spray, and PU panel foams used in refrigerators, vending machines,
roofing, wall insulation, garage doors, and cold storage applications. In addition, HFC-152a, HFC-134a,
and C02 are used to produce polystyrene sheet/board foam, which is used in food packaging and
building insulation. Low-GWP fluorinated foam blowing agents in use include HFO-1234ze(E), HFO-
1336mzz(Z), and HCFO-1233zd(E). Emissions of blowing agents occur when the foam is manufactured
as well as during the foam lifetime and at foam disposal, depending on the particular foam type.
Solvents
Chlorofluorocarbons, methyl chloroform (1,1,1 -trichloroethane), and to a lesser extent carbon
tetrachloride (CCl4) were historically used as solvents in a wide range of cleaning applications, including
precision, electronics, and metal cleaning. Since their phaseout, metal cleaning end-use applications
have primarily transitioned to non-fluorocarbon solvents and not-in-kind processes. The precision and
electronics cleaning end-uses have transitioned in part to high-GWP gases, due to their high reliability,
excellent compatibility, good stability, low toxicity, and selective solvency. These applications rely on
HFC-43-1 Omee, HFC-365mfc, HFC-245fa, and to a lesser extent, PFCs. Electronics cleaning involves
removing flux residue that remains after a soldering operation for printed circuit boards and other
contamination-sensitive electronics applications. Precision cleaning may apply to either electronic
components or to metal surfaces, and is characterized by products, such as disk drives, gyroscopes,
and optical components, that require a high level of cleanliness and generally have complex shapes,
4-178 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
small clearances, and other cleaning challenges. The use of these solvents yields fugitive emissions of
these HFCs and PFCs.
Fire Protection
Fire protection applications include portable fire extinguishers ("streaming" applications) that originally
used halon 1211, and total flooding applications that originally used halon 1301, as well as some halon
2402. Since the production and import of virgin halons were banned in the United States in 1994, the
halon replacement agent of choice in the streaming sector has been dry chemical, although HFC-236fa
is also used to a limited extent. In the total flooding sector, HFC-227ea has emerged as the primary
replacement for halon 1301 in applications that require clean agents. Other HFCs, such as HFC-23 and
HFC-125, are used in smaller amounts. The majority of HFC-227ea in total flooding systems is used to
protect essential electronics, as well as in civil aviation, military mobile weapons systems, oil/gas/other
process industries, and merchant shipping. Fluoroketone (FK-5-1-12) is also used as a low-GWP option
and 2-BTP is being used in niche applications. As fire protection equipment is tested or deployed,
emissions of these fire protection agents occur.
Methodology and Time-Series Consistency
Using a Tier 2 method in accordance with the IPCC methodological decision tree, a detailed Vintaging
Model of ODS-containing equipment and products was used to estimate the actual—versus potential—
emissions of various ODS substitutes, including HFCs, PFCs, and C02. The name of the model refers to
the fact that it tracks the use and emissions of various compounds for the annual "vintages" of new
equipment that enter service in each end-use. The Vintaging Model predicts ODS and ODS substitute
use in the United States based on modeled estimates of the quantity of equipment or products sold
each year containing these chemicals and the amount of the chemical required to manufacture and/or
maintain equipment and products over time. Emissions for each end-use were estimated by applying
annual leak rates and release profiles, which account for the lag in emissions from equipment as they
leak over time. By aggregating the data for 80 different end-uses, the model produces estimates of
annual use and emissions of each compound. Further information on the Vintaging Model is contained
in Annex 3.10.
Methodological approaches were applied to the entire time series to ensure time-series consistency
from 1990 through 2023.
Uncertainty
Given that emissions of ODS substitutes occur from thousands of different kinds of equipment and from
millions of point and mobile sources throughout the United States, emission estimates must be made
using analytical tools such as the Vintaging Model or the methods outlined in IPCC (2006). Though the
model is more comprehensive than the IPCC default methodology, significant uncertainties still exist
with regard to the levels of equipment sales, equipment characteristics, and end-use emissions profiles
that were used to estimate annual emissions for the various compounds.
The uncertainty analysis quantifies the level of uncertainty associated with the aggregate emissions
across the 80 end-uses in the Vintaging Model. In order to calculate uncertainty, functional forms were
developed to simplify some of the complex "vintaging" aspects of some end-use sectors, especially
with respect to refrigeration and air-conditioning, and to a lesser degree, fire extinguishing. These
Industrial Processes and Product Use 4-179
-------
sectors calculate emissions based on the entire lifetime of equipment, not just equipment put into
commission in the current year, thereby necessitating simplifying equations. The functional forms used
variables that included growth rates, emission factors, transition from ODSs, change in charge size as a
result of the transition, disposal quantities, disposal emission rates, and either stock (e.g., number of air
conditioning units in operation) for the current year or ODS consumption before transition to alternatives
began (e.g., in 1985 for most end-uses). Uncertainty was estimated around each variable within the
functional forms based on expert judgment, and a Monte Carlo analysis was performed.
Inputs to the ODS substitutes uncertainty model generally take on a normal distribution with a 90 to 95
percent confidence interval but do utilize other probability density functions such as a uniform or PERT
BETA distribution. The uncertainty inputs are based on conversations with industry experts and how
certain assumptions are developed in the Vintaging Model. For example, if the Vintaging Model
estimates are specifically aligned with actual reported data, then the uncertainty is decreased. This can
be seen with the unitary AC end-use where annual stock data is aligned with shipment data published
by the Air-Conditioning, Heating, and Refrigeration Institute (AHRI). The stock is assumed to be fairly
accurate and therefore, uncertainty range for the stock of unitary AC is set to an upper and lower bound
of only 2.5 percent. The most significant sources of uncertainty for the substitution of ODS source
category include the total stock of refrigerant installed in industrial process refrigeration and cold
storage equipment, as well as the charge size for technical aerosols using HFC-134a. For technical
aerosols, a triangular distribution is utilized to apply an asymmetrical range to the inventory value. This
is to account for the uncertainty that technical aerosols using HFC-134a might have higher market
penetration than what the Vintaging Model currently estimates.
The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 4-125.
Substitution of ozone depleting substances HFC and PFC emissions were estimated to be between
183.1 and 226.2 MMT C02 Eq. at the 95 percent confidence level. This indicates a range of
approximately 3.1 percent below to 19.6 percent above the emission estimate of 189.0 MMT C02 Eq.
Table 4-125: Approach 2 Quantitative Uncertainty Estimates for HFC and PFC
Emissions from ODS Substitutes (MMT C02 Eq. and Percent)
2023 Emission
Uncertainty Range Relative to Emission
Source
Gases
Estimate
Estimate3
(MMT CO2 Eq.)
(MMT CO2 Eq.)
(%)
Lower
Upper
Lower
Upper
Bound
Bound
Bound
Bound
Substitution of Ozone
Depleting
HFCs and
PFCs
189.0
183.1
226.2
-3.1%
+ 19.6%
Substances
a Range of emission estimates predicted by Monte Carlo stochastic simulation for a 95 percent confidence interval.
Uncertainty estimates were performed in a similar manner on a species basis for HFC-32, HFC-125,
HFC-134a, and HFC-143a. A discussion of these uncertainty estimates is contained in Annex 3.10.
QA/QC and Verification
For more information on the general QA/QC process applied to this source category, consistent with
Volume 1, Chapter 6 of the 2006IPCC Guidelines, see the QA/QC and Verification Procedures section in
the introduction of the IPPU chapter. Category specific QA/QC findings are described below.
4-180 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
The QA and verification process for individual gases and sources in the Vintaging Model includes review
against up-to-date market information, including equipment stock estimates, leak rates, and sector
transitions to new chemicals and technologies. In addition, comparisons against published emission
and consumption sources by gas and by source are performed when available as described further
below. Independent peer reviews of the Vintaging Model are periodically performed, including one
conducted in 2017 (EPA 2018), to confirm Vintaging Model estimates and identify updates. For the
purposes of reporting emissions to protect confidential business information (CBI), some HFCs and
PFCs are grouped into two unspecified mixes of saturated HFCs and other PFCs and HFOs. The HFCs
and PFCs within the unspecified mix of HFCs and PFCs are modeled and verified individually in the
same process as all other gases and sources in the Vintaging Model.
Data from EPA's Greenhouse Gas Reporting Program (GHGRP)113 and emissions of some fluorinated
greenhouse gases estimated for the contiguous United States by scientists at the National Oceanic and
Atmospheric Administration (NOAA) were used to perform additional quality control as specified in 2006
IPCC Guidelines for National Greenhouse Gas Inventories and the 2019 Refinement to the 2006IPCC
Guidelines for National Greenhouse Gas Inventories (IPCC 2019). These comparisons are detailed
further in Annex 3.10.
Recalculations Discussion
For the current Inventory, updates to the Vintaging Model included routine data review and updates,
specifically updating 2023 growth rates for window units to align with sales data for Energy Star- and
non-Energy Star-certified units (EPA 2024a).
The Vintaging Model's cold storage warehouse end-use was also updated to reflect refrigerated storage
space estimates published biannually from the United States Department of Agriculture (USDA). In
addition, refrigerant transitions to ammonia and C02 were added and market penetrations were updated
based on industry data. Refrigerant charge assumptions were updated for ODS and HFC refrigerants as
well as to the newly added ammonia and C02 transitions based on data from California Air Resources
Board (CARB) and USDA (EPA 2024b).
The calculation logic used to calculate post-life emissions for closed-cell foams was also corrected to
include a multiplier for the amount of blowing agent contained in the foam. This update resulted in slight
changes in emissions for domestic refrigerator and freezer insulation Rigid Polyurethane (PU) and rigid
polyurethane and polyisocyanurate (PIR) boardstock (EPA 2025).
Together, these updates increased ODS substitute emissions on average by 4.1 MMT C02 Eq. (3.2
percent) between 1990 and 2022, compared to the previous Inventory.
Planned Improvements
Future improvements to the Vintaging Model are planned for the Refrigeration and Air-conditioning, Fire
Suppression, and Aerosols sectors, contingent on available resources. Specifically, bus and train
113 For the GHGRP data, EPA verifies annual facility-level and company-level reports through a multi-step process (e.g.,
including a combination of pre-and post-submittal electronic checks and manual reviews by staff) to identify potential
errors and ensure that data submitted to EPA are accurate, complete, and consistent (EPA 2015). Based on the results
of the verification process, EPA follows up with facilities to resolve mistakes that may have occurred. The post-
submittals checks are consistent with a number of general and category-specific QC procedures, including range
checks, statistical checks, algorithm checks, and year-to-year checks of reported data.
Industrial Processes and Product Use 4-181
-------
registrations and sales published by the U.S. Federal Highway Administration (FHWA) and American
Public Transit Association (APTA) are also being reviewed against current stock estimates in the
Vintaging Model. Residential and commercial unitary air-conditioning and multi-split air-conditioning
units projected growth rates and annual sales estimates are under review to align with the most recent
available data. Flooding agent fire suppression market transitions are under review to align more closely
with industry activities. In addition, further refinement of HFC consumption in MDIs is expected from
review of data collected on HFC use for MDI production, imports, and exports. EPA expects these
revisions to be prepared for the 2026 Inventory.
As discussed above, future reporting to EPA may provide useful information for verification purposes
and possible improvements to the Vintaging Model, such as information on HFC stockpiling behaviors.
EPA has some information and expects more by late 2026 and incorporation into the 2026 or 2027
report. Should the data suggest structural changes to the model, such as the handling of stockpiles
before use, EPA expects to introduce the revised model for the 2027 or 2028 Inventory.
Several potential improvements to the Inventory were identified in the 2022 Inventory based on the
comparisons mentioned above and discussed in Annex 3.10—net supply values from reporting to EPA
and emission estimates derived from atmospheric measurements—and remain valid. To estimate HFC
emissions for just the contiguous United States, matching the coverage by the atmospheric
measurements, EPA will investigate the availability of data from Alaska, Hawaii, and U.S. territories. This
is planned for the 2026 Inventory. To improve estimates of HFC-125 and HFC-143a, further research into
the refrigeration market can be made. Research in this industry on the shift away from blends such as R-
404A or success in lowering emission rates could be used to improve the Inventory estimate. This is
planned for the 2026 Inventory. That said, for the years where both the atmospheric measurements and
the model display a roughly constant emission of HFC-143a at similar levels, the new results suggest
robust estimates for the refrigeration market. Uncertainty estimates by species has aided in
comparisons to atmospheric data. EPA continues to explore the possibility of revising the Monte Carlo
analysis to differentiate between additional species for future version of the Inventory. Reclamation
reports and additional information could be used to improve the understanding of how chemical moves
through the economy and could resolve some of the temporal effects discussed in Annex 3.10. This
would likely require revisions to the basic model structure and could be introduced for the 2027 or 2028
Inventory. The additional data from the atmospheric measurements suggests additional items to
investigate. The faster uptick in HFC-32 and HFC-125 emissions in some years, approximately 2016
through 2020, suggests additional emissions of R-410A compared to the model's estimation. Further
investigation into the average emission rate, the variability over time of the emission rate, stocks,
lifetimes, and other factors will be investigated for the next Inventory (2026).
4.26 Electrical Equipment (Source Category
2G1)
This reporting category (2G1) includes emissions from electrical equipment manufacturing and use. The
largest use of sulfur hexafluoride (SF6), both in the United States and internationally, is as an electrical
insulator and interrupter in equipment that transmits and distributes electricity (RAND 2004). The gas
has been employed by the electric power industry in the United States since the 1950s because of its
4-182 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
dielectric strength and arc-quenching characteristics. It is used in gas-insulated substations, circuit
breakers, and other switchgear. SF6 has replaced flammable insulating oils in many applications and
allows for more compact substations in dense urban areas. Another greenhouse gas emitted in much
smaller amounts by the electric power industry is tetrafluoromethane (CF4), which is sometimes mixed
with SF6 to avoid liquefaction at low temperatures (Middleton 2000). While mixed gas circuit breakers
are more common in extremely cold climates in regions outside of the United States, some U.S.
manufacturers of electrical equipment emit CF4 during the manufacturing of equipment designed to
hold the SF6/CF4 gas mixture. However, no electric power systems in the United States have reported
emissions of or equipment using CF4. SF6 emissions exceed PFC emissions from electrical equipment
manufacturing and use on both a GWP-unweighted and GWP-weighted basis.
Fugitive emissions of SF6 and CF4 can escape from gas-insulated substations and switchgear through
seals, especially from older equipment. These gases can also be released during equipment
manufacturing, installation, servicing, and disposal. Emissions of SF6 and CF4 from equipment
manufacturing and from electric power systems were estimated to be 5.1 MMT C02 Eq. (0.2 kt) in 2023.
This quantity represents a 79 percent decrease from the estimate for 1990 (see Table 4-126 and Table
4-127). There are a few potential causes for this decrease: a sharp increase in the price of SF6 during the
1990s, a growing awareness of the environmental impact of SF6 emissions through programs such as
EPA's voluntary SF6 Emission Reduction Partnership for Electric Power Systems (Partnership) and EPA's
GHGRP, regulatory drivers at the state and local levels, and research and development of alternative
gases to SF6 that can be used in gas-insulated substations. Utilities participating in the Partnership have
lowered their emission rate (kg SF6 emitted per kg of nameplate capacity) from 13 percent in 1999 to 1.0
percent in 2023, and utilities that are not Partners but that report to EPA's GHGRP have lowered their
emission rate from 4.5 percent in 2011 to 1.4 percent in 2023. SF6 emissions reported by both sets of
electric power systems to EPA's GHGRP have decreased by 50 percent from 2011 to 2023.114 However,
total emissions from electrical equipment in 2023 were higher than 2022 emissions, increasing by 4.5
percent.
Table 4-126: SF6 and CF4 Emissions from Electric Power Systems and Electrical
Equipment Manufacturers (MMT C02 Eq.)
1990
2005
2019
2020
2021
2022
2023
Electric Power Systems
24.3
11.1
5.6
5.0
5.1
4.6
4.9
Electrical Equipment Manufacturers
0.3
0.7
0.4
0.5
0.4
0.3
0.2
Total
24.6
11.8
6.0
5.5
5.5
4.9
5.1
Note: Totals may not sum due to independent rounding.
114 Analysis of emission trends from facilities reporting to EPA's GHGRP is imperfect due to an inconsistent group of
reporters year to year. A facility that has reported total non-biogenic greenhouse gas emissions below 15,000 metric
tons of carbon dioxide equivalent (MT CO2 Eq.) for three consecutive years or below 25,000 MT CO2 Eq. for five
consecutive years to EPA's GHGRP can discontinue reporting for all direct emitter subparts. For this sector, most of the
variability in the group of reporters is due to facilities exiting the GHGRP due to being below one of these thresholds;
however, facilities must re-enter the program if their emissions at a later date are above 25,000 MT CO2 Eq., which may
occur for a variety of reasons, including changes in facility size and changes in emission rates.
Industrial Processes and Product Use 4-183
-------
Table 4-127: SF6 and CF4 Emissions from Electric Power Systems and Electrical
Equipment Manufacturers (kt)
1990
2005
2019
2020
2021
2022
2023
SFe Emissions
1
1
+
+
+
+
+
CF4 Emissions
+
+
+
+
+
+
NO
+ Does not exceed 0.5 kt.
NO (Not Occurring)
Methodology and Time-Series Consistency
The estimates of emissions from electrical equipment are comprised of emissions from electric power
systems and emissions from the manufacture of electrical equipment. The methodologies for
estimating both sets of emissions are described below.
1990 through 1998 Emissions from Electric Power Systems
Emissions from electric power systems from 1990 through 1998 were estimated based on (1) the
emissions estimated for this source category in 1999, which, as discussed in the next section, were
based on the emissions reported during the first year of EPA's SF6 Emission Reduction Partnership for
Electric Power Systems (Partnership), and (2) the RAND survey of global SF6 emissions. Because most
utilities participating in the Partnership reported emissions only for 1999 through 2011, modeling was
used to estimate SF6 emissions from electric power systems for the years 1990 through 1998. To
perform this modeling, U.S. emissions were assumed to follow the same trajectory as global emissions
from this source during the 1990 through 1999 period. To estimate global emissions, the RAND survey of
global SF6 sales was used, together with the following equation for estimating emissions, which is
derived from the mass-balance equation for chemical emissions (Volume 3, Equation 7.3) in the 2006
IPCC Guidelines.115 (Although Equation 7.3 of the 2006IPCC Guidelines appears in the discussion of
substitutes for ozone-depleting substances, it is applicable to emissions from any long-lived
pressurized equipment that is periodically serviced during its lifetime.)
Equation 4-23: Estimation for SF6 Emissions from Electric Power Systems
Emissions (kilograms SF()) = SF6 purchased to refill existing equipment (kilograms) + nameplate
capacity of retiring equipment (kilograms) 116
Note that the above equation holds whether the gas from retiring equipment is released or recaptured; if
the gas is recaptured, it is used to refill existing equipment, thereby lowering the amount of SF6
purchased by utilities for this purpose.
Gas purchases by utilities and equipment manufacturers from 1961 through 2003 are available from the
RAND (2004) survey. To estimate the quantity of SF6 released or recovered from retiring equipment, the
nameplate capacity of retiring equipment in a given year was assumed to equal 81.2 percent of the
amount of gas purchased by electrical equipment manufacturers 40 years previous (e.g., in 2000, the
nameplate capacity of retiring equipment was assumed to equal 81.2 percent of the gas purchased in
115 Ideally, sales to utilities in the United States between 1990 and 1999 would be used as a model. However, this
information was not available. There were only two U.S. manufacturers of SFe during this time period, so it would not
have been possible to conceal sensitive sales information by aggregation.
116 Nameplate capacity is defined as the amount of SFe within fully charged electrical equipment.
4-184 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
1960). The remaining 18.8 percent was assumed to have been emitted at the time of manufacture. The
18.8 percent emission factor is an average of IPCC default SF6 emission rates for Europe and Japan for
1995 (IPCC 2006). The 40-year lifetime for electrical equipment is also based on IPCC (2006). The
results of the two components of the above equation were then summed to yield estimates of global SF6
emissions from 1990 through 1999.
U.S. emissions between 1990 and 1999 are assumed to follow the same trajectory as global emissions
during this period. To estimate U.S. emissions, global emissions for each year from 1990 through 1998
were divided by the estimated global emissions from 1999. The result was a time series of factors that
express each year's global emissions as a multiple of 1999 global emissions. Historical U.S. emissions
were estimated by multiplying the factor for each respective year by the estimated U.S. emissions of SF6
from electric power systems in 1999 (estimated to be 13.9 MMT C02 Eq.).
Two factors may affect the relationship between the RAND sales trends and actual global emission
trends. One is utilities' inventories of SF6 in storage containers. When SF6 prices rise, utilities are likely to
deplete internal inventories before purchasing new SF6 at the higher price, in which case SF6 sales will
fall more quickly than emissions. On the other hand, when SF6 prices fall, utilities are likely to purchase
more SF6 to rebuild inventories, in which case sales will rise more quickly than emissions. This effect
was accounted for by applying 3-year smoothing to utility SF6 sales data. The other factor that may
affect the relationship between the RAND sales trends and actual global emissions is the level of
imports from and exports to Russia and China. SF6 production in these countries is not included in the
RAND survey and is not accounted for in any another manner by RAND. However, atmospheric studies
confirm that the downward trend in estimated global emissions between 1995 and 1998 was real (see
the Uncertainty discussion below).
1999 through 2023 Emissions from Electric Power Systems
Emissions from electric power systems from 1999 to 2023 were estimated based on: (1) reporting from
utilities participating in EPA's SF6 Emission Reduction Partnership for Electric Power Systems (Partners),
which began in 1999; (2) reporting from utilities covered by EPA's GHGRP, which began in 2012 for
emissions occurring in 2011 (GHGRP-Only Reporters); (3) SF6 emissions from California estimated by
the California Air Resources Board (CARB) and (4) the relationship between utilities' reported emissions
and their transmission miles as reported in the 2001, 2004, 2007, 2010, 2013, and 2016 Utility Data
Institute (UDI) Directories of Electric Power Producers and Distributors (UDI 2001, 2004, 2007, 2010,
2013, and 2017), and 2019, 2020, 2021, 2022, and 2023 Homeland Infrastructure Foundation-Level Data
(HIFLD) (HIFLD 2019, 2020,2021, 2022, and 2023), which was applied to the electric power systems that
do not report to EPA (Non-Reporters). Total U.S. transmission mileage was interpolated between 2016
and 2019 to estimate transmission mileage of electric power systems in 2017 and 2018. (Transmission
miles are defined as the miles of lines carrying voltages above 34.5 kV).
Partners
Over the period from 1999 to 2023, Partner utilities, which for inventory purposes are defined as utilities
that either currently are or previously have been part of the Partnership,117 represented 48 percent, on
average, of total U.S. transmission miles. Partner utilities estimated their emissions using a Tier 3 utility-
117 Starting in the 1990 to 2015 Inventory, partners who had reported three years or [ess of data prior to 2006 were removed.
Most of these Partners had been removed from the list of current Partners but remained in the Inventory due to the
extrapolation methodology for non-reporting partners.
Industrial Processes and Product Use 4-185
-------
level mass balance approach (IPCC 2006). If a Partner utility did not provide data for a particular year,
emissions were interpolated between years for which data were available or extrapolated based on
Partner-specific transmission mile growth rates. In 2012, many Partners began reporting their emissions
(for 2011 and later years) through EPA's GHGRP (discussed further below) rather than through the
Partnership. In 2023, less than 1 percent of the total emissions attributed to Partner utilities were
reported through Partnership reports. Approximately 99.1 percent of the total emissions attributed to
Partner utilities were reported and verified through EPA's GHGRP.118 Overall, the emission rates reported
by Partners have decreased significantly throughout the time series.
Non-Partners
Non-Partners consist of two groups: Utilities that have reported to the GHGRP beginning in 2012
(reporting 2011 emissions) or later years (GHGRP-only Reporters) and utilities that have never reported
to the GHGRP (Non-Reporters). EPA's GHGRP requires users of SF6 in electric power systems to report
emissions if the facility has a total SF6 nameplate capacity that exceeds 17,820 pounds. (This quantity is
the nameplate capacity that would result in annual SF6 emissions equal to 25,000 metric tons of C02
equivalent at the historical emission rate reported under the Partnership). As under the Partnership,
electric power systems that report their SF6 emissions under EPA's GHGRP are required to use the Tier 3
utility-level mass-balance approach. GHGRP-Only Reporters accounted for 16 percent of U.S.
transmission miles and 15 percent of estimated U.S. emissions from electric power system in 2023.119
From 1999 through 2008, emissions from both GHGRP-only Reporters and Non-Reporters were
estimated in the same way. From 1999 through 2008, emissions were estimated using the results of a
regression analysis that correlated the 1999 emissions from Partner utilities with their 1999
transmission miles.120 The 1999 regression coefficient (emission factor) was held constant through
2008 and multiplied by the transmission miles estimated for the non-Partners for each year.
The 1999 regression equation for Non-Partners was developed based on the emissions reported by a
subset of Partner utilities who reported non-zero emissions and non-zero transmission miles
(representing approximately 50 percent of total U.S. transmission miles). The regression equation for
1999 is displayed in the equation below.
Equation 4-24: Regression Equation for Estimating SF6 Emissions of Non-Reporting
Facilities in 1999
Emissions (kg) = 0.771 x Transmission Miles
118 Only data reported as of August 19, 2024 are used in the emission estimates for the prior year of reporting. Emissions for
Partners that did not report to the Partnership or GHGRP are extrapolated for three years using a utility-specific
transmission mile growth rate. After four consecutive years of non-reporting they are included in the 'non-reporting
Partners' category. It should be noted that data reported through EPA's GHGRP must go through a verification process.
For electric power systems, verification involved a series of electronic range, completeness, and algorithm checks for
each report submitted.
119 GHGRP-reported and Partner transmission miles from a number of facilities were equal to zero with non-zero emissions.
These facilities emissions were added to the emissions totals for their respective parent companies when identifiable
and not included in the regression equation when not identifiable or applicable. Other facilities reported non-zero
transmission miles with zero emissions, or zero transmission miles and zero emissions. These facilities were not
included in the development of the regression equations (discussed further below). These emissions are already
implicitly accounted for in the relationship between transmission miles and emissions.
120 In the United States, SFe is contained primarily in transmission equipment rated above 34.5 kV.
4-186 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
The 1999 emission factor (0.77 SF6 emissions/Transmission Miles) for the non-Partners was held
constant to estimate non-Partner emissions from 2000-2008. Non-partner emissions were assumed to
decrease beginning in 2009, trending toward the regression coefficient (emission factor) calculated for
the GHGRP-only reporters based on their reported 2011 emissions and transmission miles. Emission
factors for 2009 and 2010 were linearly interpolated between the 1999 and 2011 emission factors. For
2009, the emissions of non-Partners were estimated by multiplying their transmission miles by the
interpolated 2009 emission factor (0.65 kg/transmission mile).
The 2011 regression equation was developed based on the emissions reported by GHGRP-Only
Reporters who reported non-zero emissions and non-zero transmission miles (representing
approximately 23 percent of total U.S. transmission miles). The regression equation for 2011 is displayed
below.
Equation 4-25: Regression Equation for Estimating SF6 Emissions of GHGRP-Only
Reporters in 2011
Emissions (kg) = 0.397 x Transmission Miles
For 2011 and later years, the emissions of GHGRP-only reporters were generally equated to their
reported emissions, unless they did not report. The emissions of GHGRP-only reporters that have years
of non-reporting between reporting years are gap filled by interpolating between reported values.
For 2010 and later years, the emissions of non-Reporters were estimated by multiplying their
transmission miles by the estimated 2010 emission factor (0.52 kg/transmission mile), which was held
constant from 2010 through 2023.
Off-ramping GHGRP Facilities
The GHGRP program has an "off-ramp" provision (40 CFR Part 98.2(i)) that allows facilities to stop
reporting under certain conditions. If reported total greenhouse gas emissions are below 15,000 metric
tons of carbon dioxide equivalent (MT C02 Eq.) for three consecutive years or below 25,000 MT C02 Eq.
for five consecutive years, the facility may elect to discontinue reporting. Emissions of GHGRP reporters
that have off-ramped are extrapolated for three years of non-reporting using the weighted average
growth rate in reported nameplate capacity across all utilities. After three consecutive years of non-
reporting, emissions for facilities (except those in California) that off-ramped from GHGRP were
estimated using an emissions rate derived from the reported emissions and transmission miles of
GHGRP-only reporters in the respective year. For facilities in California, a California-specific emissions
rate is used as described in the following section.
Table 4-128: GHGRP-only Average Emission Rate (kg per mile)
Year
2011
2019
2020
2021
2022
2023
Average emission rate
0.43
0.29
0.26
0.25
0.22
0.26
Table 4-129: Categorization of Utilities and Timeseries for Application of
Corresponding Emission Estimation Methodologies
Categorization of Utilities Timeseries
Partners 1999-2021
Non-Partners (GHGRP-Only) 2011 - 2021
Industrial Processes and Product Use 4-187
-------
Categorization of Utilities
Timeseries
Non-Partners (Remaining Non-Reporting Utilities)
1999-2021
Off-ramping GHGRP Facilities
2017-2021
California
CARB reports the total SF6 emissions from electrical equipment within the state of California (CARB
2023). Because California utilities are required to report their SF6 emissions to CARB even when they are
not required to report to the GHGRP, CARB's estimates of California SF6 emissions are expected to be
more accurate for the California utilities that do not report to GHGRP than the methodology described
above. As a result, the CARB SF6 emissions estimates are used as California's contribution to the
national total for 2011 -2023, except in years where CARB's estimate is smaller than the California
estimates reported to EPA or years for which CARB has not published estimates. Since CARB's
emissions estimates include emissions from facilities that do not report to GHGRP, emissions for
California GHGRP reporters that have off-ramped are not extrapolated. Specifically, CARB estimates are
used for 2011 through 2021.
For each utility with transmission mileage in California, the GHGRP or voluntarily reported emissions
attributed to California for that utility were determined using the percentage of that utility's transmission
mileage within California based on data from HIFLD. These emissions across all California utilities were
summed to find the California emissions that were reported through GHGRP or voluntarily to the EPA.
Then, if CARB's emissions estimates for the reporting year were larger than the those from GHGRP and
voluntary reporting, CARB's emissions replaced the California emissions from GHGRP and voluntary
reporting.
If CARB's emissions estimates were lower than the California emissions estimated based on GHGRP
and voluntary reporting and on the HIFLD transmission miles for California, it is assumed there is likely
an error in the CARB estimates, as this would imply negative emissions by GHGRP non-reporters. This
was the case in 2015 and 2016. For these years, the GHGRP and voluntarily reported emissions from
California are retained, and emissions from non-reporting utilities are estimated using a California-
specific SF6 emissions rate, which is based on CARB emission data. The California SF6 emissions rate of
0.42 lbs SF6 per transmission mile is found by taking the average of CARB emissions divided by the total
California transmission mileage in years where CARB estimates are larger. Emissions from California
non-reporting utilities are then found by multiplying the California SF6 emissions rate by the California
transmission mileage from non-reporting utilities. This methodology is also used if CARB has not
published emissions estimates for a particular year. CARB has not yet published estimates for 2022 or
2023.
Table 4-130: California GHGRP and Voluntarily Reported SF6 Emissions Compared to
CARB's SF6 Emissions (MMT C02 Eq.)
2011
2015
2016
2017
2018
2019
2020
2021
2022
2023
CA GHGRP and Voluntary
0.19
0.16
0.24
0.12
0.08
0.16
0.26
0.13
0.12
0.09
CARB (CARB 2023)
0.25
0.14
0.11
0.19
0.14
0.18
0.25
0.25
NE
NE
Final CA
0.25
0.21
0.29
0.19
0.14
0.18
0.30
0.25
0.17
0.16
NE (Not Estimated)
4-188 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Total Industry Emissions
Total electric power system emissions from 1999 through 2023 were determined for each year by
summing the emissions reported by or estimated for Partners, non-Partners that report to the GHGRP,
off-ramping GHGRP Facilities (non-reporters), non-reporters who eventually report to GHGRP, and non-
reporting utilities (except in California). Then, the California GHGRP and voluntarily reported emissions
are subtracted from the total and replaced with CARB's emissions (or with GHGRP and voluntarily
reported emissions plus California non-reporting utilities' emissions).
Non-Partner Transmission Miles
Data on transmission miles for each Non-Reporter for the years 2000, 2003, 2006, and 2009, 2012, and
2016 were obtained from the 2001, 2004, 2007, 2010, 2013, and 2017 UDI Directories of Electric Power
Producers and Distributors, respectively (UDI 2001, 2004, 2007, 2010, 2013, and 2017). For 2019 to
2023 non-reporter transmission mileage was derived by subtracting reported transmission mileage data
from the total U.S. transmission mileage from 2019 to 2023 HIFLD Data (HIFLD 2019, 2020, 2021, 2022,
and 2023). The following trends in transmission miles have been observed over the time series:
• The U.S. transmission system grew by over 22,000 miles between 2000 and 2003 yet declined by
almost 4,000 miles between 2003 and 2006. Given these fluctuations, periodic increases are
assumed to occur gradually. Therefore, transmission mileage was assumed to increase at an
annual rate of 1.2 percent between 2000 and 2003 and decrease by 0.20 percent between 2003
and 2006.
• The U.S. transmission system's annual growth rate grew to 1.7 percent from 2006 to 2009 as
transmission miles increased by more than 33,000 miles.
• The annual growth rate for 2009 through 2012 was calculated to be 1.4 percent as transmission
miles grew yet again by over 29,000 miles during this time period.
• The annual transmission mile growth rate for 2012 through 2016 was calculated to be 0.2
percent, as transmission miles increased by approximately 6,600 miles.
• The annual transmission mile growth rate for 2016 through 2020 was calculated to be 0.9
percent, as transmission miles increased by approximately 25,000 miles.
• The annual transmission mile growth rate for 2020 through 2022 was calculated to be 1.5
percent, as transmission miles increased by approximately 22,000 miles.
• The annual transmission mile growth rate for 2022 through 2023 was calculated to be 0.2
percent, as transmission miles increased by approximately 1,000 miles.
Transmission miles for each year for non-reporters were calculated by interpolating between UDI
reported values obtained from the 2001, 2004, 2007, 2010, 2013 and 2017 UDI directories and HIFLD
data for 2019 and subsequent years. In cases where a non-reporter previously reported the GHGRP or
the Partnership, transmission miles were interpolated between the most recently reported value and the
next available UDI value.
1990 through 2023 Emissions from Manufacture of Electrical Equipment
Three different methods were used to estimate 1990 to 2023 emissions from original electrical
equipment manufacturers (OEMs).
Industrial Processes and Product Use 4-189
-------
• OEM SF6 emissions from 1990 through 2000 were derived by assuming that manufacturing
emissions equaled 10 percent of the quantity of SF6 provided with new equipment. The 10
percent emission rate is the average of the "ideal" and "realistic" manufacturing emission rates
(4 percent and 17 percent, respectively) identified in a paper prepared under the auspices of the
International Council on Large Electric Systems (CIGRE) in February 2002 (O'Connell et al.
2002). The quantity of SF6 provided with new equipment was estimated based on statistics
compiled by the National Electrical Manufacturers Association (NEMA). These statistics were
provided for 1990 to 2000.
• OEM SF6 emissions from 2000 through 2010 were estimated by (1) interpolating between the
emission rate estimated for 2000 (10 percent) and an emission rate estimated for 2011 based on
reporting by OEMs through the GHGRP (6.1 percent), and (2) estimating the quantities of SF6
provided with new equipment for 2001 to 2010. The quantities of SF6 provided with new
equipment were estimated using Partner reported data and the total industry SF6 nameplate
capacity estimate (160.8 MMT C02 Eq. in 2010). Specifically, the ratio of new nameplate
capacity to total nameplate capacity of a subset of Partners for which new nameplate capacity
data was available from 1999 to 2010 was calculated. These ratios were then multiplied by the
total industry nameplate capacity estimate for each year to derive the amount of SF6 provided
with new equipment for the entire industry. Additionally, to obtain the 2011 emission rate
(necessary for estimating 2001 through 2010 emissions), the estimated 2011 emissions
(estimated using the third methodology listed below) were divided by the estimated total
quantity of SF6 provided with new equipment in 2011. The 2011 quantity of SF6 provided with
new equipment was estimated in the same way as the 2001 through 2010 quantities.
• OEM CF4 emissions from 1991 through 2010 were estimated by using an average ratio of
reported SF6 and CF4 emissions from 2011 through 2013. This ratio was applied to the estimated
SF6 emissions for 1991 through 2010 to arrive at CF4 emissions. CF4 emissions are estimated
starting in 1991 and assumed zero prior to 1991 based on the entry of the CF4/SF6 gas mixture
into the market (Middleton 2000).
• OEM emissions from 2011 through 2023 were estimated using the SF6 and CF4 emissions from
OEMs reporting to the GHGRP, and an assumption that these reported emissions account for a
conservatively low estimate of 50 percent of the total emissions from all U.S. OEMs (those that
report and those that do not).
• OEM SF6 emissions from facilities off-ramping from the GHGRP were determined by
extrapolation. First, emission growth rates were calculated for each reporting year for each OEM
reporting facility as well as an average emissions growth rate (2011 through 2023). Averages of
reported emissions from last three consecutive reporting years were multiplied by the average
growth rate for each off-ramping OEM to estimate emissions for the non-reporting year(s).
Methodological approaches were applied to the entire time series to ensure time-series consistency
from 1990 through 2023.
Uncertainty
To estimate the uncertainty associated with emissions of SF6 and CF4 from electrical equipment,
uncertainties associated with four quantities were estimated: (1) emissions from Partners, (2) emissions
from GHGRP-Only Reporters, (3) emissions from Non-Reporters, and (4) emissions from manufacturers
4-190 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
of electrical equipment. A Monte Carlo analysis was then applied to estimate the overall uncertainty of
the emissions estimate.
Total emissions from the SF6 Emission Reduction Partnership include emissions from both reporting
(through the Partnership or EPA's GHGRP) and non-reporting Partners. For reporting Partners, individual
Partner-reported SF6 data was assumed to have an uncertainty of +/-10 percent. Based on a Monte
Carlo analysis, the cumulative uncertainty of all Partner-reported data was estimated to be 4.9 percent.
The uncertainty associated with extrapolated or interpolated emissions from non-reporting Partners was
assumed to be 20 percent.
For GHGRP-Only Reporters, reported SF6 data was assumed to have an uncertainty of 10 percent. Based
on a Monte Carlo analysis, the cumulative uncertainty of all GHGRP-Only reported data was estimated
to be 6.5 percent.
As discussed below, EPA has substantially revised its method for estimating emissions from non-
Reporters, assuming that the average emission rate of non-Reporters has declined much more slowly
than the average emission rate of reporting facilities rather than declining at the same rate. This
assumption brings the U.S. SF6 emissions estimated in this Inventory into better agreement with the U.S.
SF6 emissions inferred from atmospheric observations. However, it must be emphasized that the actual
emission rates of non-Reporters remain unknown. It is possible that they are lower or even higher than
estimated here. One possibility is that SF6 sources other than electric power systems are contributing to
the emissions inferred from atmospheric observations, implying that the emissions from non-Reporters
are lower than estimated here. Another is that the emissions inferred from atmospheric measurements
are over- (or under-) estimated, implying that emissions from non-Reporters could be either lower or
higher than estimated here. These uncertainties are difficult to quantify and are not reflected in the
estimated uncertainty below. The estimated uncertainty below accounts only for the two sources of
uncertainty associated with the regression equations used to estimate emissions in 2019 from Non-
Reporters: (1) uncertainty in the coefficients (as defined by the regression standard error estimate), and
(2) the uncertainty in total transmission miles for Non-Reporters. Uncertainties were also estimated
regarding (1) estimates of SF6 and CF4 emissions from OEMs reporting to EPA's GHGRP, and (2) the
assumption on the percent share of OEM emissions from OEMs reporting to EPA's GHGRP.
The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 4-131. Electrical
equipment emissions were estimated to be between 4.1 and 6.1 MMT C02 Eq. at the 95 percent
confidence level, a range of approximately 20 percent below and 20 percent above the emission
estimate of 5.1 MMT C02 Eq. There is no uncertainty estimate for CF4in 2023 as CF4 emissions did not
occur in 2023 from electrical equipment.
Table 4-131: Approach 2 Quantitative Uncertainty Estimates for SF6and CF4 Emissions
from Electrical Equipment (MMT C02 Eq. and Percent)
2023 Emission
Uncertainty Range Relative to 2022 Emission
Source
Gas
Estimate
Estimate3
(MMTCO2 Eq.)
(MMT CO2 Eq.)
(%)
Lower
Upper
Lower
Upper
Bound
Bound
Bound
Bound
Electrical
Equipment
SFe
5.1
4.1
6.1
-20%
+20%
Electrical
Equipment
LL
o
NO
NE
NE
NE
NE
Industrial Processes and Product Use 4-191
-------
a Range of emission estimates predicted by Monte Carlo stochastic simulation for a 95 percent confidence interval.
In addition to the uncertainty quantified above for the 2023 estimate, there is uncertainty associated
with the emission rates of GHGRP-only facilities before 2011 and of non-Reporters throughout the time
series. As noted above in the discussion of the uncertainty of non-Reporters for 2022, these
uncertainties are difficult to quantify.
There is also uncertainty associated with using global SF6 sales data to estimate U.S. emission trends
from 1990 through 1999. However, the trend in global emissions implied by sales of SF6 appears to
reflect the trend in global emissions implied by changing SF6 concentrations in the atmosphere. That is,
emissions based on global sales declined by 29 percent between 1995 and 1998 (RAND 2004), and
emissions based on atmospheric measurements declined by 17 percent over the same period (Levin et
al. 2010).
Several pieces of evidence indicate that U.S. SF6 emissions were reduced as global emissions were
reduced. First, the decreases in sales and emissions coincided with a sharp increase in the price of SF6
that occurred in the mid-1990s and that affected the United States as well as the rest of the world. A
representative from DILO, a major manufacturer of SF6 recycling equipment, stated that most U.S.
utilities began recycling rather than venting SF6 within two years of the price rise. Finally, the emissions
reported by the one U.S. utility that reported its emissions for all the years from 1990 through 1999 under
the Partnership showed a downward trend beginning in the mid-1990s.
QA/QC and Verification
For more information on the general QA/QC process applied to this source category, consistent with
Volume 1, Chapter 6 of the 2006IPCC Guidelines, see the QA/QC and Verification Procedures section in
the introduction of the IPPU chapter and Annex 8 for more details. Category specific QC findings are
described below.
For the GHGRP data, EPA verifies annual facility-level reports through a multi-step process (e.g.,
including a combination of pre-and post-submittal electronic checks and manual reviews by staff) to
identify potential errors and ensure that data submitted to EPA are accurate, complete, and consistent
(EPA 2015).121 Based on the results of the verification process, EPA follows up with facilities to resolve
mistakes that may have occurred. The post-submittals checks are consistent with a number of general
and category-specific QC procedures including: range checks, statistical checks, algorithm checks, and
year-to-year checks of reported data and emissions.
Additionally, EPA provides additional quality control for the SF6 emissions estimates using atmospheric
derived estimates for comparison. The 2019 Refinement to the 2006 IPCC Guidelines for National
Greenhouse Gas Inventories (IPCC 2019) Volume 1: General Guidance and Reporting, Chapter 6:
Quality Assurance, Quality Control and Verification notes that atmospheric concentration
measurements can provide independent data sets as a basis for comparison with inventory estimates.
Further, it identifies fluorinated gases as particularly suited for such comparisons. The 2019 Refinement
makes this conclusion for fluorinated gases based on their lack of significant natural sources,122 their
generally long atmospheric lifetimes, their well-known loss mechanisms, and the potential
121 GHGRP Report Verification Factsheet. See https://www.epa.gov/sites/production/files/7015-
07/rtocuments/ghgrp verification factsheet.pdf.
122 See Harnisch and Eisenhauer (1998).
4-192 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
uncertainties in bottom-up inventory methods for some of their sources. Unlike non-fluorinated
greenhouse gases (C02, CH4, and N20), SF6 has no significant natural sources; therefore, the SF6
estimates derived from atmospheric measurements are driven overwhelmingly by anthropogenic
emissions. The 2019 Refinement provides guidance on conducting such comparisons (as summarized
in Table 6.2 of IPCC (2019) Volume 1, Chapter 6) and provides guidance on using such comparisons to
identify areas of improvement in national inventories (as summarized in Box 6.5 of IPCC (2019) Volume
1, Chapter 6). Emission estimates derived from atmospheric measurements of SF6 made at NOAA and
described in Hu et al. (2022) were used to perform a comparison to the inventory estimates. This
comparison resulted in changes to historical emission estimates, as more thoroughly described in the
2021 Inventory (EPA 2022). No further changes were made to the electrical equipment estimates for the
current (i.e., 1990 through 2022) Inventory based on this comparison.
Recalculations Discussion
Several updates to activity data led to recalculations of previous Inventory results. The major updates
are as follows:
• As discussed in the methodology above, emissions of GHGRP reporters that have off-ramped
are extrapolated for three years of non-reporting using a weighted average growth rate in
reported nameplate capacity across all utilities. Formerly, the industry-wide transmission mile
growth rate was used.
• Transmission mileage for off-ramping utilities after their first three years of consecutive non-
reporting was estimated by applying the national transmission mileage growth rate to the
utilities' most recent year of reported transmission mileage. Formerly, transmission miles from
UDI for off-ramping facilities were used to develop this growth rate. Updates were made to
reporter emissions where facilities had resubmitted data.
• A correction was made to CARB's estimates of SF6, which were obtained in units of Tg C02 Eq.
and converted to Tg SF6 using the IPCC AR5 global warming potential (GWP) for the previous
Inventory. However, because CARB used the IPCC AR4 GWP to calculate its Tg C02 Eq. value,
this conversion to Tg SF6+was recalculated using the IPCC AR4 GWP. This increases emissions
for 2011 to 2014, 2017 to 2019, and 2021.
• A correction was made to calculations that estimate total 2010 transmission mileage, which
excluded one facility that eventually reported to GHGRP.
In combination, these updates resulted in changes in estimated emissions over the time series between
-5.1 percent (in 2022) and +0.3 percent (in 2016).
These updates resulted in an average annual decrease of less than 0.5 MMT C02 Eq. (0.7 percent) across
the time series compared to the previous Inventory.
Planned Improvements
EPA plans to revisit the methodology for determining emissions from the manufacture of electrical
equipment, in particular, the assumption that emissions reported by OEMs account for a conservatively
low estimate of 50 percent of the total emissions from all U.S. OEMs. Additional market research will be
required to confirm or modify the assumptions regarding the portion of industry not reporting to the
Industrial Processes and Product Use 4-193
-------
GHGRP program. EPA also plans to revisit the methodology for back casting emissions, which accounts
for changes in nameplate capacity and transmission miles, but does not account for the significant
decline in emission rates occurring at the same time.
4.27 SF6 and PFCs from Other Product Use
(Source Category 2G.2)
There are a variety of other products and processes that use fluorinated greenhouse gases. This section
estimates emissions of sulfur hexafluoride (SF6) and perfluorocarbons (PFCs) from other product use
(Source Category 2G.2), including military and scientific applications. Many of these applications utilize
SF6or PFCs to exploit their unique chemical properties, such as the high dielectric strength of SF6 and
the stability of PFCs. Emission profiles from these processes may vary greatly, ranging from immediate
and unavoidable release of all of the chemical to largely avoidable, delayed release from leak-tight
products after decades of use. In addition to estimating SF6 and PFC emissions, this category also
calculates NF3 and HFC emissions not accounted for elsewhere in the Inventory (e.g. HFC-125 used in
specialized applications), HFEs, and other Fluorinated Alcohols, Ethers, Alkanes, and Acetates
emissions are noted for informational purposes, although not included in the total emission sums.
Military applications employ SF6and PFCs in many processes. For example, SF6 is used in the radar
systems of military reconnaissance planes of the Boeing E-3A type, commonly known as Airborne
Warning and Control Systems (AWACS). These systems use SF6 to prevent electric flashovers in the
hollow conductors of the antenna, where voltages can reach up to 135 kilovolts (kV). During ascent of
the planes, SF6 is automatically released from the AWACS to maintain appropriate pressure difference
between the system and the outside air. During descent, the system is automatically charged with SF6
from an SF6 container on board. Most emissions occur during ascent but may also occur from system
leakage during other phases of flight or during time on the ground. Emissions from AWACS are largely
dependent on the number of active planes and sorties (take-offs) per year.
Other uses of SF6 in military applications include the oxidation of lithium in navel torpedoes and infrared
decoys. SF6 has also been documented for use in the quieting of torpedo propellers, as well as a by-
product of the processing of nuclear material for the production of fuel and nuclear warheads.
Military electronics are believed to be a key application for PFC heat transfer fluids, particularly in areas
such as ground and airborne radar avionics, missile guidance systems, and sonar. PFCs may also be
used to cool electric motors, especially for equipment where noise reduction is a priority (e.g.,
submarines). The specific PFCs used in military applications are similar to heat transfer fluids identified
in the electronics industry (see Section 4.24). PFCs are typically contained in a closed system, so the
emissions are most likely to occur during the manufacture, maintenance, and disposal of equipment.
SF6 and PFCs are also employed in several scientific applications, such as for use in particle
accelerators. Particle accelerators can be found in university and research settings, as well as in
industrial and medical applications. SF6 is typically used as an insulating gas and is operated in a vessel
exceeding atmospheric pressure. The amount of SF6 used in particle accelerators is largely dependent
on the terminal voltage of the unit. Emissions of SF6 typically occur when SF6 is transferred to storage
tanks while maintenance is occurring, when pressure relief valves are actuated, and through slow leaks.
4-194 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
The emission and charge assumptions for industrial particle accelerators differ from those of university
and research accelerators, as discussed in the methodology below. PFCs (particularly PFC-14) may also
be used in particle accelerators as particle detectors or counters (Workman 2022).
SF6 may also be employed in other high-voltage scientific equipment, including lasers, x-ray machines,
and electron microscopes. SF6 emission estimates for this other equipment were not quantified for this
Inventory.
There is a range of unidentified processes that also use SF6and PFCs, such as R&D activities. PFCs are
likely used primarily as heat transfer fluids (HTFs). Emissions reported for these unknown activities
group under "Other Scientific Applications."
Emissions of SF6, PFCs, and other gases unaccounted for elsewhere in the Inventory from the
applications outlined above are presented in Table 4-135. Additional emissions, included for
informational purposes but not in Table 4-135, include emissions from HFEs and HCFEs, PFPMIE,
fluorinated alcohols or acetates, and other fully fluorinated compounds. For 2023, these additional
emissions are estimated to total 4,528 MT C02 Eq.
Table 4-132: SF6 and PFC Emissions from Other Product Use (MMT C02 Eq.)
Year
1990
2005
2019
2020
2021
2022
2023
SFe
0.6
0.6
0.4
0.3
+
0.2
0.5
Total AWACs
0.6
0.6
0.4
0.3
+
0.2
0.5
SFe
0.3
0.3
0.0
0.0
0.0
0.0
0.1
PFC
0.1
0.1
0.1
0.1
0.1
0.1
0.2
NF3
0.0
0.0
0.0
0.0
0.0
+
0.0
Total Other Military Applications
0.4
0.4
0.1
0.1
0.1
0.1
0.2
SFe
0.4
0.5
0.2
0.1
0.2
0.1
0.1
PFC-14
+
+
+
+
+
+
+
Total Particle Accelerators
0.4
0.5
0.2
0.1
0.2
0.1
0.1
SFe
+
+
+
0.1
0.2
0.2
0.1
PFC
+
+
+
+
+
+
+
nf3 b
+
+
+
+
+
+
0.0
HFCsab
0.0
0.0
0.0
+
0.0
0.0
0.0
Total Other Scientific Applications
+
+
+
0.1
0.2
0.2
0.1
Total Other Product Use
1.5
1.5
0.8
0.7
0.5
0.6
1.0
+ Does not exceed 0.05 MMT C02 Eq.
a HFCs emissions not accounted for elsewhere in the Inventory.
b Listed under "other product manufacture and use" in the summary tables.
Note: Totals may not sum due to independent rounding.
Table 4-133: SF6 and PFC Emissions from Other Product Use (kt)
1990
2005
2019
2020
2021
2022
2023
SFe Emissions
+
+
+
+
+
+
+
PFC Emissions
M
M
M
M
M
M
M
NF3 Emissions®
+
+
+
+
+
+
NO
HFC Emissions'^
NO
NO
NO
+
NO
NO
NO
Industrial Processes and Product Use 4-195
-------
+ Does not exceed 0.5 kt.
M (Mixture of gases)
NO (Not Occurring)
a Listed under "other product manufacture and use" in the summary table sections for these gases.
b HFCs emissions not accounted for elsewhere in the Inventory.
Methodology and Time-Series Consistency
Emissions are based primarily on data reported through the Federal Energy Management Program
(FEMP). However, the availability of data from FEMP differs across the 1990 through 2023 time series.
Consequently, additional emission estimates were made through utilizing methodologies from the
IPCC. Emissions from military applications and scientific applications were estimated separately, and
the approaches are described immediately below.
Military Applications
1990 through 2007
FEMP data was not readily available for the 1990 to 2007 period as the first reporting year was in 2008. In
2008 and later years, the United States Department of Defense (DOD) reported fugitive emissions of SF6
but did not specify the application(s) for the SF6. Thus, for years before 2008, estimated SF6 emissions
from AWACS were calculated based using the IPCC Tier 1 methodology (IPCC 2006). IPCC provides a
default emissions factor of 740kg of SF6 per plane per year. It was estimated that the U.S. AWACS fleet
was 33 planes from 1990 to 2006, 32 planes from 2007 to 2011, and 31 planes from 2012 to 2023. This
was based on the 2006 IPCC Guidelines and further research, interpolating where necessary (E-3 Sentry
(AWACS), 2015) The IPCC methodology was utilized for all years from 1990 to 2007.
Emissions for other military applications were estimated by taking the average of the emissions
estimated for other applications as described in the next section for the first four FEMP reporting years
(i.e., 2008 and 2010 through 2012) and held constant between 1990 through 2007.
2008 through 2023
For the period 2008 through 2023, DOD reported emission data through FEMP which was used to
develop estimates for SF6 and PFCs from other military applications. SF6 emission estimates developed
for AWACS using the IPPC Tier 1 methodology (see 1990 through 2007) were compared against SF6
emissions reported by DOD between 2008 and 2023. In years where SF6 emissions reported by DOD
were smaller than those estimated using the IPCC Tier 1 methodology, DOD-reported emissions were
assumed to account for total AWACS emissions; in years where DOD emissions were greater than the
calculated AWAC emissions, the remainder is assumed to be from other SF6 applications.
Emissions from PFCs, HFEs, and other perfluoro compounds are directly reported by DOD. In years
where there are data gaps from FEMP between two reporting years, expected emissions were
interpolated. When negative values were reported, EPA took the average of the negative value and the
values in the preceding and following years and applied the average to all three years. This 3-year
average was assumed to be more representative of actual emissions. In some instances, zeroes were
added in place of blanks to ensure calculated averages were accurate.
4-196 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Scientific Applications
1990 through 2007
For the period 1990 through 2007, where no reported data is available from the Department of Energy
(DOE), estimates for emissions of SF6 and PFCs from other product use at Department of Energy
Laboratories were determined by taking an average of the first five reporting years (i.e., 2008 through
2012) where data were available or an average of 2010 through 2014 where there were prominent data
gaps for 2008.
SF6 emissions from other (non-DOE) research and industrial particle accelerators in the United States
were calculated based on the IPCC Tier 1 methodology for estimating emissions from industrial and
university/research particle accelerators. Default emission factors, charge sizes, and usage rates are
provided by size and type of accelerator in the IPCC methodology. These default assumptions were
multiplied by the number of particle accelerators of each size and type estimated to be active in the
United States by year. This methodology remained the same from 1990 to 2007.
2008 through 2023
For the period 2008 through 2023, SF6 and PFC emissions from government particle accelerators and
other scientific equipment were developed using DOE-reported emissions. SF6 and PFC emissions from
particle accelerators were directly reported by DOE. Other fugitive emissions reported by DOE for SF6
were assumed to represent emissions from particle accelerators and other scientific equipment, as well
as two DOE-managed power facilities (WAPA and BPA).123 Emissions from these two facilities were
subtracted out to present only SF6 emissions from scientific equipment. Reported fugitive emissions for
PFC-14 were assumed to wholly represent particle accelerator applications. SF6 emissions from non-
government particle accelerators were estimated using the IPPC Tier 1 methodology used for 1990
through 2007.
Process emissions from other applications for SF6 and PFCs were reported by DOE for activities such as
R&D, and these emissions were summed by gas. However, the estimates presented here do not include
emissions reported for semiconductor research and manufacture, or from refrigeration and air
conditioning. Emissions from additional PFCs, HFEs, and other perfluoro compounds are directly
reported by DOE and are reported as "Other Applications." Emissions reported to FEMP were generally
calculated based on consumption data. In a number of years, negative values for emissions were
reported due to more gas being returned to supply than purchased in a given year. As for military
applications, when negative values were reported, EPA took the average of the negative value and the
values for the preceding and following years and applied the average to all three years. This 3-year
average was assumed to be more representative of actual emissions.
In years where there are data gaps between two reporting years, emissions were interpolated. In some
instances, zeroes were added in place of blanks to ensure calculated averages were accurate.
123 DOE-reported fugitive emissions for SFe and PFCs includes emissions from high-voltage scientific equipment such as
lasers, x-rays, and electron microscopes. Emissions from this equipment is included in the particle accelerators total.
Industrial Processes and Product Use 4-197
-------
Uncertainty
A quantitative uncertainty analysis of this source category was performed using the IPCC-
recommended Approach 2 uncertainty estimation methodology, the Monte Carlo stochastic simulation
technique. The Monte Carlo stochastic simulation was performed on the total emissions estimate from
other product use, represented in equation form as:
Equation 4-26: Total Emissions from Other Product Use
Total Emissions (ET)
= Military Applications SF6, PFC, and NF3Emissions (EMilitary)
+ Scientific Applications of SF6, PFC, HFC, and NF3 Emissions (ESC(e?zt^(C)
The uncertainty in the total emissions for other product use, presented in Table 4-121 below, results
from the convolution of two distributions of emissions, namely from military applications and scientific
applications. The approaches for estimating uncertainty in each of the sources are described below:
Military Applications Emission Uncertainty
The Monte Carlo stochastic simulation was performed on the emissions estimate from military
applications, represented in equation form as:
Equation 4-27: Total Emissions from Military Applications
Military Applications SF6 NF3, and PFC Emissions (EMilitary)
= Military AWACS SF6 Emissions (Eawacs,SF6,Military)
+ Other Military Applications SF6 Emissions (Eother,sf6,Military)
+ Other Military Applications PFC Emissions (Eother,pfc,Military)
+ Other Military Applications NF3 Emissions other,nf3,Military)
The uncertainty in EM/i/ta/y results from the convolution of four distributions of emissions, Eawacs,sf6,Military,
Eother,SF6,Military, Eother,nf3,Military, and Eother,pfc,Military ¦ The approaches for estimating each distribution and
combining them to arrive at the reported 95 percent confidence interval (CI) for EMilitary are described in
the remainder of this section.
The uncertainty estimate of Eawacs,sf6,Military, or SF6 emissions from AWACS, is developed based on the
number of AWACS in commission in the United States and the per-plane emission factor. The estimated
number of active planes installed with AWACS is 33 in 1990, although estimates range between 31 and
35. The estimated number of active planes installed with AWACS is 31 in 2023, although estimates range
between 29 and 33. Bounds for the planes were rounded to avoid non-whole numbers for AWAC plane
counts. The IPCC provides a per-plane emission factor of 740 kg of SF6 per plane annually and estimates
the uncertainty to have bounds of ±14 percent.
The uncertainty in Eother,sfs,Military and Eother,pfc,Military, or SF6 and PFC emissions from other military
applications, was obtained by determining the accuracy of government-reported emissions data and
reviewing the methodology the Department of Defense uses for developing inventory estimates.
The next step in estimating the uncertainty in emissions from military AWACS and other military
applications is convolving the distribution of reported emissions, emission factors, and number of
4-198 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
AWACS using Monte Carlo simulation. For this Monte Carlo simulation, the distributions of the reported
emissions and emission factors are assumed to be normally distributed, and the number of AWACS is
assumed to have a uniform distribution since this is a discrete number of planes. The uncertainty
bounds are assigned at 1.96 standard deviations around the estimated mean.
Scientific Applications Emission Uncertainty
The Monte Carlo stochastic simulation was performed on the emissions estimate from scientific
applications, represented in equation form as:
Equation 4-28: Total Emissions from Scientific Applications
Scientific Applications SF6, PFC, NF3, and HFC Emissions (ESC(e„t^(C)
= Particle Accelerators SF6 Emissions (EAccelerators,sf6,scientific )
+ Particle Accelerators PFC Emissions (EAccelerators>PFC>Military)
+ Other Scientific Applications SF6 Emissions (Eother,sF6,scientific)
+ Other Scientific Applications PFC Emissions (Eother>P¥C>scientific)
+ Other Scientific Applications PFC Emissions (E0ther,uF3,scientific)
+ Other Scientific Applications NF3 Emissions (Eother,hfc,scientific)
The uncertainty in Escientific results from the convolution of six distributions of emissions,
^¦Accelerators,SF6,Scientific, ^-Accelerators,PFC,Scientific, Mother,SF6,Scientific, Mother,PFC,Scientific* ^-Other,NFS,Scientific, and Mother,HFC,Scientific-
The approaches for estimating each distribution and combining them to arrive at the reported 95
percent confidence interval (CI) for Escientific are described in the remainder of this section.
The uncertainty estimate of EAcceierators,sF6,sdentificandi ^Accelerators,pfc,scientific, or SF6 and PFC emissions from
particle accelerators, is developed based on fugitive and process emissions reported by the Department
of Energy and emission estimates from the number active university and industrial particle accelerators
in the United States. The number of active particle accelerators in the United States for the time series
1990 through 2023 was determined using expert judgment; default emission factors and charge sizes for
particle accelerators of various sizes were provided by IPCC guidelines. Emissions of SF6 from electrical
transmission and distribution equipment were removed from total emissions estimates for this source
category, as they are reported elsewhere in the Inventory.
The uncertainty in Eother.sFe,scientific and Eother,pfc,scientific, or SF6 and PFC emissions from other scientific
applications, was obtained by determining the accuracy of government-reported emissions data and
reviewing the methodology the Department of Energy uses for developing inventory estimates.
The next step in estimating the uncertainty in emissions from particle accelerators and other scientific
applications is convolving the distribution of calculated emissions, emission factors, number of
accelerators using Monte Carlo simulation. Similarly, the distributions of the reported emissions and
emission factors for this Monte Carlo simulation are assumed to be normally distributed, and the
number of particle accelerators and other scientific applications is assumed to have a uniform
distribution since this is a discrete number of accelerators. The uncertainty bounds are assigned at 1.96
standard deviations around the estimated mean.
The emissions estimate for total U.S. SF6 and PFC emissions from other product use were estimated to
be between 0.5 and 1.5 MMT C02 Eq. at a 95 percent confidence level. This range represents 49 percent
below and 51 percent above the 2023 emission estimate of 1.0 MMT C02 Eq. for all emissions from other
Industrial Processes and Product Use 4-199
-------
product use. This range and the associated percentages apply to the estimate of total emissions rather
than those of individual gases. Uncertainties associated with individual gases will be somewhat higher
than the aggregate but were not explicitly modeled.
Table 4-134: Approach 2 Quantitative Uncertainty Estimates for SF6 and PFC
Emissions from Other Product Use (MMT C02 Eq. and Percent)
2023 Emission
Uncertainty Range Relative to Emission
Source
Gas
Estimate
Estimate3
(MMTCO2 Eq.)
(MMT CO2 Eq.)
(%)
Lower
Upper
Lower
Upper
Bound"
Bound"
Bound
Bound
SFe, PFC,
Other Product Use
HFC and
nf3
1.0
0.5
1.5
-49%
+51 %
a Range of emission estimates predicted by Monte Carlo stochastic simulation for a 95 percent confidence interval.
b Absolute lower and upper bounds were calculated using the corresponding lower and upper bounds in percentages.
QA/QC and Verification
For more information on the general QA/QC process applied to this source category, consistent with
Volume 1, Chapter 6 of the 2006IPCC Guidelines, see the QA/QC and Verification Procedures section in
the introduction of the IPPU chapter and Annex 8 for more details. As discussed in the Methodology and
Time-Series Consistency section above, some reported data included negative emission values or data
gaps. The approach for harmonizing these negative emission values and data gaps is described in more
detail in the aforementioned section.
Recalculations Discussion
Several updates to data availability led to recalculations of previous Inventory results. The major
updates are as follows:
• Gas categories additional to SF6 and PFCs not included elsewhere in the Inventory are now also
included in this category. This category calculates these SF6 and PFCs emissions described,
also calculating NF3and HFC emissions not accounted for elsewhere in the Inventory (e.g. HFC-
125), HFEs, and other Fluorinated Alcohols, Ethers, Alkanes, and Acetates emissions are noted
for informational purposes, although not included in the total emission sums.
• The previous year Inventory (1990 to 2022) estimated emissions from AWACs and other military
uses for 2022 by taking an average of the previous five reporting years (i.e., 2017 through 2021).
This current (i.e., 1990 to 2023) Inventory used FEMP data that has since become available to
estimate emissions from AWACs and other military uses for 2022 as described in the 2008
through 2023 sub-section under Military Applications above.
• The previous year Inventory (1990 to 2022) estimated emissions using DOE reported emissions
for 2022 by taking an average of the previous five reporting years (i.e., 2017 through 2021). This
current (i.e., 1990 to 2023) Inventory used FEMP data that has since become available to
estimate emissions from AWACs and other military uses for 2022 as described in the 2008
through 2023 sub-section under Scientific Applications above.
4-200 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
A correction was made to DOD [PERFLUORO COMPOUNDS, C5-18] Fugitive emission data, which was
incorrectly shifted in the years 2011 to 2013 emissions, also affecting 1990 through 2007 emissions
which rely on these 2011 to 2013 emissions, as described in the Military Applications and Scientific
Applications sections above.
A correction has been made, adding zeros into the data used so averages were properly calculated for
year when there was missing data.
A correction was made in the treatment of negative values in the FEMP data. Emissions reported to
FEMP were generally calculated based on consumption data. In a number of years, negative values for
emissions were reported due to more gas being returned to supply than purchased in a given year. When
negative values were reported, EPA took the average of that year and the proceeding and following year
and applied that value to all three years. This 3-year average was assumed to be more representative of
actual emissions.
The number of AWACS used in the national Inventory was updated from being 33 across all years, to
using an updated source that has a total of 31 AWACs for the U.S. in 2015 (E-3 Sentry (AWACS), 2015).
Interpolation was used for the years in between the 2006 and 2015 source, resulting in the following
estimates for the U.S. AWAC fleet: 33 planes from 1990 to 2006, 32 planes from 2007 to 2011, and 31
planes from 2012 to 2023.
Planned Improvements
There are currently no planned improvements identified.
4.28 Nitrous Oxide from Product Uses
(Source Category 2G3)
This reporting category (2G3) includes exhalation emissions of N20 that arise from medical applications
and evaporative emissions of N20 from use as a propellant in aerosol products primarily in food
industry. The amount of N20 that is actually emitted depends upon the specific product use or
application. Only the medical/dental and food propellant subcategories were assumed to release
emissions into the atmosphere that are not captured under another source category, and therefore
these subcategories were the only usage subcategories with emission rates. Emissions of N20 from
semiconductor manufacturing are described in Section 4.24 and reported under Source Category 2H3.
Nitrous oxide emissions were 3.8 MMT C02 Eq. (14 kt N20) in 2023 (see Table 4-135). Production of N20
stabilized during the 1990s because medical markets had found other substitutes for anesthetics, and
more medical procedures were being performed on an outpatient basis using local anesthetics that do
not require N20. The use of N20 as a propellant for whipped cream has also stabilized due to the
increased popularity of cream products packaged in reusable plastic tubs (Heydorn 1997). Small
quantities of N20 also are used in the following applications:
• Oxidizing agent and etchant used in semiconductor manufacturing;
• Oxidizing agent used with acetylene, in atomic absorption spectrometry;
• Production of sodium azide, which is used to inflate airbags;
Industrial Processes and Product Use 4-201
-------
• Fuel oxidant in auto racing; and
• Oxidizing agent in blowtorches used by jewelers and others (Heydorn 1997).
Table 4-135: N20 Emissions from N20 Product Usage (MMT C02 Eq.)
Year 1990
2005
2019 2020
2021
2022
2023
N2O Product Usage 3.8
3.8
CO
00
CO
00
3.8
3.8
3.8
Table 4-136: N20 Emissions from N20 Product Usage (kt N20)
Year 1990
2005
2019 2020
2021
2022
2023
N2O Product Usage 14
14
14 14
14
14
14
Production of N20 in 2023 was approximately 15 kt (see Table 4-137). Three N20 production facilities
currently operate in the United States (Ottinger 2021).
Table 4-137: N20 Production (kt)
Year
1990
2005
2019
2020
2021
2022
2023
Production (kt)
16
15
15
15
15
15
15
Methodology and Time-Series Consistency
Emissions from N20 product uses are calculated using a country-specific methodology that is
consistent with 2006IPCC Guidelines and based on available data. The 2006IPCC Guidelines do not
define methodological tiers for this source category. Emissions of N20 are estimated using the national
N20 production by subcategory use or application, the share of the subcategory, and the appropriate
emission rate for each category. The following equation is adapted from Equation 8.24 of the 2006 IPCC
Guidelines:
Equation 4-29: N20 Emissions from Product Use
Epu =^CxSax ERa)
where,
Epu
P
a
Sa
ERa
N20 emissions from product uses, metric tons
Total U.S. production of N20, metric tons
specific application
Share of N20 usage by application a
Emission rate for application a, percent
The share of total quantity of N20 usage by end-use represents the share of national N20 produced that
is used by the specific subcategory (e.g., anesthesia, food processing). In 2020, the medical/dental
industry used an estimated 89.5 percent of total N20 produced, followed by food processing propellants
at 6.5 percent. All other subcategories, including semiconductor manufacturing, atomic absorption
spectrometry, sodium azide production, auto racing, and blowtorches, used the remainder of the N20
produced. This subcategory breakdown changed slightly in the mid-1990s. For instance, the small share
of N20 usage in the production of sodium azide declined significantly during the 1990s. Due to the lack
of information on the specific time period of the phase-out in this market subcategory, most of the N20
4-202 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
usage for sodium azide production is assumed to have ceased after 1996, with the majority of its small
share of the market assigned to the larger medical/dental consumption subcategory (Heydorn 1997). For
1990 through 1996, N20 usage was allocated across the following subcategories: medical applications,
food processing propellant, and sodium azide production. A usage emissions rate was then applied for
each subcategory to estimate the amount of N20 emitted.
For the medical/dental subcategory, due to the poor solubility of N20 in blood and other tissues, none of
the N20 is assumed to be metabolized during anesthesia and quickly leaves the body in exhaled breath.
Therefore, an emission factor of 100 percent was used for this subcategory (IPCC 2006). For N20 used
as a propellant in pressurized and aerosol food products, none of the N20 is reacted during the process
and all of the N20 is emitted to the atmosphere, resulting in an emission factor of 100 percent for this
subcategory (IPCC 2006). For the remaining subcategories, all of the N20 is consumed or reacted during
the process, and therefore the emission rate was considered to be zero percent (Tupman 2002).
The 1990 through 1992 N20 production data were obtained from SRI Consulting's Nitrous Oxide, North
America (Heydorn 1997). Nitrous oxide production data for 1993 through 1995 were not available.
Production data for 1996 was specified as a range in two data sources (Heydorn 1997; Tupman 2002). In
particular, for 1996, Heydorn (1997) estimates N20 production to range between 13.6 and 18.1 thousand
metric tons. Tupman (2002) provided a narrower range (15.9 to 18.1 thousand metric tons) for 1996 that
falls within the production bounds described by Heydorn (1997). Tupman (2002) data are considered
more industry-specific and current; therefore, the midpoint of the narrower production range was used
to estimate N20 emissions for years 1993 through 2001 (Tupman 2002). The 2002 and 2003 N20
production data were obtained from the Compressed Gas Association Nitrous Oxide Fact Sheet and
Nitrous Oxide Abuse Hotline (CGA 2002, 2003). These data were also provided as a range. For example,
in 2003, CGA (2003) estimates N20 production to range between 13.6 and 15.9 thousand metric tons.
Due to the lack of publicly available data, production estimates for years 2004 through 2023 were held
constant at the 2003 value.
The 1996 share of the total quantity of N20 used by each subcategory was obtained from SRI
Consulting's Nitrous Oxide, North America (Heydorn 1997). The 1990 through 1995 share of total
quantity of N20 used by each subcategory was kept the same as the 1996 number provided by SRI
Consulting. The 1997 through 2001 share of total quantity of N20 usage by sector was obtained from
communication with a N20 industry expert (Tupman 2002). The 2002 and 2003 share of total quantity of
N20 usage by sector was obtained from CGA (2002, 2003). Due to the lack of publicly available data, the
share of total quantity of N20 usage data for years 2004 through 2021 was assumed to equal the 2003
value.
The emission factor for the food processing propellant industry was obtained from SRI Consulting's
Nitrous Oxide, North America (Heydorn 1997) and confirmed by a N20 industry expert (Tupman 2002).
The emission factor for all other subcategories was obtained from communication with a N20 industry
expert (Tupman 2002). The emission factor for the medical/dental subcategory was obtained from the
2006 IPCC Guidelines.
Methodological approaches were applied to the entire time series to ensure consistency in emissions
from 1990 through 2023.
Industrial Processes and Product Use 4-203
-------
Uncertainty
The overall uncertainty associated with the 2023 N20 emission estimate from N20 product usage was
calculated using the 2006IPCC Guidelines Approach 2 methodology. Uncertainty associated with the
parameters used to estimate N20 emissions include production data, total market share of each end
use, and the emission factors applied to each end use, respectively. The uncertainty associated with
N20 production data is ±25 percent, and a uniform probability density function is assigned, based on
expert judgment (RTI 2023). The uncertainty associated with the market share for the medical/dental
subcategory is ±0.56 percent, and uncertainty for the market share of food propellant subcategory is
±25 percent, both based on expert judgment (RTI 2023). Uncertainty for emission factors was assumed
to be zero, and using this suggested uncertainty provided in the 2006 IPCC Guidelines is appropriate
based on expert judgment (RTI 2023).
The results of this Approach 2 quantitative uncertainty analysis are summarized in Table 4-138. Nitrous
oxide emissions from N20 product usage were estimated to be between 2.9 and 4.6 MMT C02 Eq. at the
95 percent confidence level. This indicates a range of approximately 24 percent below to 24 percent
above the emission estimate of 3.8 MMT C02 Eq.
Table 4-138: Approach 2 Quantitative Uncertainty Estimates for N20 Emissions from
N20 Product Usage (MMT C02 Eq. and Percent)
Source
Gas
2023 Emission Estimate
(MMT CO2 Eq.)
Uncertainty Range Relative to Emission Estimate"
(MMT CO2 Eq.) (%)
Lower
Upper Lower
Upper
Bound
Bound Bound
Bound
N2O from Product
Uses
N2O
3.8
2.9
4.6 -24%
+24%
a Range of emission estimates predicted by Monte Carlo stochastic simulation for a 95 percent confidence interval.
QA/QC and Verification
General quality assurance/quality control (QA/QC) procedures were applied consistent with the U.S.
Inventory QA/QC plan, which is in accordance with Volume 1, Chapter 6 of 2006 IPCC Guidelines as
described in the introduction of the IPPU chapter (see Annex 8 for more details).
Recalculations Discussion
No recalculations were performed for the 1990 to 2022 portion of the time series.
Planned Improvements
EPA continues to advance an evaluation of alternative production statistics for updating time-series
activity data, emission factors, assumptions, etc., and a reassessment of N20 product use
subcategories that accurately represent trends. This evaluation includes conducting a literature review
of publications and research that may provide additional information on market trends and also
emission management activities within the industry. This work remains ongoing, and thus far no
additional sources of data have been found to update this category.
4-204 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Pending additional resources and planned improvement prioritization, EPA may also evaluate
production and use cycles, and the potential need to incorporate a time lag between production and
ultimate product use and resulting release of N20. Additionally, planned improvements include
considering imports and exports of N20 for product uses.
Finally, for future Inventories, EPA will re-examine data from EPA's GHGRP to improve the emission
estimates for the N20 product use subcategory. Particular attention will be made to ensure aggregated
information can be published without disclosing CBI and time-series consistency, as the facility-level
reporting data from EPA's GHGRP are not available for all inventory years as required in this Inventory.
This is a lower priority improvement given preliminary analysis indicated limited available data, and EPA
is still assessing the possibility of incorporating aggregated GHGRP CBI data to estimate emissions;
therefore, this planned improvement is still in development and not incorporated in the current
Inventory report.
4.29 Industrial Processes and Product Use
Sources of Precursor Gases
In addition to the main greenhouse gases addressed above, many industrial processes can result in
emissions of various greenhouse gas precursors. This section summarizes information on precursor
emissions, which include carbon monoxide (CO), nitrogen oxides (NOx), non-methane volatile organic
compounds (NMVOCs), ammonia (NH3), and sulfur dioxide (S02). These gases are not direct greenhouse
gases, but indirectly impact Earth's radiative balance by altering the concentrations of greenhouse
gases (e.g., ozone) and atmospheric aerosol (e.g., particulate sulfate). Combustion byproducts such as
CO and NOx are emitted from industrial applications that employ thermal incineration as a control
technology. NMVOCs, commonly referred to as "hydrocarbons," are the primary gases emitted from
most processes employing organic or petroleum-based products, and can also result from the product
storage and handling.
Accidental releases of precursors associated with product use and handling can constitute major
emissions in this category. In the United States, emissions from product use are primarily the result of
solvent evaporation, whereby the lighter hydrocarbon molecules in the solvents escape into the
atmosphere. The major categories of product uses include: degreasing, graphic arts, surface coating,
other industrial uses of solvents (e.g., electronics), dry cleaning, and non-industrial uses (e.g., uses of
paint thinner). Product usage in the United States also results in the emission of hydrofluorocarbons
(HFCs) and small amounts of hydrofluoroethers (HFEs), which are included under Substitution of Ozone
Depleting Substances and the Electronics Industry in this chapter.
Total emissions of NOx, CO, NMVOCs, NH3, and S02 from non-energy industrial processes and product
use from 1990 to 2023 are reported in Table 4-139.
Table 4-139: NOx, CO, NMVOC, NH3, and S02 Emissions from Industrial Processes and
Product Use (kt)
Gas/Source
1990 2005 2018
2019
2020
2021
2022
z
o
7741 672 I 440
391
402
390
389
Mineral Industry
160 | 200 | 114
99
99
95
95
Industrial Processes and Product Use 4-205
-------
Other Industrial Processes®
326
335
206
187
188
185
183
Metal Industry
96
58
60
52
60
57
57
Chemical Industry
192
80
59
54
55
53
53
CO
4,096
1,701
1,011
852
899
885
882
Metal Industry
182
120
106
96
95
95
95
Other Industrial Processes®
561
662
331
291
307
330
327
Mineral Industry
2,260
707
448
340
355
322
322
Chemical Industry
1,093
211
126
125
142
139
139
NMVOCs
6,982
3,668
2,996
3,364
3,505
3,403
3,403
Other Industrial Processes®
9
10
7
6
6
6
6
Chemical Industry
6,270
3,396
2,883
3,259
3,395
3,299
3,299
Mineral Industry
102
40
20
17
19
19
19
Metal Industry
601
221
86
82
85
79
79
NH3
193
117
65
57
56
56
56
Mineral Industry
+
1
2
1
1
1
1
Other Industrial Processes
23
98
41
31
34
33
33
Metal Industry
3
2
1
0
1
1
1
Chemical Industry
167
17
21
25
21
21
21
SO2
1,490
776
309
265
273
261
260
Other Industrial Processes®
166
138
25
26
28
28
28
Chemical Industry
476
256
134
120
125
119
118
Mineral Industry
566
140
53
37
38
39
39
Metal Industry
283
242
97
82
83
75
75
a Other Industrial Processes includes storage and transport, other industrial processes (manufacturing of agriculture, food, and
kindred products; wood, pulp, paper, and publishing products; rubber and miscellaneous plastic products; machinery products;
construction; transportation equipment; and textiles, leather, and apparel products), and miscellaneous sources
(catastrophic/accidental release, other combustion (structural fires), health services, repair shops, and fugitive dust). It does not
include agricultural fires or slash/prescribed burning, which are accounted for under the Field Burning of Agricultural Residues
source.
Note: Totals by gas may not sum due to independent rounding.
Source: (EPA 2023a). Emission categories from EPA (2023a) are aggregated into sectors and categories reported as shown in
Table 2-3.
Methodology and Time-Series Consistency
Emission estimates for 1990 through 2023 were obtained from data published on the National
Emissions Inventory (NEI) Air Pollutant Emissions Trends Data website (EPA 2024). For Table 4-139, NEI
reported emissions of CO, NOx, S02, NH3, and NMVOCs were recategorized from NEI Emissions
Inventory System (EIS) sectors to source categories more closely aligned with sectors and categories in
this report, based on discussions between the EPA GHG Inventory and NEI staff (see crosswalk
documented in Annex 6.3).124 EIS sectors mapped to the IPPU sector categories in this report include:
chemical and allied product manufacturing, metals processing, storage and transport, solvent
utilization, other industrial processes, and miscellaneous sources. As described in the NEI Technical
Support Documentation (TSD) (EPA 2023b), NEI emissions are estimated through a combination of
emissions data submitted directly to the EPA by state, local, and tribal air agencies, as well as additional
information added by the Agency from EPA emissions programs, such as the emission trading program,
Toxics Release Inventory (TRI), and data collected during rule development or compliance testing.
124 The NEI estimates and reports emissions from six criteria air pollutants (CAPs) and 187 hazardous air pollutants (HAPs)
in support of National Ambient Air Quality Standards. EPA reported CAP emission trends are grouped into 60 sectors
and 15 Tier 1 source categories, which broadly cover similar source categories to those presented in this chapter. For
reporting precursor emissions in common data tables, EPA has mapped and regrouped emissions of greenhouse gas
precursors (CO, NOx, SO2, and NMVOCs) from NEI's EIS sectors to better align with NIR source categories, and to ensure
consistency and completeness to the extent possible. See Annex 6.3 for more information on this mapping.
4-206 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Methodological approaches were applied to the entire time series to ensure time-series consistency
from 1990 through 2023, which are described in detail in the NEI's TSD and on EPA's Air Pollutant
Emission Trends web site (EPA 2024; EPA 2023b). A quantitative uncertainty analysis was not performed.
Industrial Processes and Product Use 4-207
-------
Agriculture
-------
5 Agriculture
Agricultural activities contribute directly to emissions of greenhouse gases through a variety of
processes. This chapter provides an assessment of methane (CH4) from enteric fermentation, livestock
manure management, rice cultivation, and field burning of agricultural residues; nitrous oxide (N20)
emissions from agricultural soil management, livestock manure management, and field burning of
agricultural residues; as well as carbon dioxide (C02) emissions from liming and urea fertilization (see
Figure 5-1). Additional C02, CH4 and N20 fluxes from agriculture-related land-use and land-use
conversion activities, such as cultivation of cropland, management on grasslands, grassland fires,
aquaculture, and conversion of forest land to cropland, are presented in the Land Use, Land-Use
Change, and Forestry (LULUCF) chapter. Carbon dioxide emissions from stationary and mobile on-farm
energy use and CH4 and N20 emissions from stationary on-farm energy use are reported in the Energy
chapter under the Industrial sector emissions. Methane and N20 emissions from mobile on-farm energy
use are reported in the Energy chapter under mobile fossil fuel combustion emissions.
Figure 5-1: 2023 Agriculture Sector Greenhouse Gas Emission Sources
Agricultural Soil Management
Enteric Fermentation
Manure Management
Rice Cultivation |
Liming
Urea Fertilization
Field Burning of Agricultural Residues
80 100 120 140 160 180 200 220
MMT CO2 Eq.
In 2023, the Agriculture sector was responsible for emissions of 595.4 MMT C02 Eq.,1 or 9.6 percent of
total U.S. greenhouse gas emissions. Emissions of N20 by agricultural soil management through
activities such as fertilizer application and other agricultural practices that increased nitrogen
availability in the soil was the largest source of U.S. N20 emissions, accounting for 76.6 percent of
national N20 emissions, and the largest source of emissions from the Agriculture sector, accounting for
49.8 percent of total sector emissions. Methane emissions from enteric fermentation and manure
management represented 27.2 percent and 9.5 percent of total CH4 emissions from anthropogenic
activities, respectively, and 31.4 and 13.8 percent of Agriculture sector emissions, respectively. Of all
domestic animal types, beef and dairy cattle were the largest emitters of CH4. Rice cultivation and field
burning of agricultural residues were minor sources of CH4. Manure management and field burning of
1 This Inventory report presents CO2 equivalent values based on the IPCC Fifth Assessment Report (AR5) GWP values. See
the Introduction chapter as well as Chapter 9 for more information.
5-2 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
agricultural residues were also small sources of N20 emissions. Urea fertilization and liming each
accounted for 0.1 percent of total C02 emissions from anthropogenic activities.
Table 5-1 and Table 5-2 present emission estimates for the Agriculture sector. Between 1990 and 2023,
C02, CH4, and N20 emissions from agricultural activities increased by 48.3 percent, 12.3 percent, and
3.5 percent respectively. Trends in sources of agricultural emissions over the 1990 to 2023 time series
are shown in Figure 5-2. From 2022 to 2023, C02 and N20 emissions increased by 25.6 percent and 1.4
percent, respectively, as a result of soil management practices, while CH4 emissions decreased by 1.6
percent, driven by lower enteric fermentation emissions associated with beef cattle.
Table 5-1: Emissions from Agriculture (MMT C02 Eq.)
Gas/Source
1990 | 2005
2019
2020
2021
2022
2023
O
O
7.1
7.9
7.2
7.9
7.5
8.4
10.5
Liming
4.7
4.4
2.2
2.9
2.4
3.2
5.3
Urea Fertilization
2.4
3.5
4.9
5.0
5.1
5.2
5.3
CH4
241.7
264.4
280.2
282.4
282.0
275.9
271.6
Enteric Fermentation
183.1
188.2
197.3
196.3
196.5
192.6
187.1
Manure Management
39.1
55.0
66.7
66.9
66.4
64.7
65.1
Rice Cultivation
18.9
20.6
15.6
18.6
18.5
18.0
18.7
Field Burning of Agricultural Residues
0.5
0.6
0.7
0.6
0.6
0.6
0.6
N20
302.7
310.2
333.4
310.1
316.3
309.0
313.3
Agricultural Soil Management
289.1
294.7
316.4
293.0
298.9
291.8
296.3
Manure Management
13.4
15.2
16.8
16.9
17.1
17.0
16.8
Field Burning of Agricultural Residues
0.2
0.2
0.2
0.2
0.2
0.2
0.2
Total
551.5
582.5 |
620.8
600.4
605.8
593.3
595.4
Note: Totals may not sum due to independent rounding.
Table 5-2: Emissions from Agriculture (kt)
Gas/Source
1990
2005
2019
2020
2021
2022
2023
O
O
7,106
7,869
7,153
7,918
7,492
8,388
10,538
Liming
4,690
4,365
2,203
2,887
2,387
3,194
5,280
Urea Fertilization
2,417
3,504
4,950
5,031
5,105
5,193
5,258
CH4
8,633
9,444
10,008
10,087
10,073
9,853
9,699
Enteric Fermentation
6,539
6,722
7,045
7,010
7,017
6,878
6,683
Manure Management
1,398
1,964
2,382
2,390
2,373
2,312
2,326
Rice Cultivation
677
735
558
664
661
642
667
Field Burning of Agricultural Residues
19
23
23
22
22
22
22
N2O
1,142
1,170
1,258
1,170
1,194
1,166
1,182
Agricultural Soil Management
1,091
1,112
1,194
1,106
1,128
1,101
1,118
Manure Management
50
57
63
64
65
64
63
Field Burning of Agricultural Residues
1
1
1
1
1
1
1
Note: Totals may not sum due to independent rounding.
Agriculture 5-3
-------
Figure 5-2: Trends in Agriculture Sector Greenhouse Gas Emission Sources
700
650
600
550
500
450
d~
LU
400
O
u
350
Z
z
300
250
200
150
100
50
0
OTHfsjn^-Lnv£>r>.oocT>Oi-ifNn^-invDrsvCOO^o-<-HrMm^-mv£)rsscoc^OTH(Nm
cncr\cncr\o\cncncncncr>ooooooooooT-HT--iTHT-i'T-iT-i'T-iT-if-iT-irMJfM (T) CT\
/v\ L/"i irt in
va KD VO
CO CO CO
LD LO LT>
5-4 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
are not estimated due to incomplete data, with the exception of urea fertilization in Puerto Rico. EPA
continues to identify and review available data on an ongoing basis to include agriculture emissions
from U.S. Territories to the extent they are occurring in future Inventories. Other minor outlying U.S.
Territories in the Pacific Islands have no permanent populations (e.g., Baker Island) and therefore EPA
assumes no agricultural activities are occurring. See Annex 5 for more information on EPA's assessment
of the sources not included in this Inventory.
5.1 Enteric Fermentation (Source Category
3A)
Methane is produced as part of normal digestive processes in animals. During digestion, microbes
resident in an animal's digestive system ferment food consumed by the animal. This microbial
fermentation process, referred to as enteric fermentation, produces CH4 as a byproduct, which can be
exhaled or eructated by the animal. The amount of CH4 produced and emitted by an individual animal
depends primarily upon the animal's digestive system, and the amount and type of feed it consumes.2
Ruminant animals (e.g., cattle, buffalo, sheep, goats, and camels) are the major emitters of CH4
because of their unique digestive system. Ruminants possess a rumen, or large "fore-stomach," in
which microbial fermentation breaks down the feed they consume into products that can be absorbed
and metabolized. The microbial fermentation that occurs in the rumen enables them to digest coarse
plant material that non-ruminant animals cannot. Ruminant animals, consequently, have the highest
CH4 emissions per unit of body mass among all animal types.
Non-ruminant animals (e.g., swine, horses, and mules and asses) also produce CH4 emissions through
enteric fermentation, although this microbial fermentation occurs in the large intestine. These non-
ruminants emit significantly less CH4 on a per-animal-mass basis than ruminants because the capacity
of the large intestine to produce CH4 is lower.
In addition to the type of digestive system, an animal's feed quality and feed intake also affect CH4
emissions. In general, lower feed quality and/or higher feed intake leads to higher CH4 emissions. Feed
intake is positively correlated to animal size, growth rate, level of activity and production (e.g., milk
production, wool growth, pregnancy, or work). Therefore, feed intake varies among animal types as well
as among different management practices for individual animal types (e.g., animals in feedlots or
grazing on pasture).
Methane emission estimates from enteric fermentation are provided in Table 5-3 and Table 5-4. Total
livestock CH4 emissions in 2023 were 187.1 MMT C02 Eq. (6,681 kt). Beef cattle remain the largest
contributor of CH4 emissions from enteric fermentation, accounting for 70 percent in 2023. Emissions
2 CO2 emissions from livestock are not estimated because annual net CO2 emissions are assumed to be zero - the CO2
photosynthesized by plants is returned to the atmosphere as respired CO2 (IPCC 2006).
Agriculture 5-5
-------
from dairy cattle in 2023 accounted for 26 percent, and the remaining methane emissions were from
swine, horses, sheep, goats, American bison, and mules and asses.3
Table 5-3: CH4 Emissions from Enteric Fermentation (MMT C02 Eq.)
Livestock Type
1990
2005
2019
2020
2021
2022
2023
Beef Cattle
132.8
139.6
141.7
140.5
140.3
137.0
131.7
Dairy Cattle
43.3
41.3
48.5
48.8
49.4
48.9
48.7
Swine
2.3
2.6
3.2
3.2
3.1
3.1
3.1
Horses
1.1
2.0
1.3
1.2
1.1
1.0
1.0
Sheep
2.9
1.5
1.3
1.3
1.3
1.3
1.3
Goats
0.6
0.7
0.7
0.7
0.7
0.7
0.7
American Bison
0.1
0.5
0.4
0.5
0.5
0.5
0.5
Mules and Asses
+
0.1
0.1
0.1
0.1
0.1
0.1
Total
183.1
188.2
197.3
196.3
196.5
192.6
187.1
Note: Totals may not sum due to independent rounding.
+ Does not exceed 0.05 MMT C02 Eq.
Table 5-4: CH4
Emissions from Enteric Fermentation (kt CH4)
Livestock Type
1990
2005
2019
2020
2021
2022
2023
Beef Cattle
4,742
4,986
5,062
5,018
5,010
4,891
4,704
Dairy Cattle
1,547
1,473
1,732
1,743
1,764
1,748
1,740
Swine
81
92
115
115
111
110
112
Horses
40
70
46
43
40
37
35
Sheep
102
55
47
47
47
46
46
Goats
23
26
25
25
25
25
25
American Bison
4
17
16
16
17
17
17
Mules and Asses
1
2
3
3
3
3
3
Total
6,539
6,722
7,045
7,010
7,017
6,878
6,683
Note: Totals may not sum due to independent rounding.
From 1990 to 2023, emissions from enteric fermentation have increased by 2.2 percent. From 2022 to
2023, emissions decreased by 2.8 percent, largely driven by a decrease in beef cattle populations. While
emissions generally follow trends in cattle populations, there are exceptions across the time series. For
example, while dairy cattle emissions increased by 13 percent over the entire time series, the
population has declined by 5.8 percent, and milk production increased by 63 percent (USDA 2023;
USDA 2024). These trends indicate that while emissions per head are increasing, emissions per unit of
product (i.e., meat, milk) are decreasing.
3 Enteric fermentation emissions from poultry are not estimated because no IPCC method has been developed for
determining enteric fermentation ChU emissions from poultry; at this time, developing a country-specific method would
require a disproportionate amount of resources given the small magnitude of this source category. Enteric fermentation
emissions from camels are not estimated because there is no significant population of camels in the United States.
Given the insignificance of estimated camel emissions in terms of the overall level and trend in national emissions, there
are no immediate improvement plans to include this emissions category in the Inventory. See Annex 5 for more
information on significance of estimated camel emissions.
5-6 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Generally, from 1990 to 1995 emissions from beef cattle increased and then decreased from 1996 to
2004. These trends were mainly due to fluctuations in beef cattle populations and increased digestibility
of feed for feedlot cattle. Beef cattle emissions generally increased from 2004 to 2007, as beef cattle
populations increased, and an extensive literature review indicated a trend toward a decrease in feed
digestibility for those years. Beef cattle emissions decreased again from 2007 to 2014, as populations
again decreased, but increased from 2015 to 2018, consistent with another increase in population over
those same years. Emissions and populations generally declined from 2018 to 2023, with a slight post-
pandemic rebound in 2021.
Emissions from dairy cattle generally trended downward from 1990 to 2004, along with an overall dairy
cattle population decline during the same period. Similar to beef cattle, dairy cattle emissions rose from
2004 to 2007 due to population increases and a decrease in feed digestibility (based on an analysis of
more than 250 dairy cow diets used by producers across the United States). Dairy cattle emissions
continued to trend upward from 2007 to 2021, generally in line with dairy cattle population changes.
Regarding trends in other animals, populations of sheep have steadily declined, with an overall
decrease of 55 percent since 1990. Horse populations peaked in 2007 and have been declining by an
average of 4 percent annually since 2007, with their current population 13 percent lower than it was in
1990. Goat populations increased by about 20 percent through 2007 followed by a steady decrease
through 2012. From 2012 through 2023, goat populations increased by 1 percent annually. Swine
populations have trended upward through most of the time series, increasing by 43 percent from 1990 to
2020. However, swine populations decreased by around 5 percent from 2020 to 2022 before increasing
again in 2023, with a net decrease of 3 percent since 2020. The population of American bison more than
quadrupled over the 1990 to 2023 time period, while the population of mules and asses increased by a
factor of five.
Methodology and Time-Series Consistency
Livestock enteric fermentation emission estimate methodologies fall into two categories: cattle and
other domesticated animals. Cattle, due to their large population, large size, and particular digestive
characteristics, account for the majority of enteric fermentation CH4 emissions from livestock in the
United States. A more detailed methodology (i.e., IPCC Tier 2) was therefore applied to estimate
emissions for all cattle. Emission estimates for other domesticated animals (horses, sheep, swine,
goats, American bison, and mules and asses) were estimated using the IPCC Tier 1 approach, as
suggested by the 2006 IPCC Guidelines (see the Planned Improvements section).
While the large diversity of animal management practices cannot be precisely characterized and
evaluated, significant scientific literature exists that provides the necessary data to estimate cattle
emissions using the IPCC Tier 2 approach. The Cattle Enteric Fermentation Model (CEFM), developed by
EPA and used to estimate cattle CH4 emissions from enteric fermentation using IPCC's Tier 2 method,
incorporates this information and other analyses of livestock population, feeding practices, and
production characteristics. For the current Inventory, CEFM results for 1990 through 2022 were carried
over from the 1990 to 2022 Inventory (i.e., 2024 Inventory) to focus resources on CEFM improvements,
and a simplified approach was used to estimate 2023 enteric emissions from cattle.
Methodological approaches consistent with the 2006 IPCC Guidelines were applied to the entire time
series to ensure consistency in emission estimates from 1990 through 2023. See Annex 3.11 for more
detailed information on the methodology and data used to calculate CH4 emissions from enteric
Agriculture 5-7
-------
fermentation. In addition, variables and the resulting emissions are also available at the state level in
Annex 3.11.
1990-2022 Inventory Methodology for Cattle
National cattle population statistics were disaggregated into the following cattle sub-populations:
• Dairy Cattle
¦ Calves
Heifer Replacements
¦ Cows
• Beef Cattle
¦ Calves
Heifer Replacements
Heifer and Steer Stockers
Animals in Feedlots (Heifers and Steer)
¦ Cows
Bulls
Calf birth rates, end-of-year population statistics, detailed feedlot placement information, and slaughter
weight data were used to create a transition matrix that models cohorts of individual animal types and
their specific emission profiles. The key variables tracked for each of the cattle population categories
are described in Annex 3.11. These variables include performance factors such as pregnancy and
lactation as well as average weights and weight gain. Annual cattle population data were obtained from
the USDA's National Agricultural Statistics Service (NASS) QuickStats database (USDA2023).
Diet characteristics were estimated by region for dairy, grazing beef, and feedlot beef cattle. These diet
characteristics were used to calculate digestible energy (DE) values (expressed as the percent of gross
energy intake digested by the animal) and CH4 conversion rates (Ym) (expressed as the fraction of gross
energy converted to CH4) for each regional population category. The IPCC recommends Ym ranges of
3.0±1.0 percent for feedlot cattle and 6.5±1.0 percent for other well-fed cattle consuming temperate-
climate feed types (IPCC 2006). Given the availability of detailed diet information for different regions
and animal types in the United States, DE and Ym values unique to the United States were developed.
The diet characterizations and estimation of DE and Ym values were based on information from state
agricultural extension specialists, a review of published forage quality studies and scientific literature,
expert opinion, and modeling of animal physiology.
The diet characteristics for dairy cattle were based on Donovan and Baldwin (1999) and an extensive
review of nearly 20 years of literature from 1990 through 2009 (see Annex 3.11 for more information).
Estimates of DE were national averages based on the feed components of the diets observed in the
literature for the following year groupings: 1990 through 1993, 1994 through 1998, 1999 through 2003,
2004 through 2006, 2007, and 2008 onward.4 Base year Ym values by region were estimated using
Donovan (1999) and Donovan and Baldwin (1999). As described in ERG (2016), a ruminant digestion
model (COWPOLL, as selected in Kebreab et al. 2008) was used to evaluate Ym for each diet evaluated
4 Due to inconsistencies in the 2003 literature values, the 2002 values were used for 2003 as well.
5-8 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
from the literature, and a function was developed to adjust regional values over time based on the
national trend. Dairy replacement heifer diet assumptions were based on the observed relationship in
the literature between dairy cow and dairy heifer diet characteristics.
For feedlot animals, the DE and Ym values used for 1990 were recommended by Johnson (1999). Values
for DE and Ym for 1991 through 1999 were linearly extrapolated based on the 1990 and 2000 data. DE
and Ym values for 2000 onwards were based on survey data in Galyean and Gleghorn (2001) and
Vasconcelos and Galyean (2007).
For grazing beef cattle, Ym values were based on Johnson (2002), DE values for 1990 through 2006 were
based on specific diet components estimated from Donovan (1999), and DE values from 2007 onwards
were developed from an analysis byArchibeque (2011), based on diet information in Preston (2010) and
USDA-APHIS:VS (2010). Weight and weight gains for cattle were estimated from Holstein (2010), Doren
et al. (1989), Enns (2008), Lippke et al. (2000), Pinchack et al. (2004), Platter et al. (2003), Skogerboe et
al. (2000), and expert opinion. See Annex 3.11 for more details on the method used to characterize
cattle diets and weights in the United States.
Calves younger than 4 months are not included in emission estimates because calves consume mainly
milk and the IPCC recommends the use of a Ym of zero for all juveniles consuming only milk. Diets for
calves aged 4 to 6 months are assumed to go through a gradual weaning from milk decreasing to 75
percent at 4 months, 50 percent at age 5 months, and 25 percent at age 6 months. The portion of the
diet made up with milk still results in zero emissions. For the remainder of the diet, beef calf DE and Ym
are set equivalent to those of beef replacement heifers, while dairy calf DE is set equal to that of dairy
replacement heifers and dairy calf Ym is provided at 4 and 7 months of age by Soliva (2006). Estimates of
Ym for 5- and 6-month-old dairy calves are linearly interpolated from the values provided for 4 and 7
months.
To estimate CH4 emissions, the population was divided into state, age, sub-type (i.e., dairy cows and
replacements, beef cows and replacements, heifer and steer stockers, heifers and steers in feedlots,
bulls, beef calves 4 to 6 months, and dairy calves 4 to 6 months), and production (i.e., pregnant,
lactating) groupings to more fully capture differences in CH4 emissions from these animal types. The
transition matrix was used to simulate the age and weight structure of each sub-type on a monthly basis
in order to more accurately reflect the fluctuations that occur throughout the year. Cattle diet
characteristics were then used in conjunction with Tier 2 equations from IPCC (2006) to produce CH4
emission factors for the following cattle types: dairy cows, beef cows, dairy replacements, beef
replacements, steer stockers, heifer stockers, steer feedlot animals, heifer feedlot animals, bulls, and
calves. To estimate emissions from cattle, monthly population data from the transition matrix were
multiplied by the calculated emission factor for each cattle type in each state. More details are provided
in Annex 3.11.
2023 Inventory Methodology for Cattle
As noted above, a simplified approach for cattle enteric emissions was used in lieu of the CEFM for the
year 2023 to focus resources on CEFM improvements. First, 2023 populations for each of the CEFM
cattle subpopulations were estimated, then these populations were multiplied by the corresponding
2022 implied emission factors developed from the CEFM for the 1990 through 2022 Inventory (i.e., 2024
Inventory). Dairy cow, beef cow, and bull populations for 2023 were based on data directly from the
USDA-NASS QuickStats database (USDA 2023, USDA 2024). Because the remaining CEFM cattle sub-
population categories do not correspond exactly to the remaining QuickStats cattle categories, 2023
Agriculture 5-9
-------
populations for these categories were estimated by extrapolating the 2022 populations based on
percent changes from 2022 to 2023 in similar QuickStats categories, consistent with Volume 1, Chapter
5 of the 2006IPCC Guidelines on time-series consistency. Table 5-5 lists the QuickStats categories used
to estimate the percent change in population for each of the CEFM categories.
Table 5-5: Cattle Sub-Population Categories for 2023 Population Estimates
CEFM Cattle Category USDA-NASS QuickStats Cattle Category
Dairy Calves Cattle, Calves
Dairy Cows
Cattle, Cows, Milk
Dairy Replacements 7-11 months
Cattle, Heifers, GE 500 lbs, Milk Replacement
Dairy Replacements 12-23 months
Cattle, Heifers, GE 500 lbs, Milk Replacement
Bulls
Cattle, Bulls, GE 500 lbs
Beef Calves
Cattle, Calves
Beef Cows
Cattle, Cows, Beef
Beef Replacements 7-11 months
Cattle, Heifers, GE 500 lbs, Beef Replacement
Beef Replacements 12-23 months
Cattle, Heifers, GE 500 lbs, Beef Replacement
Steer Stockers
Cattle, Steers, GE 500 lbs
Heifer Stockers
Cattle, Heifers, GE 500 lbs, (Excl. Replacement)
Steer Feedlot
Cattle, On Feed
Heifer Feedlot
Cattle, On Feed
Non-Cattle Livestock
Emission estimates for other animal types were based on average emission factors (Tier 1 default IPCC
emission factors) representative of entire populations of each animal type. The methodology is in
accordance with the methodological decision tree for methane emissions from enteric fermentation
(IPCC 2019). Methane emissions from these animals accounted for a minor portion of total CH4
emissions from livestock in the United States from 1990 through 2023. Additionally, the variability in
emission factors for each of these other animal types (e.g., variability by age, production system, and
feeding practice within each animal type) is less than that for cattle.
Annual livestock population data for 1990 to 2023 for sheep, swine, goats, horses, mules and asses, and
American bison were obtained for available years from USDA-NASS (USDA 2023; USDA 2019). Horse,
goat, and mule and ass population data were available for 1987,1992,1997, 2002, 2007, 2012, and
2017 (USDA 2019); the remaining years between 1990 and 2022 were interpolated and extrapolated
from the available estimates (with the exception of goat populations being held constant between 1990
and 1992). American bison population estimates were available from USDA for 2002, 2007, 2012, and
2017 (USDA 2019) and from the National Bison Association (1999) for 1990 through 1999. Additional
years were based on observed trends from the National Bison Association (1999), interpolation between
known data points, and extrapolation beyond 2017, as described in more detail in Annex 3.11. For the
simplified approach applied to the current 1990 through 2023 Inventory, 2023 values for goats, horses,
mules and asses, and American bison were determined by extrapolating 2012 through 2022 to reflect
recent trends in the animal populations. Sheep and swine 2023 national-level animal population data
were downloaded from USDA (2024).
Methane emissions from sheep, goats, swine, horses, American bison, and mules and asses were
estimated by using emission factors utilized in Crutzen et al. (1986, cited in IPCC 2006; IPCC 2019).
5-10 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
These emission factors are representative of typical animal sizes, feed intakes, and feed characteristics
in developed countries. For American bison, the emission factor for buffalo was used and adjusted
based on the ratio of live weights to the 0.75 power. The methodology is the same as that recommended
bylPCC (2006).
Uncertainty
A quantitative uncertainty analysis for this source category was performed using the IPCC-
recommended Approach 2 uncertainty estimation methodology based on a Monte Carlo stochastic
simulation technique as described in ICF (2003). These uncertainty estimates were developed for the
1990 through 2001 Inventory (i.e., Inventory published in 2003). While there are plans to update the
uncertainty to reflect recent methodological updates and forthcoming changes (see Planned
Improvements, below), at this time the uncertainty estimates were directly applied to the 2023 emission
estimates in this Inventory.
A total of 185 primary input variables (177 for cattle and 8 for non-cattle) were identified as key input
variables for the uncertainty analysis. A normal distribution was assumed for almost all activity- and
emission factor-related input variables. Triangular distributions were assigned to three input variables
(specifically, cow-birth ratios for the three most recent years included in the 2001 model run) to ensure
only positive values would be simulated. For some key input variables, the uncertainty ranges around
their estimates (used for Inventory estimation) were collected from published documents and other
public sources; others were based on expert opinion and best estimates. In addition, both endogenous
and exogenous correlations between selected primary input variables were modeled. The exogenous
correlation coefficients between the probability distributions of selected activity-related variables were
developed through expert judgment.
Among the individual cattle sub-source categories, beef cattle account for the largest amount of CH4
emissions, as well as the largest degree of uncertainty in the emission estimates—due mainly to the
difficulty in estimating the diet characteristics for grazing members of this animal group. Among non-
cattle, horses represent the largest percent of uncertainty in the uncertainty analysis last conducted in
2001 because the Food and Agricultural Organization (FAO) of the United Nations population estimates
used for horses at that time had a higher degree of uncertainty than for the USDA population estimates
used for swine, goats, and sheep. The horse populations are drawn from the same USDA source as the
other animal types,5 and therefore the uncertainty range around horses is likely overestimated. Cattle
calves, American bison, mules and asses were excluded from the initial uncertainty estimate because
they were not included in emission estimates at that time.
The uncertainty ranges associated with the activity data-related input variables were ±10 percent or
lower. However, for many emission factor-related input variables, the lower- and/or the upper-bound
uncertainty estimates were over 20 percent. The results of the quantitative uncertainty analysis are
summarized in Table 5-6. Based on this analysis, enteric fermentation CH4 emissions in 2023 were
estimated to be between 166.5 and 220.7 MMT C02 Eq. at a 95 percent confidence level, which
indicates a range of 11 percent below to 18 percent above the 2023 emission estimate of 187.1 MMT
C02 Eq.
5 The change from using FAO data to USDA data for horse populations took place during the development of the 1990
through 2011 Inventory, published in 2013.
Agriculture 5-11
-------
As a comparison to Approach 2, a quantitative uncertainty analysis for this source category was
performed using the IPCC (2006) recommended Approach 1 based on simple error propagation. Enteric
fermentation CH4 emissions in 2023 were estimated to be between 129.2 and 245.0 MMTC02 Eq.,
which indicates a range of ±31 percent above and below the 2023 emission estimate of 187.1 MMT C02
Eq. A ±10 percent uncertainty factor is applied to the activity data (e.g., animal populations), and a ±40
percent default uncertainty factor is applied to the emission factors (IPCC 2019).
Table 5-6: Approach 2 Quantitative Uncertainty Estimates for CH4 Emissions from
Enteric Fermentation (MMT C02 Eq. and Percent)
Uncertainty Range Relative to Emission Estimate3'b,c
2023 Emission
(MMT CO2 Eq.)
(%)
Estimate
Lower
Upper
Lower
Upper
Source
Gas
(MMTCO2 Eq.)
Bound
Bound
Bound
Bound
Enteric Fermentation
CH4
187.1
166.5
220.7
-11 %
+ 18%
a Range of emissions estimates predicted by Monte Carlo stochastic simulation for a 95 percent confidence interval.
b Note that the relative uncertainty range was estimated with respect to the 2001 emission estimates from the 2003 Inventory
and applied to the 2023 estimates.
c The overall uncertainty calculated in 2003 Inventory, and applied to the 2023 emission estimate, did not include uncertainty
estimates for calves, American bison, and mules and asses. Additionally, for bulls the emissions estimate was based on the
Tier 1 methodology. Since bull emissions are now estimated using the Tier 2 method, the uncertainty surrounding their
estimates is likely lower than indicated by the previous uncertainty analysis.
QA/QC and Verification
In order to ensure the quality of the emission estimates from enteric fermentation, the General (IPCC
Tier 1) and category-specific (Tier 2) Quality Assurance/Quality Control (QA/QC) procedures were
implemented consistent with the U.S. Inventory QA/QC Plan outlined in Annex 8. Category-specific or
Tier 2 QA procedures included independent review of emission estimate methodologies from previous
Inventories.
As part of the quality assurance process, average implied emissions factors for U.S. dairy and beef cattle
were developed based on CEFM output and compared to emission factors for other countries provided
by IPCC (2006). This comparison is discussed in further detail in Annex 3.11.
Over the past few years, particular importance has been placed on harmonizing the data exchange
between the enteric fermentation and manure management source categories. The current Inventory
utilizes the same transition matrix from the CEFM for estimating cattle populations and weights for both
source categories, and the CEFM is used to output volatile solids and nitrogen excretion estimates using
the diet assumptions in the model in conjunction with the energy balance equations from the IPCC
(2006). This approach facilitates the QA/QC process for both of these source categories.
Recalculations Discussion
No time-series recalculations were performed. The1990 to 2022 estimates were retained from the 1990
through 2022 Inventory (i.e., 2024 Inventory), and 2023 estimates were based on a simplified approach
that used emission factors and extrapolated population estimates for all animals (as discussed above).
5-12 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Planned Improvements
Regular annual data reviews and updates are necessary to maintain an emissions inventory that reflects
the current base of knowledge. In addition to the documented approaches currently used to address
data availability, EPA conducts the following annual assessments to identify and determine the
applicability of newer data when updating the estimates to extend time series each year and plan future
improvements:
Further research to improve the estimation of dry matter intake (as gross energy intake) using
data from appropriate production systems;
Updating input variables that are from older data sources, such as beef births by month, beef
and dairy annual calving rates, and beef cow lactation rates;
Investigating the availability of data for dairy births by month, to replace the current assumption
that births are evenly distributed throughout the year;
Investigating the availability of annual data for updating the DE, Ym, and crude protein values of
specific diet and feed components for grazing and feedlot animals (including investigating the
availability of existing models to estimate diet characteristics, as well as the use and impact of
feed additives on emissions);
Further investigation on additional sources or methodologies for estimating DE for dairy cattle,
given the many challenges in characterizing dairy cattle diets;
Further evaluation of the assumptions about weights and weight gains for beef cows, such that
trends beyond 2007 are updated, rather than held constant; and
Further evaluation of the estimated weight for dairy cows (i.e., 1,500 lbs) that is based solely on
Holstein cows as mature dairy cow weight is likely slightly overestimated, based on knowledge
of the breeds of dairy cows in the United States.
EPA, in cooperation with USDA, is currently working to update diet parameters used in the CEFM for
dairy and beef feedlot cattle. Specifically, the EPA is incorporating recent feed composition data
reported by the Institute for Feed Education and Research (IFEEDER 2021) into the latest available
animal nutrition models to develop updated DE, Ym, and crude protein values. Diet composition data
includes feed ingredients and quantities fed in both dairy and beef sectors at the state level, and these
data are being used to develop representative diets required to meet the nutritional needs of cattle in
major beef and dairy cattle states. EPA has run the updated diets through select models after
discussions with livestock experts; these include the Ruminant Nutrition System (RNS) model for beef
cattle (Tedeschi and Fox 2020) and the National Academies of Sciences, Engineering, and Medicine
(NASEM) Dairy-8 model for dairy (NASEM 2021). EPA is assessing the model outputs and their effects on
enteric fermentation emissions, and potential consequential changes to manure management or
agricultural soils estimates. Pending further review of outputs, EPA anticipates that new diet inputs will
be integrated in the next Inventory (i.e., publishing in 2026).
Depending upon the outcome of ongoing investigations, future improvement efforts for enteric
fermentation could also include some of the following options, additional to the regular updates, and
mayor may not have implications for regular updates once addressed. Many of these improvements are
major updates and may take multiple years to fully implement:
Agriculture 5-13
-------
Potentially updating to a Tier 2 methodology for other animal types (i.e., sheep, swine, goats,
horses). Efforts to move to Tier 2 will consider the emissions significance of livestock types;
Investigation of methodologies and emission factors for including enteric fermentation emission
estimates from poultry;
Comparison of the current CEFM with other models that estimate enteric fermentation
emissions for quality assurance and verification; investigation of recent research implications
suggesting that certain parameters in enteric models may be simplified without significantly
diminishing model accuracy; and
Recent changes that have been implemented to the CEFM warrant an assessment of the current
uncertainty analysis; therefore, a revision of the quantitative uncertainty surrounding emission
estimates from this source category has been initiated. EPA plans to perform this uncertainty
analysis following the completed updates to the CEFM.
5.2 Manure Management (Source Category
3B)
The treatment, storage, and transportation of livestock manure can produce anthropogenic CH4 and
N20 emissions.6 Methane is produced by the anaerobic decomposition of manure and nitrous oxide is
produced from direct and indirect pathways through the processes of nitrification and denitrification; in
addition, there are many underlying factors that can affect these resulting emissions from manure
management, as described below.
When livestock manure is stored or treated in systems that promote anaerobic conditions (e.g., as a
liquid/slurry in lagoons, ponds, tanks, or pits), the decomposition of the volatile solids component in the
manure tends to produce CH4. When manure is handled as a solid (e.g., in stacks or drylots) or
deposited on pasture, range, or paddock lands, it tends to decompose aerobically and produce C02 and
little or no CH4. Ambient temperature, moisture, and manure storage or residency time affect the
amount of CH4 produced because they influence the growth of the bacteria responsible for CH4
formation. For non-liquid-based manure systems, moist conditions (which are a function of rainfall and
humidity) can promote CH4 production. Manure composition, which varies by animal diet, growth rate,
and animal type (particularly the different animal digestive systems), also affects the amount of CH4
produced. In general, the greater the energy content of the feed, the greater the potential for CH4
emissions. However, some higher-energy feeds also are more digestible than lower quality forages,
which can result in less overall waste excreted from the animal.
As previously stated, N20 emissions are produced through both direct and indirect pathways. Direct
N20 emissions are produced as part of the nitrogen cycle through the nitrification and denitrification of
6 CO2 emissions from livestock are not estimated because annual net CO2 emissions are assumed to be zero - the CO2
photosynthesized by plants is returned to the atmosphere as respired CO2 (IPCC 2006).
5-14 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
the nitrogen in livestock dung and urine.7 There are two pathways for indirect N20 emissions. The first is
the result of the volatilization of nitrogen in manure (as NH3 and NOx) and the subsequent deposition of
these gases and their products (NH4+ and N03~) onto soils and the surface of lakes and other waters. The
second pathway is the runoff and leaching of nitrogen from manure into the groundwater below, into
riparian zones receiving drain or runoff water, or into the ditches, streams, rivers, and estuaries into
which the land drainage water eventually flows.
The production of direct N20 emissions from livestock manure depends on the composition of the
manure (manure includes both feces and urine), the type of bacteria involved in the process, and the
amount of oxygen and liquid in the manure system. For direct N20 emissions to occur, the manure must
first be handled aerobically where organic nitrogen is mineralized or decomposed to NH4 which is then
nitrified to N03 (producing some N20 as a byproduct) (nitrification). Next, the manure must be handled
anaerobically where the nitrate is then denitrified to N20 and N2 (denitrification). NOx can also be
produced during denitrification (Groffman et al. 2000; Robertson and Groffman 2015). These emissions
are most likely to occur in dry manure handling systems that have aerobic conditions, but that also
contain pockets of anaerobic conditions due to saturation. A very small portion of the total nitrogen
excreted is expected to convert to N20 in the waste management system (WMS).
Indirect N20 emissions are produced when nitrogen is lost from the system through volatilization (as
NH3 or NOx) or through runoff and leaching. The vast majority of volatilization losses from these
operations are NH3. Although there are also some small losses of NOx, there are no quantified estimates
available for use, so losses due to volatilization are only based on NH3 loss factors. Runoff losses would
be expected from operations that house animals or store manure in a manner that is exposed to
weather. Runoff losses are also specific to the type of animal housed on the operation due to
differences in manure characteristics. Little information is known about leaching from manure
management systems as most research focuses on leaching from land application systems. However,
storage systems are often designed to minimize leaching (e.g., clay soil or synthetic liners in lagoons).
Since leaching losses are expected to be minimal, leaching losses are coupled with runoff losses and
the runoff/leaching estimate provided in this chapter does not account for any leaching losses.
Estimates of CH4 emissions from manure management in 2023 were 65.1 MMT C02 Eq. (2,326 kt); in
1990, emissions were 39.1 MMT C02 Eq. (1,398 kt). This represents a 66 percent increase in emissions
from 1990. Emissions increased on average by 0.8 MMT C02 Eq. (2 percent) annually over this period.
The majority of this increase is due to dairy cattle and beef cattle manure, where emissions increased
109 and 137 percent, respectively. From 2022 to 2023, there was a 0.6 percent increase in total CH4
emissions from manure management, mainly due to an increase in swine and poultry populations.
Although a large quantity of managed manure in the United States is handled as a solid, producing little
CH4, the general trend in manure management, particularly for dairy cattle and swine (which are both
shifting towards larger facilities), is one of increasing use of liquid systems. Also, new regulations
controlling the application of manure nutrients to land have shifted manure management practices at
smaller dairies from daily spread systems to storage and management of the manure on site. In many
cases, manure management systems with the most substantial methane emissions are those
7 Direct and indirect N2O emissions from dung and urine spread onto fields either directly as daily spread or after it is
removed from manure management systems (i.e., lagoon, pit, etc.) and from livestock dung and urine deposited on
pasture, range, or paddock lands are accounted for and discussed in the agricultural soil management source category
within the Agriculture sector.
Agriculture 5-15
-------
associated with confined animal management operations where manure is handled in liquid-based
systems. Nitrous oxide emissions from manure management vary significantly between the types of
management system used and can also result in indirect emissions due to other forms of nitrogen loss
from the system (IPCC 2006).
While national dairy animal populations have decreased since 1990, some states have seen increases
in their dairy cattle populations as the industry becomes more concentrated in certain areas of the
country and the number of animals contained on each facility increases. These areas of concentration,
such as California, New Mexico, and Idaho, tend to utilize more liquid-based systems to manage (flush
or scrape) and store manure. Thus, the shift toward larger dairy cattle and swine facilities since 1990 has
translated into an increasing use of liquid manure management systems, which have higher potential
CH4 emissions than dry systems. This significant shift in both the dairy cattle and swine industries was
accounted for by incorporating state and WMS-specific CH4 conversion factor (MCF) values in
combination with the 1992,1997, 2002, 2007, 2012, and 2017 farm-size distribution data reported in the
U.S. Department of Agriculture (USDA) Census of Agriculture (USDA2019d).
In 2023, total N20 emissions from manure management were estimated to be 16.8 MMT C02 Eq. (63 kt);
in 1990, emissions were 13.4 MMT C02 Eq. (50 kt). These values include both direct and indirect N20
emissions from manure management. Nitrous oxide emissions have increased since 1990. Multiple
drivers increase N20 emissions, such as increasing nitrogen excretion rates for some animal types (see
Table A-163) and increasing numbers of animals on feedlots versus other dry systems (e.g., pasture).
Across the entire time series, the overall net effect is that N20 emissions showed a 25 percent increase
from 1990 to 2023, but recent declines in a few animal populations (e.g., beef and dairy cattle) resulted
in a 1.3 percent decrease from 2022 to 2023.
Table 5-7 and Table 5-8 provide estimates of CH4 and N20 emissions from manure management by
animal category.8
Table 5-7: CH4 and N20 Emissions from Manure Management (MMT C02 Eq.)
Gas/Animal Type
1990
2005
2019
2020
2021
2022
2023
CH4a
39.1
55.0
66.7
66.9
66.4
64.7
65.1
Dairy Cattle
16.0
26.4
34.4
34.7
34.3
33.4
33.5
Swine
17.4
22.7
24.9
24.9
24.6
23.8
24.3
Poultry
3.8
3.4
3.1
3.0
3.0
3.0
3.0
Beef Cattle
1.8
2.2
4.1
4.2
4.4
4.3
4.2
Horses
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Sheep
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Goats
+
+
+
+
+
+
+
American Bison
+
+
+
+
+
+
+
Mules and Asses
+
+
+
+
+
+
+
N2Ob
13.4
15.2
16.8
16.9
17.1
17.0
16.8
Beef Cattle
5.2
6.0
6.0
6.1
6.4
6.4
6.2
Dairy Cattle
5.5
5.5
6.2
6.2
6.3
6.2
6.1
Manure management emissions from camels are not estimated because there is no significant population of camels in
the United States. Given the insignificance of estimated camel emissions in terms of the overall level and trend in
national emissions, there are no immediate improvement plans to include this emissions category in the Inventory. See
Annex 5 for more information on significance of estimated camel emissions.
5-16 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Gas/Animal Type
1990
2005
2019
2020
2021
2022
2023
Swine
1.1
1.5
1.8
1.9
1.8
1.8
1.8
Poultry
1.3
1.8
2.3
2.3
2.3
2.3
2.3
Sheep
0.1
0.3
0.3
0.3
0.3
0.3
0.3
Horses
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Goats
+
+
+
+
+
+
+
Mules and Asses
+
+
+
+
+
+
+
American Bison0
NA
NA
NA
NA
NA
NA
NA
Total
52.5
70.2
83.5
83.8
83.6
81.7
81.9
+ Does not exceed 0.05 MMT C02 Eq.
NA (Not Available)
a Accounts for ChU reductions due to capture and destruction of ChU at facilities using anaerobic digesters.
b Includes both direct and indirect N20 emissions.
c There are no American bison N20 emissions from managed systems; American bison are maintained entirely on pasture, range,
and paddock.
Notes: N20 emissions from manure deposited on pasture, range and paddock are included in the agricultural soils management
category. Totals may not sum due to independent rounding.
Table 5-8: CH4 and N20 Emissions from Manure Management (kt)
Gas/Animal Type
1990
2005
2019
2020
2021
2022
2023
CH4a
1,398
1,964
2,382
2,390
2,373
2,312
2,326
Dairy Cattle
572
943
1,227
1,238
1,226
1,193
1,195
Swine
621
812
890
888
877
851
868
Poultry
135
123
111
109
108
108
108
Beef Cattle
63
78
148
150
157
154
149
Horses
4
5
3
3
3
2
2
Sheep
3
2
2
2
2
2
2
Goats
+
+
+
+
+
+
+
American Bison
+
+
+
+
+
+
+
Mules and Asses
+
+
+
+
+
+
+
N2Ob
50
57
63
64
65
64
63
Beef Cattle
20
23
23
23
24
24
23
Dairy Cattle
21
21
23
24
24
23
23
Swine
4
6
7
7
7
7
7
Poultry
5
7
9
9
9
9
9
Sheep
+
1
1
1
1
1
1
Horses
+
+
+
+
+
+
+
Goats
+
+
+
+
+
+
+
Mules and Asses
+
+
+
+
+
+
+
American Bison0
NA
NA
NA
NA
NA
NA
NA
+ Does not exceed 0.5 kt.
NA (Not Available)
a Accounts for ChU reductions due to capture and destruction of ChU at facilities using anaerobic digesters.
b Includes both direct and indirect N20 emissions.
° There are no American bison N20 emissions from managed systems; American bison are maintained entirely on pasture, range,
and paddock.
Notes: N20 emissions from manure deposited on pasture, range and paddock are included in the agricultural soils management
category. Totals by gas may not sum due to independent rounding.
Agriculture 5-17
-------
Methodology and Time-Series Consistency
The methodologies presented in IPCC (2006) form the basis of the CH4 and N20 emission estimates for
each animal type, including Tier 1, Tier 2, and use of the CEFM previously described for enteric
fermentation. These methodologies use:
IPCC (2019) Tier 1 default N20 emission factors and Methane conversion factors (MCFs) for dry
systems
U.S. specific MCFs for liquid systems (ERG 2001)
U.S. specific values for volatile solids (VS) production rate and nitrogen excretion rate for some
animal types, including cattle values from the CEFM
This combination of Tier 1 and Tier 2 methods was applied to all livestock animal types and follows
guidance for methodological choice presented in decision trees from the IPCC (2006). This section
presents a summary of the methodologies used to estimate CH4 and N20 emissions from manure
management. For the current Inventory, time-series results were carried over from the previous
Inventory (i.e., 2024 publication) and a simplified approach was used to estimate manure management
emissions for 2023.
See Annex 3.12 for more detailed information on the methodologies (including detailed formulas and
emission factors), data used to calculate CH4 and N20 emissions, and emission results (including input
variables and results at the state-level) from manure management.
Methane Calculation Methods
The following inputs were used in the calculation of manure management CH4 emissions for 1990
through 2022:
Animal population data (by animal type and state);
Typical animal mass (TAM) data (by animal type);
Portion of manure managed in each WMS, by state and animal type;
VS production rate (by animal type and state or United States);
Methane producing potential (B0) of the volatile solids (by animal type); and
MCF, the extent to which the CH4 producing potential is realized for each type of WMS (by state
and manure management system, including the impacts of any biogas collection efforts).
Methane emissions were estimated by first determining activity data, including animal population, TAM,
WMS usage, and waste characteristics. The activity data sources are described below:
Annual animal population data for 1990 through 2022 for all livestock types, except goats,
horses, mules and asses, and American bison were obtained from the USDA-NASS. For cattle,
the USDA populations were utilized in conjunction with birth rates, detailed feedlot placement
information, and slaughter weight data to create the transition matrix in the Cattle Enteric
Fermentation Model (CEFM) that models cohorts of individual animal types and their specific
emission profiles. The key variables tracked for each of the cattle population categories are
described in Section 5.1 and in more detail in Annex 3.11. Goat population data for 1992,1997,
2002, 2007, 2012, and 2017; horse and mule and ass population data for 1987,1992, 1997,
5-18 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
2002, 2007, 2012, and 2017; and American bison population for 2002, 2007, 2012, and 2017
were obtained from the Census of Agriculture (USDA 2019d). American bison population data
for 1990 through 1999 were obtained from the National Bison Association (1999).
The TAM is an annual average weight that was obtained for animal types other than cattle from
information in USDA's Agricultural Waste Management Field Handbook (USDA 1996), the
American Society of Agricultural Engineers, Standard D384.1 (ASAE 1998) and others (Meagher
1986; EPA 1992; Safley 2000; ERG 2003b; IPCC 2006; ERG 2010a). For a description of the TAM
data used for cattle, see Annex 3.11.
WMS usage was estimated for swine and dairy cattle for different farm size categories using
state and regional data from USDA (USDA APHIS 1996; Bush 1998; Ott 2000; USDA 2016c) and
EPA (ERG 2000a; EPA 2002a and 2002b; ERG 2018, ERG 2019). For beef cattle and poultry,
manure management system usage data were not tied to farm size but were based on other data
sources (ERG 2000a; USDA APHIS 2000; UEP 1999, ERG 2023). For other animaltypes, manure
management system usage was based on previous estimates (EPA 1992). American bison WMS
usage was assumed to be the same as not on feed (NOF) cattle, while mules and asses were
assumed to be the same as horses.
VS production rates for all cattle except for calves were calculated by head for each state and
animal type in the CEFM. VS production rates by animal mass for all other animals were
determined using data from USDA's Agricultural Waste Management Field Handbook (USDA
1996 and 2008; ERG 2010b and 2010c) and data that was not available in the most recent
Handbook were obtained from the American Society of Agricultural Engineers, Standard D384.1
(ASAE 1998) or the 2006 IPCC Guidelines (IPCC 2006). American bison VS production was
assumed to be the same as NOF bulls.
Bo was determined for each animal type based on literature values (Morris 1976; Bryant et al.
1976; Hashimoto 1981; Hashimoto 1984; EPA 1992; Hill 1982; Hill 1984).
MCFs for dry systems were set equal to default IPCC factors based on state climate for each
year (IPCC 2019). The IPCC 2019 factors are more representative of U.S. systems and reflect the
latest science. MCFs for liquid/slurry, anaerobic lagoon, and deep pit systems were calculated
based on the forecast performance of biological systems relative to temperature changes as
predicted in the van't Hoff-Arrhenius equation which is consistent with IPCC (2006) Tier 2
methodology.
Data from anaerobic digestion systems with CH4 capture and combustion were obtained from
the EPA AgSTAR Program, including information available in the AgSTAR project database (EPA
2023). Anaerobic digester emissions were calculated based on estimated methane production
and collection and destruction efficiency assumptions (ERG 2008).
For all cattle except for calves, the estimated amount of VS (kg per animal-year) managed in
each WMS for each animal type, state, and year were taken from the CEFM, assuming American
bison VS production to be the same as NOF bulls. For animals other than cattle, the annual
amount of VS (kg per year) from manure excreted in each WMS was calculated for each animal
type, state, and year. This calculation multiplied the animal population (head) by the VS
excretion rate (kg VS per 1,000 kg animal mass per day), the TAM (kg animal mass per head)
divided by 1,000, the WMS distribution (percent), and the number of days per year (365.25).
Agriculture 5-19
-------
The estimated amount of VS managed in each WMS was used to estimate the CH4 emissions (kg CH4 per
year) from each WMS. The amount of VS (kg per year) was multiplied by the B0 (m3 CH4 per kg VS), the
MCF for that WMS (percent), and the density of CH4 (kg CH4per m3 CH4). The CH4 emissions for each
WMS, state, and animal type were summed to determine the total U.S. CH4 emissions. See details in
Step 5 of Annex 3.12.
The following approach was used in the calculation of manure management CH4 emissions for 2023:
Obtained 2023 national-level animal population data: Sheep, poultry, and swine data were
downloaded from USDA-NASS QuickStats (USDA 2024a, 2024b, 2024c). Cattle populations
were obtained from the CEFM (see NIR Section 5.1 and Annex 3.11). Data for goats, horses,
bison, mules, and asses were extrapolated based on the 2012 through 2022 population values
to reflect recent trends in animal populations.
Multiplied the national populations by the animal-specific 2022 implied emission factors9 for
CH4 to calculate national-level 2023 CH4 emissions estimates by animal type. These methods
were utilized in order to maintain time-series consistency consistent with Volume 1, Chapter 5
of the 2006IPCC Guidelines.
Nitrous Oxide Calculation Methods
The following inputs were used in the calculation of direct and indirect manure management N20
emissions for 1990 through 2022:
Animal population data (by animal type and state);
TAM data (by animal type);
Portion of manure managed in each WMS (by state and animal type);
Total Kjeldahl nitrogen excretion rate (Nex);
Direct N20 emission factor (EFwms);
Indirect N20 emission factor for volatilization (EFvoiatiiization);
Indirect N20 emission factor for runoff and leaching (EFrun0ff/ieach);
Fraction of nitrogen loss from volatilization of NH3 and NOx (Fracgas); and
Fraction of nitrogen loss from runoff and leaching (Fracrunoff/ieach).
Nitrous oxide emissions were estimated by first determining activity data, including animal population,
TAM, WMS usage, and waste characteristics. The activity data sources (except for population, TAM, and
WMS, which were described above) are described below:
Nex for all cattle except for calves were calculated by head for each state and animal type in the
CEFM. Nex rates by animal mass for all other animals were determined using data from USDA's
Agricultural Waste Management Field Handbook (USDA 1996 and 2008; ERG 2010b and 2010c)
9 An implied emission factor is defined as emissions divided by the relevant measure of activity; the implied emission
factor is equal to emissions per activity data unit. For source/sink categories that are composed of several subcategories,
the emissions and activity data are summed up across all subcategories. Hence, the implied emission factors are
generally not equivalent to the emission factors used to calculate emission estimates, but are average values that could
be used, with caution, in data comparisons (UNFCCC 2017).
5-20 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
and data from the American Society of Agricultural Engineers, Standard D384.1 (ASAE 1998) and
IPCC (2006). American bison Nex were assumed to be the same as NOF bulls.10
Country-specific estimates for the fraction of nitrogen loss from volatilization (Fracgas) and runoff
and leaching (Fracrun0ff/ieach) were developed. Fracgas values were based on WMS-specific
volatilization values as estimated from EPA's National Emission Inventory - Ammonia Emissions
from Animal Agriculture Operations (EPA2005). Fracrunoff/ieaching values were based on regional
cattle runoff data from EPA's Office of Water (EPA 2002b; see Annex 3.12).
To estimate N20 emissions for cattle (except for calves), the estimated amount of nitrogen excreted (kg
per animal-year) that is managed in each WMS for each animal type, state, and year were taken from the
CEFM. For calves and other animals, the amount of nitrogen excreted (kg per year) in manure in each
WMS for each animal type, state, and year was calculated. The population (head) for each state and
animal was multiplied by TAM (kg animal mass per head) divided by 1,000, the nitrogen excretion rate
(Nex, in kg N per 1,000 kg animal mass per day), WMS distribution (percent), and the number of days per
year.
Direct N20 emissions were calculated by multiplying the amount of nitrogen excreted (kg per year) in
each WMS by the N20 direct emission factor for that WMS (EFwms, in kg N20-N per kg N) and the
conversion factor of N20-N to N20. These emissions were summed over state, animal, and WMS to
determine the total direct N20 emissions (kg of N20 per year). See details in Step 6 of Annex 3.12.
Indirect N20 emissions from volatilization (kg N20 per year) were then calculated by multiplying the
amount of nitrogen excreted (kg per year) in each WMS by the fraction of nitrogen lost through
volatilization (Fracgas) divided by 100, the emission factor for volatilization (EFvoiatiiiZation, in kg N20 per kg
N), and the conversion factor of N20-N to N20. Indirect N20 emissions from runoff and leaching (kg N20
per year) were then calculated by multiplying the amount of nitrogen excreted (kg per year) in each WMS
by the fraction of nitrogen lost through runoff and leaching (Fracrunoff/ieach) divided by 100, the emission
factor for runoff and leaching (EFrun0ff/ieach, in kg N20 per kg N), and the conversion factor of N20-N to N20.
The indirect N20 emissions from volatilization and runoff and leaching were summed to determine the
total indirect N20 emissions. See details in Step 6 of Annex 3.12.
Following these steps, direct and indirect N20 emissions were summed to determine total N20
emissions (kg N20 per year) for the years 1990 to 2022.
Methodological approaches, changes to historic data, and other parameters were applied to the entire
time series to ensure consistency in emissions estimates from 1990 through 2022. In some cases, the
activity data source changed over the time series. For example, updated WMS distribution data were
applied to 2016 for dairy cows and 2009 for swine. While previous WMS distribution data were from
another data source, EPA integrated the more recent data source to reflect the best available current
WMS distribution data for these animals. EPA assumed a linear interpolation distribution for years
between the two data sources. Refer to Annex 3.12 for more details on data sources and methodology.
The following approach was used in the calculation of manure management N20 emissions for 2023:
10 Nex of American bison on grazing lands are accounted for and discussed in the agricultural soil management source
category and included under pasture, range and paddock (PRP) emissions. Because American bison are maintained
entirely on unmanaged WMS and N2O emissions from unmanaged WMS are not included in the manure management
source category, there are no N2O emissions from American bison included in the manure management source category.
Agriculture 5-21
-------
Obtained 2023 national-level animal population data: Sheep, poultry, and swine data were
downloaded from USDA-NASS Quickstats (USDA 2024a, 2024b, 2024c). Cattle populations
were obtained from the CEFM, see Section 5.1 and Annex 3.11 (Enteric Fermentation). Data for
goats, horses, bison, mules, and asses were extrapolated based on the 2012 through 2022
population values to reflect recent trends in animal populations.
The national populations were multiplied by the animal-specific 2022 implied emission factors
for N20 (which combines both direct and indirect N20) to calculate national-level 2023 N20
emissions estimates by animal type. These methods were utilized in order to maintain time-
series consistency consistent with Volume 1, Chapter 5 of the 2006IPCC Guidelines.
Uncertainty
An analysis (ERG 2003a) was conducted for the manure management emission estimates presented in
the 1990 through 2001 Inventory (i.e., Inventory published in 2003) to determine the uncertainty
associated with estimating CH4 and N20 emissions from livestock manure management. The
quantitative uncertainty analysis for this source category was performed in 2002 through the IPCC-
recommended Approach 2 uncertainty estimation methodology, the Monte Carlo stochastic simulation
technique. The uncertainty analysis was developed based on the methods used to estimate CH4 and
N20 emissions from manure management systems. The series of equations used were condensed into a
single equation for each animal type and state. The equations for each animal group contained four to
five variables around which the uncertainty analysis was performed for each state. A normal probability
distribution was assumed for all variables in the estimation equations. While there are plans to update
the uncertainty to reflect recent manure management updates and forthcoming changes (see Planned
Improvements, below), at this time the uncertainty estimates were directly applied to the 2023 emission
estimates.
The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 5-9. Manure
management CH4 emissions in 2023 were estimated to be between 53.4 and 78.2 MMT C02 Eq. at a 95
percent confidence level, which indicates a range of 18 percent below to 20 percent above the actual
2023 emission estimate of 65.1 MMT C02 Eq. At the 95 percent confidence level, N20 emissions were
estimated to be between 14.1 and 20.8 MMT C02 Eq. (or approximately 16 percent below and 24 percent
above the actual 2023 emission estimate of 16.8 MMT C02 Eq.).
A quantitative uncertainty analysis for this source category was also performed using the IPCC (2006)
recommended Approach 1 based on simple error propagation as well. Based on this analysis, manure
management:
CH4 emissions in 2023 were estimated to be between 50.7 and 79.5 MMT C02 Eq., which
indicates a range of ±22 percent above and below the 2023 emission estimate of 65.1 MMT C02
Eq. A ±25 percent uncertainty factor is applied to the activity data (e.g., animal populations),
and a ±30 percent default uncertainty factor for Tier 1 and ±20 percent default uncertainty factor
for Tier 2 is applied to the emission factors (IPCC 2006).
N20 emissions in 2023 were estimated to be between 11.6 and 21.9 MMTC02 Eq., which
indicates a range of ±31 percent above and below the 2023 emission estimate of 16.8 MMT C02
Eq. A ±25 percent uncertainty factor is applied to the activity data (e.g., animal populations),
and a ±50 percent default uncertainty factor is applied to the emission factors (IPCC 2006).
5-22 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Combined CH4 and N20 emissions in 2023 were estimated to be between 66.6 and 97.2 MMT
C02 Eq., which indicates a range of ±19 percent above and below the 2023 emission estimate of
81.9 MMT C02 Eq. A ±25 percent uncertainty factor is applied to the activity data (e.g., animal
populations), and a ±20-50 percent default uncertainty factor is applied to the emission factors
(IPCC2006).
Table 5-9: Approach 2 Quantitative Uncertainty Estimates for CH4 and N20 (Direct and
Indirect) Emissions from Manure Management (MMT C02 Eq. and Percent)
Uncertainty Range Relative to Emission Estimate"
2023 Emission
(MMT CO2
Eq.)
(%)
Estimate
Lower
Upper
Lower
Upper
Source
Gas
(MMTCO2 Eq.)
Bound
Bound
Bound
Bound
Manure Management
CH4
65.1
53.4
78.2
-18%
+20%
Manure Management
n2o
16.8
14.1
20.8
-16%
+24%
a Range of emission estimates predicted by Monte Carlo stochastic simulation for a 95 percent confidence interval.
QA/QC and Verification
General (Tier 1) and category-specific (Tier 2) QA/QC activities were conducted consistent with the U.S.
Inventory QA/QC plan outlined in Annex 8. Tier 2 activities focused on comparing estimates for the
previous and current Inventories for N20 emissions from managed systems and CH4 emissions from
livestock manure. All errors identified were corrected. Order of magnitude checks were also conducted,
and corrections made where needed. In addition, manure nitrogen data were checked by comparing
state-level data with bottom-up estimates derived at the county level and summed to the state level.
Similarly, a comparison was made by animal and WMS type for the full time series, between national
level estimates for nitrogen excreted, both for pasture and managed systems, and the sum of county
estimates for the full time series. This was done to ensure consistency between excreted nitrogen within
the manure management sector and those data provided to the managed soils sector. All errors
identified were corrected.
Time-series data, including population, are validated by experts to ensure they are representative of the
best available U.S.-specific data. The U.S.-specific values for TAM, Nex, VS, B0, and MCF were also
compared to the IPCC default values and validated by experts. Although significant differences exist in
some instances, these differences are due to the use of U.S.-specific data and the differences in U.S.
agriculture as compared to other countries. The U.S. manure management emission estimates use the
most reliable country-specific data, which are more representative of U.S. animals and systems than
the IPCC (2006) default values.
For additional verification of the 1990 to 2023 estimates, the implied CH4 emission factors for manure
management (kg of CH4 per head per year) were compared against the default IPCC (2006) values. Table
5-10 presents the implied emission factors of kg of CH4 per head per year used for the manure
management emission estimates as well as the IPCC (2006) default emission factors. The U.S. implied
emission factors fall within the range of the IPCC (2006) default values, except in the case of sheep,
goats, and some years for horses and dairy cattle. The U.S. implied emission factors are less than the
IPCC (2006) default value for those animals due to the use of U.S.-specific data for typical animal mass
and VS excretion. There is an increase in implied emission factors for dairy cattle and swine across the
Agriculture 5-23
-------
time series. This increase reflects the dairy cattle and swine industry trend towards larger farm sizes;
large farms are more likely to manage manure as a liquid and therefore produce more CH4 emissions.
See the Recalculations for explanations for changes that affect emissions which impact these implied
emission factors.
Table 5-10: IPCC (2006) Implied Emission Factor Default Values Compared with
Calculated Values for CH4 from Manure Management (kg/head/year)
Animal Type
IPCC Default
Cm Emission Factors
(kg/head/year)a
1990
2005
2019
2020
2021
2022
2023
Dairy Cattle
48-112
29.3
53.0
65.0
65.9
65.0
64.1
64.1
Beef Cattle
1-2
0.8
0.9
1.8
1.8
1.9
1.9
1.9
Swine
10-45
11.5
13.3
11.6
11.5
11.8
11.6
11.6
Sheep
0.19-0.37
0.3
0.4
0.4
0.4
0.4
0.4
0.4
Goats
0.13-0.26
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Poultry
0.02-1.4
0.1
0.1
0.05
0.05
0.05
0.05
0.05
Horses
1.56-3.13
1.9
1.4
1.2
1.2
1.2
1.2
1.2
American Bison
NA
0.8
0.9
0.9
0.9
0.9
0.9
0.9
Mules and Asses
0.76-1.14
0.4
0.4
0.4
0.4
0.4
0.4
0.4
NA (Not Applicable)
a Ranges reflect 2006 IPCC Guidelines (Volume 4, Table 10.14) default emission factors for North America across different
climate zones.
Notes: ChU implied emission factors were not calculated for 2023 due to the simplified emissions estimation approach used to
estimate emissions for that year. 2022 values were used for 2023.
In addition, default IPCC (2006) emission factors for N20 were compared to the U.S. Inventory implied
N20 emission factors. Default N20 emission factors from the 2006 IPCC Guidelines were used to
estimate N20 emission from each WMS in conjunction with U.S.-specific Nex values. The implied
emission factors differed from the U.S. Inventory values due to the use of U.S.-specific Nex values and
differences in populations present in each WMS throughout the time series.
Recalculations Discussion
No time-series recalculations were performed. The 1990 to 2022 estimates were retained from the
previous Inventory (i.e., 2024 Inventory), and 2023 estimates were based on a simplified approach that
used emission factors and extrapolated population estimates for all animals (as discussed in the
Methodology approach section).
Planned Improvements
Regular annual data reviews and updates are necessary to maintain an emissions inventory that reflects
the current base of knowledge. In addition to the documented approaches currently used to address
data availability, EPA conducts data assessments to pursue a number of potential improvements.
EPA notes the follow improvements are likely to be implemented within the next two or three Inventory
cycles:
Ongoing improvement efforts for updating cattle diet data, as part of the Enteric Fermentation
category, are expected to impact manure management emissions estimates as well due to
5-24 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
changes to both VS excretion rates (affecting CH4) and Nex excretion rates (affecting N20). This
change results from the effect of individual dietary components (e.g., more or less fiber) on
digestion within the cattle's rumen. See Section 5.1 Planned Improvements (Enteric
Fermentation) for more details.
Potential improvements for future inventories. Many of these are major updates and considered long-
term improvements:
Investigating the updated IPCC 2079 Refinement default N20 emissions factor for anaerobic
digesters. Historically, EPA has not estimated N20 emissions from digesters as the default
guidance was no emissions. Incorporating AgSTAR data for N20 emissions, like CH4 emissions,
is a longer-term improvement for EPA.
Investigating updates to the current anaerobic digester MCFs based on IPCC (2019).
EPA is aware of the following potential updates or improvements but notes that implementation
will be based on available resources and data availability:
Updating the B0 data used in the Inventory, as data become available. EPA is conducting
outreach with counterparts from USDA as to available data and research on B0.
Comparing CH4 and N20 emission estimates with estimates from other models and more recent
studies and compare the results to the Inventory.
Comparing manure management emission estimates with on-farm measurement data to
identify opportunities for improved estimates.
Comparing VS and Nex data to literature data to identify opportunities for improved estimates.
Determining if there are revisions to the U.S.-specific method for calculating liquid systems for
MCFs based on updated guidance from the IPCC 2019 Refinement.
EPA previously began this investigation to determine the potential differences between the
methods.
As part of this review, EPA plans to investigate available data on manure temperature in
liquid systems. If these data exist, EPA would need to weigh the benefits of using those data
over ambient air, understanding that the U.S. currently has monthly ambient temperature
data available.
EPA would also continue to review the assumptions for VS carryover. The current
assumption is that anaerobic lagoons are emptied once a year in October. As the season
impacts the overall temperature, and thus emissions, EPA would like to confirm these
assumptions are still accurate for the U.S. livestock industry.
Investigating improved emissions estimate methodologies for swine pit systems with less than
one month of storage (the updated swine WMS data included this WMS category).
Improving the linkages with the Enteric Fermentation source category estimates. For future
Inventories, it may be beneficial to have the CEFM and Manure Management calculations in the
same model, as they rely on much of the same activity data and on each other's outputs to
properly calculate emissions. EPA has begun this investigation and initial development of a
combined approach to calculate emissions for these two categories.
Agriculture 5-25
-------
Continuing to investigate new sources of WMS data. EPA is collaborating with the USDAto
collect or use existing survey data for potential improvements to the Inventory.
Revising the uncertainty analysis to address changes that have been implemented to the CH4
and N20 estimates. Updates to the uncertainty analysis have begun and the plan is to align the
timing of the updated Manure Management uncertainty analysis with the uncertainty analysis
for Enteric Fermentation.
5.3 Rice Cultivation (Source Category 3C)
Most of the world's rice is grown on flooded fields (Baicich 2013) that create anaerobic conditions
leading to CH4 production through a process known as methanogenesis. Approximately 60 to 90 percent
of the CH4 produced by methanogenic bacteria in flooded rice fields is oxidized in the soil and converted
to C02 by methanotrophic bacteria. The remainder is emitted to the atmosphere (Holzapfel-Pschorn et
al. 1985; Sass et al. 1990) or transported as dissolved CH4 into groundwater and waterways (Neue et al.
1997). Methane is transported to the atmosphere primarily through the rice plants, but some CH4 also
escapes via ebullition (i.e., bubbling through the water) and to a much lesser extent by diffusion through
the water (van Bodegom et al. 2001).
Water management is arguably the most important factor affecting CH4 emissions in rice cultivation,
and improved water management has the largest potential to mitigate emissions (Yan et al. 2009).
Upland rice fields are not flooded, and therefore do not produce CH4, but large amounts of CH4 can be
emitted in continuously irrigated fields, which is the most common practice in the United States (USDA
2012). Single or multiple aeration events with drainage of a field during the growing season can
significantly reduce these emissions (Wassmann et al. 2000a), but drainage may also increase N20
emissions. Deepwater rice fields (i.e., fields with flooding depths greater than one meter, such as natural
wetlands) tend to have fewer living stems reaching the soil, thus reducing the amount of CH4 transport
to the atmosphere through the plant compared to shallow-flooded systems (Sass 2001).
Other management practices also influence CH4 emissions from flooded rice fields including rice
residue straw management and application of organic amendments, in addition to cultivar selection due
to differences in the amount of root exudates11 among rice varieties (Neue et al. 1997). These practices
influence the amount of organic matter available for methanogenesis, and some practices, such as
mulching rice straw or composting organic amendments, can reduce the amount of labile carbon and
limit CH4 emissions (Wassmann et al. 2000b). Fertilization practices also influence CH4 emissions,
particularly the use of fertilizers with sulfate, which can reduce CH4 emissions (Wassmann et al. 2000b;
Linquist et al. 2012). Other environmental variables also impact the methanogenesis process such as
soil temperature and soil type. Soil temperature regulates the activity of methanogenic bacteria, which
in turn affects the rate of CH4 production. Soil texture influences decomposition of soil organic matter
but is also thought to have an impact on oxidation of CH4 in the soil (Sass et al. 1994).
Rice is currently cultivated in 12 states, including Arkansas, California, Florida, Illinois, Kentucky,
Louisiana, Minnesota, Mississippi, Missouri, New York, Tennessee, and Texas. Soil types, rice varieties,
11 The roots of rice plants add organic material to the soil through a process called "root exudation." Root exudation is
thought to enhance decomposition of the soil organic matter and release nutrients that the plant can absorb for
production. The amount of root exudate produced by a rice plant over a growing season varies among rice varieties.
5-26 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
and cultivation practices vary across the United States, but most farmers apply fertilizers and do not
harvest crop residues. In addition, a second ratoon rice crop is sometimes grown in the Southeastern
region of the country. Ratoon crops are produced from regrowth of the stubble remaining after the
harvest of the first rice crop. Methane emissions from ratoon crops are higher than those from the
primary crops due to the increased amount of labile organic matter available for anaerobic
decomposition in the form of relatively fresh crop residue straw. Emissions tend to be higher in rice
fields if the residues have been in the field for less than 30 days before planting the next rice crop
(Lindau and Bollich 1993; IPCC 2006; Wang et al. 2013).
Overall, rice cultivation is a minor source of CH4 emissions in the United States relative to other source
categories (see Table 5-11, Table 5-12, and Figure 5-3). Most emissions occur in Arkansas, California,
Louisiana, Mississippi, Missouri, and Texas. In 2023, CH4 emissions from rice cultivation were 18.7 MMT
C02 Eq. (667 kt CH4). Annual emissions fluctuated between 1990 and 2023, which is largely due to
differences in the amount of rice harvested areas over time. There has been a marginal decrease in
emissions since 1990. Interestingly, the estimated emissions in 2023 are roughly the same as emissions
in 1990.
Table 5-11: CH4 Emissions from Rice Cultivation (MMT C02 Eq.)
Source 1990 I
2005
2019 2020
2021
2022
2023
Rice Cultivation 18.9
20.6 |
15.6 18.6
18.5
18.0
18.7
Table 5-12: CH4 Emissions from Rice Cultivation (kt CH4)
Source 1990
2005
2019 2020
2021
2022
2023
Rice Cultivation 677
735 |
558 664
661
642
667
Agriculture 5-27
-------
Figure 5-3: Annual CH4 Emissions from Rice Cultivation, 2020, Using the Tier 3
DayCent Model
Note: Only riational-scale emissions are estimated for 2021 to 2023 in this Inventory using a surrogate data method described in
the Methodology section; therefore, the fine-scale emission patterns in this map are based on the estimates for 2020.
Methodology and Time-Series Consistency
The methodology used to estimate CH4 emissions from rice cultivation is based on a combination of
IPCC Tier 1 and 3 approaches. A combination of Tier 1 and 3 methods are used across most of the time
series, while a surrogate data method has been applied to estimate national emissions for 2021 to 2023
in this Inventory due to lack of data in these years of the time series,
The Tier 3 method utilizes the DayCent process-based model to estimate CH4 emissions from rice
cultivation (Cheng et al. 2013) and has been tested in the United States (see Annex 3.13) and Asia
(Cheng et al. 2013, 2014). The model simulates hydrological conditions and thermal regimes, organic
matter decomposition, root exudation, rice plant growth and its influence on oxidation of CH4, as well as
CH4 transport through the plant and via ebullition (Cheng et al. 2013). The method captures the
influence of organic amendments and rice straw management on methanogenesis in the flooded soils,
and ratooning of rice crops with a second harvest during the growing season. In addition to CH4
emissions, DayCent simulates soil carbon stock changes and N2Q emissions (Parton et al. 1987 and
1998; Del Grosso et al. 2010) and allows for a seamless set of simulations for crop rotations that include
both rice and non-rice crops.
The Tier 1 method is applied to estimate CH4 emissions from rice when grown in rotation with crops that
are not simulated by DayCent, such as some vegetable crops (see Annex 3.13 for additional details on
DayCent). The Tier 1 method is also used for areas converted between agriculture (i.e., cropland and
5-28 inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
grassland) and other land uses, such as forest land, wetland, and settlements. In addition, the Tier 1
method is used to estimate CH4 emissions from organic soils (i.e., Histosols) and from areas with very
gravelly, cobbly, or shaley soils (greater than 35 percent by volume). The Tier 3 method using DayCent
has not been fully tested for estimating emissions associated with these conditions.
The Tier 1 method for estimating CH4 emissions from rice production utilizes a default base emission
rate and scaling factors (IPCC 2006). The base emission rate represents emissions for continuously
flooded fields with no organic amendments. Scaling factors are used to adjust the base emission rate
for water management and organic amendments that differ from continuous flooding with no organic
amendments. The method accounts for pre-season and growing season flooding; types and amounts of
organic amendments; and the number of rice production seasons within a single year (i.e., single
cropping and double-cropping with ratooning). The Tier 1 analysis is implemented in the Agriculture and
Land Use National Greenhouse Gas Inventory (ALU) software (Ogle et al. 2016).12
Rice cultivation areas are based on crop and land use histories recorded in the USDA National
Resources Inventory (NRI) survey (USDA-NRCS 2020) and extended through 2020 using the USDA-NASS
Crop Data Layer Product (USDA-NASS 2021, Johnson and Mueller 2010). The areas have been modified
in the original NRI survey through a process in which the Forest Inventory and Analysis (FIA) survey data
and the National Land Cover Dataset (Yang et al. 2018) are harmonized with the NRI data (Nelson et al.
2020). This process ensures that the land use areas are consistent across all land use categories (See
Section 6.1, Representation of the U.S. Land Base for more information).
The NRI is a statistically based sample of all non-federal land and includes approximately 604,000
survey locations in agricultural cropland and grassland for the conterminous United States and Hawaii
of which 7,888 include one or more years of rice cultivation. The Tier 3 method is used to estimate CH4
emissions from 5,998 of the NRI survey locations, and the remaining 1,890 survey locations are
estimated with the Tier 1 method. Each NRI survey location is associated with a survey weight that
allows scaling of CH4 emission to the entire land base with rice cultivation (i.e., each weight
approximates the amount of area with the same land-use/management history as the survey location).
Land-use and some management information in the NRI (e.g., crop type, soil attributes, and irrigation)
were collected on a 5-year cycle beginning in 1982, along with cropping rotation data in four out of five
years for each five-year time period (i.e., 1979 to 1982, 1984 to 1987, 1989 to 1992, and 1994 to 1997).
The NRI program began collecting annual data in 1998, with data through 2017 (USDA-NRCS 2020). For
2018 through 2020, the time series is extended with the crop data provided in USDA-NASS CDL (USDA-
NASS 2021). CDL data have a 30 to 58 m spatial resolution, depending on the year. NRI survey locations
are overlaid on the CDL in a geographic information system, and the crop types are extracted to extend
the cropping histories. The harvested rice areas in each state are presented in Table 5-13.
12 See http://www.nrel.colostate.edu/projects/AI Usoftware/.
Agriculture 5-29
-------
Table 5-13: Rice Area Harvested (1,000 Hectares)
State/Crop
1990
2005
2019
2020
2021
2022
2023
Arkansas
611
782
512
663
NE
NE
NE
California
251
237
218
224
NE
NE
NE
Florida
0
3
0
0
NE
NE
NE
Illinois
0
1
0
1
NE
NE
NE
Kentucky
0
0
0
0
NE
NE
NE
Louisiana
399
400
313
383
NE
NE
NE
Minnesota
3
6
3
0
NE
NE
NE
Mississippi
177
191
96
109
NE
NE
NE
Missouri
48
96
74
85
NE
NE
NE
New York
1
0
0
0
NE
NE
NE
Tennessee
0
1
0
0
NE
NE
NE
Texas
294
104
119
167
NE
NE
NE
Total
1,784
1,823
1,335
1,633
NE
NE
NE
NE (Not Estimated). See Methodology section for more information on how the historical time-series activity data inform the
latest emissions estimates. Area data will be updated in the next Inventory.
Note: Totals may not sum due to independent rounding.
The Southeastern states have sufficient growing periods for a ratoon crop in some years (Table 5-14). For
example, the growing season length is occasionally sufficient for ratoon crops to be grown on about two
percent of the rice fields in Arkansas. No data are available about ratoon crops in Missouri or
Mississippi, so the average amount of ratooning in Arkansas was assigned to these states. Ratoon
cropping occurs much more frequently in Louisiana (LSU 2015, for years 2000 through 2013, 2015) and
Texas (TAMU 2015, for years 1993 through 2015), averaging 32 percent and 45 percent of rice acres
planted, respectively. Florida also has a large fraction of area with a ratoon crop (49 percent). Ratoon
rice crops are not grown in California. Ratooning practices are assigned to individual NRI locations using
a hot-deck imputation method with six complete imputations for each NRI location to address
uncertainty. The method is based on random assignment of ratooning to approximate the percentages of
fields managed with ratooning provided in Table 5-14.
Table 5-14: Average Ratooned Area as Percent of Primary Growth Area (Percent)
State 1990-2020
Arkansas® 1.9%
California 0%
Floridab 45.2%
Louisiana0 39.5%
Mississippi® 37.8%
Missouri® 2.4%
Texasd 49.5%
a Arkansas: 1990-2000 (Slaton 1999 through 2001); 2001-2011 (Wilson 2002 through 2007, 2009 through 2012); 2012-2013
(Hardke 2013, 2014). Estimates of ratooningfor Missouri and Mississippi are based on the data from Arkansas.
b Florida - Ratoon: 1990-2000 (Schueneman 1997,1999 through 2001); 2001 (Deren 2002); 2002-2003 (Kirstein 2003 through
2004, 2006); 2004 (Cantens 2004 through 2005); 2005-2013 (Gonzalez 2007 through 2014).
c Louisiana: 1990-2013 (Linscombe 1999, 2001 through 2014).
d Texas: 1990-2002 (Klosterboer 1997,1999 through 2003); 2003-2004 (Stansel 2004 through 2005); 2005 (Texas Agricultural
Experiment Station 2006); 2006-2013 (Texas Agricultural Experiment Station 2007 through 2014).
5-30 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
While rice crop production in the United States includes a minor amount of land with mid-season
drainage or alternate wet-dry periods, the majority of rice growers use continuously flooded water
management systems (Hardke 2015; UCCE 2015; Hollier 1999; Way et al. 2014). Therefore, continuous
flooding was assumed in the DayCent simulations and the Tier 1 analysis. Variation in flooding can be
incorporated in future inventories if updated water management data are available.
Winter flooding is another key practice associated with water management in rice fields, and the impact
of winter flooding on CH4 emissions is addressed in the Tier 3 and Tier 1 analyses. Flooding is used to
prepare fields for the next growing season, and to create waterfowl habitat (Young 2013; Miller et al.
2010; Fleskes et al. 2005). Fitzgerald et al. (2000) suggests that as much as 50 percent of the annual
emissions may occur during winter flooding. Winter flooding is a common practice with an average of 34
percent of fields managed with winter flooding in California (Miller et al. 2010; Fleskes et al. 2005), and
approximately 21 percent of the fields managed with winter flooding in Arkansas (Wilson and Branson
2005 and 2006; Wilson and Runsick 2007 and 2008; Wilson et al. 2009 and 2010; Hardke and Wilson
2013 and 2014; Hardke 2015). No data are available on winter flooding for Texas, Louisiana, Florida,
Missouri, or Mississippi. For these states, the average amount of flooding is assumed to be similar to
Arkansas. In addition, the amount of flooding is assumed to be relatively constant over the Inventory
time series. Similar to ratooning practices, winter flooding is assigned to individual NRI locations using a
hot-deck imputation method with six complete imputations for each NRI location to address
uncertainty. The method is based on random assignment of winter flooding to approximate the
percentages of fields managed with winter flooding as discussed above.
A data splicing method is used to estimate emissions from 2021 to 2023 associated with the rice CH4
emissions for Tier 1 and 3 methods. Specifically, a linear regression model with autoregressive moving
average (ARMA) errors was used to estimate the relationship between the surrogate data and emissions
data from 1990 through 2020, which were derived using the Tier 3 methods (Brockwell and Davis 2016).
Surrogate data are based on rice commodity statistics from USDA-NASS.13 See Box 5-1 for more
information about the surrogate data method. For the Tier 1 method, a linear-time series model is used
to estimate emissions for 2021 to 2023 without surrogate data.
Box 5-1: Surrogate Data Method
An approach to extend the time series is needed to estimate emissions from rice cultivation because
there are gaps in activity data at the end of the time series. This is mainly because the National
Resources Inventory (NRI) does not release data every year, and the NRI is a key data source for
estimating greenhouse gas emissions.
A surrogate data method has been selected to impute missing emissions at the end of the time series. A
linear regression model with autoregressive moving average (ARMA) errors (Brockwell and Davis 2016) is
used to estimate the relationship between the surrogate data and the observed 1990 to 2020 emissions
data that has been compiled using the inventory methods described in this section. The model to extend
the time series is given by
Y = xp + £,
where Y is the response variable (e.g., CH4 emissions), X(B is the surrogate data that is used to predict
the missing emissions data, and s is the remaining unexplained error. Models with a variety of surrogate
13 See https://qijickstats.nass.usda.gov/.
Agriculture 5-31
-------
data were tested, including commodity statistics, weather data, or other relevant information.
Parameters are estimated from the observed data for 1990 to 2020 using standard statistical
techniques, and these estimates are used to predict the missing emissions data for 2021 to 2023.
A critical issue in using splicing methods is to adequately account for the additional uncertainty
introduced by predicting emissions with related information without compiling the full inventory. For
example, predicting CH4 emissions will increase the total variation in the emission estimates for these
specific years, compared to those years in which the full inventory is compiled. This added uncertainty
is quantified within the model framework using a Monte Carlo approach. The approach requires
estimating parameters for results in each Monte Carlo simulation for the full inventory (i.e., the surrogate
data model is refit with the emissions estimated in each Monte Carlo iteration from the full inventory
analysis with data from 1990 to 2020).
In order to ensure time-series consistency, the same methods are applied from 1990 to 2020, and data
splicing methods are used to approximate emissions for the remainder of the 2021 to 2023 time series
based on the emissions data from 1990 to 2020. The data splicing methods used in this Inventory are
consistent with the approaches described in IPCC (2006).
Uncertainty
Sources of uncertainty in the Tier 3 method include management practices, uncertainties in model
structure (i.e., algorithms and parameterization), and variance associated with the NRI sample. Sources
of uncertainty in the IPCC (2006) Tier 1 method include the emission factors, management practices,
and variance associated with the NRI sample. The total uncertainty was quantified with two variance
components (Ogle et al. 2010) that are combined using simple error propagation methods provided by
the IPCC (2006), i.e., by taking the square root of the sum of the squares of the standard deviations of
the uncertain quantities. For the first variance component, a Monte Carlo analysis was used to
propagate uncertainties in the Tier 1 and 3 methods for the management data, as well as emission
factors and model structure/parameterization, respectively. The second variance component is
quantifying uncertainty in scaling from the NRI survey to the entire area of rice cultivation, and is
computed using a standard variance estimator for a two-stage sample design (Sarndal et al. 1992). For
2021 to 2023, there is additional uncertainty propagated through the Monte Carlo analysis associated
with the surrogate data method (see Box 5-1 for information about propagating uncertainty with the
surrogate data method). The uncertainties from the Tier 1 and 3 approaches are combined to produce
the final CH4 emissions estimate using simple error propagation (IPCC 2006). Additional details on the
uncertainty methods are provided in Annex 3.13.
Rice cultivation CH4 emissions in 2023 were estimated to be between 4.7 and 32.6 MMT C02 Eq. at a 95
percent confidence level, which indicates a range of 75 percent below to 75 percent above the 2023
emission estimate of 18.7 MMT C02 Eq. (see Table 5-15).
5-32 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Table 5-15: Approach 2 Quantitative Uncertainty Estimates for CH4 Emissions from
Rice Cultivation (MMT C02 Eq. and Percent)
Uncertainty Range Relative to Emission Estimate"
2023 Emission
(MMT CO2 Eq.)
(%)
Inventory
Estimate
Lower
Upper
Lower
Upper
Source
Method
Gas
(MMT C02 Eq.)
Bound
Bound
Bound
Bound
Rice Cultivation
Tier 3
CH4
15.7
1.8
29.6
-89%
+89%
Rice Cultivation
Tier 1
ch4
3.0
2.0
4.1
-35%
+35%
Rice Cultivation
Total
ch4
18.7
4.7
32.6
-75%
+75%
a Range of emission estimates is the 95 percent confidence interval.
QA/QC and Verification
General (Tier 1) and category-specific (Tier 2) QA/QC activities were conducted consistent with the U.S.
Inventory QA/QC plan outlined in Annex 8. Quality control measures include checking input data, model
scripts, and results to ensure data are properly handled throughout the inventory process. Inventory
forms and text are reviewed and revised as needed to correct transcription errors.
Model results are compared to field measurements to verify that results adequately represent CH4
emissions. The comparisons included over 17 long-term experiments, representing about 238
combinations of management treatments across all the sites. A statistical relationship was developed
to assess uncertainties in the model structure and parameterization, adjusting the estimates for model
bias and assessing precision in the resulting estimates (methods are described in Ogle et al. 2007). See
Annex 3.13 for more information.
Recalculations Discussion
A recalculation was completed for the data splicing values that were estimated for 2021 and 2022. The
correction was needed because the rice yield and area covariates used in the previous Inventory were
not applied to the correct years (i.e., 2021 covariates were used to predict 2022 emissions, and vice
versa). As a result of these changes, C02-equivalent emissions decreased by 0.3 MMT C02 Eq., or 1.8
percent, from 2021 to 2022. The emissions estimates for the remainder of the time series are the same
as the previous Inventory.
Planned Improvements
The following are planned improvements for Rice Cultivation:
Refining the model algorithms and re-calibration of the Tier 3 DayCent model using the latest
observational data from experiments. This is a key improvement for Rice Cultivation.
Collecting more information about water management and refinement of the application to
incorporate mid-season drainage and alternate wetting and drying systems.
The earliest these improvements could be completed would be the next Inventory (i.e., 2026 publication,
1990 through 2024 Inventory), pending prioritization of resources.
Agriculture 5-33
-------
5.4 Agricultural Soil Management (Source
Category 3D)
Nitrous oxide is naturally produced in soils through the microbial processes of nitrification and
denitrification that is driven by the availability of mineral nitrogen (N) (Firestone and Davidson 1989).14
Mineral nitrogen is made available in soils through decomposition of soil organic matter and plant litter,
as well as asymbiotic fixation of nitrogen from the atmosphere.15 Several agricultural activities increase
mineral nitrogen availability in soils that lead to direct N20 emissions at the site of a management
activity (see Figure 5-4) (Mosier et al. 1998). These activities include synthetic nitrogen fertilization;
application of managed livestock manure; application of other organic materials such as biosolids (i.e.,
treated sewage sludge); deposition of manure on soils by domesticated animals in pastures, range, and
paddocks (PRP) (i.e., unmanaged manure); retention of crop residues (nitrogen-fixing legumes and non-
legume crops and forages); and drainage of organic soils16 (i.e., Histosols) (IPCC 2006). Additionally,
agricultural soil management activities, including irrigation, drainage, tillage practices, cover crops, and
fallowing of land, can influence nitrogen mineralization from soil organic matter and levels of asymbiotic
nitrogen fixation. Indirect emissions of N20 occur when nitrogen is transported from a site and is
subsequently converted to N20; there are two pathways for indirect emissions: (1) volatilization and
subsequent atmospheric deposition of applied/mineralized nitrogen, and (2) surface runoff and
leaching of applied/mineralized nitrogen into groundwater and surface water.17 Direct and indirect
emissions from agricultural lands are included in this section (i.e., cropland and grassland as defined in
Section 6.1). Nitrous oxide emissions from forest land and settlements soils are found in Sections 6.2
and 6.10, respectively.
14 Nitrification and denitrification are driven by the activity of microorganisms in soils. Nitrification is the aerobic microbial
oxidation of ammonium (NH4+) to nitrate (NO3-), and denitrification is the anaerobic microbial reduction of nitrate to N2.
Nitrous oxide is a gaseous intermediate product in the reaction sequence of nitrification and denitrification.
15 Asymbiotic nitrogen fixation is the fixation of atmospheric N2 by bacteria living in soils that do not have a direct
relationship with plants.
16 Drainage of organic soils in former wetlands enhances mineralization of nitrogen-rich organic matter, thereby increasing
N2O emissions from these soils.
17 These processes entail volatilization of applied or mineralized nitrogen as NH3 and NOx, transformation of these gases in
the atmosphere (or upon deposition), and deposition of the nitrogen primarily in the form of particulate NH4+, nitric acid
(HNO3), and NOx. In addition, hydrological processes lead to leaching and runoff of NC>3~ that is converted to N2O in
aquatic systems, e.g., wetlands, rivers, streams and lakes. Note: N2O emissions are not estimated for aquatic systems
associated with nitrogen inputs from terrestrial systems in order to avoid double-counting.
5-34 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Figure 5-4: Sources and Pathways of Nitrogen that Result in N20 Emissions from
Agricultural Soil Management
Sources and Pathways of N that Result in N2O Emissions from Agricultural Soil Management
N Volatilization
N Flows:
j \ N Inputs to
I / Managed Soils
Direct NjO
Emissions
N Volatilization
"" * and Deposition
I \ Indirect N;0
¦' Emissions
A symbiotic Fixation
Fixation of atm ospheric Nj by bacteria
Irving in soils that do not have a dii«d
relationship with plants
This graphic illustrates the sources and pathways of nitrogen that result
in direct and indirect N,0 emissions from soils using the methodologies
described in this Inventory. Emission pathways are shown with arrows.
On the lower right-hand side is a cut-away view of a representative
section of a managed soil; histosol cultivation is represented here.
Groundwater
Histosol
Cultivation
Crop Residues
Includes above- and belowground
residues for all crops (non-N andN-
twing (and from perennial forage
crops and pastires following renewal I
Mineializalion ol
Soil Organic Mattel
I fcnum
SyntheticN Fertilizers
Synthetic N fertilizer applied to Soil
Oiganic
Amendments
Includes both commeoal and
norvco/nmercist fertilizers Ge»
animal manure compost
sewage sludge ttnkag^tc)
Uiineand Dung Item
Glazing Animals
Manore deposited on pasture '®ige:
and paddock
Includes N convened to mineral form
upon decomposition of soil organic
matter
Agricultural soils produce the majority of N20 emissions in the United States. Estimated emissions in
2023 are 296.3 MMT C02 Eq. (1,118 kt) (see Table 5-16 and Table 5-17). Annual N20 emissions from
Agriculture 5-35
-------
agricultural soils are 2.5 percent greater in 2023 compared to 1990, but emissions fluctuated between
1990 and 2023 due to inter-annual variability largely associated with weather patterns, synthetic
fertilizer use, and crop production. From 1990 to 2023, cropland accounted for 68 percent of total direct
emissions on average from agricultural soil management, while grassland accounted for 32 percent. On
average, 79 percent of indirect emissions are from croplands and 21 percent from grasslands. Estimated
direct and indirect N20 emissions by sub-source category are shown in Table 5-18 and Table 5-19.
Table 5-16: N20 Emissions from Agricultural Soils (MMT C02 Eq.)
Activity
1990
2005
2019
2020
2021
2022
2023
Direct
259.2
266.2
281.6
263.5
268.5
263.3
266.8
Cropland
174.9
180.7
193.5
182.5
184.4
180.4
181.6
Grassland
84.3
85.5
88.1
81.0
84.1
82.9
85.1
Indirect
29.9
28.5
34.8
29.5
30.4
28.5
29.6
Cropland
23.6
22.4
28.0
23.3
24.2
22.3
23.1
Grassland
6.4
6.2
6.8
6.2
6.3
6.2
6.5
Total
289.1
294.7
316.4
293.0
298.9
291.8
296.3
Notes: Estimates for 2021 to 2023 are based on a data splicing method, except for other organic nitrogen amendments that are
based on a data splicing method for 2018 to 2023 and drainage of organic soils which was extended for 2023 (See Methodology
section). Totals may not sum due to independent rounding.
Table 5-17: N20 Emissions from Agricultural Soils (kt N20)
Activity
1990
2005
2019
2020
2021
2022
2023
Direct
978
1,005
1,063
994
1,013
994
1,007
Cropland
660
682
730
689
696
681
685
Grassland
318
323
332
306
318
313
321
Indirect
113
108
131
111
115
107
112
Cropland
89
84
106
88
91
84
87
Grassland
24
23
26
23
24
23
24
Total
1,091
1,112
1,194
1,106
1,128
1,101
1,118
Notes: Estimates for 2021 to 2023 are based on a data splicing method, except for other organic nitrogen amendments that are
based on a data splicing method for 2018 to 2023 and drainage of organic soils which was extended for 2023 (See Methodology
section). Totals may not sum due to independent rounding.
Table 5-18: Direct N20 Emissions from Agricultural Soils by Land Use Type and
Nitrogen Input Type (MMT C02 Eq.)
Activity
1990
2005
2019
2020
2021
2022
2023
Cropland
174.9
180.7
193.5
182.5
184.4
180.4
181.6
Mineral Soils
171.4
177.4
190.6
179.6
181.4
177.5
178.7
Synthetic Fertilizer
61.0
64.3
65.7
63.2
63.4
62.0
62.4
Organic Amendment0
11.5
12.8
14.6
14.4
14.8
14.6
14.7
Residue Nb
34.0
35.0
34.6
37.6
33.2
32.4
32.7
Mineralization and Asymbiotic Fixation
64.8
65.3
75.7
64.3
70.1
68.4
69.0
Drained Organic Soils
3.5
3.3
3.0
2.9
2.9
2.9
2.9
Grassland
84.3
85.5
88.1
81.0
84.1
82.9
85.1
Mineral Soils
81.7
83.0
85.6
78.5
81.6
80.3
82.6
Synthetic Fertilizer
+
+
+
+
+
+
+
5-36 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Activity
1990
2005
2019
2020
2021
2022
2023
PRP Manure
15.5
14.4
13.9
13.7
14.4
14.2
14.5
Managed Manure0
+
+ 1
+
+
+
+
+
Biosolids (i.e., treated Sewage Sludge)
0.2
0.4
0.4
0.4
0.4
0.4
0.4
Residue Nd
27.1
28.4
28.3
28.2
26.3
25.9
26.7
Mineralization and Asymbiotic Fixation
38.9
39.8
42.9
36.2
40.5
39.8
41.0
Drained Organic Soils
2.6
2.5 |
2.5
2.5
2.6
2.6
2.6
Total
259.2
266.2
281.6
263.5
268.5
263.3
266.8
+ Does not exceed 0.05 MMT C02 Eq.
a Organic amendment inputs include managed manure, daily spread manure, and commercial organic fertilizers (i.e., dried
blood, dried manure, tankage, compost, and other).
b Cropland residue nitrogen inputs include nitrogen in unharvested cover crops as well as harvested crops.
c Managed manure inputs include managed manure and daily spread manure amendments that are applied to grassland soils.
d Grassland residue nitrogen inputs include residual biomass, both legumes and grasses, that is ungrazed and becomes dead
organic matter.
Notes: Estimates for 2021 to 2023 are based on a data splicing method, except for other organic nitrogen amendments that are
based on a data splicing method for 2018 to 2023 and drainage of organic soils which was extended for 2023 (see Methodology
section). Totals may not sum due to independent rounding.
Table 5-19: Indirect N20 Emissions from Agricultural Soils (MMT C02 Eq.)
Activity
1990
2005
2019
2020
2021
2022
2023
Cropland
23.6
22.4
28.0
23.3
24.2
22.3
23.1
Volatilization &Atm. Deposition
6.6
7.0
7.1
7.5
7.4
7.4
7.4
Surface Leaching & Run-Off
17.0
15.4
20.9
15.8
16.8
14.9
15.8
Grassland
6.4
6.2
6.8
6.2
6.3
6.2
6.5
Volatilization &Atm. Deposition
3.4
3.4
3.2
3.1
3.2
3.2
3.3
Surface Leaching & Run-Off
2.9
2.7
3.6
3.1
3.1
3.0
3.2
Total
29.9
28.5
34.8
29.5
30.4
28.5
29.6
Notes: Estimates for 2021 to 2023 are based on a data splicing method, except for other organic nitrogen amendments that are
based on a data splicing method for 2018 to 2023 and drainage of organic soils which was extended for 2023 (See Methodology
section). Totals may not sum due to independent rounding.
Figure 5-5 and Figure 5-6 show regional patterns for direct N20 emissions. Figure 5-7 and Figure 5-8
show indirect N20 emissions from volatilization, and Figure 5-9 and Figure 5-10 show the indirect N20
emissions from leaching and runoff in croplands and grasslands, respectively.
Direct N20 emissions from croplands occur throughout all of the cropland regions but tend to be high in
the Midwestern Corn Belt Region (particularly, Illinois, Iowa, Kansas, Minnesota, Nebraska), where a
large portion of the land is used for growing highly fertilized corn and nitrogen-fixing soybean crops (see
Figure 5-5). There are high emissions from the Southeastern region, and portions of the Great Plains.
Emissions are also high in the Lower Mississippi River Basin from Missouri to Louisiana, and highly
productive irrigated areas, such as Platte River, which flows from Colorado and Wyoming through
Nebraska, Snake River Valley in Idaho, and the Central Valley in California. Direct emissions from
croplands are low in mountainous regions of the Eastern United States because only a small portion of
land is cultivated, and in much of the Western United States where rainfall and access to irrigation water
are limited, in addition to mountainous, which are generally not suitable for crop production.
Direct N20 emissions from grasslands are more evenly distributed throughout the United States
compared to emissions from cropland due to suitable areas for grazing in most regions (see Figure 5-6).
Total emissions tend be highest in the Great Plains and western United States where a large proportion
Agriculture 5-37
-------
of the land is dominated by grasslands with cattle and sheep grazing (particularly Kansas, Montana,
Nebraska, New Mexico, Oklahoma, South Dakota, Texas).
Figure 5-5: Croplands, 2020 Annual Direct N20 Emissions Estimated Using the Tier 3
DayCent Model
Note: Only national-scale emissions are estimated for 2023 using a splicing method, and therefore the fine-scale emission
patterns in this map are based on Inventory data from 2020.
5-38 inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Figure 5-6: Grasslands, 2020 Annual Direct N20 Emissions Estimated Using the Tier 3
DayCent Model
Note: Only national-scale emissions are estimated for 2023 using a splicing method, and therefore the fine-scale emission
patterns in this map are based on Inventory data from 2020.
Indirect N20 emissions from volatilization in croplands have a similar pattern as the direct N20
emissions with higher emissions in the Midwestern Corn Belt, Lower Mississippi River Basin,
Southeastern region, and parts of the Great Plains and irrigated areas of the Western United States.
Indirect N20 emissions from volatilization in grasslands are higher in the Eastern and Central United
States, along with relatively small areas scattered around the Western United States. The higher
emissions are partly due to large additions of PRP manure nitrogen, which in turn, stimulates NH3
volatilization.
Indirect N2C emissions from surface runoff and leaching of applied/mineralized nitrogen in croplands is
highest in the Midwestern Corn Belt. There are also relatively high emissions associated with nitrogen
management in the Lower Mississippi River Basin, Piedmont region of the Southeastern United States
and the Mid-Atlantic states. In addition, areas of high emissions occur in portions of the Great Plains
that have irrigated croplands with high leaching rates of applied/mineralized nitrogen. Indirect N2C
emissions from surface runoff and leaching of applied/mineralized nitrogen in grasslands are higher in
the eastern United States and coastal Northwest region. These regions have greater precipitation and
higher levels of leaching and runoff compared to arid to semi-arid regions in the Western United States.
Agriculture 5-39
-------
Figure 5-7: Croplands, 2020 Annual Indirect N20 Emissions from Volatilization Using
the Tier 3 DayCent Model
Note: Only national-scale emissions are estimated for 2023 using a splicing method, and therefore the fine-scale emission
patterns in this map are based on Inventory data from 2020.
Figure 5-8: Grasslands, 2020 Annual Indirect N20 Emissions from Volatilization Using
the Tier 3 DayCent Model
Note: Only national-scale emissions are estimated for 2023 using a splicing method, and therefore the fine-scale emission
patterns in this map are based on Inventory data from 2020.
5-40 inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Figure 5-9: Croplands, 2020 Annual Indirect N20 Emissions from Leaching and Runoff
Using the Tier 3 DayCent Model
Note: Only national-scale emissions are estimated for 2023 using a splicing method, and therefore the fine-scale emission
patterns in this map are based on Inventory data from 2020.
Note: Only national-scale emissions are estimated for 2023 using a splicing method, and therefore the fine-scale emission
patterns in this map are based on Inventory data from 2020.
Figure 5-10: Grasslands, 2020 Annual indirect N20 Emissions from Leaching and
Runoff Using the Tier 3 DayCent Model
Agriculture 5-41
-------
Methodology and Time-Series Consistency
The 2006IPCC Guidelines (IPCC 2006) divide emissions from the agricultural soil management source
category into five components, including (1) direct emissions from nitrogen additions to cropland and
grassland mineral soils from synthetic fertilizers, biosolids (i.e., treated sewage sludge), crop residues
(legume nitrogen-fixing and non-legume crops), and organic amendments; (2) direct emissions from soil
organic matter mineralization due to land use and management change; (3) direct emissions from
drainage of organic soils in croplands and grasslands; (4) direct emissions from soils due to manure
deposited by livestock on PRP grasslands; and (5) indirect emissions from soils and water from nitrogen
additions and manure deposition to soils that lead to volatilization, leaching, or runoff of nitrogen and
subsequent conversion to N20.
In this source category, the United States reports on all croplands, as well as all managed grasslands,
whereby anthropogenic greenhouse gas emissions are estimated in a manner consistent with the
managed land concept (IPCC 2006), including direct and indirect N20 emissions from asymbiotic
fixation18 and mineralization of nitrogen associated with decomposition of soil organic matter and
residues. One recommendation from IPCC (2006) that has not been completely adopted is the
estimation of emissions from grassland pasture renewal, which involves occasional plowing to improve
forage production in pastures. Currently no data are available to address pasture renewal.
In addition, estimates of N20 emissions from managed croplands and grasslands are not available for
Alaska and Hawaii except for managed manure and PRP nitrogen, and biosolid additions for Alaska, and
managed manure and PRP nitrogen, biosolids additions, and crop residue for Hawaii. There is a planned
improvement to include the additional sources of emissions in a future Inventory.
Direct N2O Emissions
The methodology used to estimate direct N20 emissions from agricultural soil management in the
United States is based on a combination of IPCC Tier 1 and 3 approaches, along with application of a
splicing method for latter years in the Inventory time series (IPCC 2006; Del Grosso et al. 2010). A Tier 3
process-based model (DayCent) is used to estimate direct emissions from a variety of crops that are
grown on mineral (i.e., non-organic) soils, as well as the direct emissions from non-federal grasslands
except for applications of biosolids (i.e., treated sewage sludge) (Del Grosso et al. 2010). The Tier 3
approach has been specifically designed and tested to estimate N20 emissions in the United States,
accountingfor more of the environmental and management influences on soil N20 emissions than the
IPCC Tier 1 method (see Box 5-2 for further elaboration). Moreover, the Tier 3 approach addresses direct
N20 emissions and soil carbon stock changes from mineral cropland soils in a single analysis. Carbon
and nitrogen dynamics are linked in plant-soil systems through biogeochemical processes of microbial
decomposition and plant production (McGill and Cole 1981). Coupling the two source categories (i.e.,
agricultural soil carbon and N20) in a single inventory analysis ensures that there is consistent activity
data and treatment of the processes, and interactions are considered between carbon and nitrogen
cycling in soils.
18 Nitrogen inputs from asymbiotic nitrogen fixation are not directly addressed in 2006 IPCC Guidelines but are a
component of the nitrogen inputs and total emissions from managed lands and are included in the Tier 3 approach
developed for this source.
5-42 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Crop and land use histories are based on the USDA National Resources Inventory (NRI) (USDA-NRCS
2020) and extended through 2020 using the USDA-NASS Crop Data Layer Product (USDA-NASS 2021;
Johnson and Mueller 2010). The areas have been modified in the original NRI survey through a process in
which the Forest Inventory and Analysis (FIA) survey data and the National Land Cover Dataset (Yang et
al. 2018) are harmonized with the NRI data (Nelson et al. 2020). This process ensures that the land use
areas are consistent across all land use categories (see Section 6.1).
The NRI is a statistically-based sample and includes 364,333 survey locations on agricultural land for
the conterminous United States that are included in the Tier 3 method. The Tier 1 approach is used to
estimate the emissions from an annual average of 239,757 locations in the NRI survey across the time
series, which are designated as cropland or grassland (discussed later in this section). The Tier 1
method is used to estimate emissions for components that are not simulated by DayCent. DayCent has
not been parametrized to simulate some crop types and soil types, as described below. Each survey
location is associated with a survey weight that allows scaling of N20 emissions from NRI survey
locations to the entire country (i.e., each survey weight is an approximation of the amount of area with
the same land-use/management history as the survey location). Each NRI survey location was sampled
on a 5-year cycle from 1982 until 1997. For cropland, data were collected in 4 out of 5 years in the cycle
(i.e., 1979 through 1982,1984 through 1987,1989 through 1992, and 1994 through 1997). In 1998, the
NRI program began collecting annual data, which are currently available through 2017 (USDA-NRCS
2020). For 2018-2020, the time series is extended with the crop data provided in USDA-NASS CDL
(USDA-NASS 2021). CDL data have a 30 to 58 m spatial resolution, depending on the year. Specifically,
NRI survey locations are overlaid on the CDL in a geographic information system, and the crop types are
extracted to extend the cropping histories for the inventory analysis.
Box 5-2: Tier 1 vs. Tier 3 Approach for Estimating N20 Emissions
The IPCC (2006) Tier 1 approach is based on multiplying activity data on different nitrogen inputs (i.e.,
synthetic fertilizer, manure, nitrogen fixation, etc.) by the appropriate default IPCC emission factors to
estimate N20 emissions on an input-by-input basis. The Tier 1 approach requires a minimal amount of
activity data, readily available in most countries (e.g., total nitrogen applied to crops); calculations are
simple; and the methodology is highly transparent. In contrast, the Tier 3 approach developed for this
Inventoryis based on application of a process-based model (i.e., DayCent) that represents the
interaction of nitrogen inputs, land use and management, as well as environmental conditions at
specific locations, such as freeze-thaw effects that generate pulses of N20 emissions (Wagner-Riddle et
al. 2017; Del Grosso et al. 2022). Consequently, the Tier 3 approach accounts for land-use and
management impacts and their interaction with environmental factors, such as weather patterns and
soil characteristics, in a more comprehensive manner, which will enhance or dampen anthropogenic
influences. However, the Tier 3 approach requires more detailed activity data (e.g., crop-specific
nitrogen fertilization rates), additional data inputs (e.g., daily weather, soil types), and considerable
computational resources and programming expertise. The Tier 3 methodology is less transparent, and
thus it is critical to evaluate the output of Tier 3 methods against measured data in order to demonstrate
that the method is an improvement over lower tier methods for estimating emissions (IPCC 2006).
Another important difference between the Tier 1 and Tier 3 approaches relates to assumptions regarding
nitrogen cycling. Tier 1 assumes that nitrogen added to a system is subject to N20 emissions only during
that year and cannot be stored in soils and contribute to N20 emissions in subsequent years. This is a
simplifying assumption that may create bias in estimated N20 emissions for a specific year. In contrast,
the process-based model in the Tier 3 approach includes the legacy effect of nitrogen added to soils in
Agriculture 5-43
-------
previous years that is re-mineralized from soil organic matter and emitted as N20 during subsequent
years.
DayCent is used to estimate N20 emissions associated with production of alfalfa hay, barley, corn,
cotton, dry beans, grass hay, grass-clover hay, lentils, oats, onions, peanuts, peas, potatoes, rice,
sorghum, soybeans, sugar beets, sunflowers, sweet potatoes, tobacco, tomatoes, and wheat; but is not
applied to estimate N20 emissions from other crops or rotations with other crops,19 such as sugarcane,
some vegetables, and perennial/horticultural crops. Areas that are converted between agriculture (i.e.,
cropland and grassland) and other land uses, such as forest land, wetland and settlements, are not
simulated with DayCent. DayCent is also not used to estimate emissions from land areas with very
gravelly, cobbly, or shaley soils in the topsoil (greater than 35 percent by volume in the top 30 cm of the
soil profile), or to estimate emissions from drained organic soils (Histosols). The Tier 3 method has not
been fully tested for estimating N20 emissions associated with these crops and rotations, land uses, as
well as organic soils or cobbly, gravelly, and shaley mineral soils. In addition, federal grassland areas are
not simulated with DayCent due to limited activity data on land use histories. For areas that are not
included in the DayCent simulations, Tier 1 methods are used to estimate emissions, including (1) direct
emissions from nitrogen inputs for crops on mineral soils that are not simulated by DayCent; (2) direct
emissions from PRP nitrogen additions on federal grasslands; (3) direct emissions for land application of
biosolids (i.e., treated sewage sludge) to soils; and (4) direct emissions from drained organic soils in
croplands and grasslands.
A splicing method is used to estimate soil N20 emissions for 2021 to 2023 at the national scale because
new activity data have not been incorporated into the analysis for those years. Specifically, linear
regression models with autoregressive moving-average (ARMA) errors (Brockwell and Davis 2016) are
used to estimate the relationship between surrogate data and the 1990 to 2020 emissions that are
derived using the Tier 3 method. For the Tier 1 method, the same modeling approach is used to estimate
emissions for 2021-2023 without surrogate data. In addition, this data splicing model is used to estimate
emissions data for 2018 to 2023 for other organic nitrogen amendments (i.e., commercial organic
fertilizer) due to a gap in the activity data during the latter part of the time series (TVA1991 through 1994;
AAPFCO 1995 through 2022). For drainage from organic soils, the data splicing method is used to
estimate emissions for 2023 since activity data are available through 2022. See Box 5-3 for more
information about the splicing method. Emission estimates for years with imputed data will be
recalculated in future Inventory reports when new NRI data and other organic amendment nitrogen data
are available.
Box 5-3: Data Splicing Method
An approach to extend the time series is needed for agricultural soil management because there are
typically activity data gaps at the end of the time series. This is mainly because the NRI survey program,
which provides critical information for estimating greenhouse gas emissions and removals, does not
release data every year.
Splicing methods have been used to impute missing data at the end of the emission time series for both
the Tier 1 and 3 methods. Specifically, a linear regression model with autoregressive moving-average
19 A small proportion of the major commodity crop production, such as corn and wheat, is included in the Tier 1 analysis
because these crops are rotated with other crops or land uses (e.g., forest lands) that are not simulated by DayCent.
5-44 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
(ARMA) errors (Brockwell and Davis 2016) is used to estimate emissions based on the emissions data
that has been compiled using the inventory methods described in this section. The model to extend the
time series is given by the equation:
Y = xp + £,
where Y is the response variable (e.g., soil nitrous oxide), X(B for the Tier 3 method contains specific
surrogate data depending on the response variable, and s is the remaining unexplained error. Models
with a variety of surrogate data were tested, including commodity statistics, weather data, or other
relevant information. The term X(B for the Tier 1 method only contains year as a predictor of emission
patterns over the time series (change in emissions per year), and therefore, is a linear time series model
with no surrogate data. Parameters are estimated using standard statistical techniques, and used in the
model described above to predict the missing emissions data.
A critical issue with splicing methods is to account for the additional uncertainty introduced by
predicting emissions without compiling the full inventory. Specifically, uncertainty will increase for years
with imputed estimates based on the splicing methods, compared to those years in which the full
inventory is compiled. This additional uncertainty is quantified within the model framework using a
Monte Carlo approach. Consequently, the uncertainty from the original inventory data is combined with
the uncertainty in the data splicing model. The approach requires estimating parameters in the data
splicing models in each Monte Carlo simulation for the full inventory (i.e., the surrogate data model is
refit with the draws of parameters values that are selected in each Monte Carlo iteration, and used to
produce estimates with inventory data). Therefore, the data splicing method generates emissions
estimates from each surrogate data model in the Monte Carlo analysis, which are used to derive
confidence intervals in the estimates for the missing emissions data. Furthermore, the 95 percent
confidence intervals are estimated using the 3 sigma rules assuming a unimodal density (Pukelsheim
1994).
Tier 3 Approach for Mineral Cropland Soils
The DayCent biogeochemical model (Parton et al. 1998; Del Grosso et al. 2001 and 2011) is used to
estimate direct N20 emissions from mineral cropland soils that are managed for production of a wide
variety of crops (see list in previous section) based on the crop histories in the 2017 NRI (USDA-NRCS
2020) and extended through 2020 using CDL (USDA-NASS 2021). Crops simulated by DayCent are grown
on approximately 85 percent of total cropland area in the United States. The model simulates net
primary productivity (NPP) using the NASA-CASA production algorithm MODIS Enhanced Vegetation
Index (EVI) products, MOD13Q1 and MYD13Q120 (Potter et al. 1993, 2007). The model simulates soil
temperature and water dynamics, using daily weather data from a 4-kilometer gridded product
developed by the PRISM Climate Group (2022), and soil attributes from the Soil Survey Geographic
Database (SSURGO) (Soil Survey Staff 2020). DayCent is used to estimate direct N20 emissions due to
mineral nitrogen available from the following sources: (1) application of synthetic fertilizers; (2)
application of livestock manure; (3) retention of crop residues in the field for nitrogen-fixing legumes and
20 Net Primary Production is estimated with the NASA-CASA algorithm for most of the cropland that is used to produce
major commodity crops in the central United States from 2000 to 2020. Other regions and years prior to 2000 are
simulated with a method that incorporates water, temperature, and moisture stress on crop production (see Metherell et
al. 1993) but does not incorporate the additional information about crop condition provided with remote sensing data.
Agriculture 5-45
-------
non-legume crops and subsequent mineralization of nitrogen during microbial decomposition (i.e.,
leaving residues in the field after harvest instead of burning or collecting residues); (4) mineralization of
nitrogen from decomposition of soil organic matter; and (5) asymbiotic fixation.
Management activity data from several sources supplement the activity data from the NRI. The USDA-
NRCS Conservation Effects and Assessment Project (CEAP) provides data on a variety of cropland
management activities, and is used to inform the inventory analysis about tillage practices, mineral
fertilization, manure amendments, cover crop management, as well as planting and harvest dates
(USDA-NRCS 2022; USDA-NRCS 2018; USDA-NRCS 2012). CEAP data are collected at a subset of NRI
survey locations, and currently provide management information from approximately 2002 to 2006 and
2013 to 2016. These data are combined with other datasets in an imputation analysis. This imputation
analysis is comprised of three steps: a) determine the trends in management activity across the time
series by combining information from several datasets (discussed below); b) use Gradient Boosting
(Friedman 2001) to determine the likely management practice at a given NRI survey location; and c)
assign management practices from the CEAP survey to the specific NRI locations using a predictive
mean matching method for certain variables that are adapted to reflect the trending information (Little
1988, van Buuren 2012). Gradient boosting is a machine learning technique used in regression and
classification tasks, among others. It combines predictions from multiple weak prediction models and
outperforms many complicated machine learning algorithms. It makes the best predictions at specific
NRI survey locations or at state or region level models. The predictive mean matching method identifies
the most similar management activity recorded in the CEAP surveys that match the prediction from the
gradient boosting algorithm. The matching ensures that imputed management activities are realistic for
each NRI survey location, and not odd or physically unrealizable results that could be generated by the
gradient boosting. There are six complete imputations of the management activity data using these
methods.
To determine trends in mineral fertilization and manure amendments, CEAP data are combined with
information on fertilizer use and rates by crop type for different regions of the United States from the
USDA Economic Research Service. The data collection program was known as the Cropping Practices
Surveys through 1995 (USDA-ERS 1997), and is now part of data collection known as the Agricultural
Resource Management Surveys (ARMS) (USDA-ERS 2020). Additional data on fertilization practices are
compiled through other sources particularly the National Agricultural Statistics Service (USDA-NASS
1992, 1999, 2004). To determine the trends in tillage management, CEAP data are combined with
Conservation Technology Information Center data between 1989 and 2004 (CTIC 2004) and OpTIS Data
Product21 for 2008 to 2020 (Hagen et al. 2020). The CTIC data are adjusted for long-term adoption of no-
till agriculture (Towery 2001). For cover crops, CEAP data are combined with information from USDA
Census of Agriculture (USDA-NASS 2012, 2017) and the OpTIS22 data (Hagen et al. 2020). It is assumed
that cover crop management was minimal prior to 1990 and the rates increased linearly over the decade
to the levels of cover crop management in the CEAP survey.
The IPCC method considers crop residue nitrogen inputs and nitrogen mineralized from soil organic
matter as activity data. However, they are not treated as activity data in DayCent simulations because
residue production, symbiotic nitrogen fixation (e.g., legumes), mineralization of nitrogen from soil
organic matter, and asymbiotic nitrogen fixation are internally generated by the model as part of the
21 OpTIS data on tillage practices provided by Regrow Agriculture, Inc.
22 OpTIS data on cover crop management provided by Regrow Agriculture, Inc.
5-46 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
simulation. In other words, DayCent accounts for the influence of symbiotic nitrogen fixation,
mineralization of nitrogen from soil organic matter and crop residue retained in the field, and asymbiotic
nitrogen fixation on N20 emissions, but these are not model inputs.
The N20 emissions from crop residues are reduced by approximately 3 percent (the assumed average
burned portion for crop residues in the United States) to avoid double counting associated with non-C02
greenhouse gas emissions from agricultural residue burning. Estimated levels of residue burning are
based on state inventory data (ILENR 1993; Oregon Department of Energy 1995; Noller 1996; Wisconsin
Department of Natural Resources 1993; Cibrowski 1996).
Uncertainty in the emission estimates from DayCent is associated with input uncertainty due to missing
management data in the NRI survey that is imputed from other sources; model uncertainty due to
incomplete specification of carbon and nitrogen dynamics in the DayCent model parameters and
algorithms; and sampling uncertainty associated with the statistical design of the NRI survey.
Uncertainty is estimated with two variance components (Ogle et al., 2010). The first variance
component quantifies the uncertainty in management activity data, model structure and
parameterization. To assess this uncertainty, carbon and nitrogen dynamics at each NRI survey location
are simulated six times using the imputation product and other model driver data. Uncertainty in
parameterization and model algorithms are determined using a structural uncertainty estimator derived
from fitting a linear mixed-effect model (Ogle et al. 2007; Del Grosso et al. 2010). The data is combined
in a Monte Carlo stochastic simulation with 1,000 iterations for 1990 through 2020. For each iteration,
there is a random selection of management data from the imputation product (select one of the six
imputations), and random selection of parameter values and random effects for the linear mixed-effect
model (i.e., structural uncertainty estimator). The second variance component quantifies uncertainty in
scaling from the NRI survey to the entire land base. The second variance component is computed using
the replicate weights provided with the NRI survey data, and a standard variance estimator for a two-
stage sample design (Sarndal etal. 1992). The two variance components are summed to quantify the
total uncertainty and produce confidence intervals associated with the estimated emissions.
In order to ensure time-series consistency, the DayCent model is applied from 1990 to 2020, and a data
splicing method is used to approximate emissions for 2021 to 2023 based on the pattern in emissions
data from 1990 to 2020 (See Box 5-3). The pattern is determined by using a linear regression model with
autoregressive moving-average (ARMA) errors and surrogate data, including corn and soybean yields
from USDA-NASS statistics,23 and weather data from the PRISM Climate Group (PRISM 2022). The
method is based on a linear extrapolation of trends, which is a standard data splicing method for
estimating emissions at the end of a time series (IPCC 2006). The time series will be updated with the
Tier 3 method in the future as new activity data are incorporated into the analysis.
Nitrous oxide emissions from managed agricultural lands are the result of interactions among
anthropogenic activities (e.g., nitrogen fertilization, manure application, tillage) and other driving
variables, such as weather and soil characteristics. These factors influence key processes associated
with nitrogen dynamics in the soil profile, including immobilization of nitrogen by soil microbial
organisms, decomposition of organic matter, plant uptake, leaching, runoff, and volatilization, as well as
the processes leading to N20 production (nitrification and denitrification). It is not possible to partition
N20 emissions into each anthropogenic activity directly from model outputs due to the complexity of
23 See https://qijickstats.nass.usda.gov/.
Agriculture 5-47
-------
the interactions (e.g., N20 emissions from synthetic fertilizer applications cannot be distinguished from
those resulting from manure applications). To approximate emissions by activity, the amount of
synthetic nitrogen fertilizer added to the soil, or mineral nitrogen made available through decomposition
of soil organic matter and plant litter, as well as asymbiotic fixation of nitrogen from the atmosphere, is
determined for each nitrogen source and then divided by the total amount of mineral nitrogen in the soil
according to the DayCent model simulation. For 2021 -2023, the contribution of each nitrogen source is
based on the average of values that are estimated for 2018 to 2020. The percentages are then multiplied
by the total of direct N20 emissions in order to approximate the portion attributed to nitrogen
management practices. This approach is only an approximation because it assumes that all nitrogen
made available in soil has an equal probability of being released as N20, regardless of its source, which
is unlikely to be the case (Delgado et al. 2009). However, this approach allows for further disaggregation
of emissions by source of nitrogen, which is valuable information and is analogous to the IPCC (2006)
Tier 1 method, in that it associates portions of the total soil N20 emissions with individual sources of
nitrogen.
Tier 1 Approach for Mineral Cropland Soils
The IPCC (2006) Tier 1 methodology is used to estimate direct N20 emissions for mineral cropland soils
that are not simulated by DayCent (e.g., DayCent has not been parametrized to simulate all crop types
and some soil types such as Histosols). For the Tier 1 method, estimates of direct N20 emissions from
nitrogen applications are based on mineral soil N that is made available from the following practices: (1)
the application of synthetic commercial fertilizers; (2) application of managed manure and non-manure
commercial organic fertilizers; and (3) decomposition and mineralization of nitrogen from above- and
below-ground crop residues in agricultural fields (i.e., crop biomass that is not harvested). Non-manure
commercial organic amendments are only included in the Tier 1 analysis because these data are not
available at the county-level, which is necessary for the DayCent simulations. Consequently, all
commercial organic fertilizer, as well as manure that is not added to crops in the DayCent simulations,
are included in the Tier 1 analysis. The following sources are used to derive activity data:
A process-of-elimination approach is used to estimate synthetic nitrogen fertilizer additions for
crop areas that are not simulated by DayCent. The total amount of fertilizer used on farms has
been estimated at the county-level by the USGS using sales records from 1990 to 2012 (Brakebill
and Gronberg 2017). For 2013 through 2017, fertilizer sales data from AAPFCO (AAPFCO 2013
through 2022) 24 after adjusting for the proportion of on-farm application to determine the
amount applied to crops. The amount of fertilizer applied after 2017 is estimated using the data
splicing method described in Box 5-4 for the linear time series model. Then the portion of
fertilizer applied to crops and grasslands simulated by DayCent is subtracted from the on-farm
sales data (see Tier 3 Approach for mineral cropland soils and direct N20 emissions from
grassland soils sections for information on data sources), and the remainder of the total
fertilizer used on farms is assumed to be applied to crops that are not simulated by DayCent. At
a minimum, 3 percent of state-level on-farm fertilizer sales are assumed to be applied to
cropland in the Tier 1 method.
24 The fertilizer consumption data in AAPFCO are recorded in "fertilizer year" totals, (i.e., July to June), but are converted to
calendar year totals. This is done by assuming that approximately 35 percent of fertilizer usage occurred from July to
December and 65 percent from January to June (TVA 1992b).
5-48 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Similarly, a process-of-elimination approach is used to estimate manure nitrogen additions for
crops that are not simulated by DayCent. The total amount of manure available for land
application to soils has been estimated with methods described in the manure management
section (Section 5.2) and annex (Annex 3.12). The amount of manure nitrogen applied in the Tier
3 approach to crops and grasslands is subtracted from total annual manure nitrogen available
for land application (see Tier 3 Approach for mineral cropland soils and direct N20 emissions
from grassland soils sections for information on data sources). This difference is assumed to be
applied to crops that are not simulated by DayCent.
Commercial organic fertilizer additions are based on organic fertilizer consumption statistics
through 2017,25 which are converted from mass of fertilizer to units of nitrogen using average
organic fertilizer nitrogen content, ranging between 2.3 to 4.2 percent across the time series
(TVA1991 through 1994; AAPFCO 1995 through 2022). Commercial fertilizers include dried
manure and biosolids (i.e., treated sewage sludge), but the amounts are removed from the
commercial fertilizer data to avoid double counting26 with the manure nitrogen dataset
described above and the biosolids (i.e., treated sewage sludge) amendment data discussed
later in this section.
Crop residue nitrogen is derived by combining amounts of above- and below-ground biomass,
which are determined based on NRI crop area data (USDA-NRCS 2020), as extended using the
CDL data (USDA-NASS 2021), crop production yield statistics (USDA-NASS 2023), dry matter
fractions (IPCC 2006), linear equations to estimate above-ground biomass given dry matter crop
yields from harvest (IPCC 2006), ratios of below-to-above-ground biomass (IPCC 2006), and
nitrogen contents of the residues (IPCC 2006). Nitrogen inputs from residue were reduced by 3
percent to account for average residue burning portions in the United States.
The total amounts of soil mineral nitrogen from applied synthetic and organic fertilizers, manure
nitrogen additions and crop residues are multiplied by the IPCC (2006) default emission factor to derive
an estimate of direct N20 emissions using the Tier 1 method. Further elaboration on the methodology
and data used to estimate N20 emissions from mineral soils are described in Annex 3.13.
In order to ensure time-series consistency, the Tier 1 methods are applied from 1990 to 2020 for most
sources, and a data splicing method is used to approximate emissions for 2021 to 2023 based on the
emission patterns between 1990 and 2020 (see Box 5-3). The exceptions include crop residue nitrogen
and biosolid (i.e., treated sewage sludge) amendments to grasslands which are estimating using the Tier
1 method for 1990 to 2023 with no data splicing method; other organic nitrogen fertilizers (i.e.,
commercial fertilizers) are estimated with a data splicing method for 2018 to 2023 due to a gap in the
activity data during the latter part of the time series (TVA 1991 through 1994; AAPFCO 1995 through
2022); and drainage of organic soils are estimated with the data splicing method for 2023 since activity
data are not available after 2022. For data splicing, the emission pattern is determined by using a linear
25 Soil N2O emissions are imputed using data splicing methods for commercial fertilizers, i.e., other organic fertilizers, after
2017 because the activity data are not available.
26 Commercial organic fertilizers include dried blood, tankage, compost, and other, but the dried manure and biosolids
(i.e., treated sewage sludge) are also included in other datasets in this Inventory. Consequently, the proportions of dried
manure and biosolids, which are provided in the reports (TVA 1991 through 1994; AAPFCO 1995 through 2022), are used
to estimate the nitrogen amounts in dried manure and biosolids. To avoid double counting, the resulting nitrogen
amounts for dried manure and biosolids are subtracted from the total nitrogen in commercial organic fertilizers before
estimating emissions usingtheTier 1 method.
Agriculture 5-49
-------
regression model with autoregressive moving-average (ARMA) errors. The method is based on a linear
extrapolation of trends, which is a standard data splicing method for estimating emissions at the end of
a time series (IPCC 2006). As with the Tier 3 method, the time series that is based on the splicing
methods will be recalculated in a future Inventory report with updated activity data.
Tier 1 and 3 Approaches from Mineral Grassland Soils
As with N20 emissions from croplands, the Tier 3 process-based approach with application of the
DayCent model and Tier 1 method described in IPCC (2006) are combined to estimate emissions from
non-federal grasslands and PRP manure nitrogen additions for federal grasslands, respectively.
Grassland includes pasture and rangeland that produce grass or mixed grass/legume forage primarily for
livestock grazing. Rangelands are extensive areas of native grassland that are not intensively managed,
while pastures are seeded grassland (possibly following tree removal) that may also have additional
management, such as irrigation, fertilization, or inter-seeding legumes. DayCent is used to simulate N20
emissions from NRI survey locations (USDA-NRCS 2020) on non-federal grasslands resulting from
manure deposited by livestock directly onto pastures and rangelands (i.e., PRP manure), nitrogen
fixation from legume seeding, managed manure amendments (i.e., manure other than PRP manure such
as daily spread or manure collected from other animal waste management systems such as lagoons
and digesters), and synthetic fertilizer application. Other nitrogen inputs are simulated within the
DayCent framework, including nitrogen input from mineralization due to decomposition of soil organic
matter and nitrogen inputs from senesced grass litter, as well as asymbiotic fixation of nitrogen from the
atmosphere. The simulations used the same weather, soil, and synthetic nitrogen fertilizer data as
discussed under the Tier 3 Approach in the mineral cropland soils section. Synthetic nitrogen
fertilization rates are based on data from the Carbon Sequestration Rural Appraisals (CSRA) conducted
by the USDA-NRCS (USDA-NRCS, unpublished data). The CSRA was a solicitation of expert knowledge
from USDA-NRCS staff throughout the United States to support the Inventory. Biological nitrogen
fixation is simulated within DayCent, and therefore is not an input to the model.
Manure nitrogen deposition from grazing animals in PRP systems (i.e., PRP manure nitrogen) is a key
input of nitrogen to grasslands. The amounts of PRP manure nitrogen applied on non-federal grasslands
for each NRI survey location are based on the amount of nitrogen excreted by livestock in PRP systems
that is estimated in the manure management section (see Section 5.2 and Annex 3.12). The total
amount of nitrogen excreted in each county is divided by the grassland area to estimate the nitrogen
input rate associated with PRP manure. The resulting rates are a direct input into the DayCent
simulations. The nitrogen input is subdivided between urine and dung based on a 50:50 split. DayCent
simulations of non-federal grasslands accounted for approximately 71 percent of total PRP manure
nitrogen in aggregate across the country.27 The remainder of the PRP manure nitrogen in each state is
assumed to be excreted on federal grasslands, and the N20 emissions are estimated using the IPCC
(2006) Tier 1 method.
Biosolids (i.e., treated sewage sludge) are assumed to be applied on grasslands.28 Application of
biosolids is estimated from data compiled by EPA (1993, 1999, 2003), McFarland (2001), and NEBRA
(2007) (see Section 7.2 for a detailed discussion of the methodology for estimating treated sewage
27 A small amount of PRP nitrogen (less than 1 percent) is deposited in grazed pasture that is in rotation with annual crops
and is reported in the grassland N2O emissions.
28 A portion of biosolids may be applied to croplands, but there is no national dataset to disaggregate the amounts between
cropland and grassland.
5-50 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
sludge available for land application application). Biosolids data are only available at the national scale,
and it is not possible to associate application with specific soil conditions and weather at NRI survey
locations. Therefore, DayCent could not be used to simulate the influence of biosolids on N20
emissions from grassland soils, and consequently, emissions from biosolids are estimated using the
IPCC (2006) Tier 1 method.29
Soil N20 emission estimates from DayCent are adjusted using a structural uncertainty estimator
accounting for uncertainty in model algorithms and parameter values (Del Grosso et al. 2010). There is
also sampling uncertainty for the NRI survey that is quantified with replicate sampling weights
associated with the survey, as discussed for Tier 3 method associated with mineral cropland soils. N20
emissions for the PRP manure nitrogen deposited on federal grasslands and applied biosolids nitrogen
are estimated using the Tier 1 method by multiplying the nitrogen input by the default emission factor.
Emissions from manure nitrogen are estimated at the state level and aggregated to the entire country,
but emissions from biosolids nitrogen are calculated exclusively at the national scale. Further
elaboration on the methodology and data used to estimate N20 emissions from mineral soils are
described in Annex3.13.
Soil N20 emissions and 95 percent confidence intervals are estimated for each year between 1990 and
2020 based on the Tier 1 and 3 methods, except for biosolids and crop residue. In order to ensure time-
series consistency, emissions from 2021 to 2023 are estimated using a data splicing method that
consists of a linear regression model with autoregressive moving-average (ARMA) errors. The method is
based on a linear extrapolation of trends, which is a standard data splicing method for estimating
emissions at the end of a time series (IPCC 2006). As with croplands, estimates for 2021 to 2023 will be
recalculated in a future Inventory when the activity data are updated. Biosolids application and crop
residue nitrogen data are compiled through 2023 in this Inventory, and therefore soil N20 emissions are
estimated using the Tier 1 method for all years without application of the splicing method.
Tier 1 Approach for Drainage of Organic Soils in Croplands and Grasslands
The IPCC (2006) Tier 1 method is used to estimate direct N20 emissions due to drainage of organic soils
in croplands and grasslands at a state scale. State-scale estimates of the total area of drained organic
soils are obtained from the 2017 NRI (USDA-NRCS 2020), and extended through 2022 using CDL (USDA-
NASS 2021) and the Forest Inventory and Analysis (FIA) survey data, which is harmonized with the NRI
data (Nelson et al. 2020). Organic soils are identified using soils data from the Soil Survey Geographic
Database (SSURGO) (Soil Survey Staff 2020). The IPCC climate region map is used to subdivide areas
into temperate and tropical climates according to the climate classification from IPCC (2006). To
estimate annual emissions, the total temperate area is multiplied by the IPCC default emission factor
for temperate regions, and the total tropical area is multiplied by the IPCC default emission factor for
tropical regions (IPCC 2006). In order to ensure time-series consistency, a data splicing method is used
to estimate emissions in 2023 based on a linear regression model with autoregressive moving-average
(ARMA) errors. The method is based on a linear extrapolation of trends, which is a standard data splicing
method for estimating emissions at the end of a time series (IPCC 2006). Estimates for 2023 will be
recalculated in a future Inventory when the activity data are updated.
29 Data on biosolids in the United States is shared by the team which compiles the Waste Chapter of this Inventory to
ensure consistency of activity data across sectors.
Agriculture 5-51
-------
Total Direct N20 Emissions from Cropland and Grassland Soils
Annual direct emissions from the Tier 1 and 3 approaches for mineral and drained organic soils
occurring in both croplands and grasslands are summed to obtain the total direct N20 emissions from
agricultural soil management (see Table 5-16 and Table 5-17). Further elaboration on the methodology
and data used to estimate soil N20 emissions are described in Annex 3.13.
Indirect N2O Emissions Associated with Nitrogen Management in
Cropland and Grasslands
Indirect N20 emissions occur when synthetic nitrogen applied or made available through anthropogenic
activity is transported from the soil either in gaseous or aqueous forms and later converted into N20.
There are two pathways leading to indirect emissions. The first pathway results from volatilization of
nitrogen as NOx (nitrogen oxides) and NH3 (ammonia) following application of synthetic fertilizer, organic
amendments (e.g., manure, biosolids), and deposition of PRP manure. Nitrogen made available from
mineralization of soil organic matter and residue, including nitrogen incorporated into crops and forage
from symbiotic nitrogen fixation, and input of nitrogen from asymbiotic fixation also contributes to
volatilized nitrogen emissions. Volatilized nitrogen can be returned to soils through atmospheric
deposition, and a portion of the deposited nitrogen is emitted to the atmosphere as N20. The second
pathway occurs via leaching and runoff of soil nitrogen (primarily in the form of N03", i.e., nitrate) that is
made available through anthropogenic activity on managed lands, including organic and synthetic
fertilization, organic amendments, mineralization of soil organic matter and residue, and inputs of
nitrogen into the soil from asymbiotic fixation. Nitrate is subject to denitrification in water bodies, which
leads to N20 emissions. Regardless of the eventual location of the indirect N20 emissions, the
emissions are assigned to the original source of the nitrogen for Inventory purposes, which here includes
croplands and grasslands.
Tier 1 and 3 Approaches for Indirect N20 Emissions from Atmospheric Deposition of
Volatilized Nitrogen
The Tier 3 DayCent model and IPCC (2006) Tier 1 methods are combined to estimate the amount of
nitrogen that is volatilized and eventually emitted as N20. DayCent is used to estimate nitrogen
volatilization for land areas whose direct emissions are simulated with DayCent (i.e., most commodity
and some specialty crops and most grasslands). The nitrogen inputs included are the same as
described for direct N20 emissions in the Tier 3 approach for mineral cropland and grassland soils
sections. Nitrogen volatilization from all other areas is estimated using the Tier 1 method with default
IPCC fractions for nitrogen subject to volatilization (i.e., synthetic and manure nitrogen on croplands not
simulated by DayCent, other organic nitrogen inputs (i.e., commercial fertilizers), PRP manure nitrogen
excreted on federal grasslands, and biosolids [i.e., treated sewage sludge] application on grasslands).
The IPCC (2006) default emission factor is multiplied by the amount of volatilized nitrogen generated
from both DayCent and Tier 1 methods to estimate indirect N20 emissions occurringwith re-deposition
of the volatilized nitrogen from 1990-2020 (see Table 5-19). A linear regression model with autoregressive
moving-average (ARMA) errors, described in Box 5-3, is applied to estimate emissions from 2021 to 2023
based on the emission patterns from 1990 to 2020. The method is based on a linear extrapolation of
trends, which is a standard data splicing method for estimating emissions at the end of a time series
(IPCC 2006). Further elaboration on the methodology and data used to estimate indirect N20 emissions
are described in Annex 3.13.
5-52 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Tier 1 and 3 Approaches for Indirect N20 Emissions from Leaching/Runoff
As with the calculations of indirect emissions from volatilized nitrogen, the Tier 3 DayCent model and
IPCC (2006) Tier 1 method are combined to estimate the amount of nitrogen that is subject to leaching
and surface runoff into water bodies, and eventually emitted as N20. DayCent is used to simulate the
amount of nitrogen transported from lands in the Tier 3 Approach. Nitrogen transport from all other
areas is estimated using the Tier 1 method and the IPCC (2006) default factor for the proportion of
nitrogen subject to leaching and runoff associated with nitrogen applications on croplands that are not
simulated by DayCent, applications of biosolids on grasslands, other organic N fertilizer applications,
crop residue nitrogen inputs, and PRP manure nitrogen excreted on federal grasslands.
For both the DayCent Tier 3 and IPCC (2006) Tier 1 methods, N03" leaching is assumed to be an
insignificant source of indirect N20 in cropland and grassland systems in arid regions, as discussed in
IPCC (2006). In the United States, the threshold for significant N03" leaching is based on the potential
evapotranspiration (PET) and rainfall amount, similar to IPCC (2006), and is assumed to be negligible in
regions where the amount of precipitation does not exceed 80 percent of PET (Note: All irrigated systems
are assumed to have significant amounts of leaching of nitrogen even in drier climates).
For leaching and runoff data estimated by the Tier 3 and Tier 1 approaches, the IPCC (2006) default
emission factor is used to estimate indirect N20 emissions that occur in groundwater and waterways
(see Table 5-19). Further elaboration on the methodology and data used to estimate indirect N20
emissions are described in Annex 3.13.
In order to ensure time-series consistency, indirect soil N20 emissions are estimated using the Tier 1
and 3 approaches from 1990 to 2020 and then a linear regression model with autoregressive moving-
average (ARMA) errors, described in Box 5-3, is applied to estimate emissions from 2021 to 2023 based
on the emission patterns from 1990 to 2020. The method is based on a linear extrapolation of trends,
which is a standard data splicing method for estimating emissions at the end of a time series (IPCC
2006). As with the direct N20 emissions, the time series will be recalculated in a future Inventory when
new activity data are incorporated into the analysis.
Uncertainty
Uncertainty is estimated for each of the following five components of N20 emissions from agricultural
soil management: (1) direct emissions simulated by DayCent; (2) the components of indirect emissions
(nitrogen volatilized and leached or runoff) simulated by DayCent; (3) direct emissions estimated with
the IPCC (2006) Tier 1 method; (4) the components of indirect emissions (nitrogen volatilized and
leached or runoff) estimated with the IPCC (2006) Tier 1 method; and (5) indirect emissions estimated
with the IPCC (2006) Tier 1 method. Uncertainty in direct emissions as well as the components of
indirect emissions that are estimated from DayCent are derived from two variance components (Ogle et
al. 2010). For the first component, a Monte Carlo Analysis (consistent with IPCC Approach 2) is used to
address uncertainties in management activity data as well as model parameterization and structure (Del
Grosso et al. 2010). The second variance component is quantifying uncertainty in scaling from the NRI
survey to the entire land base, and computed using a standard variance estimator for a two-stage
sample design (Sarndal ef al. 1992). The two variance components are combined using simple error
propagation methods provided by the IPCC (2006), i.e., by taking the square root of the sum of the
squares of the standard deviations of the uncertain quantities. For 2021 to 2023 (as well as 2018 to 2023
for other organic nitrogen fertilizers and 2023 for drainage of organic soils) there is additional uncertainty
Agriculture 5-53
-------
propagated through the Monte Carlo Analysis associated with the splicing method (See Box 5-3) except
for the Tier 1 method for biosolids and crop residue nitrogen inputs, which do not use the data splicing
method.
Simple error propagation methods (IPCC 2006) are used to derive confidence intervals for direct
emissions estimated with the IPCC (2006) Tier 1 method, the proportion of volatilization and leaching or
runoff estimated with the IPCC (2006) Tier 1 method, and indirect N20 emissions. Uncertainty in the
splicing method is also included in the error propagation for 2021 -2023 (see Box 5-3). Additional details
on the uncertainty methods are provided in Annex 3.13. Table 5-20 shows the combined uncertainty for
soil N20 emissions. The estimated direct soil N20 emissions range from 28 percent below to 28 percent
above the 2023 emission estimate of 266.8 MMT C02 Eq. The combined uncertainty for indirect soil N20
emissions ranges from 52 percent below to 124 percent above the 2023 estimate of 29.6 MMT C02 Eq.
Table 5-20: Quantitative Uncertainty Estimates of N20 Emissions from Agricultural
Soil Management in 2023 (MMT C02 Eq. and Percent)
Uncertainty Range Relative to Emission Estimate
2023 Emission
(MMT CO2 Eq.)
(%)
Estimate
Lower
Upper
Lower
Upper
Source
Gas
(MMT CO2 Eq.)
Bound
Bound
Bound
Bound
Direct Soil N2O Emissions
N2O
266.8
191.7
341.8
-28%
+28%
Indirect Soil N2O Emissions
N2O
29.6
14.3
66.1
-52%
+ 124%
Note: Due to lack of data, uncertainties in PRP manure nitrogen production, other organic fertilizer amendments, and biosolids
(i.e., treated sewage sludge) amendments to soils are currently treated as certain. These sources of uncertainty will be included
in a future Inventory (IPCC 2006).
Additional uncertainty is associated with an incomplete estimation of N20 emissions from managed
croplands and grasslands in Hawaii and Alaska. The Inventory currently includes the N20 emissions
from managed manure and PRP nitrogen, and biosolid additions for Alaska and managed manure and
PRP nitrogen, biosolid additions, and crop residue for Hawaii. Land areas used for agriculture in Alaska
and Hawaii are small relative to major crop commodity states in the conterminous United States, so the
emissions are likely to be minor for the other sources of nitrogen (e.g., synthetic fertilizer and crop
residue inputs. Regardless, there is a planned improvement to include the additional sources of
emissions in a future Inventory.
QA/QC and Verification
General (Tier 1) and category-specific (Tier 2) QA/QC activities were conducted consistent with the U.S.
Inventory QA/QC plan outlined in Annex 8. DayCent results for N20 emissions and N03" leaching are
compared with field data representing various cropland and grassland systems, soil types, and climate
patterns (Del Grosso et al. 2005; Del Grosso et al. 2008), and further evaluated by comparing the model
results to emission estimates produced using the IPCC (2006) Tier 1 method for the same sites. Nitrous
oxide measurement data for cropland are available for 64 sites with 769 observations of management
practice effects, and measurement data for grassland are available for 12 sites with 88 observations of
management practice effects. Nitrate leaching data are available for 14 sites, representing 432
observations of management practice effects. In general, DayCent predicted N20 emission and nitrate
5-54 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
leaching for these sites reasonably well. See Annex 3.13 for more detailed information about the
comparisons.
Computational processes and activity data have been checked that are used to conduct the inventory
and quantify uncertainty. An error was identified in the uncertainty estimation associated with the
activity data from the USDA-NRCS National Resources Inventory. The error propagation with these
activity data had not been estimated correctly. Additionally, an error was identified in the data splicing
results for emissions estimates associated with synthetic nitrogen fertilizer N additions. For organic soil
areas, an error was found in the land representation data that had been entered into the computational
system incorrectly. All of these errors were corrected. Links between spreadsheets have also been
checked, updated, and corrected as needed.
Recalculations Discussion
Several improvements have been implemented in this Inventory leading to recalculations, including a)
incorporation of a revised time series of data for manure nitrogen amendments to soils and
pasture/range/paddock manure nitrogen additions to soils, b) revised surrogate data for application of
the data splicing methods, and c) corrections to the uncertainty analysis associated with the activity
data from the NRI, which had been identified through QA/QC. The combined impact from these changes
resulted in an average annual increase in emissions of 0.6 MMT C02 Eq., or 0.2 percent, from 1990 to
2022 relative to the previous Inventory.
Planned Improvements
Several planned improvements are underway associated with improving the DayCent biogeochemical
model. These are near-term improvements that will be implemented in the 2026 Inventory publication,
at the earliest, and include:
Incorporating a better representation of plant phenology, particularly senescence events
following grain filling in crops.
Calibrating the model to improve crop parameters associated with temperature and water stress
effects on plant production.
Calibrating the nitrogen submodule and more accurately predict nitrogen-gas losses and nitrate
leaching rates. Experimental study sites will continue to be added for quantifying model
structural uncertainty with priority given to studies that have continuous (daily) measurements
of N20 (e.g., Scheer et al. 2013).
Additional DayCent improvements include:
Simulating crop residue burning in the DayCent model based on the amount of crop residues
burned according to the data that is used in the Field Burning of Agricultural Residues source
category (see Section 5.7), which will impact the estimated nitrogen inputs from crop residues
that result in N20 emissions reported as part of this source category. This improvement will lead
to greater consistency in the methods across sources, ensuring mass balance of carbon and
nitrogen in the Inventory analysis. This improvement will likely be implemented for the 2027
Inventory, at the earliest.
Agriculture 5-55
-------
Other improvements are longer-term priorities. The earliest these improvements would be incorporated
is the 1990 through 2025 Inventory (\.e. 2027 Inventory), pending improvement progress and
prioritization of resources. These improvements include:
Developing estimates for all sources of nitrogen in Alaska and Hawaii for agricultural soil
management as part of Tier 1 calculations, which currently only addresses managed manure
nitrogen and PRP nitrogen, and biosolids additions for grasslands in both states, in addition to
crop residue nitrogen inputs for Hawaii. The implementation of this improvements relies on the
identification of a suitable cropland data layer to provide necessary activity data to estimate
these additional sources of emissions in Alaska.
Incorporating the Tier 1 emission factor for N20 emissions from drained organic soils by using
the revised factors in the 2013 Supplement to the 2006IPCC Guidelines for National
Greenhouse Gas Inventories: Wetlands (IPCC 2014).
Tier 1 method associated with estimating soil N20 emissions from nitrogen mineralization due to
soil organic matter decomposition that is accelerated with land use conversions to cropland
and grassland.
Reviewing available data on biosolids (i.e., treated sewage sludge) application to improve the
distribution of biosolids application on croplands, grasslands and settlements.
5.5 Liming (Source Category 3G)
Crushed limestone (CaC03) and dolomite (CaMg(C03)2) are added to soils by land managers to increase
soil pH (i.e., to reduce acidification). Carbon dioxide emissions occur as these compounds react with
hydrogen ions in soils. The rate of degradation of applied limestone and dolomite depends on the soil
conditions, soil type, climate regime, and whether limestone or dolomite is applied. Emissions from
limestone and dolomite that are used in industrial processes (e.g., cement production, glass
production, etc.) are reported in the IPPU chapter. Emissions from liming of soils have fluctuated
between 1990 and 2023 in the United States, ranging from 2.2 MMT C02 Eq. to 6.0 MMT C02 Eq. across
the entire time series. In 2023, liming of soils in the United States resulted in emissions of 5.3 MMT C02
Eq. (1.4 MMT C), representing a 13 percent increase in emissions since 1990 (see Table 5-21 and Table
5-22). The trend is driven by variation in the amount of limestone and dolomite applied to soils over the
time period.
Table 5-21: Emissions from Liming (MMT C02 Eq.)
Source
1990
2005
2019
2020
2021
2022
2023
Limestone
4.1
3.9
1.9
2.5
2.0
2.9
5.0
Dolomite
0.6
0.4
0.3
0.4
0.4
0.3
0.3
Total
4.7
4.4
2.2
2.9
2.4
3.2
5.3
Note: Totals may not sum due to independent rounding.
5-56 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Table 5-22: Emissions from Liming (MMT C)
Source
1990
2005
2019
2020
2021
2022
2023
Limestone
1.1
1.1
0.5
0.7
0.5
0.8
1.4
Dolomite
0.2
0.1
0.1
0.1
0.1
0.1
0.1
Total
1.3
1.2
0.6
0.8
0.7
0.9
1.4
+ Does not exceed 0.05 MMT C.
Note: Totals may not sum due to independent rounding.
Methodology and Time-Series Consistency
Carbon dioxide emissions from application of limestone and dolomite to soils were estimated using a
Tier 2 methodology consistent with IPCC (2006). The annual amounts of limestone and dolomite, which
are applied to soils (see Table 5-22), were multiplied by C02 emission factors from West and McBride
(2005). The same methods were applied to the entire time series to ensure time-series consistency from
1990 through 2023. These country-specific emission factors (0.059 metric ton C/metric ton limestone,
0.064 metric ton C/metric ton dolomite) are lower than the IPCC default emission factors because they
account for the portion of carbonates that are transported from soils through hydrological processes
and eventually deposited in ocean basins (West and McBride 2005). This analysis of lime dissolution is
based on studies in the Mississippi River basin, where the vast majority of lime application occurs in the
United States (West 2008). Moreover, much of the remaining lime application is occurring under similar
precipitation regimes, and so the emission factors are considered a reasonable approximation for all
lime application in the United States (West 2008) (see Box 5-4).
The annual application rates of limestone and dolomite were derived from estimates and industry
statistics provided in the U.S. Geological Survey (USGS) Minerals Yearbook (Tepordei 1994 through 2015;
Willett 2007a, 2007b, 2009, 2010, 2011 a, 2011 b, 2013a, 2014, 2015, 2016, 2017, 2020a, 2020b, 2022a,
2022b, 2022c, 2022d, 2023a, 2023b, 2024a, 2024b), as well as preliminary data that will eventually be
published in the Minerals Yearbook for the latter part of the time series (Willett 2024c). Data for the final
year of the inventory is based on the Mineral Industry Surveys, as discussed below (USGS 2023). The
U.S. Geological Survey (USGS; U.S. Bureau of Mines prior to 1997) compiled production and use
information through surveys of crushed stone manufacturers. However, manufacturers provided
different levels of detail in survey responses so the estimates of total crushed limestone and dolomite
production and use were divided into three components: (1) production by end-use, as reported by
manufacturers (i.e., "specified" production); (2) production reported by manufacturers without end-uses
specified (i.e., "unspecified" production); and (3) estimated additional production by manufacturers
who did not respond to the survey (i.e., "estimated" production).
Box 5-4: Comparison of the Tier 2 U.S. Inventory Approach and IPCC (2006) Default
Approach
Emissions from liming of soils were estimated using a Tier 2 methodology based on emission factors
specific to the United States that are lower than the IPCC (2006) default emission factors. Most lime
application in the United States occurs in the Mississippi River basin, or in areas that have similar soil
and rainfall regimes as the Mississippi River basin. Under these conditions, a significant portion of
dissolved agricultural lime leaches through the soil into groundwater. Groundwater moves into channels
and is transported to larger rivers and eventually the ocean where CaC03 precipitates to the ocean floor
(West and McBride 2005). The U.S.-specific emission factors (0.059 metric ton C/metric ton limestone
Agriculture 5-57
-------
and 0.064 metric ton C/metric ton dolomite) are about half of the IPCC (2006) emission factors (0.12
metric ton C/metric ton limestone and 0.13 metric ton C/metric ton dolomite). For comparison, the
2023 U.S. emission estimate from liming of soils is 5.3 MMT C02 Eq. using the country-specific factors.
In contrast, emissions would be estimated at 10.7 MMT C02 Eq. using the IPCC (2006) default emission
factors.
Data on "specified" limestone and dolomite amounts were used directly in the emission calculation
because the end use is provided by the manufacturers and can be used to directly determine the
amount applied to soils. However, it is not possible to determine directly how much of the limestone
and dolomite is applied to soils for manufacturer surveys in the "unspecified" and "estimated"
categories. For these categories, the amounts of crushed limestone and dolomite applied to soils were
determined by multiplying the percentage of total "specified" limestone and dolomite production that is
applied to soils, by the total amounts of "unspecified" and "estimated" limestone and dolomite
production. In other words, the proportion of total "unspecified" and "estimated" crushed limestone
and dolomite that was applied to soils is proportional to the amount of total "specified" crushed
limestone and dolomite that was applied to soils.
In addition, data were not available for 1990 and 1992 on the fractions of total crushed stone production
that were limestone and dolomite, and on the fractions of limestone and dolomite production that were
applied to soils. To estimate the 1990 and 1992 data, a set of average fractions were calculated using
the 1991 and 1993 data. These average fractions were applied to the quantity of "total crushed stone
produced or used" as reported for 1990 and 1992 in the 1994 Minerals Yearbook (Tepordei 1996). Data
for 2023 were provided by USGS in a personal communication containing preliminary tables for the 2023
Minerals Yearbook (Willett 2024c).
The primary source for limestone and dolomite activity data is the Minerals Yearbook, published by the
Bureau of Mines through 1996 and by the USGS from 1997 to the present. In 1994, the "Crushed Stone"
chapter in the Minerals Yearbook began rounding (to the nearest thousand metric tons) quantities for
total crushed stone produced or used. It then reported revised (rounded) quantities for each of the years
from 1990 to 1993. In order to minimize the inconsistencies in the activity data, these revised production
numbers have been used in all of the subsequent calculations.
Table 5-23: Applied Minerals (MMT)
Mineral
1990
2005
2019
2020
2021
2022
2023
Limestone
19.0
18.1
8.9
11.6
9.3
13.6
23.2
Dolomite
2.4
1.9
1.2
1.6
1.6
1.1
1.1
Uncertainty
Uncertainty regarding the amount of limestone and dolomite applied to soils was estimated at ±15
percent with normal densities (Tepordei 2003; Willett 2013b). Analysis of the uncertainty associated
with the emission factors included the fraction of lime dissolved by nitric acid versus the fraction that
reacts with carbonic acid, and the portion of bicarbonate that leaches through the soil and is
transported to the ocean. Uncertainty regarding the time associated with leaching and transport was not
addressed in this analysis, but is assumed to be a relatively small contributor to the overall uncertainty
(West 2005). The probability distribution functions for the fraction of lime dissolved by nitric acid and the
5-58 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
portion of bicarbonate that leaches through the soil were represented as triangular distributions
between ranges of zero and 100 percent of the estimates. The uncertainty surrounding these two
components largely drives the overall uncertainty. The emission factor distributions were truncated at
zero so that emissions were not less than zero.
A Monte Carlo (Approach 2) uncertainty analysis was applied to estimate the uncertainty in C02
emissions from liming. The results of the Approach 2 quantitative uncertainty analysis are summarized
in Table 5-24. Carbon dioxide emissions from carbonate lime application to soils in 2023 were estimated
to be between 0.81 and 10.00 MMT C02 Eq. at the 95 percent confidence level. This confidence interval
represents a range of 85 percent below to 89 percent above the 2023 emission estimate of 5.3 MMT C02
Eq. Some carbon in the carbonate lime applied to agricultural soils is not emitted to the atmosphere due
to the dominance of the carbonate lime dissolving in carbonic acid rather than nitric acid (West and
McBride 2005).
Table 5-24: Approach 2 Quantitative Uncertainty Estimates for C02 Emissions from
Liming (MMT C02 Eq. and Percent)
Uncertainty Range Relative to Emission Estimate"
2023 Emission
(MMT CO2 Eq.)
(%)
Estimate
Lower
Upper
Lower
Upper
Source
Gas
(MMT CO2 Eq.)
Bound
Bound
Bound
Bound
Liming
C02
5.3
0.81
10.00
-85%
+89%
a Range of emission estimates predicted by Monte Carlo stochastic simulation for a 95 percent confidence interval.
QA/QC and Verification
A source-specific QA/QC plan for liming has been developed and implemented, consistent with the U.S.
Inventory QA/QC plan outlined in Annex 8. The quality control effort focused on the Tier 1 procedures for
this Inventory. Errors were identified in the activity data covering the years 1993 to 2022, including state
level application of crushed stone on agricultural lands and national estimates of crushed stone usage.
These errors were resolved by cross-referencing the data with national and state level applications of
limestone and dolomite in the Mineral Yearbooks for 1993-2022. For the 1990 through 2024 Inventory,
EPA plans to review and compare data on crushed limestone and dolomite production and obtained
through the GHGRP to activity data from USGS that are currently used as activity data for this source
category. This review will focus on ensuring no double-counting and/or under-counting of crushed stone
between these two sectors.
Recalculations Discussion
Limestone and dolomite application data were updated with the most recent publications and personal
communications from Willett, J.C. (2018-2024). Additionally, data for the 1993-2023 time series were
also revised based on a quality control review with published data available through USGS. The
emissions for 2005, 2016 and 2021 increased by 0.3 percent, 3.9 percent and <0.1 percent, respectively,
relative to the previous Inventory. However, the change over the entire time series was less than 0.1
percent relative to the previous Inventory.
Agriculture 5-59
-------
Planned Improvements
There are no planned improvements for estimating C02 emissions from liming. To ensure completeness,
EPA plans to review available agricultural fertilizer application industry reports to investigate the use of
other carbon-containing fertilizers beyond liming and urea application to identify any additional carbon-
containing fertilizers that constitute a significant source of C02 emissions and should therefore be
captured in the Inventory.
5.6 Urea Fertilization (Source Category 3H)
The use of urea (CO(NH2)2) as a fertilizer leads to greenhouse gas emissions through the release of C02
that was fixed during the production of urea. In the presence of water and urease enzymes, urea that is
applied to soils as fertilizer is converted into ammonium (NH4+), hydroxyl ion (OH), and bicarbonate
(HC03 ). The bicarbonate then evolves into C02 and water. Emissions from urea fertilization in the United
States were 5.3 MMT C02 Eq. (1.4 MMT C) in 2023 (Table 5-25 and Table 5-26). Carbon dioxideemissions
have increased by 118 percent between 1990 and 2023 due to an increasing amount of urea that is
applied to soils. The variation in emissions across the time series is driven by differences in the amounts
of fertilizer applied to soils each year. Carbon dioxide emissions associated with urea used for non-
agricultural purposes are reported in the IPPU chapter (Section 4.6).
Table 5-25: C02 Emissions from Urea Fertilization (MMT C02 Eq.)
Source 1990 2005 | 2019 2020
2021
2022
2023
Urea Fertilization 2.4 | 3.5 | 4.9 5.0
5.1
5.2
5.3
Table 5-26: C02 Emissions from Urea Fertilization (MMT C)
Source 1990
2005
2019 2020
2021
2022
2023
Urea Fertilization 0.7 |
I 1-0 I
1.3 1.4
1.4
1.4
1.4
Methodology and Time-Series Consistency
Carbon dioxide emissions from the application of urea to agricultural soils were estimated using the
IPCC (2006) Tier 1 methodology following the 2006IPCC Guidelines Figure 11.5 decision tree for C02
emissions from urea fertilization.30 The method assumes that carbon in the urea is released after
application to soils and converted to C02.
The annual amounts of urea applied to croplands (see Table 5-27) were derived from the state-level
fertilizer sales data provided in Commercial Fertilizer reports (TVA1991 through 1994; AAPFCO 1995
through 2022).31 These amounts were multiplied by the default IPCC (2006) emission factor (0.20 metric
30 2006 IPCC Guidelines Volume 4, Chapter 11, Figure 11.5 (page 11.33).
31 The amount of urea consumed for non-agricultural purposes in the United States is reported in the Industrial Processes
and Product Use chapter, Section 4.6 Urea Consumption for Non-Agricultural Purposes. Data are shared between the
Agriculture sector and IPPU sector compilation teams to ensure consistency and no double-counting of urea.
5-60 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
tons of carbon per metric ton of urea), which is equal to the carbon content of urea on an atomic weight
basis. National estimates from urea fertilization also include emissions from Puerto Rico.
Fertilizer sales data are reported in fertilizer years (July previous year through June current year), so a
calculation was performed to convert the data to calendar years (January through December).
According to monthly fertilizer use data (TVA 1992b), 35 percent of total fertilizer used in any fertilizer
year is applied between July and December of the previous calendar year, and 65 percent is applied
between January and June of the current calendaryear.
Fertilizer sales data for the 2018 through 2023 fertilizer years were not available for this Inventory.
Therefore, urea applications for 2018 through 2023 were estimated using linear regression with
autoregressive moving average (ARMA) errors using data from 1990 to 2017 at the state scale.
State-level estimates of C02 emissions from the application of urea to agricultural soils were summed
to estimate total emissions for the entire United States. The fertilizer year data is then converted into
calendaryear (Table 5-27) data using the method described above.
Table 5-27: Applied Urea (MMT)
Source
1990 I 2005 1 2019
2020
2021
2022
2023
Urea Fertilizer®
3.3 | 4.8 | 6.7
6.9
7.0
7.1
7.2
aThese numbers represent amounts applied to all agricultural land, including cropland remaining cropland, land converted to
cropland, grassland remaining grassland, land converted to grassland, settlements remaining settlements, land converted to
settlements, forest land remaining forest land and land converted to forest land, as it is not currently possible to apportion the
data by land-use/conversion category.
In order to ensure time-series consistency, the same methods are applied from 1990 to 2023, and data
splicing methods are used to approximate activity data used to derive emissions estimates for the 2018
through 2023 time series based on the emissions data from 1990 to 2017.
Uncertainty
An Approach 2 Monte Carlo analysis is conducted as described by the IPCC (2006). The largest source
of uncertainty is the default emission factor, which assumes that 100 percent of the carbon in CO(NH2)2
applied to soils is emitted as C02. The uncertainty surrounding this factor incorporates the possibility
that some of the carbon may not be emitted to the atmosphere, and therefore the uncertainty range is
set from 50 percent emissions to the maximum emission value of 100 percent using a triangular
distribution. In addition, urea consumption data have uncertainty that is represented as a normal
density. Due to the highly skewed distribution of the resulting emissions from the Monte Carlo
uncertainty analysis, the estimated emissions are based on the analytical solution to the equation, and
the confidence interval is approximated based on the values at 2.5 and 97.5 percentiles.
Carbon dioxide emissions from urea fertilization of agricultural soils in 2023 are estimated to be
between 3.01 and 5.42 MMT C02 Eq. at the 95 percent confidence level. This indicates a range of 43
percent below to 3 percent above the 2023 emission estimate of 5.3 MMT C02 Eq. (Table 5-28).
Agriculture 5-61
-------
Table 5-28: Quantitative Uncertainty Estimates for C02 Emissions from Urea
Fertilization (MMT C02 Eq. and Percent)
Uncertainty Range Relative to Emission Estimate"
2023 Emission
(MMT CO2 Eq.)
(%)
Estimate
Lower
Upper
Lower
Upper
Source
Gas
(MMT CO2 Eq.)
Bound
Bound
Bound
Bound
Urea Fertilization
C02
5.3
3.01
5.42
-43%
+3%
a Range of emission estimates predicted by Monte Carlo stochastic simulation for a 95 percent confidence interval.
There are additional uncertainties that are not quantified in this analysis. There is uncertainty
surrounding the assumptions underlying conversion of fertilizer years to calendar years. These
uncertainties are negligible over multiple years because an over- or under-estimated value in one
calendar year is addressed with a corresponding increase or decrease in the value for the subsequent
year. In addition, there is uncertainty regarding the fate of carbon in urea that is incorporated into
solutions of urea ammonium nitrate (UAN) fertilizer. Emissions of C02 from UAN applications to soils are
not estimated in the current Inventory (see Planned Improvements).
QA/QC and Verification
A source-specific QA/QC plan for Urea Fertilization has been developed and implemented, consistent
with the U.S. Inventory QA/QC plan. No quality control problems were discovered in this process except
a correction to the emissions factor value in documentation tables.
Recalculations Discussion
The data splicing method was updated from a linear extrapolation with 5 years of data to a linear
regression with autoregressive moving average errors (ARMA) that allows for linear drift in order to
estimate values for 2016 through 2023. This update led to an average decrease in emissions for the
years 2016 through 2022 of 0.1 MMT C02 Eq., or 1.4 percent. The rest of the 1990 through 2015 time
series was not affected by this change.
Planned Improvements
The following are potential improvements for Urea Fertilization:
Incorporating Urea Ammonium Nitrate (UAN) in the estimation of urea C02 emissions. Activity
data for UAN have been identified, but additional information is needed to fully incorporate this
type of fertilizer into the analysis, which will be completed in a future Inventory.
Exploring alternative sources of activity data to estimate emissions from urea application to
reduce the time period over which an extrapolation is applied to complete the most recent years
of the time series. This improvement is being conducted with USDA. This improvement would
potentially be implemented in the 1990 through 2025 Inventory published in 2027, at the
earliest.
5-62 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
5.7 Field Burning of Agricultural Residues
(Source Category 3F)
Crop production creates large quantities of agricultural crop residues, which farmers manage in a
variety of ways. For example, crop residues can be left in the field and possibly incorporated into the soil
with tillage; collected and used as fuel, animal bedding material, supplemental animal feed, or
construction material; composted and applied to soils; transported to landfills; or burned in the field.
The 2006IPCC Guidelines does not consider field burning of crop residues to be a net source of COz
emissions because it is assumed the carbon released to the atmosphere as COz during burning is
reabsorbed during the next growing season by the crop (IPCC 2006). However, crop residue burning is a
net source of CH4, NzO, CO, and NOx, which are released during combustion.
In the United States, field burning of agricultural residues occurs in southeastern states, the Great
Plains, and the Pacific Northwest (McCarty 2011). The primary crops that are managed with residue
burning include corn, cotton, lentils, rice, soybeans, sugarcane and wheat (McCarty 2009). In 2023, CH4
and NzO emissions from field burning of agricultural residues were 0.6 MMT COz Eq. (22 kt) and 0.2 MMT
COz Eq. (1 kt), respectively (Table 5-29 and Table 5-30). Annual emissions of CH4 and NzO have
increased from 1990 to 2023 by 15 percent and 16 percent, respectively. The increase in emissions over
time is partly due to higher yielding crop varieties with larger amounts of residue production and fuel
loads, but also linked with an increase in the area burned for some crop types.
Table 5-29: CH4 and N20 Emissions from Field Burning of Agricultural Residues (MMT
C02 Eq.)
Gas/Crop Type
1990
2005
2019
2020
2021
2022
2023
ch4
0.5
0.6
0.7
0.6
0.6
0.6
0.6
Sugarcane
0.1
0.2
0.2
0.1
0.1
0.1
0.1
Wheat
0.2
0.2
0.1
0.1
0.1
0.1
0.1
Maize
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Rice
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Soybeans
+
+
+
+
+
+
+
Cotton
+
+
+
+
+
+
+
Sorghum
+
+
+
+
+
+
+
Other Small Grains
+
+
+
+
+
+
+
Peanuts
+
+
+
+
+
+
+
Legume Hay
+
+
+
+
+
+
+
Barley
+
+
+
+
+
+
+
Oats
+
+
+
+
+
+
+
Grass Hay
+
+
+
+
+
+
+
Tobacco
+
+
+
+
+
+
+
Sunflower
+
+
+
+
+
+
+
Peas
+
+
+
+
+
+
+
Vegetables
+
+
+
+
+
+
+
Potatoes
+
+
+
+
+
+
+
Dry Beans
+
+
+
+
+
+
+
Agriculture 5-63
-------
Gas/Crop Type
1990
2005
2019
2020
2021
2022
2023
Sugarbeets
+
+
+
+
+
+
+
Lentils
+
+
+
+
+
+
+
Chickpeas
+
+
+
+
+
+
+
n2o
0.2
0.2
0.2
0.2
0.2
0.2
0.2
Wheat
0.1
0.1
+
+
+
+
+
Maize
+
+
+
+
+
+
+
Sugarcane
+
+
+
+
+
+
+
Rice
+
+
+
+
+
+
+
Soybeans
+
+
+
+
+
+
+
Cotton
+
+
+
+
+
+
+
Peanuts
+
+
+
+
+
+
+
Other Small Grains
+
+
+
+
+
+
+
Legume Hay
+
+
+
+
+
+
+
Sorghum
+
+
+
+
+
+
+
Grass Hay
+
+
+
+
+
+
+
Barley
+
+
+
+
+
+
+
Oats
+
+
+
+
+
+
+
Potatoes
+
+
+
+
+
+
+
Peas
+
+
+
+
+
+
+
Tobacco
+
+
+
+
+
+
+
Sugarbeets
+
+
+
+
+
+
+
Sunflower
+
+
+
+
+
+
+
Vegetables
+
+
+
+
+
+
+
Dry Beans
+
+
+
+
+
+
+
Lentils
+
+
+
+
+
+
+
Chickpeas
+
+
+
+
+
+
+
Total
0.7
0.8
0.9
0.8
0.8
0.8
0.8
+ Does not exceed 0.05 MMT C02 Eq.
Notes: Totals by gas may not sum due to independent rounding. Crops are listed in descending order based on the magnitude of
emissions for the most recent year of available data.
Table 5-30: CH4, N20, CO, and NOx Emissions from Field Burning of Agricultural
Residues (kt)
Gas/Crop Type
1990
2005
2019
2020
2021
2022
2023
ch4
19
23
23
22
22
22
22
Sugarcane
4
6
6
5
5
5
5
Wheat
6
6
5
5
5
5
5
Maize
2
4
5
5
5
5
5
Rice
3
3
3
2
3
3
3
Soybeans
1
2
2
2
2
2
2
Cotton
1
2
1
1
1
1
1
Sorghum
+
+
+
+
+
+
+
Other Small Grains
+
+
+
+
+
+
+
Peanuts
+
+
+
+
+
+
+
Legume Hay
~
-
+
+
+
+
5-64 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Gas/Crop Type
1990
2005
2019
2020
2021
2022
2023
Barley
+
+
+
+
+
+
+
Oats
+
+
+
+
+
+
+
Grass Hay
+
+
+
+
+
+
+
Tobacco
+
+
+
+
+
+
+
Sunflower
+
+
+
+
+
+
+
Peas
+
+
+
+
+
+
+
Vegetables
+
+
+
+
+
+
+
Potatoes
+
+
+
+
+
+
+
Dry Beans
+
+
+
+
+
+
+
Sugarbeets
+
+
+
+
+
+
+
Lentils
+
+
+
+
+
+
+
Chickpeas
+
+
+
+
+
+
+
n2o
1
1
1
1
1
1
1
Wheat
+
+
+
+
+
+
+
Maize
+
+
+
+
+
+
+
Sugarcane
+
+
+
+
+
+
+
Rice
+
+
+
+
+
+
+
Soybeans
+
+
+
+
+
+
+
Cotton
+
+
+
+
+
+
+
Peanuts
+
+
+
+
+
+
+
Other Small Grains
+
+
+
+
+
+
+
Legume Hay
+
+
+
+
+
+
+
Sorghum
+
+
+
+
+
+
+
Grass Hay
+
+
+
+
+
+
+
Barley
+
+
+
+
+
+
+
Oats
+
+
+
+
+
+
+
Potatoes
+
+
+
+
+
+
+
Peas
+
+
+
+
+
+
+
Tobacco
+
+
+
+
+
+
+
Sugarbeets
+
+
+
+
+
+
+
Sunflower
+
+
+
+
+
+
+
Vegetables
+
+
+
+
+
+
+
Dry Beans
+
+
+
+
+
+
+
Lentils
+
+
+
+
+
+
+
Chickpeas
+
+
+
+
+
+
+
CO
407
480
468
446
445
443
443
NOx
16
18
18
17
17
17
17
+ Does not exceed 0.5 kt C02 Eq.
Notes: Totals may not sum due to independent rounding. Crops are listed in descending order based on the magnitude of
emissions for the most recent year of available data.
Methodology and Time-Series Consistency
A country-specific Tier 2 method is used to estimate greenhouse gas emissions from field burning of
agricultural residues from 1990 to 2014 (for more details comparing the country-specific approach to
the IPCC (2006) default approach, see Box 5-5), and a data splicing method using linear regression with
Agriculture 5-65
-------
auto-regressive moving average (ARMA) errors is applied to complete the emissions time series from
2015 to 2023. The exception is sugarcane for which emissions have been estimated from 1990 to 2020,
with 2021 through 2023 estimated with the data splicing method. The following equation is used to
estimate the amounts of carbon (C) and nitrogen (N) released (R„ where /' is C or N) from burning.
Equation 5-1: Elemental C or N Released through Oxidation of Crop Residues
Rt = CP x RCR x DMF x Ftx FB x CE
where,
Crop Production (CP)
Residue: Crop Ratio (RCR)
Dry Matter Fraction (DMF)
Fraction C or N (Ft)
Fraction Burned (FB)
Combustion Efficiency (CE)
Area Burned (AB)
Crop Area Harvested (CAH)
FB =
AB
CAH
Annual production of crop, by state, kt crop production
Amount of residue produced per unit of crop production, kt
residue/kt crop production
Amount of dry matter per unit of residue biomass for a crop, kt
residue dry matter/ kt residue biomass
Fraction of C or N per unit of dry matter for a crop, kt C or N /kt
residue dry matter
Proportion of residue biomass consumed, unitless
Proportion of residue actually burned, unitless
Total area of crop burned, by state, ha
Total area of crop harvested, by state, ha
Crop production data are available by state and year from USDA-NASS (2019) for 22 crops that are
burned in the conterminous United States, including maize, rice, wheat, barley, oats, other small grains,
sorghum, cotton, grass hay, legume hay, peas, sunflower, tobacco, vegetables, chickpeas, dry beans,
lentils, peanuts, soybeans, potatoes, sugarbeets, and sugarcane.32 Crop area data are based on the
2015 and 2017 National Resources Inventories (NRI) (USDA-NRCS 2018; USDA-NRCS 2020). To estimate
total crop production, the crop yield data from USDA Quick Stats (USDA-NASS 2019) are multiplied by
the area data for these crops from the NRI survey. The production data for the crop types are presented
in Table 5-31. Alaska and Hawaii are not included in the current analysis, but there is a planned
improvement to estimate residue burning emissions for these two states in a future Inventory.
The amount of elemental carbon or nitrogen released through oxidation of the crop residues is used in
the following equation to estimate the amount of CH4, CO, N20, and NOx emissions (Eg, where g is the
specific gas, i.e., CH4, CO, N20, and NOx) from the field burning of agricultural residues:
32 Kentucky bluegrass (produced on farms for turf grass installations) may have small areas of burning that are not captured
in the sample of locations that were used in the remote sensing analysis (see Planned Improvements).
5-66 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Equation 5-2: Emissions from Crop Residue Burning
Eg = X EFg x CF
where,
Emission ratio (EFg) = emission ratio by gas, g CH4-C or CO-C/g C released, or g N20-N or
NOx-N/g N released
Conversion Factor (CF) = conversion by molecular weight ratio of CH4-C to C (16/12), CO-C to C
(28/12), N20-N to N (44/28), or NOx-N to N (30/14)
Box 5-5: Comparison of Tier 2 U.S. Inventory Approach and IPCC (2006) Default
Approach
Emissions from Field Burning of Agricultural Residues are calculated using a Tier 2 methodology that is
based on the method developed by the IPCC/UNEP/OECD/IEA (1997). The rationale for using the
IPCC/UNEP/OECD/IEA (1997) approach rather than the method provided in the 2006 IPCC Guidelines is
as follows: (1) the equations from both guidelines rely on the same underlying variables (though the
formats differ); (2) the IPCC (2006) equation was developed to be broadly applicable to all types of
biomass burning, and, thus, is not specific to agricultural residues; (3) the IPCC (2006) method provides
emission factors based on the dry matter content rather than emission rates related to the amount of
carbon and nitrogen in the residues; and (4) the IPCC (2006) default factors are provided only for four
crops (corn, rice, sugarcane, and wheat) while this Inventory includes emissions from twenty-one crops.
A comparison of the methods in the current Inventory and the default IPCC (2006) approach was
undertaken for 2014 to determine the difference in estimates between the two approaches. To estimate
greenhouse gas emissions from field burning of agricultural residues using the IPCC (2006)
methodology, the following equation—cf. IPCC (2006) Equation 2.27—was used with default factors and
country-specific values for mass of fuel.
Equation 5-3: Estimation of Greenhouse Gas Emissions from Fire
Emissions (kt) = AB x MB x Cf x Gef x 10
where,
Area Burned (AB)
Mass of Fuel (MB)
Total area of crop burned (ha)
U.S.- Specific Values using NASS Statistics33 (metric tons dry
matter)
IPCC (2006) default combustion factor with fuel biomass
consumption (metric tons dry matter ha-1)
IPCC (2006) emission factor (g kg-1 dry matter burnt)
The IPCC (2006) Tier 1 method approach resulted in 21 percent lower emissions of CH4 and 40 percent
lower emissions of N20 compared to this Inventory. In summary, the IPCC/UNEP/OECD/IEA (1997)
method is considered more appropriate for U.S. conditions because it is more flexible for incorporating
country-specific data. Emissions are estimated based on specific carbon and nitrogen content of the
Combustion Factor (Cf)
Emission Factor (Gef)
33 NASS yields are used to derive mass of fuel values because IPCC (2006) only provides default values for 4 of the 21
crops included in the Inventory.
Agriculture 5-67
-------
fuel, which is converted into CH4, CO, N20 and NOx, compared to IPCC (2006) approach that is based on
dry matter rather than elemental composition.
Table 5-31: Agricultural Crop Production (kt of Product)
Crop
1990
2005
2010
2018
2019
2020
Maize
296,065
371,256
398,618
NE
NE
NE
Rice
9,543
11,751
11,976
NE
NE
NE
Wheat
79,805
68,077
68,530
NE
NE
NE
Barley
9,281
5,161
3,942
NE
NE
NE
Oats
5,969
2,646
2,364
NE
NE
NE
Other Small Grains
2,651
2,051
1,803
NE
NE
NE
Sorghum
23,687
14,382
14,052
NE
NE
NE
Cotton
4,605
6,106
4,638
NE
NE
NE
Grass Hay
44,150
49,880
46,761
NE
NE
NE
Legume Hay
90,360
91,819
85,813
NE
NE
NE
Peas
51
660
839
NE
NE
NE
Sunflower
1,015
1,448
1,212
NE
NE
NE
Tobacco
1,154
337
470
NE
NE
NE
Vegetables
+
1,187
1,469
NE
NE
NE
Chickpeas
+
5
+
NE
NE
NE
Dry Beans
467
1,143
1,461
NE
NE
NE
Lentils
+
101
254
NE
NE
NE
Peanuts
1,856
2,176
1,925
NE
NE
NE
Soybeans
56,612
86,980
95,198
NE
NE
NE
Potatoes
18,924
20,026
19,279
NE
NE
NE
Sugarbeets
24,951
25,635
33,336
NE
NE
NE
Sugarcane
26,047
38,928
34,252
36,680
37,361
42,400
+ Absolute value does not exceed 0.05 MMT C02Eq.
NE (Not Estimated)
Note: The amount of crop production has not been compiled for 2015 to 2021 so a data splicing method is used to estimate
emissions for this portion of the time series.
The area burned is determined based on an analysis of remote sensing products (McCarty et al. 2009,
2010, 2011). The presence of fires has been analyzed at 3,600 survey locations in the NRI from 1990 to
2002 with LANDFIRE data products developed from 30 m Landsat imagery (LANDFIRE 2008), and from
2003 through 2014 using 1 km Moderate Resolution Imaging Spectroradiometer imagery (MODIS) Global
Fire Location Product (MCD14ML), combining observations from Terra and Aqua satellites (Giglio et al.
2006). A sample of states are included in the analysis with high, medium and low burning rates for
agricultural residues, including Arkansas, California, Florida, Indiana, Iowa and Washington. The area
burned is determined directly from the analysis for these states for all crops, with the exception of
sugarcane as discussed later in this section.
For other states within the conterminous United States, the area burned for the 1990 through 2014
portion of the time series is estimated from a logistical regression model that has been developed from
the data collected from the remote sensing products for the six states. The logistical regression model is
used to predict occurrence of fire events. Several variables are tested in the logistical regression
5-68 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
including a) the historical level of burning in each state (high, medium or low levels of burning) based on
an analysis by McCarty et al. (2011), b) year that state laws limit burning of fields, in addition to c) mean
annual precipitation and mean annual temperature from a 4- kilometer gridded product from the PRISM
Climate Group (2015). A K-fold model fitting procedure is used due to low frequency of burning and
likelihood that outliers could influence the model fit. Specifically, the model is trained with a random
selection of sample locations and evaluated with the remaining sample. This process is repeated ten
times to select a model that is most common among the set often, and avoid models that appear to be
influenced by outliers due to the random draw of survey locations for training the model. In order to
address uncertainty, a Monte Carlo analysis is used to sample the parameter estimates for the logistical
regression model and produce one thousand estimates of burning for each crop in the remaining forty-
two states included in this Inventory. State-level area burned data are divided by state-level crop area
data to estimate the percent of crop area burned by crop type for each state. Table 5-32 shows the
resulting percentage of crop residue burned at the national scale by crop type. State-level estimates are
also available upon request.
Table 5-32: U.S. Average Percent Crop Area Burned by Crop (Percent)
Crop
1990
2005
2010
2018
2019
2020
Maize
+
+
+
NE
NE
NE
Rice
12%
11 %
12%
NE
NE
NE
Wheat
3%
3%
2%
NE
NE
NE
Barley
1%
1%
1%
NE
NE
NE
Oats
1%
1%
1%
NE
NE
NE
Other Small Grains
5%
4%
4%
NE
NE
NE
Sorghum
1%
1%
1%
NE
NE
NE
Cotton
7%
10%
9%
NE
NE
NE
Grass Hay
+
+
+
NE
NE
NE
Legume Hay
+
+
+
NE
NE
NE
Peas
1%
1%
1%
NE
NE
NE
Sunflower
+
+
+
NE
NE
NE
Tobacco
1%
1%
1%
NE
NE
NE
Vegetables
+
+
+
NE
NE
NE
Chickpeas
+
+
+
NE
NE
NE
Dry Beans
+
+
+
NE
NE
NE
Lentils
+
1%
1%
NE
NE
NE
Peanuts
5%
5%
5%
NE
NE
NE
Soybeans
1%
1%
1%
NE
NE
NE
Potatoes
+
+
+
NE
NE
NE
Sugarbeets
+
NE
NE
NE
Sugarcane
6%
5%
6%
4%
6%
4%
+ Does not exceed 0.5 percent.
NE (Not Estimated)
The method for estimating burned area of sugarcane is similar to the approach for other crops. Areas
with sugarcane production are identified in the 2017 USDA NRI survey (USDA-NRCS 2020) based on
Agriculture 5-69
-------
Cropland Data Layer (USDA-NASS 2021).34 EPA uses the MODIS burned area product from 2002 to 2020
to identify NRI survey locations with sugarcane production that have residue burning, similar to the
process for other crops described above (Giglio et al. 2015). However, area of residue burning for
sugarcane was estimated for 1990 to 2001 using a linear extrapolation of the area burned from 2002 to
2020, instead of analyzing the remote sensing data for this portion of the time series. This approach is a
common data splicing method for filling data gaps in time series (IPCC 2006).
Additional parameters are needed to estimate emissions from the area that has residue burning,
including residue: crop ratios, dry matter fractions, carbon fractions, nitrogen fractions and combustion
efficiency. Residue: crop product mass ratios, residue dry matter fractions, and the residue N contents
are obtained from several sources (IPCC 2006 and sources at bottom of Table 5-33). The residue carbon
contents for all crops are based on IPCC (2006) default value for herbaceous biomass. The combustion
efficiency is assumed to be 90 percent for all crop types (IPCC/UNEP/OECD/IEA 1997). See Table 5-33
for a summary of the crop-specific conversion factors. Emission ratios and mole ratio conversion factors
for all gases are based on the Revised 1996 IPCC Guidelines (IPCC/UNEP/OECD/IEA 1997) (see Table
5-34).
Table 5-33: Parameters for Estimating Emissions from Field Burning of Agricultural
Residues
Residue/
Dry Matter
Carbon
Nitrogen
Combustion
Crop
Crop Ratio
Fraction
Fraction
Fraction
Efficiency (Fraction)
Maize
0.707
0.56
0.47
0.01
0.90
Rice
1.340
0.89
0.47
0.01
0.90
Wheat
1.725
0.89
0.47
0.01
0.90
Barley
1.181
0.89
0.47
0.01
0.90
Oats
1.374
0.89
0.47
0.01
0.90
Other Small Grains
1.777
0.88
0.47
0.01
0.90
Sorghum
0.780
0.60
0.47
0.01
0.90
Cotton
7.443
0.93
0.47
0.01
0.90
Grass Hay
0.208
0.90
0.47
0.02
0.90
Legume Hay
0.290
0.67
0.47
0.01
0.90
Peas
1.677
0.91
0.47
0.01
0.90
Sunflower
1.765
0.88
0.47
0.01
0.90
Tobacco
0.300
0.87
0.47
0.01
0.90
Vegetables
0.708
0.08
0.47
0.01
0.90
Chickpeas
1.588
0.91
0.47
0.01
0.90
Dry Beans
0.771
0.90
0.47
0.01
0.90
Lentils
1.837
0.91
0.47
0.02
0.90
Peanuts
1.600
0.94
0.47
0.02
0.90
Soybeans
1.500
0.91
0.47
0.01
0.90
Potatoes
0.379
0.25
0.47
0.02
0.90
Sugarbeets
0.196
0.22
0.47
0.02
0.90
Sugarcane
0.410
0.25
0.47
0.02
0.90
34 USDA-NRI program aggregates sugarcane with other crops, but areas planted with sugarcane are identified in the USDA-
NASS Crop Data Layer.
5-70 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Cotton: Combined sources (Heitholt et a 1.1992; Halevy 1976; Wells and Meredith 1984; Sadras and Wilson 1997; Pettigrewand
Meredith 1997; Torbert and Reeves 1994; Gerik et a 1.1996; Brouder and Cassmen 1990; Fritschi et al. 2003; Pettigrewet a I.
2005; Bouquet and Breitenbeck2000; Mahroni and Aharonov 1964; Bange and Milroy 2004; Hollifield et al. 2000; Mondino et
al. 2004; Wallach etal. 1978).
Lentils: IPCC (2006), Table 11.2; Beans & pulses.
Peas: IPCC (2006), Table 11.2; values are for Beans & pulses.
Peanuts: IPCC (2006); Table 11.2; Root ratio and belowground N content values are for Root crops, other.
Sugarbeets: IPCC (2006); Table 11.2; values are for Tubers.
Sunflower: IPCC (2006), Table 11.2; values are for Grains.
Sugarcane: combined sources (Wiedenfels 2000, Dua and Sharma 1976; Singels & Bezuidenhout 2002; Stirling etal. 1999;
Sitompul et al. 2000).
Tobacco: combined sources (Beyaert 1996; Moustakas and Ntzanis 2005; Crafts-Brandner et al. 1994; Hopkinson 1967; Crafts-
Brandner et al. 1987).
Vegetables (Combination of carrots, lettuce/cabbage, melons, onions, peppers and tomatoes):
Carrots: McPharlin etal. (1992); Gibberd et al. (2003); Reid and English (2000); Peach etal. (2000); see IPCC Tubers for R:S and N
fraction.
Lettuce, cabbage: combined sources (Huett and Dettman 1991; De Pinheiro Henriques & Marcelis 2000; Huett and Dettman
1989; Peach etal. 2000; Kage etal. 2003; Tan et al. 1999; Kumar et al. 1994; MacLeod et al. 1971; Jacobs et al. 2004; Jacobs et
al. 2001; Jacobs et al. 2002); values from IPCC Grains used for N fraction.
Melons: Valantin et al. (1999); squash for R:S; IPCC Grains for N fraction.
Onion: Peach et al. (2000), Halvorson et al. (2002); IPCC (2006) Tubers for N fraction.
Peppers: combined sources (Costa and Gianquinto 2002; Marcussi et al. 2004; Tadesse et al. 1999; Diaz-Perez et al. 2008); IPCC
Grains for N fraction.
Tomatoes: Scholberg et al. (2000a,b); Akintoye et al. (2005); values for AGR-N and BGR-N are from Grains.
Notes: Chickpeas: IPCC (2006), Table 11.2; values are for Beans & pulses.
Table 5-34: Greenhouse Gas Emission Ratios and Conversion Factors
Gas Emission Ratio Conversion Factor
CH4:C 0.005a 16/12
CO:C 0.060s 28/12
N20:N 0.007b 44/28
NOx:N 0.121b 30/14
a Mass of C compound released (units of C) relative to mass of total C released from burning (units of C).
b Mass of N compound released (units of N) relative to mass of total N released from burning (units of N).
To ensure time-series consistency, the same method is applied from 1990 to 2014 because new activity
data on the burned areas have not been analyzed for 2015 to 2023 for individual crops. The exception is
sugarcane in which burned areas have not been analyzed for 2021 to 2023. To complete the emissions
time series, a linear regression model with autoregressive moving-average (ARMA) errors is used to
estimate the trend in emissions over time from 1990 through 2014, and the trend is used to approximate
the CH4, N20, CO and NOx emissions (Brockwell and Davis 2016). This method is consistent with data
splicing methods described in IPCC (2006). The Tier 2 method described previously will be applied to
recalculate emissions for the latter part of the time series in a future Inventory.
Uncertainty
Emissions are estimated using a linear regression model with autoregressive moving-average (ARMA)
errors for 2023. The linear regression ARMA model produced estimates of the upper and lower bounds to
quantify uncertainty, and the results are summarized in Table 5-35. Methane emissions from field
burning of agricultural residues in 2023 are between 0.53 and 0.71 MMT C02 Eq. at a 95 percent
confidence level. This indicates a range of 14 percent below and 14 percent above the 2023 emission
estimate of 0.6 MMT C02 Eq. Nitrous oxide emissions are between 0.17 and 0.24 MMT C02 Eq., or
approximately 18 percent below and 18 percent above the 2023 emission estimate of 0.2 MMT C02 Eq.
Agriculture 5-71
-------
Table 5-35: Approach 2 Quantitative Uncertainty Estimates for CH4 and N20 Emissions
from Field Burning of Agricultural Residues (MMT C02 Eq. and Percent)
2023
Emission
Estimate
Source Gas (MMT CO2 Eq.)
Uncertainty Range Relative to Emission Estimate"
(MMT CO2 Eq.)
(%)
Lower Upper
Bound Bound
Lower Upper
Bound Bound
Field Burning of ^ Q 6
Agricultural Residues
Field Burning of _
N2O 0.2
Agricultural Residues
0.53 0.71
0.17 0.24
-14% +14%
-18% +18%
a Range of emission estimates predicted by Monte Carlo stochastic simulation for a 95 percent confidence interval.
Due to data limitations, there are additional uncertainties in agricultural residue burning, particularly the
potential omission of burning associated with Kentucky bluegrass (produced on farms for turf grass
installation).
QA/QC and Verification
A source-specific QA/QC plan for field burning of agricultural residues is implemented with Tier 1
analyses, consistent with the U.S. Inventory QA/QC plan outlined in Annex 8. Quality control measures
included checking input data, model scripts, and results to ensure data are properly handled throughout
the inventory process. Inventory forms and text are reviewed and revised as needed to correct
transcription errors. An error was identified in the application of the data splicing methods to estimate
total CO and NOx emissions for sugarcane. This error was corrected by using separate ARMA models to
estimate CO and NOx emissions for sugarcane from 2021 to 2023.
Recalculations Discussion
Recalculations have been conducted for this Inventory by altering the way that sugarcane emissions are
estimated for 2021 to 2023. Previously, a linear extrapolation method was used for sugarcane CO and
NOx emissions. This year, a linear regression method with auto-regressive moving average (ARMA) errors
was implemented to estimate sugarcane emissions, consistent with the other crops. As a result of this
change, total CO emissions decreased by 35.4 kt, or 7.4 percent, for 2021 and by 58 kt, or 11.6 percent,
for 2022 compared to the previous Inventory. In addition, NOx emissions decreased by 1.4 kt, or 7.5
percent, for 2021, and by 2.3 kt, or 12 percent, for 2022 compared to the previous Inventory. Precursors
from sectors included in this report are further summarized and reported in Chapter 2.3.
Planned Improvements
The following are long-term planned improvements for Field Burning of Agricultural Residues:
Linking agricultural residue burning with the Tier 3 methods that are used in several other source
categories, as described in the Planned Improvement section for Agricultural Soil Management.
This method is based on simulating burning events directly within the DayCent process-based
model framework using information derived from remote sensing fire products as described in
the Methodology chapter section.
5-72 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Land Use, Land-Use
Change, and Forestry
(LULUCF)
-------
6 Land Use, Land Use Change and
Forestry
This chapter provides an assessment of the greenhouse gas fluxes resulting from land use and land-use
change in the United States.1 The Intergovernmental Panel on Climate Change's 2006IPCC Guidelines
for National Greenhouse Gas Inventories (IPCC 2006) recommends quantifying fluxes according to
changes within and conversions between all land use types including: forest land, cropland, grassland,
wetlands, and settlements (as well as other land).
The greenhouse gas flux from forest land remaining forest land is reported for all forest ecosystem
carbon (C) pools (i.e., aboveground biomass, belowground biomass, dead wood, litter, and mineral and
organic soils), harvested wood pools, and non-C02 emissions from forest fires, the application of
synthetic nitrogen fertilizers to forest soils, and the draining of organic soils. Fluxes from land converted
to forest land are included for aboveground biomass, belowground biomass, dead wood, litter, and
carbon stock changes from mineral soils, while carbon stock changes from drained organic soils and all
non-C02 emissions from land converted to forest land are included in the fluxes from forest land
remaining forest land as it is not currently possible to separate these fluxes by conversion category (e.g.,
grassland converted to forestland).
Fluxes are reported for four agricultural land use/land-use change categories: cropland remaining
cropland, land converted to cropland, grassland remaining grassland, and land converted to grassland.
The reported greenhouse gas fluxes from these agricultural lands include changes in soil organic carbon
stocks in mineral and organic soils due to land use and management, and for the subcategories of forest
land converted to cropland and forest land converted to grassland, the changes in aboveground
biomass, belowground biomass, dead wood, and litter carbon stocks are also reported. The greenhouse
gas flux from grassland remaining grassland also includes estimates of non-C02 emissions from
grassland fires occurring on both grassland remaining grassland and land converted to grassland.
Fluxes from wetlands remaining wetlands include changes in carbon stocks and methane (CH4) and
nitrous oxide (N20) emissions from managed peatlands, aboveground and belowground biomass, dead
organic matter, soil carbon stock changes and CH4 emissions from coastal wetlands, as well as N20
emissions from aquaculture. In addition, CH4 emissions from reservoirs and other constructed
waterbodies are included for the subcategory flooded land remaining flooded land. Estimates for land
converted to wetlands include aboveground and belowground biomass, dead organic matter and soil
carbon stock changes, and CH4emissions from land converted to vegetated coastal wetlands. Carbon
Dioxide (C02) and CH4 emissions are included for reservoirs and other constructed waterbodies under
the subcategory land converted to flooded land. See Section 6.1 for additional information on wetlands
included in this Inventory.
Fluxes from settlements remaining settlements include changes in carbon stocks from organic soils,
N20 emissions from nitrogen fertilizer additions to soils, and C02fluxes from settlement trees and
landfilled yard trimmings and food scraps. The reported greenhouse gas flux from land converted to
1 The term "flux" is used to describe the exchange of CO2 to and from the atmosphere, with net flux of CO2 being either
positive or negative depending on the overall balance. Removal and long-term storage of CO2 from the atmosphere is also
referred to as "carbon sequestration."
6-2 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
settlements includes changes in carbon stocks in mineral and organic soils due to land use and
management for all land-use conversions to settlements, and the carbon stock changes in aboveground
biomass, belowground biomass, dead wood, and litter are also included for the subcategory forest land
converted to settlements.
In 2023, the Land Use, Land-Use Change, and Forestry (LULUCF) sector resulted in a net increase in
carbon stocks (i.e., net C02 removals) of 1,000.5 MMT C02 Eq. This represents an offset of
approximately 15.2 percent of total (i.e., gross) greenhouse gas emissions in 2023. Emissions of CH4and
N20 from LULUCF activities in 2023 were 54.7 and 5.9 MMT C02 Eq., respectively, and combined
represent 1.0 percent of total greenhouse gas emissions.3 In 2023, the overall net flux from LULUCF
resulted in a removal of 939.9 MMT C02 Eq. Emissions, removals and net greenhouse gas flux from
LULUCF are summarized in Figure 6-1 and Table 6-1 by land use and category, and Table 6-2 and Table
6-3 by gas in MMT C02 Eq. and kt, respectively. Trends in LULUCF sources and sinks over the 1990 to
2023 time series are shown in Figure 6-2.
Existing forests, including forest land remaining forest land and settlement trees, account for
approximately 75.4 percent of the total LULUCF carbon stock change.
Flooded land remaining flooded land was the largest source of non-C02 emissions from LULUCF in
2023, accounting for 75.6 percent of the LULUCF sector non-C02 emissions. Non-C02 emissions from
forest fires are the second largest source of LULUCF sector emissions, accounting for 10.3 percent of
LULUCF non-C02emissions in 2023; and have increased 14.8 percent since 1990.
2 LULUCF carbon stock change is the net carbon stock change from the following categories: forest land remaining forest
land, land converted to forest land, cropland remaining cropland, land converted to cropland, grassland remaining
grassland, land converted to grassland, wetlands remaining wetlands, land converted to wetlands, settlements
remaining settlements, and land converted to settlements.
3 LULUCF emissions include the CFU and N2O emissions reported for peatlands remaining peatlands, forest fires, drained
organic soils, grassland fires, and coastal wetlands remaining coastal wetlands; CFU emissions from land converted to
coastal wetlands, flooded land remaining flooded land, and land converted to flooded land; and N2O emissions from
forest soils and settlement soils.
Land Use, Land-Use Change, and Forestry 6-3
-------
Figure 6-1: 2023 LULUCF Chapter Greenhouse Gas Sources and Sinks
Forest Land Remaining Forest Land
Settlements Remaining Settlements
Land Converted to Forest Land
Cropland Remaining Cropland
Wetlands Remaining Wetlands
Non-CCh Emissions from Peatlands Remaining Peatlands
Non-CCh Emissions from Drained Organic Soils
Cm Emissions from Land Converted to Coastal Wetlands
CH4 Emissions from Land Converted to Flooded Land
Land Converted to Wetlands
N2O Emissions from Forest Soils
Non-CCh Emissions from Grassland Fires
N2O Emissions from Settlement Soils
Non-CCh Emissions from Coastal Wetlands Remaining Coastal Wetlands
Non-CCh Emissions from Forest Fires
Land Converted to Grassland
Grassland Remaining Grassland
Land Converted to Cropland
Non-CCh Emissions from Flooded Land Remaining Flooded Land
Land Converted to Settlements
(880)
I Carbon Stock Change
! Non-CCh Emissions
l< 0-51
< 0.5|
I< 0.51
]< 0.51
l< 0.51
l< 0.51
(250) (200) (150) (100) (50)
MMT CCh Eq.
0 50 100
Note: Parentheses in horizontal axis indicate net sequestration.
Figure 6-2: Trends in Emissions and Removals (Net C02 Flux) from Land Use, Land-Use
Change, and Forestry
600
I Forest Land Remaining Forest Land
Land Converted to Forest Land
1 Land Converted to Cropland
Land Converted to Wetlands
I Settlements Remaining Settlements I
Cropland Remaining Cropland
Grassland Remaining Grassland
Land Converted to Grassland
l Land Converted to Settlements
Wetlands Remaining Wetlands
¦ Net Emissions (Sources and Sinks)
400
O
u
co Ln co
"" s s
s
200
minim
S
s J
CT\ rr> VV
cm cn in
£ b\ P\ ffi £
ai m oi
-200
-400
-600
¦¦¦¦Hi
-800
-1,000
-1,200
o^rMro-^-Ln^Df^oDaio
cr> cn cn cn en cn cn cn o~> o
i—I rH rH t—I t-H i—i y—I i—l rH 1—1 (NJ
OOOOOOOOOi-l'-H
ooooooooooo
NMfMNfMrNfM^rNJNfM
rMro^-LnvOfxcocriO'-HrMm
oooooooooooo
(NNrJfNrNJfMNNfMfNrMlN
6-4 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Table 6-1: Emissions and Removals (Net Flux) from Land Use, Land-Use Change, and
Forestry (MMT C02 Eq.)
Land-Use Category
1990
2005
2019
2020
2021
2022
2023
Forest Land Remaining Forest Land
(1049.3)
(932.8)
(867.4)
(898.0)
(881.0)
(827.6)
(873.3)
Changes in Forest Carbon Stocks®
(1054.9)
(950.0)
(877.1)
(926.5)
(907.9)
(842.4)
(880.0)
Non-C02 Emissions from Forest Firesb
5.4
16.7
9.3
28.0
26.4
14.3
6.2
N2O Emissions from Forest Soils0
0.1
0.4
0.4
0.4
0.4
0.4
0.4
Non-C02 Emissions from Drained
Organic Soilsd
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Land Converted to Forest Land
(103.6)
(103.6)
(103.9)
(103.8)
(103.8)
(103.8)
(103.8)
Changes in Forest Carbon Stocks®
(103.6)
(103.6)
(103.9)
(103.8)
(103.8)
(103.8)
(103.8)
Cropland Remaining Cropland
1.0
(31.0)
(19.3)
(8.7)
(31.9)
(31.6)
(30.5)
Changes in Mineral and Organic Soil
Carbon Stocks
1.0
(31.0)
(19.3)
(8.7)
(31.9)
(31.6)
(30.5)
Land Converted to Cropland
48.5
35.5
31.4
29.2
34.9
35.0
35.6
Changes in all Ecosystem Carbon
Stocks'
48.5
35.5
31.4
29.2
34.9
35.0
35.6
Grassland Remaining Grassland
24.2
24.5
28.5
16.8
11.2
13.7
22.7
Changes in Mineral and Organic Soil
Carbon Stocks
24.0
23.7
28.2
15.8
10.2
13.1
22.0
Non-C02 Emissions from Grassland
Fires8
0.2
0.8
0.3
1.1
0.9
0.6
0.7
Land Converted to Grassland
35.6
21.9
20.9
24.1
19.9
20.9
20.9
Changes in all Ecosystem Carbon
Stocks'
35.6
21.9
20.9
24.1
19.9
20.9
20.9
Wetlands Remaining Wetlands
38.5
40.9
39.7
39.7
39.7
39.7
39.7
Changes in Organic Soil Carbon Stocks
in Peatlands
1.1
1.1
0.6
0.6
0.5
0.6
0.6
Non-C02 Emissions from Peatlands
Remaining Peatlands
+
+
+
+
+
+
+
Changes in Biomass, DOM, and Soil
Carbon Stocks in Coastal Wetlands
(10.8)
(10.1)
(11.1)
(11.1)
(11.1)
(11.1)
(11.1)
CH4 Emissions from Coastal Wetlands
Remaining Coastal Wetlands
4.2
4.2
4.3
4.3
4.3
4.3
4.3
N2O Emissions from Coastal Wetlands
Remaining Coastal Wetlands
0.1
0.2
0.1
0.1
0.1
0.1
0.1
CH4 Emissions from Flooded Land
Remaining Flooded Land
43.9
45.5
45.8
45.8
45.8
45.8
45.8
Land Converted to Wetlands
6.8
1.9
0.7
0.7
0.7
0.7
0.6
Changes in Biomass, DOM, and Soil
Carbon Stocks in Land Converted to
Coastal Wetlands
0.5
0.5
{+)
{+)
{+)
{+)
{+)
CH4 Emissions from Land Converted to
Coastal Wetlands
0.3
0.3
0.2
0.2
0.2
0.2
0.2
Changes in Land Converted to Flooded
Land
3.4
0.7
0.3
0.3
0.3
0.3
0.3
CH4 Emissions from Land Converted to
Flooded Land
2.7
0.5
0.2
0.2
0.2
0.2
0.2
Land Use, Land-Use Change, and Forestry 6-5
-------
Land-Use Category
1990
2005
2019
2020
2021
2022
2023
Settlements Remaining Settlements
(109.1)
(115.2)
(131.4)
(131.7)
(132.1)
(132.1)
(131.7)
Changes in Organic Soil Carbon Stocks
9.9
10.1
14.6
15.1
15.6
16.0
16.4
Changes in Settlement Tree Carbon
Stocks
(96.5)
(117.0)
(135.4)
(136.6)
(137.6)
(138.4)
(139.0)
N2O Emissions from Settlement Soilsh
2.1
3.1
2.5
2.5
2.5
2.5
2.5
Changes in Yard Trimming and Food
Scrap Carbon Stocks in Landfills
(24.5)
(11.4)
(13.1)
(12.8)
(12.5)
(12.3)
(11.7)
Land Converted to Settlements
69.5
89.0
81.4
80.3
79.7
79.8
79.8
Changes in all Ecosystem Carbon
Stocks'
69.5
89.0
81.4
80.3
79.7
79.8
79.8
LULUCF Emissions'
59.1
71.8
63.2
82.6
81.0
68.6
60.6
CH4
54.4
60.9
56.1
69.0
67.8
59.6
54.7
N2O
4.7
10.9
7.0
13.7
13.1
9.0
5.9
LULUCF Carbon Stock Change1
(1,096.9) |
(1,040.7)
(982.6)
(1,034.2)
(1,043.8)
(973.9)
(1,000.5)
LULUCF Sector Net Total"
(1,037.9)
(968.9)
(919.4)
(951.6)
(962.9)
(905.3)
(939.9)
+ Absolute value does not exceed 0.05 MMT C02 Eq.
a Includes the net changes to carbon stocks stored in all forest ecosystem pools (estimates include carbon stock changes from
drained organic soils from both forest land remaining forest land and land converted to forest land) and harvested wood
products.
b Estimates include ChUand N20 emissions from fires on both forest land remaining forest land and land converted to forest
land.
c Estimates include N20 emissions from N fertilizer additions on both forest land remaining forest land and land converted to
forest land.
d Estimates include ChUand N20 emissions from drained organic soils on both forest land remaining forest land and land
converted to forest land. Carbon stock changes from drained organic soils are included with the forest land remaining forest
land forest ecosystem pools.
e Includes the net changes to carbon stocks stored in all forest ecosystem pools.
' Includes changes in mineral and organic soil carbon stocks for all land-use conversions to cropland, grassland, and
settlements. Also includes aboveground/belowground biomass, dead wood, and litter carbon stock changes for conversion of
forest land to cropland, grassland, and settlements.
g Estimates include ChUand N20 emissions from fires on both grassland remaining grassland and land converted to grassland.
h Estimates include N20 emissions from nitrogen fertilizer additions on both settlements remaining settlements and land
converted to settlements because it is not possible to separate the activity data at this time.
LULUCF emissions include the ChUand N20 emissions reported for peatlands remaining peatlands, forest fires, drained organic
soils, grassland fires, and coastal wetlands remaining coastal wetlands; ChU emissions from land converted to coastal
wetlands, flooded land remainingflooded land, and land converted to flooded land; and N20 emissions from forest soils and
settlement soils.
' LULUCF carbon stock change includes any carbon stock gains and losses from all land use and land-use conversion categories.
kThe LULUCF sector net total is the net sum of all LULUCF ChUand N20 emissions to the atmosphere plus LULUCF net carbon
stock changes in units of MMT C02 Eq.
Notes: Totals may not sum due to independent rounding. Parentheses indicate net sequestration.
The carbon stock changes and emissions of CH4 and N20 from LULUCF are summarized in Table 6-2
(MMTCO2 Eq.) and Table 6-3 (kt).
Total net carbon sequestration in the LULUCF sector decreased by approximately 8.8 percent between
1990 and 2023. This decrease was primarily due to a decline in the rate of net carbon accumulation in
forest land, as well as an increase in emissions from land converted to settlements.4The declining
carbon sink in forest land is due to a combination of factors including an aging forest land base,
increasing C02 and non-C02 emissions from fires, and other increasing disturbances in some regions.
4 Carbon sequestration estimates are net figures. The carbon stock in a given pool fluctuates due to both gains and losses.
When losses exceed gains, the carbon stock decreases, and the pool acts as a source. When gains exceed losses, the
carbon stock increases, and the pool acts as a sink; also referred to as net carbon sequestration or removal.
6-6 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Flooded land remaining flooded land was the largest source of CH4 emissions from LULUCF in 2023,
totaling 45.8 MMT C02 Eq. (1,636 kt of CH4). In 2023, coastal wetlands remaining coastal wetlands
resulted in CH4 emissions of 4.3 MMT C02 Eq. (155 kt of CH4). Forest fires resulted in CH4 emissions of
3.8 MMT C02 Eq. (135 kt of CH4).
The largest source of LULUCF N20 emissions in 2023 was fertilizer application to settlement soils,
totaling to 2.5 MMT C02 Eq. (10 kt of N20). This represents an increase of 23.1 percent since 1990. The
second largest source of LULUCF N20 emissions from forest fires in 2023 was 2.4 MMT C02 Eq. (9 kt
N20), an increase of 8.0 percent since 1990. Additionally, the application of synthetic fertilizers to forest
soils in 2023 resulted in N20 emissions of 0.4 MMT C02 Eq. (2 kt of N20). Nitrous oxide emissions from
fertilizer application to forest soils have increased by 455.1 percent since 1990 but still account for a
relatively small portion of overall emissions.
Table 6-2: Emissions and Removals from Land Use, Land-Use Change, and Forestry by
Gas (MMT C02 Eq.)
Gas/Land-Use Category
1990
2005
2019
2020
2021
2022
2023
Carbon Stock Change (C02)a
(1,096.9)
(1,040.7)
(982.6)
(1,034.2)
(1,043.8)
(973.9)
(1,000.5)
Forest Land Remaining Forest Land
(1,054.9)
(950.0)
(877.1)
(926.5)
(907.9)
(842.4)
(880.0)
Land Converted to Forest Land
(103.6)
(103.6)
(103.9)
(103.8)
(103.8)
(103.8)
(103.8)
Cropland Remaining Cropland
1.0
(31.0)
(19.3)
(8.7)
(31.9)
(31.6)
(30.5)
Land Converted to Cropland
48.5
35.5
31.4
29.2
34.9
35.0
35.6
Grassland Remaining Grassland
24.0
23.7
28.2
15.8
10.2
13.1
22.0
Land Converted to Grassland
35.6
21.9
20.9
24.1
19.9
20.9
20.9
Wetlands Remaining Wetlands
(9.8)
(8.98)
(10.51)
(10.54)
(10.58)
(10.54)
(10.53)
Land Converted to Wetlands
3.8
1.1
0.3
0.3
0.3
0.3
0.3
Settlements Remaining Settlements
(111.1)
(118.3)
(133.9)
(134.2)
(134.6)
(134.7)
(134.3)
Land Converted to Settlements
69.5
89.0
81.4
80.3
79.7
79.8
79.8
CH4
54.4
60.9
56.1
69.0
67.8
59.6
54.7
Forest Land Remaining Forest Land:
Forest Firesb
3.2
9.9
5.5
17.9
16.8
8.8
3.8
Forest Land Remaining Forest Land:
Drained Organic Soils0
+
+
+
+
+
+
+
Grassland Remaining Grassland:
Grassland Firesd
0.1
0.4
0.2
0.6
0.5
0.3
0.4
Wetlands Remaining Wetlands:
Flooded Land Remaining Flooded
Land
43.9
45.5
45.8
45.8
45.8
45.8
45.8
Wetlands Remaining Wetlands:
Coastal Wetlands Remaining Coastal
Wetlands
4.2
4.2
4.3
4.3
4.3
4.3
4.3
Wetlands Remaining Wetlands:
Peatlands Remaining Peatlands
+
+
+
+
+
+
+
Land Converted to Wetlands: Land
Converted to Flooded Lands
2.7
0.5
0.2
0.2
0.2
0.2
0.2
Land Converted to Wetlands: Land
Converted to Coastal Wetlands
0.3
0.3
0.2
0.2
0.2
0.2
0.2
N20
4.7
10.9
7.0
13.7
13.1
9.0
5.9
Forest Land Remaining Forest Land:
2.3
6.8
3.8
10.1
9.6
5.5
2.4
Land Use, Land-Use Change, and Forestry 6-7
-------
Gas/Land-Use Category
1990
2005
2019
2020
2021
2022
2023
Forest Firesb
Forest Land Remaining Forest Land:
Forest Soils®
0.1
0.4
0.4
0.4
0.4
0.4
0.4
Forest Land Remaining Forest Land:
Drained Organic Soils0
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Grassland Remaining Grassland:
Grassland Firesd
0.1
0.4
0.1
0.5
0.4
0.3
0.3
Wetlands Remaining Wetlands:
Coastal Wetlands Remaining Coastal
Wetlands
0.1
0.2
0.1
0.1
0.1
0.1
0.1
Wetlands Remaining Wetlands:
Peatlands Remaining Peatlands
+
+
+
+
+
+
+
Settlements Remaining Settlements:
Settlement Soils'
2.1
3.1
2.5
2.5
2.5
2.5
2.5
LULUCF Carbon Stock Change"
(1,096.9)
(1,040.7)
(982.6)
(1,034.2)
(1,043.8)
(973.9)
(1,000.5)
LULUCF Emissions9
59.1
71.8
63.2
82.6
81.0
68.6
60.6
LULUCF Sector Net Totalh
(1,037.9)
(968.9)
(919.4)
(951.6)
(962.9)
(905.3)
(939.9)
+ Absolute value does not exceed 0.05 MMT C02 Eq.
a LULUCF carbon stock change is the net carbon stock change from the following categories: forest land remaining forest land,
land converted to forest land, cropland remaining cropland, land converted to cropland, grassland remaining grassland, land
converted to grassland, wetlands remaining wetlands, land converted to wetlands, settlements remaining settlements, and
land converted to settlements.
b Estimates include ChUand N20 emissions from fires on both forest land remaining forest land and land converted to forest
land.
c Estimates include ChUand N20 emissions from drained organic soils on both forest land remaining forest land and land
converted to forest land.
d Estimates include ChUand N20 emissions from fires on both grassland remaining grassland and land converted to grassland.
e Estimates include N20 emissions from nitrogen fertilizer additions on both forest land remaining forest land and land
converted to forest land.
f Estimates include N20 emissions from nitrogen fertilizer additions on both settlements remaining settlements and land
converted to settlements.
g LULUCF emissions include the ChU and N20 emissions reported for peatlands remaining peatlands, forest fires, drained
organic soils, grassland fires, and coastal wetlands remaining coastal wetlands; ChU emissions from flooded land remaining
flooded land, land converted to flooded land, and land converted to coastal wetlands; and N20 emissions from forest soils and
settlement soils.
h The LULUCF sector net total is the net sum of all LULUCF ChUand N20 emissions to the atmosphere plus LULUCF net carbon
stock changes in units of MMT C02 Eq.
Notes: Totals may not sum due to independent rounding. Parentheses indicate net sequestration.
Table 6-3: Emissions and Removals from Land Use, Land-Use Change, and Forestry by
Gas (kt)
Gas/Land-Use Category
1990
2005
2019
2020
2021
2022
2023
Carbon Stock Change (C02)a
(1,096,934)
(1,040,673)
(982,581)
(1,034,188)
(1,043,812
(973,902)
(1,000,542)
Forest Land Remaining Forest
Land
(1,054,891)
(949,962)
(877,137)
(926,477)
(907,912)
(842,391)
(879,957)
Land Converted to Forest Land
(103,607)
(103,590)
(103,876)
(103,850)
(103,847)
(103,837)
(103,833)
Cropland Remaining Cropland
1,040
(31,040)
(19,330)
(8,729)
(31,896)
(31,566)
(30,541)
Land Converted to Cropland
48,519
35,486
31,430
29,219
34,898
35,025
35,584
Grassland Remaining Grassland
23,991
23,684
28,209
15,761
10,247
13,058
22,013
Land Converted to Grassland
35,553
21,950
20,857
24,065
19,883
20,945
20,856
Wetlands Remaining Wetlands
(9,770)
(8,984)
(10,509)
(10,535)
(10,582)
(10,543)
(10,531)
6-8 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Gas/Land-Use Category
1990
2005
2019
2020
2021
2022
2023
Land Converted to Wetlands
3850
1125
289
291
285
290
293
Settlements Remaining
Settlements
(111,129)
(118,296)
(133,891)
(134,217)
(134,592)
(134,663)
(134,273)
Land Converted to Settlements
69,509
88,955
81,378
80,284
79,705
79,779
79,847
CH4
1,943
2,174
2,005
2,464
2,423
2,129
1,953
Forest Land Remaining Forest
Land:
Forest Firesb
113
355
195
639
601
313
135
Forest Land Remaining Forest
Land: Drained Organic Soils0
1
1
1
1
1
1
1
Grassland Remaining Grassland:
Grassland Firesd
4
15
6
20
18
12
14
Wetlands Remaining Wetlands:
Flooded Land Remaining
Flooded Land
1,569
1,625
1,635
1,635
1,636
1,636
1,636
Wetlands Remaining Wetlands:
Coastal Wetlands Remaining
Coastal Wetlands
149
151
153
154
154
154
155
Wetlands Remaining Wetlands:
Peatlands Remaining Peatlands
+
+
+
+
+
+
+
Land Converted to Wetlands:
Land Converted to Flooded
Lands
96
17
8
8
7
7
7
Land Converted to Wetlands:
Land Converted to Coastal
Wetlands
10
10
7
7
6
6
6
N2O
18
41
27
52
49
34
22
Forest Land Remaining Forest
Land:
Forest Firesb
9
26
14
38
36
21
9
Forest Land Remaining Forest
Land: Forest Soils®
+
2
2
2
2
2
2
Forest Land Remaining Forest
Land: Drained Organic Soils0
+
+
+
+
+
+
+
Grassland Remaining Grassland:
Grassland Firesd
+
1
1
2
2
1
1
Wetlands Remaining Wetlands:
Coastal Wetlands Remaining
Coastal Wetlands
+
1
1
1
1
1
1
Wetlands Remaining Wetlands:
Peatlands Remaining Peatlands
+
+
+
+
+
+
+
Settlements Remaining
Settlements: Settlement Soils'
8
12
9
9
10
10
10
+ Absolute value does not exceed 0.5 kt.
a LULUCF carbon stock change is the net carbon stock change from the following categories: forest land remainingforest land,
land converted to forest land, cropland remaining cropland, land converted to cropland, grassland remaining grassland, land
converted to grassland, wetlands remaining wetlands, land converted to wetlands, settlements remaining settlements, and
land converted to settlements.
b Estimates include ChUand N20 emissions from fires on both forest land remainingforest land and land converted to forest
land.
c Estimates include ChUand N20 emissions from drained organic soils on both forest land remainingforest land and land
Land Use, Land-Use Change, and Forestry 6-9
-------
converted to forest land.
d Estimates include ChUand N20 emissions from fires on both grassland remaining grassland and land converted to grassland.
e Estimates include N20 emissions from nitrogen fertilizer additions on both forest land remaining forest land and land converted
to forest land.
f Estimates include N20 emissions from nitrogen fertilizer additions on both settlements remaining settlements and land
converted to settlements.
Notes: Parentheses indicate net sequestration. Totals may not sum due to independent rounding.
Each year, some emission and sink estimates in the LULUCF sector of the Inventory are recalculated
and revised with improved methods and/or data. In general, recalculations are made to the U.S.
greenhouse gas emissions and removals estimates either to incorporate new methodologies or, most
commonly, to update recent historical data. These improvements are implemented consistently for
2023 and across the previous Inventory's time series (i.e., 1990 to 2022) to ensure that the trend is
accurate. This Inventory implements three significant updates, among others. First, new data products
were incorporated into the land representation analysis which improved the accuracy for Alaska and
Hawaii. Second, the National Scale Volume Biomass (NSVB) methodology has been fully implemented
in forest lands, with increases to understory aboveground biomass, understory belowground biomass,
and, most significantly, downed dead wood in this Inventory. This contributes an average additional 70.3
MT C02e Eq. per year (8.0 percent) over the time series for forest land remaining forest land. Third, this
Inventory includes estimates of perennial woody biomass and perennial crop biomass carbon stock
changes and biomass carbon stock changes from croplands and lands converted to and from croplands
which were not included in previous inventories. Together, these and other updates increased total
carbon sequestration estimates by an annual average of 64.7 MMT C02 Eq. (6.5 percent) and decreased
total non-C02 emissions by an annual average of 2.2 MMT C02 Eq. (3.5 percent) across the time series,
compared to the previous Inventory (i.e., 1990 to 2022). For more information on specific
methodological updates, please see the Recalculations Discussion within the respective category
section of this chapter.
Emissions and removals reported in the LULUCF chapter include those from all states including tribal
lands within states; however, for Hawaii and Alaska some emissions and removals from land use and
land-use change are not included in most cases due to challenges with data availability (see chapter
sections on Uncertainty and Planned Improvements for more details). In addition, U.S. Territories are not
included for most categories primarily due to data availability. EPA continues to review available data on
an ongoing basis to include emissions and removals from U.S. Territories in future Inventories to the
extent they are occurring (e.g., see Box 6-1). See Annex 5 for more information on EPA's assessment of
the emissions and removals not included in this Inventory.
6.1 Representation of the U.S. Land Base
A national land use representation system that is consistent and complete, both temporally and
spatially, is needed in order to assess land use and land-use change status and the associated
greenhouse gas fluxes over the Inventory time series. This system should be consistent with IPCC
(2006), such that an inventory should: (1) describe the methods and definitions used to determine areas
of managed and unmanaged lands in the country (Table 6-4), (2) describe and apply a consistent set of
definitions for land-use categories over the entire national land base and time series (i.e., such that
increases in the land areas within particular land-use categories are balanced by decreases in the land
areas of other categories unless the national land base is changing) (Table 6-5), and (3) account for
6-10 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
greenhouse gas fluxes on all managed lands. The IPCC (2006, Vol. IV, Chapter 1) considers all
anthropogenic greenhouse gas emissions and removals associated with land use and management to
occur on managed land, and all emissions and removals on managed land should be reported based on
this guidance (see IPCC (2010), Ogle et al. (2018) for further discussion). Consequently, managed land
serves as a proxy for anthropogenic emissions and removals. This proxy is intended to provide a
practical framework for conducting an inventory, even though some of the greenhouse gas emissions
and removals on managed land are influenced by natural processes that may or may not be interacting
with the anthropogenic drivers. This section of the Inventory has been developed to be consistent with
this guidance. While the 2019 Refinement to the 2006 IPCC Guidelines for National Greenhouse Gas
Inventories (IPCC 2019) provides guidance for factoring out natural emissions and removals, the United
States does not apply this guidance and estimates all emissions/removals on managed land regardless
of whether the driver was natural.
Four databases are used to track land management in the United States and are used as the basis to
classify United States land area into the thirty-six IPCC land use and land-use change categories (Table
6-5) (IPCC 2006). The four primary databases are the U.S. Department of Agriculture (USDA) National
Resources Inventory (NRI),5 the USDA Forest Service (USFS) Forest Inventory and Analysis (FIA)6
Database, the Multi-Resolution Land Characteristics Consortium (MRLC) National Land Cover Dataset
(NLCD),7 and the National Oceanic and Atmospheric Administration Coastal Change Analysis Program
(C-CAP).8 See Table 6-6 for an overview of the land area databases used to characterize land use in
federal and non-federal lands in the conterminous United States, Alaska, and Hawaii.
The total land area included in the United States Inventory is 936 million hectares across the 50 states.9
Approximately 886 million hectares of this land base is considered managed and 50 million hectares is
unmanaged, a distribution that has remained stable over the time series of the Inventory (Table 6-5). In
2023, the United States had a total of 279 million hectares of managed forest land (0.4 percent decrease
compared to 1990). There are 160 million hectares of cropland (8.3 percent decrease compared to
1990), 340 million hectares of managed grassland (0.2 percent increase compared to 1990), 39 million
hectares of managed wetlands (3.6 percent increase compared to 1990), 47 million hectares of
settlements (41.8 percent increase compared to 1990), and 21 million hectares of managed other land
(1.7 percent decrease compared to 1990) (Table 6-5). In addition, some carbon stock changes are not
currently estimated for the entire managed land base, which leads to discrepancies between the
managed land area data presented here and in the subsequent sections of the Inventory (e.g., grassland
remaining grassland within interior Alaska).1011 Planned improvements are under development to
estimate carbon stock changes and greenhouse gas emissions on all managed land and to ensure
5 NRI data are available at https://www.nrcs.usda.gov/nri.
6 FIA data are available at https://research.fs.usda.gOv/programs/fia#data-and-tools.
7 NLCD data are available at https://www.mrlc.gov and MRLC is a consortium of several U.S. government agencies.
8 C-CAP data are available at https://coast.noaa.gov/digitalcoast/tools/lca.html.
9 The current land representation does not include areas from U.S. Territories, but there are planned improvements to
include these regions in future Inventories. U.S. Territories represent approximately 0.1 percent of the total land base for
the United States. See Box 6-1.
10 Other discrepancies occur because the coastal wetlands analysis is based on another land use product (NOAA C-CAP)
that is not currently incorporated into the land representation analysis for this section which relies primarily on the NRI
and NLCD for wetland areas, with the exception of Hawaii where C-CAP is used to account for land use on federal lands,
EPA anticipates addressing these discrepancies in future Inventories.
11 These "managed area" discrepancies also occur in data tables.
Land Use, Land-Use Change, and Forestry 6-11
-------
consistency between the total area of managed land in the land-representation description and the
remainder of the Inventory.
Dominant land uses vary by region, largely due to climate patterns, soil types, geology, proximity to
coastal regions, and historical settlement and economic patterns (Figure 6-3). Forest land tends to be
more common in the eastern United States, mountainous regions of the western United States, and
Alaska. Cropland is concentrated in the mid-continent region of the United States, and grassland is
more common in the western United States and Alaska. Wetlands are fairly ubiquitous throughout the
United States, though they are more common in the upper Midwest and eastern portions of the country,
as well as coastal regions. Settlements are more concentrated along the coastal margins and in the
eastern states.
Table 6-4: Managed and Unmanaged Land Area by Land-Use Categories for All 50
States (Thousands of Hectares)
Land Use Categories
1990
2005
2019
2020
2021
2022
2023
Managed Lands
886,533
886,530
886,531
886,531
886,531
886,531
886,531
Forest
280,523
280,064
279,748
279,626
279,538
279,484
279,430
Croplands
174,495
165,626
160,692
160,102
160,061
160,016
159,960
Grasslands
339,358
341,423
340,101
340,529
340,361
340,179
340,034
Settlements
33,410
40,167
46,291
46,627
46,937
47,162
47,370
Wetlands
37,826
38,648
38,947
38,950
39,007
39,090
39,173
Other
20,921
20,602
20,751
20,697
20,626
20,600
20,564
Unmanaged Lands
49,708
49,711
49,710
49,710
49,710
49,710
49,710
Forest
8,773
8,795
8,851
8,853
8,854
8,855
8,856
Croplands
0
0
0
0
0
0
0
Grasslands
25,943
26,002
26,096
26,095
26,094
26,092
26,091
Settlements
0
0
0
0
0
0
0
Wetlands
4,258
4,196
4,070
4,069
4,069
4,070
4,070
Other
10,734
10,718
10,693
10,693
10,693
10,693
10,693
Total Land Areas
936,241
936,241
936,241
936,241
936,241
936,241
936,241
Forest
289,296
288,859
288,599
288,478
288,392
288,340
288,287
Croplands
174,495
165,626
160,692
160,102
160,061
160,016
159,960
Grasslands
365,301
367,425
366,197
366,624
366,454
366,271
366,125
Settlements
33,410
40,167
46,291
46,627
46,937
47,162
47,370
Wetlands
42,084
42,844
43,017
43,020
43,076
43,160
43,242
Other
31,655
31,320
31,444
31,390
31,319
31,293
31,257
Note: Totals may not sum due to independent rounding.
6-12 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Table 6-5: Land Use and Land-Use Change for the U.S. Managed Land Base for All 50
States (Thousands of Hectares)
Land Use & Land-Use
Change Categories"
1990
2005
2019
2020
2021
2022
2023
Total Forest Land
280,523
280,064
279,748
279,626
279,538
279,484
279,430
FF
279,447
278,860
278,557
278,443
278,368
278,316
278,268
CF
209
141
86
77
75
74
73
GF
767
951
973
973
970
973
973
WF
15
23
18
16
15
15
15
SF
11
18
21
20
19
20
22
OF
75
72
94
97
91
87
80
Total Cropland
174,495
165,626
160,692
160,102
160,061
160,016
159,960
CC
162,273
150,411
149,504
149,816
150,582
151,267
151,617
FC
170
77
62
60
61
63
65
GC
11,673
14,623
10,758
9,910
9,126
8,412
8,017
WC
119
178
98
86
80
75
71
SC
75
102
105
101
97
94
89
OC
186
235
166
129
115
107
102
Total Grassland
339,358
341,423
340,101
340,529
340,361
340,179
340,034
GG
330,060
317,339
320,576
321,736
322,861
323,761
324,509
FG
570
1,657
4,181
4,221
4,230
3,962
3,695
CG
8,177
17,745
13,491
13,212
12,213
11,456
10,879
WG
168
466
172
155
135
126
119
SG
43
526
189
138
99
92
88
OG
341
3,689
1,491
1,066
822
781
744
Total Wetlands
37,826
38,648
38,947
38,950
39,007
39,090
39,173
WW
37,271
36,626
37,824
37,966
38,150
38,303
38,475
FW
37
71
83
81
75
73
70
CW
145
638
310
262
222
188
166
GW
326
1,169
501
419
350
322
261
SW
0
38
14
10
2
2
2
OW
47
107
216
212
209
204
198
Total Settlements
33,410
40,167
46,291
46,627
46,937
47,162
47,370
SS
30,553
31,434
41,609
42,460
43,185
43,745
44,281
FS
293
466
440
437
425
418
411
CS
1,231
3,605
1,727
1,529
1,366
1,228
1,094
GS
1,276
4,375
2,349
2,063
1,831
1,650
1,468
WS
4
59
25
18
14
14
13
OS
54
229
141
120
115
108
102
Total Other Land
20,921
20,602
20,751
20,697
20,626
20,600
20,564
OO
20,188
17,036
18,310
18,536
18,753
18,821
18,937
FO
50
77
99
100
104
108
94
CO
287
603
582
540
489
444
404
GO
371
2,764
1,543
1,309
1,072
1,023
927
WO
22
101
206
203
198
195
191
Land Use, Land-Use Change, and Forestry 6-13
-------
Land Use & Land-Use
Change Categories"
1990
2005
2019
2020
2021
2022
2023
SO
2
22 |
11
10
10
10
10
Grand Total
886,533
886,530 |
886,531
886,531
886,531
886,531
886,531
a The abbreviations a re "F" for forest land, "C'for cropland, "G" for grassland, "W" for wetlands, "S"for settlements, and "0"for
other lands. Lands remaining in the same land-use category are identified with the land-use abbreviation given twice (e.g., "FF"
is forest land remaining forest land), and land-use change categories are identified with the previous land use abbreviation
followed by the new land-use abbreviation (e.g., "CF" is Cropland Converted to Forest Land).
Notes: All land areas reported in this table are considered managed. A planned improvement is underway to deal with an
exception for wetlands, which based on the definitions for the current U.S. Land Representation assessment includes both
managed and unmanaged lands. U.S. Territories have not been classified into land uses and are not included in the U.S. Land
Representation Assessment. See the Planned Improvements section for discussion on plans to include U.S. Territories in
future Inventories. In addition, carbon stock changes are not currently estimated for the entire land base, which leads to
discrepancies between the managed land area data presented here and in the subsequent sections of the Inventory {see land
use chapters e.g., Forest Land Remaining Forest Land for more information). Totals may not sum due to independent rounding.
6-14 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Figure 6-3: Percent of Total Land Area for Each State in the General Land Use
Categories for 2023
Croplands Forest Lands
Grasslands
Other Lands
Settlements
Wetlands
Percent
10 30 50 100
Land Use, Land-Use Change, and Forestry 6-15
-------
Methodology and Time-Series Consistency
IPCC (2006) describes three approaches for representing land areas. Approach 1 provides data on the
total area for each individual land use category, but does not provide detailed information on transfer of
land area between categories following land-use change and is not spatially explicit other than at the
national or regional level. With Approach 1, total net conversions between categories can be detected,
but not the individual changes (i.e., additions and/or losses) between the land-use categories that led to
those net changes. Approach 2 introduces tracking of individual land-use changes between the
categories (e.g., forest land converted to cropland, cropland converted to forest land, and grassland
converted to cropland), using survey samples or other forms of data, but does not provide spatially-
explicit location data. Approach 3 extends Approach 2 by providing spatially-explicit location data, such
as surveys with spatially identified sample locations and maps obtained from remote sensing products.
The three approaches are not presented as hierarchical tiers and are not mutually exclusive.
According to IPCC (2006), the approach or mix of approaches selected by an inventory agency should
reflect calculation needs and national circumstances. For this analysis, the NRI, FIA, NLCD, and C-CAP
have been combined to provide a complete representation of land use for managed lands. These data
sources are described in more detail later in this section. NRI, FIA, NLCD, and C-CAP are Approach 3
data sources that provide spatially-explicit representations of land use and land-use conversions. Lands
are treated as remaining in the same category (e.g., cropland remaining cropland) if a land-use change
has not occurred in the last 20 years, consistent with the IPCC guidelines (2006). Otherwise, the land is
classified in a land-use change category based on the current use and most recent use before
conversion to the current use (e.g., cropland converted to forest land).
Definitions of Land Use in the United States
Managed and Unmanaged Land
The United States definition of managed land is similar to the general definition of managed land
provided by the IPCC (2006), but with some additional elaboration to reflect national circumstances.
Based on the following definitions, most lands in the United States are classified as managed:
• Managed Land: Land is considered managed if direct human intervention has influenced its
condition. Direct intervention occurs mostly in areas accessible to human activity and includes
altering or maintaining the condition of the land to produce commercial or non-commercial
products or services; to serve as transportation corridors or locations for buildings, landfills, or
other developed areas for commercial or non-commercial purposes; to extract resources or
facilitate acquisition of resources; or to provide social functions for personal, community, or
societal objectives where these areas are readily accessible to society.12
12 Wetlands, as specified by IPCC (2006), are only considered managed if they are created through human activity, such as
dam construction, or the water level is artificially altered by human activity. Distinguishing between managed and
unmanaged wetlands in the United States is difficult due to limited data availability. Wetlands are not characterized
within the NRI with information regarding water table management or origin (i.e., constructed rather than natural origin).
Therefore, unless wetlands are converted into cropland or grassland, it is not possible to know if they are artificially
created or if the water table is managed based on the use of NRI data. As a result, most wetlands are reported as
managed with the exception of wetlands in remote areas of Alaska, but emissions from managed wetlands are only
reported for coastal regions, flooded lands (e.g., reservoirs) and peatlands where peat extraction occurs due to
6-16 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
• Unmanaged Land: All other land is considered unmanaged. Unmanaged land is largely
comprised of areas inaccessible to society due to the remoteness of the locations. Though
these lands may be influenced indirectly by human actions such as atmospheric deposition of
chemical species produced in industry or C02 fertilization, they are not influenced by a direct
human intervention. Designated unmanaged lands in the Inventory include some forests,
grasslands, and wetlands in remote interior Alaska.
In addition, land that is previously managed remains in the managed land base for 20 years before re-
classifying the land as unmanaged in order to account for legacy effects of management on carbon
stocks.13 Unmanaged land is also re-classified as managed overtime if anthropogenic activity is
introduced into the area based on the definition of managed land.
Land-Use Categories
As with the definition of managed lands, IPCC (2006) provides general non-prescriptive definitions for
the six main land-use categories: forest land, cropland, grassland, wetlands, settlements and other
land. In order to reflect national circumstances, country-specific definitions have been developed,
based predominantly on criteria used in the land-use surveys for the United States. Specifically, the
definition of forest land is based on the FIA definition of forest,14 while definitions of cropland,
grassland, and settlements are based on the NRI.15The definitions for other land and wetlands are
based on the IPCC (2006) definitions for these categories.
• Forest Land: A land-use category that includes areas at least 120 feet (36.6 meters) wide and at
least one acre (0.4 hectare) in size with at least ten percent cover (or equivalent stocking) by live
trees including land that formerly had such tree cover and that will be naturally or artificially
regenerated. Trees are woody plants having a more or less erect perennial stem(s) capable of
achieving at least 3 inches (7.6 cm) in diameter at breast height, or 5 inches (12.7 cm) diameter
at root collar, and a height of 16.4 feet (5 m) at maturity in situ. Forest land includes all areas
recently having such conditions and currently regenerating or capable of attaining such
condition in the near future. Forest land also includes transition zones, such as areas between
forest and non-forest lands, that have at least ten percent cover (or equivalent stocking) with live
trees and forest areas adjacent to urban and built-up lands. Unimproved roads and trails,
streams, and clearings in forest areas are classified as forest if they are less than 120 feet (36.6
m) wide or an acre (0.4 ha) in size. However, land is not classified as forest land if completely
surrounded by urban or developed lands, even if the criteria are consistent with the tree area
and cover requirements for forest land. These areas are classified as settlements. In addition,
forest land does not include land that is predominantly under an agricultural land use (Nelson et
al. 2020).
insufficient activity data to estimate emissions and limited resources to improve the Inventory. See the Planned
Improvements section of the Inventory for future refinements to the wetland area estimates.
13 There are examples of managed land transitioning to unmanaged land in the United States. For example, in 2018,100
hectares of managed grassland converted to unmanaged because data indicated that no further grazing occurred.
Livestock data are collected annually by the Department of Agriculture, and no livestock had occurred in the area since
the mid-1970s, and therefore there was no longer active management through livestock grazing. The area is also remote,
at least 10 miles from roads and settlements, and therefore the land was no longer managed based on the
implementation criteria.
14 See https://research.fs.usda.gov/understory/nationwide-forest-inventory-field-guide . section 2.0 Condition Class.
15 See https://www.nrcs.usda.gov/nri.
Land Use, Land-Use Change, and Forestry 6-17
-------
• Cropland: A land-use category that includes areas used for the production of adapted crops for
harvest; this category includes both cultivated and non-cultivated lands. Cultivated crops
include row crops or close-grown crops and also pasture in rotation with cultivated crops. Non-
cultivated cropland includes continuous hay, perennial crops (e.g., orchards) and horticultural
cropland. Cropland also includes land with agroforestry, such as alley cropping and
windbreaks,16 if the dominant use is crop production, assuming the stand orwoodlot does not
meet the criteria for forest land. Lands in temporary fallow or enrolled in conservation reserve
programs (i.e., set-asides17) are also classified as cropland, as long as these areas do not meet
the forest land criteria. Roads through cropland, including interstate highways, state highways,
other paved roads, gravel roads, dirt roads, and railroads are excluded from cropland area
estimates and are, instead, classified as settlements.
• Grassland: A land-use category on which the plant cover is composed principally of grasses,
grass-like plants (i.e., sedges and rushes), forbs, or shrubs suitable for grazing and browsing,
and includes both pastures and native rangelands. This includes areas where practices such as
clearing, burning, chaining, and/or chemicals are applied to maintain the grass vegetation. Land
is also categorized as grassland if there have been three or fewer years of continuous hay
production.18 Savannas, deserts, and tundra are considered grassland. Drained wetlands are
considered grassland if the dominant vegetation meets the plant cover criteria for grassland.
Woody plant communities of low forbs, shrubs and woodlands, such as sagebrush, mesquite,
chaparral, mountain shrubland, and pinyon-juniper, are also classified as grassland if they do
not meet the criteria for forest land. Grassland includes land managed with agroforestry
practices, such as silvopasture and windbreaks, if the land is principally grass, grass-like plants,
forbs, and shrubs suitable for grazing and browsing, and assuming the stand or woodlot does
not meet the criteria for forest land. Roads through grassland, including interstate highways,
state highways, other paved roads, gravel roads, dirt roads, and railroads are excluded from
grassland and are, instead, classified as settlements.
• Wetlands: A land-use category that includes land covered or saturated by water for all or part of
the year, in addition to lakes, reservoirs, and rivers. In addition, all coastal wetlands are
considered managed regardless of whether the water level is changed or if they were created by
human activity. Certain areas that fall under the managed wetlands definition are included in
other land uses based on the IPCC guidance and national circumstances, including lands that
are flooded for most or just part of the year in croplands (e.g., rice cultivation and cranberry
production), grasslands (e.g., wet meadows dominated by grass cover) and forest lands (e.g.,
riparian forests near waterways). See Section 6.8 for more information.
• Settlements: A land-use category representing developed areas consisting of units equal to or
greater than 0.25 acres (0.1 ha) that includes residential, industrial, commercial, and
institutional land; construction sites; public administrative sites; railroad yards; cemeteries;
16 Currently, there is no data source to account for biomass carbon stock change associated with woody plant growth and
losses in alley cropping systems and windbreaks in cropping systems, although these areas are included in the cropland
land base.
17 A set-aside is cropland that has been taken out of active cropping and converted to some type of vegetative cover,
including, for example, native grasses or trees, but is still classified as cropland based on national circumstances.
18 Areas with four or more years of continuous hay production are cropland because the land is typically more intensively
managed with cultivation, greater amounts of inputs, and other practices. Occasional harvest of hay from grasslands
typically does not involve cultivation or other intensive management practices.
6-18 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
airports; golf courses; sanitary landfills; sewage treatment plants; water control structures and
spillways; parks within urban and built-up areas; and highways, railroads, and other
transportation facilities. Also included are all tracts that may meet the definition of forest land,
and tracts of less than ten acres (4.05 ha) that may meet the definitions for cropland, grassland,
or other land but are completely surrounded by urban or built-up land, and so are included in
the settlements category. Rural transportation corridors located within other land uses (e.g.,
forest land, cropland, and grassland) are also included in settlements.
• Other Land: Aland-use category that includes bare soil, rock, ice, and all land areas that do not
fall into any of the other five land-use categories. Following the guidance provided by the IPCC
(2006), carbon stock changes and non-C02 emissions are not estimated for other lands
because these areas are largely devoid of biomass, litter and soil carbon pools. However,
carbon stock changes and non-C02 emissions should be estimated for land converted to other
land during the first 20 years following conversion to account for legacy effects.
Land Use Data Sources: Description and Application to
U.S. Land Area Classification
U.S. Land Use Data Sources
The four main sources for land use data in the United States are the NRI, FIA, NLCD, and C-CAP (Table
6-6). These data sources are combined to account for land use in all 50 states. FIA and NRI data are
used when available for an area because these surveys contain additional information on management,
site conditions, crop types, biometric measurements, and other data that are needed to estimate
carbon stock changes, N20, and CH4 emissions on those lands. If NRI and FIA data are not available for
an area, however, then either the NLCD or C-CAP product is used to represent the land use. Sources of
land use data included in the land representation in this Inventory are consistent with those included in
the previous Inventory.
Table 6-6: Data Sources Used to Determine Land Use and Land Area for the
Conterminous United States, Hawaii, and Alaska
NRI
FIA
NLCD
C-CAP
Forest Land
Conterminous United States
Non-Federal
•
Federal
•
Hawaii
Non-Federal
•
Federal
•
Alaska
Non-Federal
•
Federat
Land Use, Land-Use Change, and Forestry 6-19
-------
NRI FIA NLCD C-CAP
Croplands, Grasslands, Other Lands, Settlements, and Wetlands
Conterminous United States
Non-Federal
•
Federal
•
Hawaii
Non-Federal
•
Federal
•
Alaska
Non-Federal
•
Federal
•
National Resources Inventory
For the Inventory, the N Rl is the official source of data for land use and land-use change on non-federal,
non-forest lands in the conterminous United States and Hawaii, and is also used to determine the total
land base for the conterminous United States and Hawaii. The NRI is a statistically-based survey
conducted by the USDA Natural Resources Conservation Service and is designed to assess soil, water,
and related environmental resources on non-federal lands. The NRI has a stratified multi-stage sampling
design, where primary sample units are stratified on the basis of county and township boundaries
defined by the United States Public Land Survey (Nusser and Goebel 1997). Within a primary sample
unit (typically a 160 acre [64.75 ha] square quarter-section), three sample points are selected according
to a restricted randomization procedure. Each point in the survey is assigned an area weight (expansion
factor) based on other known areas and land use information (Nusser and Goebel 1997, USDA NRCS
2023). The NRI survey utilizes data obtained from remote sensing imagery and site visits in order to
provide detailed information on land use and management, particularly for croplands and grasslands
(i.e., agricultural lands), and is used as the basis to account for carbon stock changes in agricultural
lands (except federal grasslands). The NRI survey was conducted every five years between 1982 and
1997, but shifted to annualized data collection in 1998. The land use between five-year periods from
1982 and 1997 are assumed to be the same for a five-year time period if the land use is the same at the
beginning and end of the five-year period (note: most of the data have the same land use at the
beginning and end of the five-year periods). If the land use had changed during a five-year period, then
the change is assigned at random to one of the five years. For crop histories, years with missing data are
estimated based on the sequence of crops grown during years preceding and succeeding a missing year
in the NRI history. This gap-filling approach allows for development of a full time series of land use data
for non-federal lands in the conterminous United States and Hawaii. This Inventory incorporates data
through 2017 from the NRI. The land use patterns are assumed to remain the same from 2018 through
2023 for this Inventory, but the time series will be updated when new data become available and are
integrated into the land representation analysis.
Forest Inventory and Analysis
The FIA program, conducted by the USFS, is the official source of data on forest land area and
management data for the Inventory and is another statistically-based survey for the United States. The
Forest Inventory and Analysis engages in a hierarchical system of sampling, with sampling categorized
as Phases 1 through 3, in which sample points for each consecutive phase are subsets of the previous
phase. Phase 1 refers to collection of remotely-sensed data (either aerial or satellite imagery) primarily
to classify land into forest or non-forest and to identify landscape patterns like fragmentation and
urbanization. Phase 2 is the collection of field data on a network of ground plots that enable
6-20 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
classification and summarization of area, tree, and other attributes associated with forest land uses.
Phase 3 plots are a subset of Phase 2 plots where data on indicators of forest health are measured. Data
from all three phases are also used to estimate carbon stock changes for forest land. Historically, FIA
inventory surveys have been conducted periodically, with all plots in a state being measured at a
frequency of every five to ten years. A new national plot design and annual sampling design was
introduced by the FIA program in 1998 and is now used in all states. Annualized sampling means that a
portion of plots throughout each state is sampled each year, with the goal of measuring all plots once
every five to seven years in the eastern United States and once every ten years in the western United
States. See Annex 3.14 for the specific survey data available by state. The most recent year of available
data varies state by state (range of most recent data is from 2019 through 2023; see Table A-192 in Annex
3.14).
National Land Cover Dataset and Coastal Change Analysis Program
As noted above, while the NRI survey sample covers the conterminous United States and Hawaii, land
use data are only collected on non-federal lands. Gaps exist in the land representation when the NRI
and FIA datasets are combined, such as federal grasslands operated by Bureau of Land Management
(BLM), USDA, and National Park Service, as well as Alaska.19 The NLCD is used to account for land use
on federal lands in the conterminous United States, in addition to non-forest federal and non-federal
lands in Alaska. C-CAP is used to account for land use on federal lands in Hawaii.
NLCD products provide land-cover for 1992, 2001, 2004, 2006, 2008, 2011, 2013, 2016, 2019, and 2021
in the conterminous United States (Yang et al. 2018; Fry et al. 2011; Homer et al. 2007, 2015, Dewitz,
2023), and also for Alaska in 2001, 2011, and 2016. A NLCD change product is not available for Hawaii
because data are only available for one year (2001), therefore high-resolution C-CAP products are used
to provide multiple points in time. C-CAP provides land-cover for 2005 on all islands, 2010 on Hawai'i,
Kaua'i, Maui, Moloka'i, and Ni'ihau, and 2011 on Lana'i and O'ahu. Only a single date is available on
Kaho'olawe (2005). The NLCD products are based primarily on Landsat Thematic Mapper imagery at a
30-meter resolution while the C-CAP products are based on Quickbird or WorldView2 imagery at a 2.4-
meter resolution. The land-cover categories in both products have been aggregated into the 36 IPCC
land-use. The land-use patterns are assumed to remain the same after the last year of data in the time
series, which is 2005, 2010, or 2011 for Hawaii, 2021 for the conterminous United States, and 2016 for
Alaska. The time series will be updated when new data are released.
For the conterminous United States, the aggregated maps of IPCC land-use categories obtained from
the NLCD products were used in combination with the NRI database to represent land use and land-use
change for federal lands, with the exception of forest lands, which are based on FIA. Specifically, NRI
survey locations designated as federal lands were assigned a land use/land-use change category based
on the NLCD maps that had been aggregated into the IPCC categories. This analysis addressed shifts in
land ownership across years between federal or non-federal classes as represented in the NRI survey
(i.e., the ownership is classified for each survey location in the NRI). The sources of these additional
data are discussed in subsequent sections of the report.
19 The NRI survey program does not include U.S. Territories with the exception of non-federal lands in Puerto Rico. The FIA
program recently began implementing surveys of forest land in U.S. Territories (i.e. includes American Samoa, Guam,
Northern Mariana Islands, U.S. Virgin Islands, and Puerto Rico) and those data will be used in the years ahead.
Furthermore, NLCD does not include coverage for all U.S. Territories.
Land Use, Land-Use Change, and Forestry 6-21
-------
Managed Land Designation
Lands are designated as managed in the United States based on the definition provided earlier in this
section. The following criteria are used in order to apply the definition in an analysis of managed land:
• All croplands and settlements are designated as managed so only grassland, forest land,
wetlands or other lands may be designated as unmanaged land;20
• All forest lands with active fire protection are considered managed;
• All forest lands designated for timber harvests are considered managed;
• All grasslands are considered managed at a county scale if there are grazing livestock in the
county;
• Other areas are considered managed if accessible based on the proximity to roads and other
transportation corridors, and/or infrastructure;
• Protected lands maintained for recreational and conservation purposes are considered
managed (i.e., managed by public and/or private organizations);
• Lands with active and/or past resource extraction are considered managed; and
• Lands that were previously managed but subsequently classified as unmanaged remain in the
managed land base for 20 years following the conversion to account for legacy effects of
management on carbon stocks.
The analysis of managed lands, based on the criteria listed above, is conducted using a geographic
information system (Ogle et al. 2018). Lands that are used for crop production or settlements are
determined from the NLCD (Fry et al. 2011; Homer et al. 2007; Homer et al. 2015). Forest lands with
active fire management are determined from maps of federal and state management plans from the
National Atlas (U.S. Department of Interior 2005) and Alaska Interagency Fire Management Council
(1998). It is noteworthy that all forest lands in the conterminous United States have active fire
protection, and are therefore designated as managed regardless of accessibility or other criteria. In
addition, forest lands with timber harvests are designated as managed based on county-level estimates
of timber products in the U.S. Forest Service Timber Products Output Reports (U.S. Department of
Agriculture 2012). Timber harvest data lead to additional designation of managed forest land in Alaska.
The designation of grasslands as managed is based on grazing livestock population data at the county
scale from the USDA National Agricultural Statistics Service (U.S. Department of Agriculture 2015).
Accessibility is evaluated based on a 10-km buffer surrounding road and train transportation networks
using the ESRI Data and Maps product (ESRI 2008), and a 10-km buffer surrounding settlements using
NLCD.
Lands maintained for recreational purposes are determined from analysis of the Protected Areas
Database (U.S. Geological Survey 2012). The Protected Areas Database includes lands protected from
conversion of natural habitats to anthropogenic uses and describes the protection status of these lands.
Lands are considered managed that are protected from development if the regulations allow for
extractive or recreational uses or suppression of natural disturbance (e.g., forest lands with active fire
protection). Lands that are protected from development and not accessible to human intervention,
20 ALL wetlands are considered managed in this Inventorywltb the exception of remote areas in ALaska. Distinguishing
between managed and unmanaged wetLands in the conterminous United States and Hawaii is difficult due to limited
data availability. Wetlands are not characterized within the NRI with information regarding water table management.
Regardless, a planned improvement is underway to subdivide managed and unmanaged wetlands.
6-22 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
including no suppression of disturbances or extraction of resources, are not included in the managed
land base.
Multiple data sources are used to determine lands with active resource extraction: Alaska Oil and Gas
Information System (Alaska Oil and Gas Conservation Commission 2009), Alaska Resource Data File
(U.S. Geological Survey 2012), Active Mines and Mineral Processing Plants (U.S. Geological Survey
2005), and Coal Production and Preparation Report (U.S. Energy Information Administration 2011). A
buffer of 3,300 and 4,000 meters is established around petroleum extraction and mine locations,
respectively, to account for the footprint of operation and impacts of activities on the surrounding
landscape. The buffer size is based on visual analysis of disturbance to the landscape for approximately
130 petroleum extraction sites and 223 mines. After applying the criteria identified above, the resulting
managed land area is overlaid on the NLCD to estimate the area of managed land by land use for both
federal and non-federal lands in Alaska. The remaining land represents the unmanaged land base. The
resulting spatial product is also used to identify NRI survey locations that are considered managed and
unmanaged for the conterminous United States and Hawaii.21
Approach for Combining Data Sources
The managed land base in the United States has been classified into the 36 IPCC land use/land-use
conversion categories (Table 6-5) using definitions developed to meet national circumstances, while
adheringto IPCC guidelines (20 06).22 In practice, the land was initially classified into land-use
subcategories within the NRI, FIA, and NLCD or C-CAP datasets, and then aggregated into the 36 broad
land use and land-use change categories identified in IPCC (2006).
All three datasets used in the conterminous United States (NRI, FIA, and NLCD) provide information on
forest land areas, but the area data from FIA serve as the official dataset for forest land. Therefore,
another step in the analysis is to address the inconsistencies in the representation of the forest land
among the three databases. NRI and FIA have different criteria for classifying forest land in addition to
different sampling designs, leading to discrepancies in the resulting estimates of forest land area on
non-federal land in the conterminous United States. Similarly, there are discrepancies between the
NLCD and FIA data for defining and classifying forest land on federal lands. Any change in forest land
area in the NRI and NLCD also requires a corresponding change in other land use areas because of the
dependence between the forest land area and the amount of land designated as other land uses, such
as the amount of grassland, cropland, and wetlands (i.e., areas for the individual land uses must sum to
the total managed land area of the country).
FIA is the main database for forest statistics, and consequently, the NRI and NLCD are adjusted to
achieve consistency with FIA estimates of forest land in the conterminous United States. Adjustments
are made in the forest land remaining forest land, land converted to forest land, and forest land
converted to other uses (i.e., grassland, cropland, settlements, other lands, and wetlands). All
adjustments are made at the state scale to address the discrepancies in areas associated with forest
land and conversions to and from forest land. There are three steps in this process. The first step
involves adjustments to land converted to forest land (i.e., grassland, cropland, settlements, other
lands, and wetlands converted to forest land), followed by a second step in which there are adjustments
21 The exception is cropland and settlement areas in the NRI, which are classified as managed, regardless of the managed
land base obtained from the spatial analysis described in this section.
22 Definitions are provided in the previous section.
Land Use, Land-Use Change, and Forestry 6-23
-------
in forest land converted to another land use (i.e., grassland, cropland, settlements, other lands, and
wetlands), and the last step is to adjust forest land remaining forest land.
In the first step, land converted to forest land in the NRI and NLCD are adjusted to match the state-level
estimates in the FIA data for non-federal and federal land converted to forest land, respectively. FIA data
have not provided specific land-use categories that are converted to forest land in the past, but rather a
sum of all land converted to forest land.23 The NRI and NLCD provide information on specific land-use
conversions, such as grassland converted to forest land. Therefore, adjustments at the state level to NRI
and NLCD are made proportional to the amount of specific land-use conversions into forest land for the
state, prior to any further adjustments. For example, if 50 percent of the land-use change to forest land
is associated with grassland converted to forest land in a state according to NRI or NLCD, then half of
the discrepancy with FIA data in the area of land converted to forest land is addressed by increasing or
decreasing the area in grassland converted to forest land. Moreover, any increase or decrease in
grassland converted to forest land in NRI or NLCD is addressed by a corresponding change in the area of
grassland remaining grassland, so that the total amount of managed area is not changed within an
individual state. Since the sum of all land converted to forest land is used to adjust specific land-use
conversions into forest land for the state-level estimates in the NRI and NLCD, there is the potential for
differences in area estimates in states where specific land-use conversions into forest land do not exist
in the FIA data.
In the second step, state-level areas are adjusted in the NRI and NLCD to address discrepancies with
FIA data for forest land converted to other uses. Similar to land converted to forest land, FIA have not
provided information on the specific land-use changes in the past,24 so areas associated with forest
land conversion to other land uses in NRI and NLCD are adjusted proportional to the amount of area in
each conversion class in these datasets. Since the sum of all forest land converted to other uses is used
to adjust specific land-use conversions out of forest land for the state-level estimates in the NRI and
NLCD, there is the potential for differences in area estimates in states where a specific land-use
conversion out of forest land does not exist in the FIA data.
In the final step, the area of forest land remaining forest land in each state according to the NRI and
NLCD is adjusted to match the FIA estimates for non-federal and federal land, respectively. It is
assumed that the majority of the discrepancy in forest land remaining forest land is associated with
less-precise estimates of grassland remaining grassland and wetlands remainingwetlands in the NRI
and NLCD. This step also assumes that there are no changes in the land-use conversion categories.
Therefore, corresponding adjustments are made in the area estimates of grassland remaining grassland
and wetlands remainingwetlands from the NRI and NLCD. This adjustment balances the change in
forest land remaining forest land area, which ensures no change in the overall amount of managed land
within an individual state. The adjustments are based on the proportion of land within each of these
land-use categories at the state level according to NRI and NLCD (i.e., a higher proportion of grassland
led to a larger adjustment in grassland area).
The modified NRI data are then aggregated to provide the land use and land-use change data for non-
federal lands in the conterminous United States, and the modified NLCD data are aggregated to provide
the land use and land-use change data for federal lands. Data for all land uses in Hawaii are based on
23 The FIA program has started to collect data on the specific land uses that are converted to forest land, which will be
further investigated and incorporated into a future Inventory.
24 The FIA program has started to collect data on the specific land uses following conversion from forest land, which will be
further investigated and incorporated into a future Inventory.
6-24 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
NRI for non-federal lands and on C-CAP for federal lands, and are not adjusted to be consistent with FIA
forest land area. Land use data in Alaska are based on the NLCD data after adjusting this dataset to be
consistent with forest land areas in the FIA (Table 6-6). The result is land use and land-use change data
for the conterminous United States, Hawaii, and Alaska.
A summary of the details on the approach used to combine data sources for each land use are
described below.
• Forest Land: Land representation for both non-federal and federal forest lands in the
conterminous United States and Alaska are based on the FIA. The FIA is used as the basis for
both forest land area data as well as to estimate carbon stocks and fluxes on forest land in the
conterminous United States and Alaska. The FIA does have survey plots in Alaska that are used
to determine the carbon stock changes, and the associated area data for this region are
harmonized with NLCD using the methods described above. NRI is used in the current report to
provide forest land areas on non-federal lands in Hawaii, and C-CAP is used for federal lands. In
Hawaii and the U.S. Territories, FIA data are being collected; these data were used to compile
area estimates and emissions and removals for forest land in this Inventory.
• Cropland: Cropland is classified using the NRI, which covers all non-federal lands within 49
states (excluding Alaska), including state and local government-owned land as well as tribal
lands. The NRI is used as the basis for both cropland area data as well as to estimate soil carbon
stocks and fluxes on cropland. The NLCD is used to determine cropland area and soil carbon
stock changes on federal lands in the conterminous United States while C-CAP is used in
Hawaii. The NLCD is also used to determine croplands in Alaska, but carbon stock changes are
not estimated for this region in the current Inventory.
• Grassland: Grassland on non-federal lands is classified using the NRI within 49 states (excluding
Alaska), including state and local government-owned land as well as tribal lands. The NRI is
used as the basis for both grassland area data as well as to estimate soil carbon stocks and
non-C02 greenhouse emissions on grassland. Grassland area and soil carbon stock changes are
determined using the classification provided in the NLCD for federal land within the
conterminous United States. The NLCD is also used to estimate the areas of federal and non-
federal grasslands in Alaska, and C-CAP is used for the federal grasslands in Hawaii, but the
current Inventory does not include carbon stock changes in these areas.
• Wetlands: The NRI captures wetlands on non-federal lands within 49 states (excluding Alaska).
The land representation data for federal wetlands in the conterminous United States and
wetlands in Alaska are based on the NLCD, while C-CAP is used on federalwetlands in
Hawaii.25
• Settlements: The NRI captures non-federal settlement area in 49 states (excluding Alaska). If
areas of forest land or grassland under ten acres (4.05 ha) are contained within settlements or
urban areas, they are classified as settlements (urban) in the NRI database. If these parcels
exceed the ten-acre (4.05 ha) threshold and are grassland, they are classified as grassland by
NRI. Regardless of size, a forested area is classified as non-forest by FIA if it is located within an
25 This analysis does not distinguish between managed and unmanaged wetlands except for remote areas in Alaska, but
there is a planned improvement to subdivide managed and unmanaged wetlands for the entire land base.
Land Use, Land-Use Change, and Forestry 6-25
-------
urban area. Land representation for settlements on federal lands in the coterminous United
States and Alaska is based on the NLCD, federal lands in Hawaii use C-CAP.
• Other Land: Any land that is not classified into one of the previous five land-use categories is
categorized as other land using the NRI for non-federal areas in the conterminous United States
and Hawaii and using the NLCD or C-CAP for the federal lands in all regions of the United States
and for non-federal lands in Alaska.
Some lands can be classified into one or more categories due to multiple uses that meet the criteria of
more than one definition. However, a ranking has been developed for assignment priority in these cases.
The ranking process is from highest to lowest priority based on the following order:
Settlements > Cropland > Forest Land > Grassland > Wetlands > Other Land
Settlements are given the highest assignment priority because they are extremely heterogeneous with a
mosaic of patches that include buildings, infrastructure, and travel corridors, but also open grass areas,
forest patches, riparian areas, and gardens. The latter examples could be classified as grassland, forest
land, wetlands, and cropland, respectively, but when located in close proximity to settlement areas,
they tend to be managed in a unique manner compared to non-settlement areas. Consequently, these
areas are assigned to the settlements land-use category. Cropland is given the second assignment
priority, because cropping practices tend to dominate management activities on areas used to produce
food, forage, or fiber. The consequence of this ranking is that crops in rotation with pasture are classified
as cropland, and land with woody plant cover that is used to produce crops (e.g., orchards) is classified
as cropland, even though these areas may also meet the definitions of grassland or forest land,
respectively. Similarly, wetlands are considered croplands if they are used for crop production, such as
rice or cranberries. Forest land occurs next in the priority assignment because traditional forestry
practices tend to be the focus of the management activity in areas with woody plant cover that are not
croplands (e.g., orchards) or settlements (e.g., housing subdivisions with significant tree cover).
Grassland occurs next in the ranking, while wetlands and then other land complete the list.
The assignment priority does not reflect the level of importance for quantifying greenhouse gas
emissions and removals on managed land, but is intended to classify all areas into a discrete land-use
category. Currently, the IPCC does not make provisions in the guidelines for assigning land to multiple
uses. For example, a wetland is classified as forest land if the area has sufficient tree cover to meet the
stocking and stand size requirements. Similarly, wetlands are classified as cropland if they are used for
crop production, such as rice, or as grassland if they are composed principally of grasses, grass-like
plants (i.e., sedges and rushes), forbs, or shrubs suitable for grazing and browsing. Regardless of the
classification, emissions and removals from these areas should be included in the Inventory if the land
is considered managed, and therefore impacted by anthropogenic activity in accordance with the
guidance provided by the IPCC (2006).
QA/QC and Verification
The land base obtained from the NRI, FIA, NLCD, and C-CAP was compared to the Topological^
Integrated Geographic Encoding and Referencing (TIGER) Survey (U.S. Census Bureau 2010). The United
States Census Bureau gathers data on the population and economy and has a database of land areas
for the country. The area estimates of land-use categories, based on NRI, FIA, NLCD, and C-CAP, are
obtained from remote sensing data instead of the land survey approach used by the United States
Census Survey. The Census does not provide a time series of land-use change data or land
6-26 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
management information, which is needed for estimating greenhouse gas emissions from land use and
land-use change. Regardless, the Census does provide sufficient information to provide a quality
assurance check on the Inventory data. There are 46 million more hectares of land in the United States
according to the Census, compared to the total area estimate of 936 million hectares obtained from the
combined NRI, FIA, NLCD, and C-CAP data, a 4.8 percent difference. Much of this difference is
associated with open water in coastal regions and the Great Lakes, which is included in the TIGER
Survey of the Census, but not included in the land representation using the NRI, FIA, NLCD, and C-CAR
There is only a 0.4 percent difference when open water in coastal regions is removed from the TIGER
data. General QC procedures for data gathering and data documentation also were applied consistent
with the QA/QC and verification procedures described in Annex 8.
Recalculations Discussion
The land representation estimates were recalculated from the previous Inventory with the following
datasets: a) updated FIA data from 1990 to 2022 for the conterminous United States and Alaska, b) NRI
data from 1990 to 2017 for the conterminous United States and Hawaii, c) NLCD data for the
conterminous United States from 2001 through 2021 and Alaska from 2001 through 2016, and d) C-CAP
data for Hawaii in 2005 and 2010 or 2011. There were two changes in methods that resulted in small
changes between this Inventory and the previous Inventory. First, on FIA plots in interior Alaska that have
not yet been measured by a field crew, the LANDFIRE Existing Vegetation Type (EVT)26 data product was
used in conjunction with NLCD to define the plot as forest or non-forest land, specifically on those plots
that intersected with the 'Woody Wetlands' NLCD land cover category. Previously, all of those plots were
defined as forest land. This year those plots were also intersected with the EVT data product to
determine if the primary vegetation was trees or shrubs. This led to a 0.6 percent decrease in estimates
of FIA forest land area in Alaska over the time series relative to the previous Inventory (1990 through
2022), which in turn led to a 0.4 percent increase in grassland area and 0.9 percent increase in wetland
area over the time series in this Inventory. Second, because there is only a single year of NLCD available
in Hawaii, it was replaced with C-CAP data to classify federal NRI plots because it provided two years for
a majority of the State. There was little land use change estimated from the C-CAP data on those federal
NRI plots resulting in minor changes in land representation estimates for the state in this Inventory.
Planned Improvements
The following planned improvement are under review and/or being researched for incorporation into
future inventories:
• Harmonizing NRI and FIA sampling frames to improve consistency and facilitate estimation
using multi-frame sampling. Research is underway for this improvement. This includes
development of a common land use classification schema between the two land inventories
that can be used in the harmonization process. These steps will allow for population estimation
exclusive of auxiliary information (e.g., NLCD). The multi-frame sample will also serve as
reference data for the development of spatially explicit and spatially continuous map products
for each year in the Inventory time series. This is a medium- to long-term improvement.
26 LANDFIRE Existing Vegetation Type data are available at https://landfire.gov/vegetation/evt.
Land Use, Land-Use Change, and Forestry 6-27
-------
• Fully incorporating area data by land-use type for U.S. Territories. Although most of the managed
land in the United States is included in the current land use data for the conterminous United
States, Alaska, and Hawaii, a complete reporting of all lands in the United States, including U.S.
Territories, is a key goal for the near future. An initial assessment of data sources for land use
area data for U.S. Territories by land-use category are provided in Box 6-1. In addition, this
Inventory includes forest land areas estimated for American Samoa, Guam, Hawaii, Northern
Marianas Islands, U.S. Virgin Islands, and Puerto Rico using periodic inventories from the FIA
program. These estimates are included in the forest land category, and the methods for
compiling these area estimates and the associated carbon stocks and fluxes and integration of
these estimates into the land representation will be refined to compensate for data limitations
in the time series while also taking advantage of new data and data products. See Box 6-1.
• Reconciling wetlands (coastal wetlands and flooded lands) area estimates used to calculate
emissions and removals with data provided in the NRI, FIA and NLCD used to develop the land
representation. Methods in the 2013 Supplement to the 2006 Guidelines for National
Greenhouse Gas Inventories: Wetlands (IPCC 2014) have been applied to estimate emissions
and removals from coastal wetlands; specifically, greenhouse gas emissions from coastal
wetlands have been developed for the Inventory using the NOAA C-CAP land-cover product. The
NOAA C-CAP product is not used directly in the land representation analysis outside of Hawaii.
Estimates from flooded lands are also included in this Inventory, but data are not directly used
in the land representation analysis at this time.
¦ In addition, the current Inventory does not include a classification of managed and
unmanaged wetlands, except for remote areas in Alaska. Consequently, there is a planned
improvement to classify managed and unmanaged wetlands for the conterminous United
States and Hawaii, and more detailed wetlands datasets will be evaluated and integrated
into the analysis to meet this objective. EPA recently initiated an effort, coordinating with an
interagency group and other partners, to assess the representation of all wetlands in the
Inventory. More details will be included in a future Inventory.
• Quantifying uncertainty with land-use and land-use change area estimates. These are needed
as an input into the overall Inventory uncertainty analysis. This is a medium-term improvement
that will be addressed in a future Inventory.
Box 6-1: Preliminary Estimates of Land Use in U.S. Territories
Several programs have developed land-cover maps for U.S. Territories using remote sensing imagery,
including the Gap Analysis Program, Caribbean Land Cover project, National Land Cover Dataset
(NLCD), USFS Pacific Islands Imagery Project, and the National Oceanic and Atmospheric
Administration (NOAA) Coastal Change Analysis Program (C-CAP). Land-cover data can be used to
inform a land use classification if there is a time series to evaluate the dominant practices. For example,
land that is principally used for timber production with tree cover over most of the time series is
classified as forest land even if there are a few years of grass dominance following timber harvest. These
products were reviewed and evaluated for use in the national Inventory as a step towards implementing
a planned improvement to include U.S. Territories in the land representation for the Inventory.
Recommendations are to use the NOAA C-CAP Regional Land Cover Database for the smaller island
Territories (U.S. Virgin Islands, Guam, Northern Marianas Islands, and American Samoa) because this
6-28 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
program is ongoing and therefore will be continually updated. The C-CAP product does not cover the
entire territory of Puerto Rico, so the NLCD was used for this area. Results are presented below (in
hectares). The total land area of all U.S. Territories is 1.05 million hectares, representing 0.1 percent of
the total land base for the United States (see Table 6-7).
Table 6-7: Total Land Area (Hectares) by Land Use Category for U.S. Territories
Puerto Rico
U.S. Virgin
Islands
Guam
Northern
Marianas
Islands
American
Samoa
Total
Cropland
19,712
138
236
289
389
20,764
Forest Land
404,004
13,107
24,650
25,761
15,440
482,962
Grasslands
299,714
12,148
15,449
13,636
1,830
342,777
Other Land
5,502
1,006
1,141
5,186
298
13,133
Settlements
130,330
7,650
11,146
3,637
1,734
154,496
Wetlands
24,525
4,748
1,633
260
87
31,252
Total
883,788
38,796
54,255
48,769
19,777
1,045,385
Note: Totals may not sum due to independent rounding.
6.2 Forest Land Remaining Forest Land
(Source Category 4A1)
Changes in Forest Carbon Stocks (Source Category 4A1)
Delineation of Carbon Pools
For estimating carbon stocks or stock change (flux), carbon in forest ecosystems can be divided into the
following five storage pools (IPCC 2006):
• Aboveground biomass, which includes all living biomass above the soil including stem, stump,
branches, bark, seeds, and foliage. This category includes live understory.
• Belowground biomass, which includes all living biomass of coarse living roots greater than 2
millimeters (mm) diameter.
• Dead wood, which includes all non-living woody biomass either standing, lying on the ground
(but not including litter), or in the soil.
• Litter, which includes all duff, humus, and fine woody debris above the mineral soil as well as
woody fragments with diameters of up to 7.5 cm.
• Soil organic carbon (SOC), including all organic material in soil to a depth of 1 meter but
excluding the coarse roots of the belowground pools. Organic (e.g., peat and muck) soils have a
minimum of 12 to 20 percent organic matter by mass and develop under poorly drained
Land Use, Land-Use Change, and Forestry 6-29
-------
conditions of wetlands. All other soils are classified as mineral soil types and typically have
relatively low amounts of organic matter.
In addition, there are two harvested wood pools included when estimating carbon flux:
• Harvested wood products (HWP) in use.
• HWP in solid waste disposal sites (SWDS).
Forest Carbon Cycle
Carbon is continuously cycled among the previously defined carbon storage pools and the atmosphere
as a result of biogeochemical processes in forests (e.g., photosynthesis, respiration, decomposition,
and disturbances such as fires or pest outbreaks) and anthropogenic activities (e.g., harvesting,
thinning, and replanting). As trees photosynthesize and grow, carbon is removed from the atmosphere
and stored in living tree biomass. As trees die and otherwise deposit litter and debris on the forest floor,
carbon is released to the atmosphere and is also transferred to the litter, dead wood, and soil pools by
organisms that facilitate decomposition.
The net change in forest carbon is not equivalent to the net flux between forests and the atmosphere
because timber harvests do not cause an immediate flux of all harvested biomass carbon to the
atmosphere. Instead, harvesting transfers a portion of the carbon stored in wood to a "product pool."
Once in a product pool, the carbon is emitted over time as C02 in the case of decomposition and as
C02, CH4, N20, CO, and NOxwhen the wood product combusts. The rate of emission varies considerably
among different product pools. For example, if timber is harvested to produce energy, combustion
releases carbon immediately, and these emissions are reported for information purposes in the Energy
sector while the harvest (i.e., the associated reduction in forest carbon stocks) and subsequent
combustion are implicitly estimated in the Land Use, Land-Use Change, and Forestry (LULUCF) sector
(i.e., the portion of harvested timber combusted to produce energy does not enter the HWP pools).
Conversely, if timber is harvested and used as lumber in a house, it may be many decades or even
centuries before the lumber decays and carbon is released to the atmosphere. If wood products are
disposed of in SWDS, the carbon contained in the wood may be released many years or decades later or
may be stored almost permanently in the SWDS. These latter fluxes, with the exception of CH4 from
wood in SWDS, which is included in the Waste sector, are also estimated in the LULUCF sector.
Net Change in Carbon Stocks within Forest Land of the United States
This section describes the general method for quantifying the net changes in carbon stocks in the five
carbon storage pools and two harvested wood pools (a more detailed description of the methods and
data is provided in Annex 3.14). The underlying methodology for determining carbon stock and stock
change relies on data from the nationwide forest inventory (NFI) conducted by the Forest Inventory and
Analysis (FIA) program within the USDA Forest Service. The annual NFI is implemented across all U.S.
forest lands within the conterminous 48 states, Alaska, Puerto Rico, and the U.S. Virgin Islands, and
periodic inventories are available for Hawaii and some of the other U.S. Territories. The methods for
estimation and monitoring are continuously improved and these improvements are reflected in the
carbon estimates (Domke et al. 2022; Westfall et al. 2024). First, in the conterminous 48 states and
coastal southeast and southcentral Alaska, the total carbon stocks are estimated for each carbon
storage pool at the individual NFI plot, next the annual net changes in carbon stocks for each pool at the
population level are estimated, and then the changes in stocks are summed for all pools to estimate
6-30 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
total net flux at the population level (e.g., U.S. state). Changes in carbon stocks from disturbances, such
natural disturbances (e.g., wildfires, insects/disease, wind) or harvesting, are included in the net
changes (see Box 6-2 for more information). For instance, an inventory conducted after a fire implicitly
includes only the carbon stocks remaining on the NFI plot. The IPCC (2006) recommends estimating
changes in carbon stocks from forest lands according to several land-use types and conversions,
specifically forest land remaining forest land and land converted to forest land, with the former being
lands that have been forest lands for 20 years or longer and the latter being lands (i.e., croplands,
grassland, wetlands, settlements and other lands) that have been converted to forest lands for less than
20 years.
The methods and data used to delineate forest carbon stock changes by these two categories continue
to improve and in order to facilitate this delineation, a combination of estimation approaches was used
to compile estimates in this Inventory. Methods for compiling carbon stocks and stock changes on
forest land in interior Alaska are different from those used for estimation in the conterminous U.S. and
coastal Alaska due to the recency of the operational FIA inventory in that region and differences in
sampling protocols (see Annex 3.14 for more details). Finally, estimates of carbon stocks and stock
changes on forest land in Hawaii and the U.S. Territories of American Samoa, Guam, Northern Mariana
Islands, Puerto Rico, and the U.S. Virgin Islands (hereafter referred to as the U.S. Territories) are also
included in this Inventory. The FIA program has conducted annual inventories in parts of Puerto Rico
(Mainland, Vieques, Culebra) and the U.S. Virgin Islands and periodic inventories in Hawaii, American
Samoa, Guam, Northern Mariana Islands, and Puerto Rico (Mona Island). These inventories in
combination with published estimates of carbon stocks, stock changes, and IPCC (2019) default
estimates were used to compile estimates of carbon stocks and stock changes on forest land for these
regions (see Annex 3.14 for more details).
Forest Area in the United States
Approximately 32 percent of the managed U.S. land area is estimated to be forested based on the U.S.
definition of forest land as provided in Section 6.1. All annual and periodic NFI plots included in the
public FIA database as of August 2024 (which includes data collected through 2023 - note that the
COVID 19 pandemic resulted in delays in data collection in many states) were used in this Inventory. The
NFIs from the conterminous United States (USDA Forest Service 2024a, 2024b), Alaska, Hawaii, and the
U.S. Territories comprise an estimated 280 million hectares of forest land that are considered managed
and are included in the current Inventory. Some differences also exist in forest land area estimates from
the latest update to the Resources Planning Act (RPA) Assessment (Oswalt et al. 2019) and the forest
land area estimates included in this report, which are based on the annual and periodic NFI data
through 2023 for all states (USDA Forest Service 2024b; Nelson et al. 2020). The methods for compiling
area estimates for Hawaii and the U.S. Territories in this section are different from those in Section 6.1
because they do not rely on FIA data. Also, it is not possible to separate forest land remaining forest land
from land converted to forest land in Wyoming because of the split annual cycle method used for
population estimation (see Annex 3.14). This prevents harmonization of forest land in Wyoming with the
NRI/NLCD method used in Section 6.1. Agroforestry systems that meet the definition of forest land are
also not currently included in the current Inventory since they are not explicitly inventoried (i.e.,
classified as an agroforestry system) by either the FIA program or the Natural Resources Inventory (NRI)
of the USDA Natural Resources Conservation Service (Perry et al. 2005).
Land Use, Land-Use Change, and Forestry 6-31
-------
An estimated 67 percent (208 million hectares) of U.S. forests in Alaska, Hawaii and the conterminous
United States are classified as timberland, meaning they meet minimum levels of productivity and have
not been removed from production. Approximately ten percent of Alaska forest land and 73 percent of
forest land in the conterminous United States are classified as timberland. Of the remaining non-
timberland in the conterminous United States, Alaska, and Hawaii, nearly 33 million hectares are
reserved forest lands (withdrawn by law from management for production of wood products) and 102
million hectares are lower productivity forest lands (Oswalt et al. 2019). Historically, the timberlands in
the conterminous United States have been more frequently or intensively surveyed than the forest lands
removed from production because they do not meet the minimum level of productivity.
Since the late 1980s, gross forest land area in Alaska, Hawaii, and the conterminous United States has
increased by about 13 million hectares (Oswalt et al. 2019). The southern region of the United States
contains the most forest land (Figure 6-4). A substantial portion of this accrued forest land is from the
conversion of abandoned croplands to forest (e.g., Woodall et al. 2015b). Estimated forest land area in
the conterminous United States and Alaska represented in this Inventory is stable, but there are
substantial conversions as described in 6.1 and each of the land conversion sections for each land-use
category (e.g., land converted to cropland, land converted to grassland). The major influences on the net
carbon flux from forest land across the 1990 to 2023 time series are management activities, natural
disturbance, particularly wildfire, and the ongoing impacts of current and previous land-use
conversions. These activities affect the net flux of carbon by altering the amount of carbon stored in
forest ecosystems and also the area converted to forest land. For example, intensified management of
forests that leads to an increased rate of growth of aboveground biomass (and possible changes to the
other carbon storage pools) may increase the eventual biomass density of the forest, thereby increasing
the uptake and storage of carbon in the aboveground biomass pool.27 Though harvesting forests
removes much of the carbon in aboveground biomass (and possibly changes carbon density in other
pools), on average, the estimated volume of annual net growth in aboveground tree biomass in the
conterminous United States is essentially twice the volume of annual removals on timberlands (Oswalt
et al. 2019). The net effects of forest management and changes in forest land remaining forest land are
captured in the estimates of carbon stocks and fluxes presented in this section.
27 The term "biomass density" refers to the mass of live vegetation per unit area. It is usually measured on a dry-weight
basis. Species-specific carbon fractions are used to convert dry biomass to carbon (Westfall et al. 2024).
6-32 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Figure 6-4: Area by Region for Forest Land Remaining Forest Land in the United States
(1990-2023)
100—,
$ 80-
TO
o
CD
o
ro
o
cc
"So
(D
60-
40-
20-
' South
• North
Pacific
Coast
Rocky
Mountain
1 | I I I I | I I I I | I I I i | i i i i | i i i i | I I I I | I I I
1990 1995 2000 2005 2010 2015 2020
Year
Forest Carbon Stocks and Stock Change
In the forest land remaining forest land category, forest management practices, the regeneration of
forest areas cleared more than 20 years prior to the inventory year, and timber harvesting have resulted
in net removal (i.e., net sequestration or accumulation) of carbon each year from 1990 through 2023.
The rate of forest clearing in the 17th century following European settlement had slowed by the late 19th
century. Through the later part of the 20th century, many areas of previously forested land in the United
States were allowed to revert to forests or were actively reforested. The impacts of these land-use
changes still influence carbon fluxes from these forest lands. More recently, the 1970s and 1980s saw a
resurgence of federally sponsored forest management programs (e.g., the Forestry Incentive Program)
and soil conservation programs (e.g., the Conservation Reserve Program), which have focused on tree
planting, improving timber management activities, combating soil erosion, and converting marginal
cropland to forests. In addition to forest regeneration and management, forest harvests and natural
disturbance have also affected net carbon fluxes. Because most of the timber harvested from U.S. forest
land is used in wood products, and many discarded wood products are disposed of in SWDS rather than
by incineration, substantial quantities of carbon in harvested wood are transferred to these long-term
storage pools rather than being released rapidly to the atmosphere (Skog 2008). By maintaining current
Land Use, Land-Use Change, and Forestry 6-33
-------
harvesting practices and regeneration activities on forest lands, along with continued input of harvested
wood into the HWP pool, carbon stocks in the forest land remaining forest land category are likely to
continue to increase in the near term, though possibly at a slower rate.
Changes in carbon stocks in the forest ecosystem and harvested wood pools associated with forest land
remaining forest land were estimated to result in net removal of 880.0 MMT C02Eq. (240.0 MMT carbon)
in 2023 (Table 6-8, Table 6-9, Table A-202, Table A-203 and state-level estimates in Table A-206). The
estimated net uptake of carbon in the Forest Ecosystem was 789.1 MMT C02 Eq. (215.2 MMT carbon) in
2023 (Table 6-8 and Table 6-9). The majority of this uptake in 2023, 493.4 MMT C02 Eq. (134.6 MMT
carbon), was from aboveground biomass. Overall, estimates of average carbon density in forest
ecosystems (including all pools) increased consistently over the time series with an average of
approximately 210.5 MT carbon ha1 from 1990 to 2023. This was calculated by dividing the forest
ecosystem carbon stock estimates by the forest land area estimates for every year (see Table 6-10 and
Table A-204) and then calculating the mean across the entire time series, i.e., 1990 through 2023. The
increasing forest ecosystem carbon density, when combined with relatively stable forest area, results in
net carbon accumulation over time. However, due to an aging forest land base, increases in the
frequency and severity of disturbances in forests in some regions, among other drivers of change, forest
carbon density is increasing at a slower rate resulting in an overall decline in the sink strength of forest
land remaining forest land in the United States. Aboveground live biomass is responsible for the majority
of net carbon uptake among all forest ecosystem pools (Figure 6-5). These increases may be influenced
in some regions by reductions in carbon density or forest land area due to natural disturbances (e.g.,
wildfire, weather, insects/disease), particularly in Alaska. The inclusion of all managed forest land in
Alaska has increased the interannual variability in carbon stock change estimates over the time series,
and much of this variability can be attributed to severe fire years (e.g., 2022). The distribution of carbon
in forest ecosystems in Alaska is substantially different from forests in the conterminous United States.
In Alaska, more than nine percent of forest ecosystem carbon is stored in the litter carbon pool whereas
in the conterminous United States, less than seven percent of the total ecosystem carbon stocks are in
the litter pool. Much of the litter material in forest ecosystems is combusted during fire (IPCC 2006)
leading to substantial carbon losses in this pool during severe fire years (Figure 6-5, Table A-211).
The estimated net accumulation of carbon in the HWP pool, i.e., the balance of additions from the
transfer of harvested wood from the forest ecosystem and losses from the current decay of wood
harvested in the past, was 90.9 MMT C02 Eq. (24.8 MMT carbon) in 2023 (Table 6-8, Table 6-9, Tables A-
197 through Table A-199). The majority of this uptake, 63.5 MMT C02 Eq. (17.3 MMT carbon), was from
solid wood and paper in SWDS. Products in use accounted for an estimated 27.4 MMT C02 Eq. (7.5 MMT
carbon) in 2023. Harvested wood estimates are based on results from annual surveys (see Annex 3.14,
Table A-196) and models (see Methodology section).
For drained organic soils, all estimated fluxes and stocks include both forest land remaining forest land
and land converted to forest land. CH4 and N20 emissions are also calculated for this pool (see Table
6-26 and Table 6-27, and Methodology section below on emissions from drained organic soils).
Forest land area varies slightly (less than 0.5 million hectares) between this section and Section 6.1 due
to the use of FIA estimates for Hawaii and U.S. Territories in this section. Differences also exist because
forest land area estimates are based on the latest NFI data through 2023, and woodland areas
previously included as forest land have been separated and included in the grassland categories in this
Inventory. Also, it is not possible to separate forest land remaining forest land from land converted to
forest land in Wyoming because of the split annual cycle method used for population estimation which
6-34 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
prevents harmonization of forest land in Wyomingwith the NRI/NLCD method used in Section 6.1. See
Annex 3.14, Table A-205 for annual differences between the forest area reported in Section 6.1 and
Section 6.2. Forest ecosystem estimates in this section do not include agroforestry or settlement trees.
Estimates for managed forest land in interior Alaska, Hawaii, and the U.S. Territories were compiled
using the gain-loss method as described in Annex 3.14.
Table 6-8: Net C02 Flux from Forest Ecosystem Pools in Forest Land Remaining Forest
Land and Harvested Wood Pools (MMT C02 Eq.)
Carbon Pool
1990 | 2005 | 2019
2020
2021
2022
2023
Forest Ecosystem
(931.1) I
(843.9) I
(790.2)
(829.6)
(813.2)
(749.6)
(789.1)
Aboveground Biomass
(605.4)
(554.3)
(512.8)
(519.4)
(510.1)
(487.3)
(493.4)
Belowground Biomass
(117.3)
(107.6)
(101.1)
(101.4)
(99.8)
(95.9)
(96.4)
Dead Wood
(210.1)
(203.5)
(202.2)
(205.4)
(203.3)
(197.5)
(199.7)
Litter
(3.3)
17.9
25.6
(3.9)
(0.5)
30.6
(0.0)
Soil (Mineral)
4.9
3.2
(0.6)
(0.5)
(0.5)
(0.5)
(0.5)
Soil (Organic)
(0.7)
(0.4)
0.2
0.2
0.2
0.2
0.2
Drained Organic Soil
0.8
0.8
0.8
0.8
0.8
0.8
0.8
Harvested Wood
(123.8)
(106.0)
(86.9)
(96.8)
(94.7)
(92.8)
(90.9)
Products in Use
(54.8)
(42.6)
(22.6)
(32.3)
(30.4)
(28.8)
(27.4)
SWDS
(69.0) |
(63.4)
(64.3)
(64.5)
(64.3)
(63.9)
(63.5)
Total Net Flux
(1,054.9) |
(950.0) |
(877.1)
(926.5)
(907.9)
(842.4)
(880.0)
Notes: Parentheses indicate net carbon uptake (i.e., a net removal of carbon from the atmosphere). Total net flux is an estimate
of the actual net flux between the total forest carbon pool and the atmosphere. Totals may not sum due to independent
rounding.
Table 6-9: Net Carbon Flux from Forest Ecosystem Pools in Forest Land Remaining
Forest Land and Harvested Wood Pools (MMT C)
Carbon Pool
1990
2005 |
2019
2020
2021
2022
2023
Forest Ecosystem
(253.9)
(230.2)
(215.5)
(226.3)
(221.8)
(204.4)
(215.2)
Aboveground Biomass
(165.1)
(151.2)
(139.9)
(141.7)
(139.1)
(132.9)
(134.6)
Belowground Biomass
(32.0)
(29.4)
(27.6)
(27.7)
(27.2)
(26.2)
(26.3)
Dead Wood
(57.3)
(55.5)
(55.2)
(56.0)
(55.4)
(53.9)
(54.5)
Litter
(0.9)
4.9
7.0
(1.1)
(0.1)
8.4
(0.0)
Soil (Mineral)
1.3
0.9
(0.2)
(0.1)
(0.1)
(0.1)
(0.1)
Soil (Organic)
(0.2)
(0.1)
0.1
0.1
0.1
0.0
0.0
Drained Organic Soil
0.2
0.2
0.2
0.2
0.2
0.2
0.2
Harvested Wood
(33.8)
(28.9)
(23.7)
(26.4)
(25.8)
(25.3)
(24.8)
Products in Use
(14.9)
(11.6)
(6.2)
(8.8)
(8.3)
(7.9)
(7.5)
SWDS
(18.8) |
(17.3)
(17.5)
(17.6)
(17.5)
(17.4)
(17.3)
Total Net Flux
(287.7)
(259.1) |
(239.2)
(252.7)
(247.6)
(229.7)
(240.0)
Notes: Parentheses indicate net carbon uptake (i.e., a net removal of carbon from the atmosphere). Total net flux is an estimate
of the actual net flux between the total forest carbon pool and the atmosphere. Totals may not sum due to independent
rounding.
Land Use, Land-Use Change, and Forestry 6-35
-------
Stock estimates for forest ecosystem and harvested wood carbon storage pools are presented in Table
6-10. Together, the estimated aboveground biomass and soil carbon pools account for a large proportion
of total forest ecosystem carbon stocks.28
Table 6-10: Forest Area (1,000 ha) and Carbon Stocks in Forest Land Remaining Forest
Land and Harvested Wood Pools (MMT C)
1990
2005
2020
2021
2022
2023
2024
Forest Area (1,000 ha)
281,632
280,734
280,358
280,307
280,250
280,195
280,139
Carbon Pools (MMT C)
Forest Ecosystem
54,984
58,693
62,148
62,374
62,596
62,801
63,016
Aboveground Biomass
12,614
15,009
17,212
17,354
17,493
17,626
17,761
Belowground Biomass
2,240
2,703
3,133
3,160
3,188
3,214
3,240
Dead Wood
2,743
3,596
4,439
4,495
4,551
4,604
4,659
Litter
3,724
3,736
3,722
3,723
3,723
3,714
3,714
Soil (Mineral)
28,132
28,114
28,108
28,108
28,108
28,108
28,108
Soil (Organic)
5,531
5,534
5,534
5,534
5,534
5,534
5,534
Harvested Wood
1,895
2,353
2,694
2,721
2,747
2,772
2,797
Products in Use
1,249
1,447
1,530
1,538
1,547
1,555
1,562
SWDS
646
906
1,165
1,182
1,200
1,217
1,235
Total C Stock
56,879
61,046
64,842
65,095
65,343
65,573
65,813
Notes: Totals may not sum due to independent rounding. Population estimates compiled using FIA data are assumed to
represent stocks as of January 1 of the inventory year. Flux is the net annual change in stock. Thus, an estimate of flux for 2023
requires estimates of carbon stocks for 2023 and 2024.
28 See Annex 3.14, Table A-205 for annual differences between the forest area reported in Section 6.1 Representation of the
U.S. Land Base and Section 6.2 Forest Land Remaining Forest Land.
6-36 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Figure 6-5: Estimated Net Annual Changes in Carbon Stocks for All Carbon Pools in
Forest Land Remaining Forest Land in the United States (1990-2023)
25 -|
¦55
c
"O
c
I-
£
cn
c
'g
tc
E
£
"O
c
OS
o-
-25-
-50-
-75-
-100-
-125-
-150-
-175-
-200-
-225-
-250-
-275-
-300-
All forest ecosystem pools
Aboveground biomass
¦ Belowground biomass
Dead wood
¦ Litter
Soil (mineral)
i i | i i i i | i
2005 2010
Year
i i i | i i
2015
1 I 1
2020
• Soil (organic)
Drained organic soil
¦ Harvested wood products (HWP)
Products in use
Solid waste disposal sites
Total net change
(forest ecosystem + HWP)
Box 6-2: C02 Emissions from Forest Fires
As stated previously, the forest inventory approach implicitly includes all carbon losses due to
disturbances such as forest fires, because only carbon remaining in the forest is estimated. Net carbon
stock change is estimated by subtracting consecutive carbon stock estimates. A forest fire disturbance
removes carbon from the forest. The inventory data from the NFI on which net carbon stock estimates
are based already reflect this carbon loss. Therefore, estimates of net annual changes in carbon stocks
for U.S. forest land already includes C02 emissions from forest fires occurring in the conterminous
states (48 states), Hawaii, Puerto Rico, and Guam as well as the portion of managed forest lands in
Alaska. Because it is of interest to quantify the magnitude of C02 emissions from fire disturbance, these
separate estimates are highlighted here. Note that these C02 estimates are based on the same
methodology as applied for the non-C02 greenhouse gas emissions from forest fires that are also
quantified in a separate section below for consistency with IPCC Guidance.
Emissions estimates are developed using IPCC (2019) methodology and based on U.S.-specific data
and models to quantify the primary fire-specific components: area burned; availability and
combustibility of fuel; fire severity (or consumption); and C02 and non-C02 emissions. Estimated C02
emissions for fires on forest lands in the United States for 2023 are 55.5 MMT C02 per year (Table 6-11).
This estimate is an embedded component of the net annual forest carbon stock change estimates
provided previously (i.e., Table 6-9), but this separate approach to estimating C02 emissions is
Land Use, Land-Use Change, and Forestry 6-37
-------
necessary in order to associate these emissions with fire. See the discussion in Annex 3.14 for more
details on this methodology (Smith et al. 2024). Note that in Alaska, a portion of the forest lands are
considered unmanaged, therefore the estimates for Alaska provided in Table 6-11 include only managed
forest land within the state, which is consistent with carbon stock change estimates provided above.
Table 6-11: Estimates of C02 (MMT per Year) Emissions3 from Forest Fires in the
Conterminous 48 States, Hawaii, Puerto Rico, Guam, and Alaska
1990 2005 2019 2020 2021 2022 2023
CO2 emitted from fires on forest land in 11.2 I 26.8 I 20.9 227.7 210.2 65.3 41.1
the Conterminous 48 States, Hawaii,
Puerto Rico, and Guam (MMTyr1)
CO2 emitted from fires on managed 40.2 127.9 66.2 1.1 7.1 60.4 14.5
forest land in Alaska (MM Tyr1)
Total CO2 emitted (MMTyr1) 51.4 154.7 87.1 228.8 217.3 125.7 55.5
aThese emissions have already been included in the estimates of net annual changes in carbon stocks, which include the
amount sequestered minus any emissions, including the assumption that combusted wood may continue to decay through time.
Note: Totals may not sum due to independent rounding.
Methodology and Time-Series Consistency
The methodology described herein is consistent with the 2006IPCC Guidelines for National Inventories.
Forest ecosystem carbon stocks and net annual carbon stock change were determined according to the
stock-difference method for the conterminous United States and coastal southeast and southcentral
Alaska, which involved applying carbon estimation factors to annual forest inventories across time to
obtain carbon stocks and then subtracting between the years to obtain the stock change. The gain-loss
method was used to estimate carbon stocks and net annual carbon stock changes in interior Alaska,
Hawaii, and the U.S. Territories. The approaches for estimating carbon stocks and stock changes on
forest land remaining forest land are described in Annex 3.14. All annual and periodic NFI plots available
in the public FIA database (USDA Forest Service 2024b) were used in the current Inventory. Additionally,
NFI plots established and measured in 2014 as part of a pilot inventory in interior Alaska were also
included in this Inventory as were plots established and measured since 2015 as part of the operational
NFI in interior Alaska. Some of the data from the pilot and operational NFI in interior Alaska are notyet
available in the public FIA database. Only plots which meet the definition of forest land (see Section 6.1)
are measured in the NFI; as part of the pre-field process in the FIA program, all plots or portions of plots
(i.e., conditions) are classified into a land-use category. This land use information on each forest and
non-forest plot was used to estimate forest land area and land converted to and from forest land over
the time series. The estimates in this section of the report are based on land use information from the
NFI and they may differ from the other land-use categories where area estimates reported in the Land
Representation were not updated (see Section 6.1). Further, managed forest land area estimates for
Hawaii and the U.S. Territories were compiled using FIA data in this section which is different from how
estimates for these lands were compiled in Section 6.1 (see Annex 3.14 for details on differences).
To implement the stock-difference approach, forest land conditions in the conterminous United States
and coastal Alaska were observed on NFI plots at time t0 and at a subsequent time ti=t0+s, where s is the
time step (time measured in years) and is indexed by discrete (e.g., 5 year) forest age classes. The
inventory from t0 to ti was then projected to 2023. This projection approach requires simulating changes
6-38 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
in the age-class distribution resulting from forest aging and disturbance events and then applying
carbon density estimates for each age class to obtain population estimates for the nation. In cases
where there are ti estimates in the last year (e.g., 2023) of the NFI no projections are necessary for those
plots.
To implement the gain-loss approach in interior Alaska, forest land conditions in Alaska were observed
on NFI plots from 2014 to 2023. Plot-level data from the NFI were harmonized with auxiliary data
describing climate, forest structure, disturbance, and other site-specific conditions to develop non-
parametric models to predict carbon stocks by forest ecosystem carbon pool as well as fluxes over the
entire inventory period, 1990 to 2023. First, carbon stocks for each forest ecosystem carbon pool were
predicted for the year 2016 for all NFI plot locations (each plot representing 12,015 ha). Next, the
chronosequence of sampled NFI plots and auxiliary information (e.g., climate, forest structure,
disturbance, and other site-specific data) were used to predict annual gains and losses for each forest
ecosystem carbon pool. The annual gains and losses were then combined with the stock estimates and
disturbance information to compile plot- and population-level carbon stocks and fluxes for each year
from 1990 to 2023.
To implement the gain-loss approach in Hawaii and the U.S. Territories, a combination of Tier 1 and Tier 2
methods were applied. All forest land conditions were observed on annual and periodic NFI plots from
2001 to 2019 (see Annex 3.14 for specific inventories included for each Island). Plot-level data from the
NFI were harmonized with data describing ecological zone (FAO 2010), soil attributes (Johnson and Kern
2003; Deenik and McClellan, 2007, IPCC 2019), and dead wood and litter carbon stocks (Oswalt et al.
2008; IPCC 2019). Only estimates of carbon stocks in live trees were consistently available in the NFI for
Hawaii and the U.S. Territories for each inventory. These estimates were used to obtain average annual
carbon stock change estimates for above and belowground live trees which were applied to each forest
plot to capture growth, harvest removals, and mortality. The carbon stocks and annual stock change
estimates were compared with country-specific estimates (Oswalt et al. 2008; Selmants et al. 2017),
and IPCC (2019) default estimates to ensure they were consistent with other sources. There were limited
data available on disturbances and management activities on NFI plots over the times series so Tier 1
methods were applied for dead wood and litter. It was assumed that the average transfer rate into dead
wood and litter pools is equivalent to the average transfer rate out of the dead organic matter pool so
there are no net carbon stock changes included for these pools in the time series (IPCC 2006). Similarly,
given data limitations on forest soils and changes on NFI plots over the time series, a Tier 1 approach
was also used for soil carbon with country-specific estimates (Johnson and Kern 2003) and IPCC (2019)
defaults used to estimate soil carbon stocks with no net carbon stock change reported.
To estimate carbon stock changes in harvested wood, estimates were based on factors such as the
allocation of wood to various primary and end-use products as well as half-life (the time at which half of
the amount placed in use will have been discarded from use) and expected disposition (e.g., product
pool, SWDS, combustion). An overview of the different methodologies and data sources used to
estimate the carbon in forest ecosystems within the conterminous United States and Alaska and
harvested wood products for all of the United States is provided below. See Annex 3.14 for details and
additional information related to the methods and data.
Methodological recalculations were applied to the entire time series to ensure time-series consistency
from 1990 through 2023. Details on the emission/removal trends and methodologies through time are
described in more detail in the Introduction and Methodology sections.
Land Use, Land-Use Change, and Forestry 6-39
-------
Forest Ecosystem Carbon from Forest Inventory
The United States applied the compilation approach described in Woodall et a I. (2015a) for the current
Inventory which removes the older periodic inventory data, which may be inconsistent with annual
inventory data, from the estimation procedures. This approach enables the attribution of forest carbon
accumulation by forest growth, land-use change, and natural disturbances such as fire. Development
will continue on a system that attributes changes in forest carbon to disturbances and delineates land
converted to forest land from forest land remaining forest land. As part of this development, carbon pool
science will continue and will be expanded to improve the estimates of carbon stock transfers from
forest land to other land uses and include techniques to better identify land-use change (see the
Planned Improvements section below).
The annual FIA inventory system does not extend into the 1970s, necessitating the adoption of a system
to estimate carbon stocks prior to the establishment of the annual forest inventory. The estimation of
carbon stocks prior to the annual national forest inventory consisted of a modeling framework
comprised of a forest dynamics module (age transition matrices) and a land use dynamics module (land
area transition matrices). The forest dynamics module assesses forest uptake, forest aging, and
disturbance effects (e.g., disturbances such as wind, fire, and floods identified by foresters on inventory
plots). The land use dynamics module assesses carbon stock transfers associated with afforestation
and deforestation (Woodall et al. 2015b). Both modules are developed from land use area statistics and
carbon stock change or carbon stock transfer by age class. The required inputs are estimated from more
than 625,000 forest and non-forest observations recorded in the FIA national database (U.S. Forest
Service 2024a, b, c). Model predictions prior to the annual inventory period are constructed from the
estimation system using the annual estimates. The estimation system is driven by the annual forest
inventory system conducted by the FIA program (Frayer and Furnival 1999; Bechtold and Patterson 2005;
Westfall et al. 2024; USDA Forest Service 2024d, 2024a). The FIA program relies on a rotating panel
statistical design with a sampling intensity of one 674.5 m2 ground plot per 2,403 ha of land and water
area. A five or seven-panel design, with 20 percent or 14.3 percent of the field plots typically measured
each year within a state, is used in the eastern United States, and a ten-panel design, with typically ten
percent of the field plots measured each year within a state, is used in the western United States. The
interpenetrating hexagonal design across the U.S. landscape enables the sampling of plots at various
intensities in a spatially and temporally unbiased manner. Typically, tree and site attributes are
measured with higher sample intensity while other ecosystem attributes such as downed dead wood
are sampled during summer months at lower intensities. The first step in incorporating FIA data into the
estimation system is to identify annual and periodic inventory datasets by state and U.S. Territory.
Inventories include data collected on permanent inventory plots on forest lands and were organized as
separate datasets, each representing a complete inventory, or survey, of an individual state at a
specified time. Many of the annual inventories reported for states are represented as "moving window"
averages, which mean that a portion—but not all—of the previous year's inventory is updated each year
(USDA Forest Service 2024d). Forest carbon estimates are organized according to these state surveys,
and the frequency of surveys varies by state.
Using this FIA data, separate estimates were prepared for the five carbon storage pools identified by
IPCC (2006) as described above. All estimates for the conterminous United States and Alaska were
based on data collected from the extensive array of permanent, annual forest inventory plots and
associated models (e.g., live tree belowground biomass) in the United States (USDA Forest Service
2024b, 2024c). Carbon conversion factors were applied at the disaggregated level of each inventory plot
6-40 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
and then appropriately expanded to population estimates. Only live (and in some cases) standing dead
wood estimates are available in the annual and periodic FIA inventories in Hawaii and the U.S.
Territories. For this reason, a combination of approaches was used to obtain estimates for all carbon
pools for the time series in these locations.
Carbon in Biomass
Live tree carbon pools include aboveground and belowground (coarse root) biomass of live trees with
diameter at breast height (dbh) of at least 2.54 cm at 1.37 m above the litter. Separate estimates were
made for above- and belowground biomass components. Over the last decade, the USDA Forest
Service's FIA program and collaborators from universities and industry have been developing a national
methodology for the prediction of individual-tree volume, biomass, and carbon content. The resulting
methodology is referred to as the National-Scale Volume and Biomass (NSVB) framework (Westfall et al.
2024). The previous methodology used was the Component Ratio Method (CRM) (Woodall et al. 2010).
While CRM was nationally consistent, tree biomass was still based on the volume predicted by regional
models and tree carbon was assumed to be 50-percent of biomass, regardless of species. Hence, the
need for NSVB, a nationally consistent methodology for compatible predictions of tree volume,
biomass, and carbon content (Westfall et al. 2024).
The NSVB covers timber tree species in the conterminous United States and coastal Alaska. All other
trees (i.e., trees that are woodland species and trees within Pacific and Caribbean Islands) use regional
models for volume and biomass, with updated carbon fractions (when available). While NSVB did not
directly update models for trees that are considered woodland species or trees within the Pacific (USDA
Forest Service 2022a, b) and Caribbean Islands (collectively referred to hereafter as "non-NSVB trees"),
volume, biomass, and carbon estimates for these trees changed compared to the CRM. For non-NSVB
trees, the standardization of tree defects and how variables are reported (i.e., whether models for total-
stem or merchantable-bole volumes are available) may be reflected as differences in volume estimates.
Additionally, biomass estimates for non-NSVB trees are based on regional biomass models and no
longer are adjusted as they were under the CRM. Finally, updates to carbon fractions (when available)
and calculation of aboveground biomass are reflected in aboveground and belowground biomass
carbon estimates (see Annex 3.14 for more details).
Understory vegetation is a minor component of biomass, which is defined in the FIA program as all
biomass of undergrowth plants in a forest, including woody shrubs and trees less than 2.54 cm dbh. For
this Inventory, it was assumed that ten percent of total understory carbon mass is belowground (Smith
et al. 2006). Estimates of carbon density were based on information in Birdsey (1996) and tree biomass
estimates from the FIA database. Understory biomass represented over one percent of carbon in
biomass, but its contribution rarely exceeded two percent of the total carbon stocks or stock changes
across all forest ecosystem carbon pools each year.
Carbon in Dead Organic Matter
Dead organic matter is calculated as three separate pools—standing dead trees, downed dead wood,
and litter—with carbon stocks estimated from sample data or from models as described below. The
standing dead tree carbon pool includes aboveground and belowground (coarse root) biomass for trees
of at least 2.54 cm dbh. Calculations followed the basic methods applied to live trees (Westfall eta I.
2024) with additional modifications to account for decay and structural loss (Harmon et al. 2011).
Downed dead wood estimates are based on measurement of a subset of NFI plots for downed dead
Land Use, Land-Use Change, and Forestry 6-41
-------
wood (Domke et al. 2013; Woodall and Monleon 2008; Woodall et al. 2013). Downed dead wood is
defined as pieces of dead wood greater than 7.5 cm diameter, at transect intersection, that are not
attached to live or standing dead trees. This includes stumps and roots of harvested trees. As described
in the planned improvements section of the previous Inventory (1990 through 2022), the modeling
framework used to estimate downed dead wood within the dead organic matter carbon pool had not
been implemented. In this Inventory, the downed dead wood modeling framework has been
implemented to ensure consistency between the Inventory and the FIA database (Westfall et al. 2024);
see Recalculations Discussion for more information. To facilitate the downscaling of downed dead wood
carbon estimates from the state-wide population estimates to individual plots, downed dead wood
models specific to regions and forest types within each region are used. Litter carbon is the pool of
organic carbon (also known as duff, humus, and fine woody debris) above the mineral soil and includes
woody fragments with diameters of up to 7.5 cm. A subset of NFI plots are measured for litter carbon. A
modeling approach, using litter carbon measurements from NFI plots (Domke et al. 2016), was used to
estimate litter carbon for every NFI plot used in the estimation framework.
Carbon in Forest Soil
Soil is the largest terrestrial carbon sink with much of that carbon in forest ecosystems. The FIA program
has been measuring soil attributes as part of the annual inventory since 2001, and has amassed an
extensive inventory of soil measurement data on forest land in the conterminous United States and
coastal Alaska (O'Neill et al. 2005). Observations of mineral and organic soil carbon on forest land from
the FIA program and the International Soil Carbon Monitoring Network were used to develop and
implement a model framework that enabled the prediction of mineral and organic (i.e., undrained
organic soils) soil carbon to a depth of 100 cm from empirical measurements collected on sample plots
at a depth of 20 cm and included site-, stand-, and climate-specific variables thatyield predictions of
soil carbon stocks specific to forest land in the United States (Domke et al. 2017). This approach
allowed for separation of mineral and organic soils, the latter also referred to as Histosols, in the forest
land remaining forest land category. Note that mineral and organic (i.e., undrained organic soils) soil
carbon stock changes are reported to a depth of 100 cm for forest land remaining forest land to remain
consistent with previous quantification efforts in this category, however for consistency across land-use
categories, mineral (e.g., cropland, grassland, settlements) soil carbon is reported to a depth of 30 cm in
Section 6.3. Estimates of carbon stock changes from organic soils shown in Table 6-8 and Table 6-9
include the emissions from drained organic forest soils, and the methods used to develop these
estimates can be found in the Drained Organic Soils section below.
Harvested Wood Carbon
Estimates of the HWP contribution to forest carbon sinks and emissions (hereafter called "HWP
contribution") were based on methods described in Skog (2008) using the WOODCARB II model. These
methods are based on IPCC (2006) guidance for estimating the HWP contribution. IPCC (2006) provides
methods that allow for quantification of the HWP contribution using one of several different
methodological approaches: Production, stock change and atmospheric flow, as well as a default
method that assumes there is no change in HWP carbon stocks (see Annex 3.14 for more details about
each approach). The United States uses the production approach to report HWP contribution. Under the
production approach, carbon in exported wood was estimated as if it remains in the United States, and
carbon in imported wood was not included in the estimates. Though reported U.S. HWP estimates are
based on the production approach, estimates resulting from use of the two alternative approaches, the
6-42 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
stock change and atmospheric flow approaches, are also presented for comparison (see Annex 3.14).
Annual estimates of change were calculated by tracking the annual estimated additions to and removals
from the pool of products held in end uses (i.e., products in use such as housing or publications) and the
pool of products held in SWDS. The carbon loss from harvest is reported in the forest ecosystem
component of the forest land remaining forest land and land converted to forest land sections and for
informational purposes in the Energy sector, but the non-C02 emissions associated with biomass
energy are included in the Energy sector emissions (see Chapter 3). EPA includes HWP within the forest
chapter because forests are the source of wood that goes into the HWP estimates.
Solidwood products include lumber and panels. End-use categories for solidwood include single and
multifamily housing, alteration and repair of housing, and other end uses. There is one product category
and one end-use category for paper. Additions to and removals from pools were tracked beginning in
1900, with the exception of additions of softwood lumber to housing, which began in 1800. Solidwood
and paper product production and trade data were taken from USDA Forest Service and USDC Bureau of
the Census, among other sources (Hair and Ulrich 1963; Hair 1958; USDC Bureau of Census 1976;
Ulrich 1985, 1989; Steer 1948; AF&PA 2006a, 2006b; Howard 2003, 2007; Howard and Jones 2016;
Howard and Liang 2019; AF&PA 2021; AF&PA 2023; FAO 2023). Estimates for disposal of products
reflects the change over time in the fraction of products discarded to SWDS (as opposed to burning or
recycling) and the fraction of SWDS that were in sanitary landfills versus dumps.
There are five annual HWP variables that were used in varying combinations to estimate HWP
contribution using any one of the three main approaches listed above. These are:
• (1 A) annual change of carbon in wood and paper products in use in the United States,
• (1B) annual change of carbon in wood and paper products in SWDS in the United States,
• (2A) annual change of carbon in wood and paper products in use in the United States and other
countries where the wood came from trees harvested in the United States,
• (2B) annual change of carbon in wood and paper products in SWDS in the United States and
other countries where the wood came from trees harvested in the United States,
• (3) Carbon in imports of wood, pulp, and paper to the United States,
• (4) Carbon in exports of wood, pulp and paper from the United States, and
• (5) Carbon in annual harvest of wood from forests in the United States.
The sum of variables 2Aand 2Byielded the estimate for HWP contribution under the production
estimation approach. A key assumption for estimating these variables that adds uncertainty in the
estimates was that products exported from the United States and held in pools in other countries have
the same half-lives for products in use, the same percentage of discarded products going to SWDS, and
the same decay rates in SWDS as they would in the United States.
Uncertainty
A quantitative uncertainty analysis placed bounds on the flux estimates for forest ecosystems through a
combination of sample-based and model-based approaches to uncertainty estimation for forest
ecosystem C02 flux using IPCC Approach 1 (Table 6-12 and Table A-206 for state-level uncertainties). A
Monte Carlo stochastic simulation of the methods described above, and probabilistic sampling of
Land Use, Land-Use Change, and Forestry 6-43
-------
carbon conversion factors, were used to determine the HWP uncertainty using IPCC Approach 2. See
Annex 3.14 for additional information. The 2023 net annual change for forest carbon stocks was
estimated to be between -959.5 and -801.3 MMT C02 Eq. around a central estimate of-880.0 MMT C02
Eq. at a 95 percent confidence level. This includes a range of -864.4 to -713.7 MMT C02 Eq. around a
central estimate of-789.1 MMT C02 Eq. for forest ecosystems and -116.4 to -68.4 MMT C02 Eq. around
a central estimate of -90.9 MMT C02 Eq. for HWP.
Table 6-12: Quantitative Uncertainty Estimates for Net C02 Flux from Forest Land
Remaining Forest Land: Changes in Forest Carbon Stocks (MMT C02 Eq. and Percent)
Uncertainty Range Relative to Flux Estimate
2023 Flux
(MMT CO2 Eq.)
(%)
Estimate
Lower
Upper
Lower
Upper
Source
Gas
(MMT CO2 Eq.)
Bound
Bound
Bound
Bound
Forest Ecosystem C Pools®
C02
(789.1)
(864.4)
(713.7)
-9.6%
+9.6%
Harvested Wood Productsb
CO2
(90.9)
(116.4)
(68.4)
-28.1%
+24.7%
Total Forest
CO2
(880.0)
(959.5)
(801.3)
-9.0%
+8.9%
a Range of flux estimates predicted through a combination of sample-based and model-based uncertainty for a 95 percent
confidence interval, IPCC Approach 1.
b Range of flux estimates predicted by Monte Carlo stochastic simulation for a 95 percent confidence interval, IPCC Approach 2.
Notes: Parentheses indicate negative values or net uptake. Totals may not sum due to independent rounding.
QA/QC and Verification
The FIA program has conducted consistent forest surveys based on extensive statistically-based
sampling of most of the forest land in the conterminous U.S., dating back to 1952. The FIA program
includes numerous quality assurance and quality control (QA/QC) procedures, including calibration
among field crews, duplicate surveys of some plots, and systematic checking of recorded data. Because
of the statistically-based sampling, the large number of survey plots, and the quality of the data, the
survey databases developed by the FIA program form a strong foundation for carbon stock estimates.
Field sampling protocols, summary data, and detailed inventory databases are archived and are publicly
available (USDA Forest Service 2024d).
General quality control procedures were used in performing calculations to estimate carbon stocks
based on survey data. For example, the carbon datasets, which include inventory variables such as
areas and volumes, were compared to standard inventory summaries such as the forest resource
statistics of Oswalt et al. (2019) or selected population estimates generated from the FIA database,
which are available at an FIA internet site (USDA Forest Service 2024b). Agreement between the carbon
datasets and the original inventories is important to verify accuracy of the data used.
Estimates of the HWP variables and the HWP contribution under the production estimation approach
use data from USDC Bureau of the Census and USDA Forest Service surveys of production and trade,
among other sources (Hair and Ulrich 1963; Hair 1958; USDC Bureau of Census 1976; Ulrich 1985,
1989; Steer 1948; AF&PA 2006a, 2006b; Howard 2003, 2007; Howard and Jones 2016; Howard and Liang
2019; AF&PA 2021; AF&PA 2023; FAO 2023). Factors to convert wood and paper to units of carbon are
based on estimates by industry and USDA Forest Service published sources (see Annex 3.14). The
WOODCARB II model uses estimation methods suggested by IPCC (2006). Estimates of annual carbon
change in solid wood and paper products in use were calibrated to meet two independent criteria. The
6-44 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
first criterion is that the WOODCARB II model estimate of carbon in houses standing in 2001 needs to
match an independent estimate of carbon in housing based on U.S. Census and USDA Forest Service
survey data. Meeting the first criterion resulted in an estimated half-life of about 80 years for single
family housing built in the 1920s, which is confirmed by other U.S. Census data on housing. The second
criterion is that the WOODCARB II model estimate of wood and paper being discarded to SWDS needs
to match EPA estimates of discards used in the Waste sector each year over the period 1990 to 2000
(EPA 2006). These criteria help reduce uncertainty in estimates of annual change in carbon in products
in use in the United States and, to a lesser degree, reduce uncertainty in estimates of annual change in
carbon in products made from wood harvested in the United States. In addition, WOODCARB II landfill
decay rates have been validated by ensuring that estimates of CH4 emissions from landfills based on
EPA (2006) data are reasonable in comparison to CH4 estimates based on WOODCARB II landfill decay
rates.
Recalculations Discussion
There were several methodological improvements implemented for some carbon pools in the previous
Inventory which have now been fully implemented in this Inventory resulting in changes between the
previous Inventory (1990 through 2022) and this Inventory.
First, there were new FIA data included for several states, in some cases, multiple years of new data in
this Inventory resulting from delays that occurred due to the COVID-19 pandemic. Delays still exist in
some states so it is possible that multiple years of data may be available in the years ahead leading to
small changes in forest ecosystem carbon stocks and stock changes throughout the time series (Table
6-13). These changes are most notable in the conterminous United States. There was also new burned
area data incorporated in the conterminous United States and Alaska for the most recent years in the
time series which contributed to changes in estimates, particularly in interior Alaska (Table 6-14 and
Table 6-16).
Second, there were large changes in the dead wood carbon pool and, to a lesser extent, the
aboveground and belowground biomass pools between this Inventory and the previous Inventory. These
changes can be attributed to the implementation of the NSVB modeling framework (Westfall et a I. 2024)
in the previous Inventory for estimating aboveground biomass carbon in live and standing dead trees
which resulted in changes in the aboveground and belowground biomass, dead wood, and litter pools.
While estimates of aboveground live tree biomass, belowground live tree biomass, standing dead wood,
and litter pools were updated in the previous Inventory, understory aboveground biomass, understory
belowground biomass, and downed dead wood estimates were compiled in advance of the
implementation of the NSVB modeling framework in the previous Inventory, so changes to these pool
estimates resulting from those methodological improvements were not included in the previous
Inventory. The NSVB methods have been fully implemented in this Inventory for forest lands remaining
forest lands, resulting in a relatively large increase in downed dead wood carbon stocks and stock
changes which are part of the dead wood carbon pool and understory aboveground and belowground
carbon stocks and stock changes which are a part of the aboveground and belowground carbon pools
(Table 6-13, Table 6-15).
Next, there was an improvement in land classification methods for wetlands in interior Alaska described
in the Representation of the U.S. Land Base (Section 6.1) that resulted in changes in the carbon stocks
and stock change estimates between this Inventory and the previous Inventory. Specifically, for NFI
Land Use, Land-Use Change, and Forestry 6-45
-------
plots in interior Alaska that have not yet been measured by an FIA field crew, the LANDFIRE Existing
Vegetation Type (EVT) data product was used in conjunction with NLCD to define the plot as forest or
non-forest land for those plots that intersected with the "Woody Wetlands" NLCD land cover category. In
the previous Inventory all of those plots were defined as forest land. This year those plots were also
intersected with the EVT data product to determine if the primary vegetation was trees or shrubs. This
led to a small decrease in estimates of FIA forest land area in interior Alaska over the time series in this
Inventory relative to the previous Inventory and an associated decrease in the carbon stocks (Table 6-14)
and changes in the carbon stock changes on forest land (Table 6-16).
Table 6-13: Recalculations of Forest Area (1,000 ha) and Carbon Stocks in Forest Land
Remaining Forest Land and Harvested Wood Pools (MMT C)
2023 Estimate,
Previous Inventory
2023 Estimate,
Current Inventory
2024 Estimate,
Current Inventory
Forest Area (1000 ha)
281,725
280,195
280,139
Carbon Pools (Stocks) (MMT C)
Forest Ecosystem
62,320
62,801
63,016
Aboveground Biomass
17,757
17,626
17,761
Belowground Biomass
3,233
3,214
3,240
Dead Wood
3,184
4,604
4,659
Litter
3,761
3,714
3,714
Soil (Mineral)
28,401
28,108
28,108
Soil (Organic)
5,983
5,534
5,534
Harvested Wood
2,772
2,772
2,797
Products in Use
1,555
1,555
1,562
SWDS
1,217
1,217
1,235
Total Stock
65,092
65,573
65,813
Note: Totals may not sum due to independent rounding.
Table 6-14: Recalculations of Forest Area (1,000 ha) and Carbon Stocks in Forest Land
Remaining Forest Land (MMT C) in Interior Alaska
2023 Estimate,
Previous Inventory
2023 Estimate,
Current Inventory
2024 Estimate,
Current Inventory
Forest Area (1000 ha)
25,758
24,148
24,148
Carbon Pools (Stocks) (MMT C)
Forest Ecosystem
9,334
8,557
8,560
Aboveground Biomass
663
603
607
Belowground Biomass
97
90
90
Dead Wood
171
191
191
Litter
1,240
1,165
1,164
Soil (Mineral)
2,304
2,080
2,080
Soil (Organic)
4,859
4,429
4,429
Note: Totals may not sum due to independent rounding.
6-46 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Table 6-15: Recalculations of Net Carbon Flux from Forest Ecosystem Pools in Forest
Land Remaining Forest Land and Harvested Wood Pools (MMT C)
Carbon Pool (MMT C)
2022 Estimate,
Previous Inventory
2022 Estimate,
Current Inventory
2023 Estimate,
Current Inventory
Forest Ecosystem
(189.3)
(204.4)
(215.2)
Aboveground Biomass
(134.1)
(132.9)
(134.6)
Belowground Biomass
(26.4)
(26.2)
(26.3)
Dead Wood
(35.8)
(53.9)
(54.5)
Litter
7.2
8.4
(0.0)
Soil (Mineral)
(0.3)
(0.1)
(0.1)
Soil (Organic)
(0.0)
0.0
0.0
Drained organic soil
0.2
0.2
0.2
Harvested Wood
(25.3)
(25.3)
(24.8)
Products in Use
(7.9)
(7.9)
(7.5)
SWDS
(17.4)
(17.4)
(17.3)
Total Net Flux
(214.6)
(229.7)
(240.0)
Note: Totals may not sum due to independent rounding. Parentheses indicate net sequestration.
Table 6-16: Recalculations of Net Carbon Flux from Forest Ecosystem Pools in Forest
Land Remaining Forest Land (MMT C) in Interior Alaska
Carbon Pool (MMT C)
2022 Estimate,
Previous Inventory
2022 Estimate,
Current Inventory
2023 Estimate,
Current Inventory
Forest Ecosystem
8.9
11.0
(3.0)
Aboveground Biomass
(0.2)
0.4
(3.5)
Belowground Biomass
(0.0)
0.1
(0.5)
Dead Wood
1.1
1.1
0.0
Litter
8.0
9.4
0.9
Soil (Mineral)
0.0
0.0
0.0
Soil (Organic)
0.0
0.0
0.0
Notes: Totals may not sum due to independent rounding. Parentheses indicate net sequestration.
Planned Improvements
Reliable estimates of forest carbon stocks and changes across the diverse ecosystems of the United
States require a high level of investment in both annual monitoring and associated analytical
techniques. Development of improved monitoring/reporting techniques is a continuous process that
occurs simultaneously with the annual Inventory. Planned improvements can be broadly assigned to the
following categories: development of a robust estimation and reporting system, individual carbon pool
estimation, coordination with other land-use categories, and periodic and annual inventory data
incorporation.
While this Inventory includes carbon change by forest land remaining forest land and land converted to
forest land and carbon stock changes for all IPCC pools in these two categories, there are many
improvements that are still necessary:
Land Use, Land-Use Change, and Forestry 6-47
-------
• The estimation approach used for the conterminous United States in the current Inventory tor
the forest land category operates at the state scale, whereas previously the western United
States and southeast and southcentral coastal Alaska operated at a regional scale. While this is
an improvement over previous inventories and led to improved estimation and separation of
land-use categories in the current Inventory, including coastal Alaska, research is underway to
leverage all FIA data (periodic and annual inventories) and auxiliary information (i.e., remotely
sensed information) to operate at finer spatial and temporal scales. As in previous Inventories,
emissions and removals associated with natural (e.g., wildfire, insects, and disease) and human
(e.g., harvesting) disturbances are implicitly included in the report given the design of the annual
NFI, but not explicitly estimated. In addition to integrating auxiliary information into the
estimation framework and leveraging all NFI plot measurements, alternative estimators are also
being evaluated which will eliminate latency in population estimates from the NFI, improve
annual estimation and characterization of interannual variability, facilitate attribution of fluxes
to particular activities, and allow for streamlined harmonization of NFI data with auxiliary data
products. This will also facilitate separation of prescribed and wildfire emissions in future
reports. The transparency and repeatability of estimation and reporting systems will be
improved through the dissemination of open-source code (e.g., R programming language) in
concert with the public availability of the periodic and annual NFI (USDA Forest Service 2024b).
Also, several FIA database processes are being institutionalized to increase efficiency and
QA/QC in reporting and further improve transparency, completeness, consistency, accuracy,
and availability of data. Finally, a combination of approaches was used to estimate uncertainty
associated with carbon stock changes in the forest land remaining forest land category in this
report. There is research underway investigating more robust approaches to estimate total
uncertainty (Clough et al. 2016), which will be considered in future Inventories.
• Components of various carbon pools, such as carbon in belowground biomass (Russell et al.
2015) and understory vegetation (Russell et al. 2014; Johnson et al. 2017), are being explored
but may require additional investment in field inventories, beyond those incorporated in the
NSVB approach, before improvements can be realized in the Inventory report.
• The foundation of forest carbon estimation is the annual NFI. The ongoing annual surveys by the
FIA program are expected to improve the accuracy and precision of forest carbon estimates as
new state surveys become available (USDA Forest Service 2024b). With the exception of
Wyoming (which will have sufficient remeasurements in the years ahead), all other states in the
conterminous United States and coastal Alaska now have sufficient annual NFI data to
consistently estimate carbon stocks and stock changes for the future using the state-level
compilation system. The FIA program continues to install permanent plots in interior Alaska as
part of the operational NFI, and as more plots are added to the NFI, they will be used to improve
estimates for all managed forest land in Alaska. Estimates of carbon stocks and stock changes
for Hawaii and the U.S. Territories were included in this Inventory using Tier 1 and Tier 2
methods. The methods used to include all managed forest land in the conterminous United
States will be used in future Inventories for Hawaii and U.S. Territories as additional forest
carbon data become available (only a small number of plots from Hawaii are currently available
from the annualized sampling design). To that end, research is underway to incorporate all NFI
information (both annual and periodic data) and the dense time series of remotely sensed data
in multiple inferential frameworks for estimating greenhouse gas emissions and removals as
well as change (i.e., disturbance or land-use changes) detection and attribution across the
6-48 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
entire time-series and all managed forest land in the United States. Leveraging this auxiliary
information will aid the efforts to improve estimates for interior Alaska, Hawaii, and the U.S.
Territories, as well as the entire inventory system. In addition to fully inventorying all managed
forest land in the United States, the more intensive sampling (i.e., more samples) of fine woody
debris, litter, and soil organic carbon on a subset of NFI plots continues and will substantially
improve spatial and temporal resolution of carbon pools (Westfall et a I. 2013) as this
information becomes available. Increased sample intensity of some carbon pools and using
annualized sampling data as it becomes available for those states currently not reporting are
planned for future Inventories. The NFI sampling frame extends beyond the forest land-use
category (e.g., woodlands, which fall into the grasslands land-use category, and urban areas,
which fall into the settlements land-use category) with inventory-relevant information for trees
outside of forest land. These data will be utilized as they become available in the NFI.
Non-C02 Emissions from Forest Fires
Emissions of non-C02 gases from forest fires were estimated using U.S.-specific data and models for
annual area of forest burned, fuel, consumption, and emission consistent with latest IPCC methods
(2019). The IPCC 2019 Refinements were implemented to reflect the latest methodological guidance
and science. In 2023, emissions from this source were estimated to be 3.8 MMT C02 Eq. of CH4 and 2.4
MMT C02 Eq. of N20 (Table 6-17; kt units provided in Table 6-18). The estimates of non-C02 emissions
from forest fires include the conterminous 48 states, Hawaii, Puerto Rico, Guam and managed forest
land in Alaska (Ogle et al. 2018) because the fire data in use with the current methods identifies fires on
these areas within the interval 1990 through 2023.
Table 6-17: Non-C02 Emissions from Forest Fires (MMT C02 Eq.)a
Gas
1990
2005
2019
2020
2021
2022
2023
CH4
3.2
9.9
5.5
17.9
16.8
8.8
3.8
n2o
2.3
6.8
3.8
10.1
9.6
5.5
2.4
Total
5.4
16.7
9.3
28.0
26.4
14.3
6.2
aThese estimates include non-C02 emissions from forest fires on forest land remaining forest land and land converted to forest
land.
Note: Totals may not sum due to independent rounding.
Table 6-18: Non-C02 Emissions from Forest Fires (kt)a
Gas
1990
2005
2019
2020
2021
2022
2023
CH4
113
355
195
639
601
313
135
n2o
9
26
14
38
36
21
9
CO
2,947
9,189
5,044
13,280
12,732
7,375
3,114
NOx
45
132
78
223
206
117
56
aThese estimates include non-C02 emissions from forest fires on forest land remaining forest land and land converted to forest
land.
Methodology and Time-Series Consistency
Non-C02 emissions from forest fires—primarily CH4 and N20 emissions—were calculated consistent
with IPCC (2019) methodology, which represent updates of the IPCC (2006) guidance on quantifying fire
emissions. For the conterminous United States and Alaska, estimates were developed with U.S.-
specific data and models on area burned, fuel, consumption, and emissions as provided through the
Land Use, Land-Use Change, and Forestry 6-49
-------
Wildland Fire Emissions Inventory System calculator (WFEIS, French et al. 2011, 2014, Smith et al.
2024). However, these fire emissions models did not extend to include Hawaii, Puerto Rico, or Guam, so
forest fire estimates for these areas relied on Tier 1 emissions factors (IPCC 2019). Spatial definitions of
wildland burned areas were the starting point for all estimates, from WFEIS or Tier 1. The three burned
area datasets in use are the MonitoringTrends in Burn Severity (MTBS, Eidenshinket al. 2007), MODIS
burned area mapping (MODIS MCD64A1 V6.1, Giglio et al. 2018), and Wildland Fire Interagency
Geospatial Service (WFIGS) fire perimeters (WFIGS 2023). The MTBS data available for this report (MTBS
2023) included fires from 1990 through 2022 for all states and Puerto Rico. The MODIS-based records
include 2001 through 2023 for the 50 states. The WFIGS-based records for 2021 through 2023 included
all states plus Puerto Rico and Guam. Note that N20 emissions are not included in WFEIS calculations;
the emissions provided here are based on the average N20 to C02 ratio of 0.000166 (Larkin et al. 2014;
IPCC 2019). See the emissions from forest fires section in Annex 3.14 for further details on all fire-
related emissions calculations for forests. Consistent use of available data sources, data processing,
and calculation methods were applied to the entire time series to ensure time-series consistency from
1990 through 2023.
Uncertainty
Uncertainty estimates for non-C02 emissions from forest fires are based on a Monte Carlo (IPCC
Approach 2) approach to propagate variability among the alternate WFEIS annual estimates per state.
Uncertainty in parts of the WFEIS system are not currently quantified. Among potential sources for
future analysis are burned areas from MTBS, WFIGS, or MODIS, the fuels models or the Consume model
(Prichard et al. 2014). See Annex 3.14 for the quantities and assumptions employed to define and
propagate uncertainty. The results of the Approach 2 quantitative uncertainty analysis are summarized
in Table 6-19.
Table 6-19: Quantitative Uncertainty Estimates of Non-C02 Emissions from Forest
Fires (MMT C02 Eq. and Percent)3
2023 Emission
Estimate
Source Gas (MMTC02Eq.)
Uncertainty Range Relative to Emission Estimate11
(MMT CO2 Eq.)
(%)
Lower Upper
Bound Bound
Lower Upper
Bound Bound
Non-CC>2 Emissions from Forest Fires ChU 3.8
Non-CC>2 Emissions from Forest Fires N2O 2.4
3.0 4.6
1.9 3.0
-22% +22%
-22% +22%
a These estimates include non-C02 emissions from forest fires on forest land remaining forest land and land converted to forest
land.
b Range of flux estimates predicted by Monte Carlo stochastic simulation for a 95 percent confidence interval.
QA/QC and Verification
Tier 1 and Tier 2 QA/QC activities were conducted consistent with the U.S. QA/QC plan. Source-specific
quality control measures for estimating non-C02 emissions from forest fires included checking input
data, documentation, and calculations to ensure data were properly handled through the inventory
process and results were consistent with values expected from those calculations. The QA/QC
procedures did not reveal any inaccuracies or incorrect input values.
6-50 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Recalculations Discussion
The methods used in the current (1990 through 2023) Inventory to compile estimates of non-C02
emissions from forest fires represent a slight change relative to the previous (1990 through 2022)
Inventory. The basic components of calculating forest fire emissions (IPCC 2019) remain unchanged,
but the WFEIS-based estimates use the North American Wildland Fuels Database as one source of fuel
for the burn simulation Consume; the use of median fuel levels for these estimates represents a slight
change in approach relative to past such applications of WFEIS, which previously applied mean values
(Smith etal. 2024).
Planned Improvements
Improvements are planned for developing better fire and site-specific estimates for forest fires,
including:
• Improving on the Tier-1 factors currently employed for Puerto Rico and Guam.
• Addressing the best use of WFEIS, better resolution of uncertainty as discussed above, and
identification of burned areas that are not currently captured by the burn records in use.
N20 Emissions from N Additions to Forest Soils
Of the synthetic nitrogen (N) fertilizers applied to soils in the United States, no more than one percent is
applied to forest soils. Application rates are similar to those occurring on cropland soils, but in any given
year, only a small proportion of total forested land receives N fertilizer. This is because forests are
typically fertilized only twice during their approximately 40-year growth cycle (once at planting and once
midway through their life cycle). While the rate of N fertilizer application for the area of forests that
receives N fertilizer in any given year is relatively high, the annual application rate is quite low over the
entire area of forest land.
N additions to soils result in direct and indirect N20 emissions. Direct emissions occur on-site due to
the N additions. Indirect emissions result from fertilizer N that is transformed and transported to another
location through volatilization in the form of ammonia (NH3) and nitrogen oxide (NOx), in addition to
leaching and runoff of nitrates (N03), and later converted into N20 at off-site locations. The indirect
emissions are assigned to forest land because the management activity leading to the emissions
occurred in forest land.
Direct soil N20 emissions from forest land remaining forest land and land converted to forest land29 in
2023 were 0.3 MMT C02 Eq. (1.2 kt), and the indirect emissions were 0.1 MMT C02 Eq. (0.4 kt). Total
emissions for 2023 were 0.4 MMT C02 Eq. (1.5 kt) and have increased by 455 percent from 1990 to 2023.
Total forest soil N20 emissions are summarized in Table 6-20.
29 The N2O emissions from [and converted to forest [and are inc[uded with forest [and remaining forest [and because it is not
currency possib[e to separate the activity data by [and use conversion category.
Land Use, Land-Use Change, and Forestry 6-51
-------
Table 6-20: N20 Fluxes from Soils in Forest Land Remaining Forest Land and Land
Converted to Forest Land (MMT C02 Eq. and kt N2Q)
1990
2005
2019
2020
2021
2022
2023
Direct N2O Fluxes from Soils
MMT CO2 Eq.
0.1
0.3
0.3
0.3
0.3
0.3
0.3
ktN20
0.2
1.2
1.2
1.2
1.2
1.2
1.2
Indirect N2O Fluxes from Soils
MMT CO2 Eq.
+
0.1
0.1
0.1
0.1
0.1
0.1
ktN20
0.1
0.4
0.4
0.4
0.4
0.4
0.4
Total (MMT CO2 Eq.)
0.1
0.4
0.4
0.4
0.4
0.4
0.4
Total (kt N2O)
0.3
1.5
1.5
1.5
1.5
1.5
1.5
+ Does not exceed 0.05 MMT C02 Eq. or 0.05 kt.
Notes: Totals may not sum due to independent rounding. The N20 emissions from land converted to forest land are included with
forest land remaining forest land because it is not currently possible to separate the activity data by land use conversion
category.
Methodology and Time-Series Consistency
The IPCC Tier 1 approach is used to estimate N20 from soils within forest land remaining forest land and
land converted to forest land. According to USDA Forest Service statistics for 1996 (USDA Forest Service
2001), approximately 75 percent of trees planted are for timber, and about 60 percent of national total
harvested forest area is in the southeastern United States. Although southeastern pine plantations
represent the majority of fertilized forests in the United States, this Inventory also incorporated N
fertilizer application to commercial Douglas-fir stands in western Oregon and Washington. For the
Southeast, estimates of direct N20 emissions from fertilizer applications to forests are based on the
area of pine plantations receiving fertilizer in the southeastern United States and estimated application
rates (Albaugh et al. 2007; Fox et al. 2007). Fertilizer application is rare for hardwoods and therefore not
included in the inventory (Binkley et al. 1995). For each year, the area of pine receiving N fertilizer is
multiplied by the weighted average of the reported range of N fertilization rates (121 lbs. N per acre).
Area data for pine plantations receiving fertilizer in the Southeast are not available for 2005 through
2023, so data from 2004 are used for these years. For commercial forests in Oregon and Washington,
only fertilizer applied to Douglas-fir is addressed in the inventory because the vast majority
(approximately 95 percent) of the total fertilizer applied to forests in this region is applied to Douglas-fir
stands (Briggs 2007). Estimates of total Douglas-fir area and the portion of fertilized area are multiplied
to obtain annual area estimates of fertilized Douglas-fir stands. Similar to the Southeast, data are not
available for 2005 through 2023, so data from 2004 are used for these years. The annual area estimates
are multiplied by the typical rate used in this region (200 lbs. N per acre) to estimate total N applied
(Briggs 2007), and the total N applied to forests is multiplied by the IPCC (2006) default emission factor
of one percent to estimate direct N20 emissions.
For indirect emissions, the volatilization and leaching/runoff N fractions for forest land are calculated
using the IPCC default factors of 10 percent and 30 percent, respectively The amount of N volatilized is
multiplied by the IPCC default factor of one percent for the portion of volatilized N that is converted to
N20 off-site. The amount of N leached/runoff is multiplied by the IPCC default factor of 0.075 percent for
the portion of leached/runoff N that is converted to N20 off-site. The resulting estimates are summed to
obtain total indirect emissions.
6-52 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
The same method is applied in all years of this Inventory to ensure time-series consistency from 1990
through 2023.
Uncertainty
The amount of N20 emitted from forests depends not only on N inputs and fertilized area, but also on a
large number of variables, including organic carbon availability, oxygen gas partial pressure, soil
moisture content, pH, temperature, and tree planting/harvesting cycles. The effect of the combined
interaction of these variables on N20 flux is complex and highly uncertain. IPCC (2006) does not
incorporate any of these variables into the default methodology, except variation in estimated fertilizer
application rates and estimated areas of forested land receiving nitrogen fertilizer. All forest soils are
treated equivalently under this methodology. Furthermore, only applications of synthetic nitrogen
fertilizers to forest are captured in this Inventory, so applications of organic nitrogen fertilizers are not
estimated. However, the total quantity of organic nitrogen inputs to soils in the United States is included
in the Inventory within the agricultural soil management source category (Section 5.4) and settlements
remaining settlements (Section 6.10).
Uncertainties exist in the fertilization rates, annual area of forest lands receiving fertilizer, and the
emission factors. Fertilization rates are assigned a default level30 of uncertainty at ±50 percent, and area
receiving fertilizer is assigned a ±20 percent according to expert knowledge (Binkley 2004). IPCC (2006)
provided estimates for the uncertainty associated with direct and indirect N20 emission factor for
synthetic N fertilizer application to soils.
Uncertainty is quantified using simple error propagation methods (IPCC 2006). The results of the
quantitative uncertainty analysis are summarized in Table 6-21. Direct N20 fluxes from soils in 2023 are
estimated to be between 0.04 and 1.1 MMT C02 Eq. at a 95 percent confidence level. This indicates a
range of 86 percent below and 250 percent above the emission estimate of 0.3 MMT C02 Eq. for 2023.
Indirect N20 emissions in 2023 are 0.1 MMTC02 Eq. and have a range are between 0.01 and 0.4 MMT
C02 Eq., which is 94 percent below to 267 percent above the emission estimate for 2023.
Table 6-21: Quantitative Uncertainty Estimates of N20 Fluxes from Soils in Forest Land
Remaining Forest Land and Land Converted to Forest Land (MMT C02 Eq. and Percent)
2023 Emission
Estimate
Source Gas (MMTC02Eq.)
Uncertainty Range Relative to Emission Estimate
(MMT CO2 Eq.)
(%)
Lower Upper
Bound Bound
Lower Upper
Bound Bound
Forest Land Remaining Forest Land
Direct N2O Fluxes from Soils N2O 0.3
Indirect N2O Fluxes from Soils N2O 0.1
+ 1.1
+ 0.4
-86% +250%
-94% +267%
+ Does not exceed 0.05 MMT C02 Eq.
Note: Totals may not sum due to independent rounding.
30 Uncertainty is unknown for the fertilization rates so a conservative value of ±50 percent is used in the analysis.
Land Use, Land-Use Change, and Forestry 6-53
-------
QA/QC and Verification
The spreadsheet containing fertilizer applied to forests and calculations for N20 and uncertainty ranges
are checked and verified based on the sources of these data consistent with the U.S. Inventory QA/QC
plan, which is in accordance with Volume 1, Chapter 6 of the 2006IPCC Guidelines (see Annex 8 for
more details).
Recalculations Discussion
No recalculations were performed for the current Inventory.
Planned Improvements
There are currently no planned improvements identified.
Emissions from Drained Organic Soils on Forest Land31
Drained organic soils on forest land are identified separately from other forest soils largely because
mineralization of the exposed or partially dried organic material results in continuous C02 and N20
emissions (IPCC 2006). In addition, the 2013 Supplement to the 2006 IPCC Guidelines for National
Greenhouse Gas Inventories: Wetlands (IPCC 2014) calls for estimating CH4 emissions from these
drained organic soils and the ditch networks used to drain them.
Organic soils are identified on the basis of thickness of organic horizon and percent organic matter
content. All organic soils are assumed to have originally been wet, and drained organic soils are further
characterized by drainage or the process of artificially lowering the soil water table, which exposes the
organic material to drying and the associated emissions described in this section. The land base
considered here is drained inland organic soils that are coincident with forest area as identified by the
NFI of the USDA Forest Service (USDA Forest Service 2024b).
The estimated area of drained organic soils on forest land is 70,849 ha and did not change over the time
series based on the data used to compile the estimates in the current Inventory. These estimates are
based on permanent plot locations of the NFI (USDA Forest Service 2024b) coincident with mapped
organic soil locations (STATSG02 2016), which identifies forest land on organic soils. Forest sites that
are drained are not explicitly identified in the data, but for this estimate, planted forest stands on sites
identified as mesic or xeric (which are identified in USDA Forest Service 2024c, 2024d) are labeled
"drained organic soil" sites.
Land use, region, and climate are broad determinants of emissions as are more site-specific factors
such as nutrient status, drainage level, exposure, or disturbance. Current data are limited in spatial
precision and thus lack site specific details. At the same time, corresponding emissions factor data
specific to U.S. forests are similarly lacking. Tier 1 estimates are provided here following IPCC (2014).
Total annual non-C02 emissions on forest land with drained organic soils in 2023 are estimated as 0.1
MMT C02 Eq. peryear (Table 6-22; kt units provided in Table 6-23).
31 Estimates of CO2 emissions from drained organic soils are described in this section but reported in Table 6-8 and Table
6-9 for both Forest land remaining forest land and land converted to forest land in order to allow for reporting of all carbon
stock changes on forest lands in a complete and comprehensive manner.
6-54 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
The Tier 1 methodology provides methods to estimate emissions of C02 from three pathways: direct
emissions primarily from mineralization; indirect, or off-site, emissions associated with dissolved
organic carbon releasing C02 from drainage waters; and emissions from (peat) fires on organic soils.
Data about forest fires specifically located on drained organic soils are not currently available; as a
result, no corresponding estimate is provided here. Non-C02 emissions provided here include CH4 and
N20. Methane emissions generally associated with anoxic conditions do occur from the drained land
surface, but the majority of these emissions originate from ditches constructed to facilitate drainage at
these sites. Emission of N20 can be significant from these drained organic soils in contrast to the very
low emissions from wet organic soils.
Table 6-22: Non-C02 Emissions from Drained Organic Forest Soilsa'b (MMT C02 Eq.)
Source
1990
2005
2019
2020
2021
2022
2023
CH4
+
+
+
+
+
+
+
n2o
0.1
0.1
0.1
0.1
0. 1
0.1
0.1
Total
0.1
0.1
0.1
0.1
0.1
0.1
0.1
+ Does not exceed 0.05 MMT C02 Eq.
aThis table includes estimates from forest land remaining forest land and land converted to forest land.
b Estimates of C02 emissions from drained organic soils are described in this section but reported in Table 6-8 and Table 6-9 for
both forest land remaining forest land and land converted to forest land in order to allow for reporting of all carbon stock
changes on forest lands in a complete and comprehensive manner.
Note: Totals may not sum due to independent rounding.
Table 6-23: Non-C02 Emissions from Drained Organic Forest Soilsa'b(kt)
Source
1990
2005
2019
2020
2021
2022
2023
CH4
1
1
1
1
1
1
1
n2o
+
+
+
+
+
+
+
+ Does not exceed 0.5 kt.
aThis table includes estimates from forest land remaining forest land and land converted to forest land.
b Estimates of C02 emissions from drained organic soils are described in this section but reported in Table 6-8 and Table 6-9 for
both forest land remaining forest land and land converted to forest land in order to allow for reporting of all carbon stock
changes on forest lands in a complete and comprehensive manner.
Methodology and Time-Series Consistency
The Tier 1 methods for estimating C02, CH4 and N20 emissions from drained inland organic soils on
forest lands follow IPCC (2006), with extensive updates and additional material presented in the 2013
Supplement to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories: Wetlands (I PCC
2014). With the exception of quantifying area of forest on drained organic soils, which is user-supplied,
all quantities necessary for Tier 1 estimates are provided in Chapter 2, Drained Inland Organic Soils of
IPCC (2014).
Estimated area of drained organic soils on forest land is 70,849 ha based on analysis of the permanent
NFI of the USDA Forest Service and did not change over the time series. The most recent plot data per
state within the inventories were used in a spatial overlay with the STATSG02 (2016) soils data, and
forest plots coincident with the soil order histosolwere selected as having organic soils. Information
specific to identifying "drained organic" are not in the inventory data so an indirect approach was
employed here. Specifically, artificially regenerated forest stands (inventory field STDORGCD=1) on
mesic orxeric sites (inventory field 11
-------
this selection, forest area and sampling error for forest on drained organic sites are based on the
population estimates developed within the inventory data for each state (USDA Forest Service 2024d).
Eight states, all temperate forests (including pine forest in northern Florida, which largely display
characteristics of temperate forests), were identified as having drained organic soils (Table 6-24).
Table 6-24: States identified as having Drained Organic Soils, Area of Forest on
Drained Organic Soils, and Sampling Error
State
Forest on Drained Organic Soil (1,000 ha)
Sampling Error (68.3% as ± Percentage of Estimate)
Florida
2.4
79
Georgia
3.7
71
Michigan
18.7
34
Minnesota
30.2
19
North Carolina
1.3
99
Virginia
2.3
102
Washington
2.1
101
Wisconsin
10.1
30
Total
70.8
14
Note: Totals may not sum due to independent rounding.
The Tier 1 methodology provides methods to estimate emissions for three pathways of carbon emission
as C02. Note that subsequent mention of equations and tables in the remainder of this section refer to
Chapter 2 of IPCC (2014). The first pathway-direct C02 emissions-is calculated according to Equation
2.3 and Table 2.1 as the product of forest area and emission factor for temperate drained forest land.
The second pathway—indirect, or off-site, emissions—is associated with dissolved organic carbon
(DOC) releasing C02 from drainage waters according to Equation 2.4 and Table 2.2, which represent a
default composite of the three pathways for this flux: (1) the flux of DOC from natural (undrained)
organic soil; (2) the proportional increase in DOC flux from drained organic soils relative to undrained
sites; and (3) the conversion factor for the part of DOC converted to C02 after export from a site. The
third pathway—emissions from (peat) fires on organic soils—assumes that the drained organic soils
burn in a fire, but not any wet organic soils. However, this Inventory currently does not include emissions
for this pathway because data on the combined fire and drained organic soils information are not
available at this time; this may become available in the future with additional analysis.
Non-C02 emissions, according to the Tier 1 method, include methane (CH4), nitrous oxide (N20), and
carbon monoxide (CO). Emissions associated with peat fires include factors for CH4 and CO in addition
to C02, but fire estimates are assumed to be zero for the current Inventory, as discussed above.
Methane emissions generally associated with anoxic conditions do occur from the drained land surface,
but the majority of these emissions originate from ditches constructed to facilitate drainage at these
sites. From this, two separate emission factors are used, one for emissions from the area of drained
soils and a second for emissions from drainage ditch waterways. Calculations are conducted according
to Equation 2.6 and Tables 2.3 and 2.4, which includes the default fraction of the total area of drained
organic soil which is occupied by ditches. Emissions of N20 can be significant from these drained soils
in contrast to the very low emissions from wet organic soils. Calculations are conducted according to
Equation 2.7 and Table 2.5, which provide the estimate as kg N per year.
6-56 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Methodological calculations were applied to the entire set of estimates for 1990 through 2023. Year-
specific data are not available. Estimates are based on a single year and applied as the annual
estimates over the interval.
Uncertainty
Uncertainties are based on the sampling error associated with forest area of drained organic soils and
the uncertainties provided in the Chapter 2 (IPCC 2014) emissions factors (Table 6-25). The estimates
and resulting quantities representing uncertainty are based on the IPCC Approach 1-error propagation.
However, probabilistic sampling of the distributions defined for each emission factor produced a
histogram result that contained a mean and 95 percent confidence interval. The primary reason for this
approach was to develop a numerical representation of uncertainty with the potential for combining
with other forest components. The methods and parameters applied here are identical to previous
inventories, but input values were resampled for this Inventory, which results in minor changes in the
number of significant digits in the resulting estimates, relative to past values. The total non-C02
emissions in 2023 from drained organic soils on forest land remaining forest land and land converted to
forest land were estimated to be between zero and 0.150 MMT C02 Eq. around a central estimate of
0.068 MMT C02 Eq. at a 95 percent confidence level.
Table 6-25: Quantitative Uncertainty Estimates for Non-C02 Emissions on Drained
Organic Forest Soils (MMT C02 Eq. and Percent)3
Uncertainty Range Relative to Emission Estimate
(MMT CO2 Eq.)
(%)
2023 Emission Estimate
Lower
Upper
Lower
Upper
Source
(MMTCO2 Eq.)
Bound
Bound
Bound
Bound
CH4
+
+
+
-69%
+82%
n2o
0.1
+
0.1
-118%
+ 132%
Total
0.1
+
0.2
-107%
+121%
+ Does not exceed 0.05 MMT C02 Eq.
a Range of flux estimates predicted through a combination of sample-based and IPCC defaults for a 95 percent confidence
interval, IPCC Approach 1.
Note: Totals may not sum due to independent rounding.
QA/QC and Verification
IPCC (2014) guidance cautions of a possibility of double counting some of these emissions. Specifically,
the off-site emissions of dissolved organic carbon from drainage waters may be double counted if soil
carbon stock and change is based on sampling and this carbon is captured in that sampling. Double
counting in this case is unlikely since plots identified as drained were treated separately in this chapter.
Additionally, some of the non-C02 emissions may be included in either the wetlands or sections on N20
emissions from managed soils. These paths to double counting emissions are unlikely here because
these issues are taken into consideration when developing the estimates and this chapter is the only
section directly including such emissions on forest land.
Recalculations Discussion
No recalculations were performed for the current Inventory.
Land Use, Land-Use Change, and Forestry 6-57
-------
Planned Improvements
Additional data will be compiled to update estimates of forest areas on drained organic soils as new
reports and geospatial products become available. For example, current and recent past estimates are
based on drained organic soils identified in a limited number of the conterminous states; if forests on
drained organic soils are identified in additional areas including Alaska, Hawaii, Puerto Rico, or Guam,
they will be included in future Inventories.
6.3 Land Converted to Forest Land
(Source Category 4A2)
The carbon stock change estimates for land converted to forest land that are provided in this Inventory
include all forest land in an inventory year that had been in another land use(s) during the previous 20
years.32 For example, cropland or grassland converted to forest land during the past 20 years would be
reported in this category. Converted lands are in this category for 20 years as recommended in the 2006
IPCC Guidelines (IPCC 2006), afterwhich they are classified as forest land remaining forest land.
Estimates of carbon stock changes from all pools (i.e., aboveground and belowground biomass, dead
wood, litter and soils), as recommended by IPCC (2006), are included in the land converted to forest
land category of this Inventory.
33
Area of Land Converted to Forest in the United States
Land conversion to and from forests has occurred regularly throughout United States history. The 1970s
and 1980s saw a resurgence of federally sponsored forest management programs (e.g., the Forestry
Incentive Program) and soil conservation programs (e.g., the Conservation Reserve Program), which
have focused on tree planting, improving timber management activities, combating soil erosion, and
converting marginal cropland to forests. Recent analyses suggest that net accumulation of forest area
continues in areas of the United States, in particular the northeastern United States (Woodall et al.
2015b). Specifically, the annual conversion of land from other land-use categories (i.e., cropland,
grassland, wetlands, settlements, and other lands) to forest land resulted in a fairly continuous net
annual accretion of forest land area from over the time series at an average rate of 1.0 million ha year1.
Over the 20-year conversion period used in the land converted to forest land category, the conversion of
grassland to forest land resulted in the largest source of carbon transfer and uptake, accounting for
approximately 37 percent of the uptake annually. Estimated carbon uptake has remained relatively stable
32 The annual NFI data used to compile estimates of carbon transfer and uptake in this section are based on 5- to 10-yr
remeasurements so the exact conversion period was limited to the remeasured data over the time series.
33 The estimates reported in this section only include the 48 conterminous states in the United States. Land use
conversions to forest land in Alaska are currently included in the Forest Land Remaining Forest Land section because
currently there is insufficient data to separate the changes and estimates for Hawaii were not included because there is
insufficient NFI data to support inclusion at this time. Also, it is not possible to separate forest land remaining forest land
from land converted to forest land in Wyoming because of the split annual cycle method used for population estimation,
this prevents harmonization of forest land in Wyoming with the NRI/NLCD method used in Section 6.1. See Annex 3.14,
Table A-205 for annual differences between the forest area reported in Section 6.1 and Section 6.3.
6-58 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
over the time series across all conversion categories (see Table 6-32). The net flux of carbon from all forest
pool stock changes in 2023 was-103.8 MMT C02 Eq. (-28.3 MMT C) (see Table 6-32 and Table 6-33).
Mineral soil carbon stocks increased slightly over the time series for land converted to forest land. The
small gains are associated with cropland converted to forest land, settlements converted to forest land,
and other land converted to forest land. Much of this conversion is from soils that are more intensively
used under annual crop production or settlement management, or are conversions from other land,
which has little to no soil carbon. In contrast, grassland converted to forest land leads to a loss of soil
carbon across the time series, which negates some of the gain in soil carbon with the other land-use
conversions. Managed pasture to forest land is the most common conversion. This conversion leads to
a loss of soil carbon because pastures are mostly improved in the United States with fertilization and/or
irrigation, which enhances carbon input to soils relative to typical forest management activities.
Table 6-26: Net C02 Flux from Forest Carbon Pools in Land Converted to Forest Land
by Land Use Change Category (MMT C02 Eq.)
Land Use/Carbon Pool
1990 2005
2019
2020
2021
2022
2023
Cropland Converted to Forest Land
(18.2)
(18.1)
(17.9)
(17.8)
(17.8)
(17.8)
(17.8)
Aboveground Biomass
(9.9)
(9.9)
(9.8)
(9.8)
(9.8)
(9.8)
(9.8)
Belowground Biomass
(1.7)
(1.7)
(1.7)
(1.7)
(1.7)
(1.7)
(1.7)
Dead Wood
(3.2)
(3.1)
(3.1)
(3.1)
(3.1)
(3.1)
(3.1)
Litter
(3.2)
(3.2)
(3.2)
(3.2)
(3.2)
(3.2)
(3.2)
Mineral Soil
(0.2)
(0.2)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
Grassland Converted to Forest Land
(38.3)
(38.5)
(38.9)
(38.9)
(38.9)
(38.9)
(38.9)
Aboveground Biomass
(22.2)
(22.3)
(22.5)
(22.5)
(22.5)
(22.5)
(22.5)
Belowground Biomass
(2.6)
(2.7)
(2.7)
(2.7)
(2.7)
(2.7)
(2.7)
Dead Wood
(5.9)
(5.9)
(6.0)
(6.0)
(6.0)
(6.0)
(6.0)
Litter
(7.8)
(7.8)
(7.9)
(7.9)
(7.9)
(7.9)
(7.9)
Mineral Soil
0.2
0.2
0.2
0.2
0.2
0.2
0.2
Other Land Converted to Forest Land
(5.3)
(5.4)
(5.6)
(5.6)
(5.6)
(5.6)
(5.6)
Aboveground Biomass
(2.1)
(2.2)
(2.2)
(2.2)
(2.2)
(2.2)
(2.2)
Belowground Biomass
(0.3)
(0.3)
(0.3)
(0.3)
(0.3)
(0.3)
(0.3)
Dead Wood
(1.0)
(1.0)
(1.0)
(1.0)
(1.0)
(1.0)
(1.0)
Litter
(1.1)
(1.1)
(1.1)
(1.1)
(1.1)
(1.1)
(1.1)
Mineral Soil
(0.7)
(0.8)
(0.9)
(0.9)
(0.9)
(0.9)
(0.9)
Settlements Converted to Forest Land
(33.0)
(32.7)
(32.5)
(32.5)
(32.5)
(32.5)
(32.5)
Aboveground Biomass
(18.7)
(18.6)
(18.4)
(18.4)
(18.4)
(18.4)
(18.4)
Belowground Biomass
(3.2)
(3.1)
(3.1)
(3.1)
(3.1)
(3.1)
(3.1)
Dead Wood
(6.2)
(6.2)
(6.1)
(6.1)
(6.1)
(6.1)
(6.1)
Litter
(4.9)
(4.8)
(4.8)
(4.8)
(4.8)
(4.8)
(4.8)
Mineral Soil
(0.0)
(0.0)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
Wetlands Converted to Forest Land
(8.8)
(8.9)
(9.0)
(9.0)
(9.0)
(9.0)
(9.0)
Aboveground Biomass
(4.6)
(4.7)
(4.7)
(4.7)
(4.7)
(4.7)
(4.7)
Belowground Biomass
(0.8)
(0.8)
(0.8)
(0.8)
(0.8)
(0.8)
(0.8)
Dead Wood
(1.7)
(1.7)
(1.7)
(1.7)
(1.7)
(1.7)
(1.7)
Litter
(1.7)
(1.7)
(1.7)
(1.7)
(1.7)
(1.7)
(1.7)
Mineral Soil
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Total Aboveground Biomass Flux
(57.6)
(57.6)
(57.6)
(57.6)
(57.6)
(57.6)
(57.6)
Total Belowground Biomass Flux
(8.6)
(8.6)
(8.6)
(8.6)
(8.6)
(8.6)
(8.6)
Total Dead Wood Flux
(17.9)
(17.9)
(18.0)
(18.0)
(18.0)
(18.0)
(18.0)
Land Use, Land-Use Change, and Forestry 6-59
-------
Land Use/Carbon Pool
1990 I
2005
2019
2020
2021
2022
2023
Total Litter Flux
(18.7)
(18.7)
(18.8)
(18.7)
(18.7)
(18.7)
(18.7)
Total Mineral Soil Flux
(0.8)
(0.8)
(0.9)
(0.9)
(0.9)
(0.9)
(0.9)
Total Flux
(103.6) |
(103.6)
(103.9)
(103.8)
(103.8)
(103.8)
(103.8)
Notes: Totals may not sum due to independent rounding. Parentheses indicate net uptake. Forest ecosystem carbon stock
changes from land conversion in interior Alaska, Hawaii, and the U.S. Territories are currently included in the forest land
remainingforest land section because there is insufficient data to separate the changes at this time. It is not possible to separate
forest land remainingforest land from land converted to forest land in Wyoming because of the split annual cycle method used
for population estimation, this prevents harmonization of forest land in Wyoming with the NRI/NLCD method used in Section 6.1.
See Annex 3.14, Table A-205 for annual differences between the forest area reported in Section 6.1 and Section 6.3. The forest
ecosystem carbon stock changes from land conversion do not include trees on non-forest land (e.g., agroforestry systems and
settlement areas—see Section 6.10 for estimates of carbon stock change from settlement trees). It is not possible to separate
emissions from drained organic soils between forest land remainingforest land and land converted to forest land so estimates
for all organic soils are included in Table 6-8 of the Forest Land Remaining Forest Land section of the Inventory.
Table 6-27: Net Carbon Flux from Forest Carbon Pools in Land Converted to Forest
Land by Land Use Change Category (MMT C)
Land Use/Carbon Pool
1990
2005
2019
2020
2021
2022
2023
Cropland Converted to Forest Land
(5.0)
(4.9)
(4.9)
(4.9)
(4.9)
(4.9)
(4.9)
Aboveground Biomass
(2.7)
(2.7)
(2.7)
(2.7)
(2.7)
(2.7)
(2.7)
Belowground Biomass
(0.5)
(0.5)
(0.5)
(0.5)
(0.5)
(0.5)
(0.5)
Dead Wood
(0.9)
(0.9)
(0.8)
(0.8)
(0.8)
(0.8)
(0.8)
Litter
(0.9)
(0.9)
(0.9)
(0.9)
(0.9)
(0.9)
(0.9)
Mineral Soil
(0.1)
(0.1)
(0.0)
(0.0)
(0.0)
(0.0)
(0.0)
Grassland Converted to Forest Land
(10.4)
(10.5)
(10.6)
(10.6)
(10.6)
(10.6)
(10.6)
Aboveground Biomass
(6.1)
(6.1)
(6.1)
(6.1)
(6.1)
(6.1)
(6.1)
Belowground Biomass
(0.7)
(0.7)
(0.7)
(0.7)
(0.7)
(0.7)
(0.7)
Dead Wood
(1.6)
(1.6)
(1.6)
(1.6)
(1.6)
(1.6)
(1.6)
Litter
(2.1)
(2.1)
(2.1)
(2.1)
(2.1)
(2.1)
(2.1)
Mineral Soil
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Other Land Converted to Forest Land
(1.4)
(1.5)
(1.5)
(1.5)
(1.5)
(1.5)
(1.5)
Aboveground Biomass
(0.6)
(0.6)
(0.6)
(0.6)
(0.6)
(0.6)
(0.6)
Belowground Biomass
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
Dead Wood
(0.3)
(0.3)
(0.3)
(0.3)
(0.3)
(0.3)
(0.3)
Litter
(0.3)
(0.3)
(0.3)
(0.3)
(0.3)
(0.3)
(0.3)
Mineral Soil
(0.2)
(0.2)
(0.2)
(0.3)
(0.3)
(0.3)
(0.3)
Settlements Converted to Forest Land
(9.0)
(8.9)
(8.9)
(8.9)
(8.9)
(8.9)
(8.9)
Aboveground Biomass
(5.1)
(5.1)
(5.0)
(5.0)
(5.0)
(5.0)
(5.0)
Belowground Biomass
(0.9)
(0.9)
(0.8)
(0.8)
(0.8)
(0.8)
(0.8)
Dead Wood
(1.7)
(1.7)
(1.7)
(1.7)
(1.7)
(1.7)
(1.7)
Litter
(1.3)
(1.3)
(1.3)
(1.3)
(1.3)
(1.3)
(1.3)
Mineral Soil
(0.0)
(0.0)
(0.0)
(0.0)
(0.0)
(0.0)
(0.0)
Wetlands Converted to Forest Land
(2.4)
(2.4)
(2.4)
(2.4)
(2.4)
(2.4)
(2.4)
Aboveground Biomass
(1.3)
(1.3)
(1.3)
(1.3)
(1.3)
(1.3)
(1.3)
Belowground Biomass
(0.2)
(0.2)
(0.2)
(0.2)
(0.2)
(0.2)
(0.2)
Dead Wood
(0.5)
(0.5)
(0.5)
(0.5)
(0.5)
(0.5)
(0.5)
Litter
(0.5)
(0.5)
(0.5)
(0.5)
(0.5)
(0.5)
(0.5)
Mineral Soil
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Total Aboveground Biomass Flux
(15.7)
(15.7)
(15.7)
(15.7)
(15.7)
(15.7)
(15.7)
Total Belowground Biomass Flux
(2.4)
(2.4)
(2.4)
(2.3)
(2.3)
(2.3)
(2.3)
Total Dead Wood Flux
(4.9)
(4.9)
(4.9)
(4.9)
(4.9)
(4.9)
(4.9)
Total Litter Flux
(5.1)
(5.1)
(5.1)
(5.1)
(5.1)
(5.1)
(5.1)
Total Mineral Soil Flux
(0.2)
(0.2)
(0.2)
(0.2)
(0.3)
(0.2)
(0.2)
Total Flux
(28.3)
(28.3)
(28.3)
(28.3)
(28.3)
(28.3)
(28.3)
6-60 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
+ Absolute value does not exceed 0.05 MMT C.
Notes: Totals may not sum due to independent rounding. Parentheses indicate net uptake. Forest ecosystem carbon stock
changes from land conversion in interior Alaska, Hawaii, and the U.S. Territories are currently included in the forest land
remainingforest land section because there is not sufficient data to separate the changes at this time. It is not possible to
separate forest land remainingforest land from land converted to forest land in Wyoming because of the split annual cycle
method used for population estimation, this prevents harmonization of forest land in Wyoming with the NRI/NLCD method used
in Section 6.1. See Annex 3.14, Table A-205 for annual differences between the forest area reported in Section 6.1 and Section
6.3. The forest ecosystem carbon stock changes from land conversion do not include trees on non-forest land (e.g., agroforestry
systems and settlement areas—see Section 6.10 for estimates of carbon stock change from settlement trees). It is not possible
to separate emissions from drained organic soils between forest land remainingforest land and land converted to forest land so
estimates for organic soils are included in Table 6-9 and Table 6-10 of the Forest Land Remaining Forest Land section of the
Inventory.
Methodologyand Time-Series Consistency
The following section includes a description of the methodology used to estimate stock changes in all
forest carbon pools for land converted to forest land. Nationwide Forest Inventory (NFI) data and IPCC
(2006) defaults for reference carbon stocks were used to compile separate estimates for the five carbon
storage pools. Estimates for aboveground and belowground biomass, dead wood and litter were based
on data collected from the extensive array of permanent, annual NFI plots and associated models (e.g.,
live tree belowground biomass estimates) in the United States (USDA Forest Service 2024b, 2024c).
Carbon conversion factors were applied at the individual plot and then appropriately expanded to state
population estimates, which are summed to provide the national estimate. To ensure consistency in the
land converted to forest land category where carbon stock transfers occur between land-use categories,
all soil estimates are based on methods from Ogle et al. (2003, 2006) and IPCC (2006).
The methods used for estimating carbon stocks and stock changes in the land converted to forest land
are consistent with those used for forest land remaining forest land. For land-use conversion, IPCC
(2006) default biomass carbon stock values were applied in the year of conversion on individual plots to
estimate the C stocks removed due to land-use conversion from croplands and grasslands. There is no
biomass loss data or IPCC (2006) defaults to include transfers, losses, or gains of carbon in the year of
the conversion for other land use (i.e., other lands, settlements, wetlands) conversions to forest land so
these were incorporated for these conversion categories. All annual NFI plots included in the public FIA
database as of August 2024 were used in this Inventory Forest land conditions were observed on NFI
plots at time t0 and at a subsequent time ti=t0+s, where s is the time step (time measured in years) and is
indexed by discrete (e.g., 5 year) forest age classes. The inventory from t0 was then projected from ti to
2023. This projection approach requires simulating changes in the age-class distribution resulting from
forest aging and disturbance events and then applying carbon density estimates for each age class to
obtain population estimates for the nation.
Carbon in Biomass
Live tree carbon pools include aboveground and belowground (coarse root) biomass of live trees with
diameter at breast height (dbh) of at least 2.54 cm at 1.37 m above the forest floor. Separate estimates
were made for above and belowground biomass components. Over the last decade, the USDA Forest
Service's FIA program and collaborators from universities and industry have been developing a national
methodology for the prediction of individual-tree volume, biomass, and carbon content. The resulting
methodology is referred to as the National-Scale Volume and Biomass (NSVB) framework. The previous
methodology used was the Component Ratio Method (CRM) (Woodall et al. 2010). While CRM was
nationally consistent, tree biomass was still based on the volume predicted by regional models and tree
Land Use, Land-Use Change, and Forestry 6-61
-------
carbon was assumed to be 50-percent of biomass, regardless of species. Hence, the need for NSVB, a
nationally consistent methodology for compatible predictions of tree volume, biomass, and carbon
content (Westfall et al. 2024).
The NSVB covers timber tree species in the conterminous United States and coastal Alaska. All other
trees (i.e., trees that are woodland species and trees within Pacific and Caribbean Islands) use regional
models for volume and biomass, with updated carbon fractions (when available). While NSVB did not
directly update models for trees that are considered woodland species or trees within the Pacific (USDA
Forest Service 2022a, b) and Caribbean Islands (collectively referred to hereafter as "non-NSVB trees"),
volume, biomass, and carbon estimates for these trees changed compared to the CRM. For non-NSVB
trees, the standardization of tree defects and how variables are reported (i.e., whether models for total-
stem or merchantable-bole volumes are available) may be reflected as differences in volume estimates.
Additionally, biomass estimates for non-NSVB trees are based on regional biomass models and no
longer are adjusted as they were under the CRM. Finally, updates to carbon fractions (when available)
and calculation of aboveground biomass are reflected in aboveground and belowground biomass
carbon estimates (see Annex 3.14 for more details). Understory vegetation is a minor component of
biomass and is defined as all biomass of undergrowth plants in a forest, including woody shrubs and
trees less than 2.54 cm dbh. For the current Inventory, it was assumed that 10 percent of total
understory carbon mass is belowground (Smith et al. 2006). Estimates of understory carbon density
were based on information in Birdsey (1996) and biomass estimates from Jenkins et al. (2003).
Understory biomass represented over one percent of carbon in biomass, but its contribution rarely
exceeded 2 percent of the total.
Biomass losses associated with conversion from grassland and cropland to forest land were assumed to
occur in the year of conversion. To account for these losses, IPCC (2006) defaults for aboveground and
belowground biomass on grasslands and aboveground biomass on croplands were subtracted from
sequestration in the year of the conversion. As previously discussed, for all other land use (i.e., other
lands, settlements, wetlands) conversions to forest land no biomass loss data were available, and no
IPCC (2006) defaults currently exist to include transfers, losses, or gains of carbon in the year of the
conversion, so none were incorporated for these conversion categories. As defaults or country-specific
data become available for these conversion categories, they will be incorporated.
Carbon in Dead Organic Matter
Dead organic matter was initially calculated as three separate pools—standing dead trees, downed
dead wood, and litter—with carbon stocks estimated from sample data or from models. The standing
dead tree carbon pool includes aboveground and belowground (coarse root) biomass for trees of at
least 2.54 cm dbh. Calculations followed the basic method applied to live trees (Westfall et al. 2024,
Woodall et al. 2011 a) with additional modifications to account for decay and structural loss (Harmon et
al. 2011). Downed dead wood estimates are based on measurement of a subset of NFI plots for downed
dead wood (Domke et al. 2013; Woodall and Monleon 2008; Woodall et al. 2013). Downed dead wood is
defined as pieces of dead wood greater than 7.5 cm diameter, at transect intersection, that are not
attached to live or standing dead trees. This includes stumps and roots of harvested trees. To facilitate
the downscaling of downed dead wood carbon estimates from the state-wide population estimates to
individual plots, downed dead wood models specific to regions and forest types within each region are
used. Litter carbon is the pool of organic carbon (also known as duff, humus, and fine woody debris)
above the mineral soil and includes woody fragments with diameters of up to 7.5 cm. A subset of NFI
6-62 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
plots are measured for litter carbon. A modeling approach, using litter carbon measurements from NFI
plots (Domke et al. 2016) was used to estimate litter carbon for every NFI plot used in the estimation
framework. Dead organic matter carbon stock estimates are included for all land-use conversions to
forest land.
Mineral Soil Carbon Stock Changes
A Tier 2 method is applied to estimate mineral soilC stock changes for land converted to forest land
(Ogle et al. 2003, 2006; IPCC 2006). For this method, land is stratified by climate, soil types, land use,
and land management activity, and then assigned reference carbon levels and factors for the forest land
and the previous land use. The difference between the stocks is reported as the stock change under the
assumption that the change occurs over 20 years. Reference C stocks have been estimated from data in
the National Soil Survey Characterization Database (USDA-NRCS 1997), and U.S.-specific stock change
factors have been derived from published literature (Ogle et al. 2003, 2006). Land use and land-use
change patterns are determined from a combination of the Forest Inventory and Analysis Dataset (FIA)
and the 2017 National Resources Inventory (NRI) (USDA-NRCS 2020). The areas have been modified in
the NRI survey through a process in which the FIA survey data and the National Land Cover Dataset
(NLCD; Yang et al. 2018) are harmonized with the NRI data. This process ensures that the land use areas
are consistent across all land-use categories (see Section 6.1 for more information). Note that soil C in
this Inventory is reported to a depth of 100 cm in the forest land remaining forest land category (Domke
et al. 2017) while other land-use categories report soil C to a depth of 30 cm. However, to ensure
consistency in the land converted to forest land category where C stock transfers occur between land-
use categories, soil C estimates were based on a 30 cm depth using methods from Ogle et al. (2003,
2006) and IPCC (2006). See Annex 3.13 for more information about this method.
In order to ensure time-series consistency, the same methods are applied from 1990 to 2020 so that
changes reflect anthropogenic activity and not methodological adjustments. Mineral soil organic C
stock changes from 2021 to 2023 are estimated using a linear extrapolation method described in Box
6-3 of the Methodology section in Cropland Remaining Cropland. The extrapolation is based on a linear
regression model with moving-average (ARMA) errors using the 1990 to 2020 emissions data and is a
standard data splicing method for estimating emissions at the end of a time series if activity data are not
available (IPCC 2006). TheTier 2 method described previously will be applied to recalculate the 2021 to
2023 emissions in a future Inventory.
Uncertainty
A quantitative uncertainty analysis placed bounds on the flux estimates for land converted to forest land
through a combination of sample-based and model-based approaches to uncertainty for forest
ecosystem C02 Eq. flux (IPCC Approach 1). Uncertainty estimates for forest pool carbon stock changes
were developed using the same methodologies as described in the forest land remaining forest land
section for aboveground and belowground biomass, dead wood, and litter. The exception was when
IPCC default estimates were used for reference carbon stocks in certain conversion categories (i.e.,
cropland converted to forest land and grassland converted to forest land). In those cases, the
uncertainties associated with the IPCC (2006) defaults were included in the uncertainty calculations.
IPCC Approach 2 was used to propagate errors with estimation of mineral soils carbon stock changes
for land-use conversions, and is described in the cropland remaining cropland section.
Land Use, Land-Use Change, and Forestry 6-63
-------
Uncertainty estimates are presented in
Table 6-28 for each land conversion category and carbon pool. Uncertainty estimates were obtained
using a combination of sample-based and model-based approaches for all non-soil carbon pools (IPCC
Approach 1) and a Monte Carlo approach (IPCC Approach 2) was used for mineral soil. Uncertainty
estimates were combined using the error propagation model (IPCC Approach 1). The combined
uncertainty for all carbon stocks in land converted to forest land ranged from 11 percent below to 11
percent above the 2023 carbon stock change estimate of -103.8 MMT C02 Eq.
Table 6-28: Quantitative Uncertainty Estimates for Forest Carbon Pool Stock Changes
(MMT C02 Eq. per Year) in 2023 from Land Converted to Forest Land by Land Use
Change
Land Use/Carbon Pool
2023 Flux
Estimate
(MMT C02 Eq.)
Uncertainty Range Relative to Flux Range"
(MMT CO2 Eq.)
(%)
Lower
Bound
Upper
Bound
Lower
Bound
Upper
Bound
Cropland Converted to Forest Land
(17.8)
(26.4)
(9.3)
-48%
48%
Aboveground Biomass
(9.8)
(18.1)
(1.5)
-85%
85%
Belowground Biomass
(1.7)
(2.7)
(0.6)
-62%
62%
Dead Wood
(3.1)
(4.3)
(1.9)
-39%
39%
Litter
(3.2)
(4.3)
(2.1)
-34%
34%
Non-federal Mineral Soils
(0.1)
(0.3)
0.1
-170%
170%
Federal Mineral Soils
0.0
(0.1)
0.1
-9161%
9161%
Grassland Converted to Forest Land
(38.9)
(41.4)
(36.4)
-6% 6%
Aboveground Biomass
(22.5)
(23.9)
(21.2)
-6%
6%
Belowground Biomass
(2.7)
(3.0)
(2.4)
-10%
10%
Dead Wood
(6.0)
(6.2)
(5.9)
-2%
2%
Litter
(7.9)
(8.4)
(7.4)
-6%
6%
Non-federal Mineral Soils
0.2
(0.1)
0.5
-138%
138%
Federal Mineral Soils
0.0
(0.1)
0.1
-839%
839%
Other Lands Converted to Forest Land
(5.6)
(8.1)
(3.2)
-43%
43%
Aboveground Biomass
(2.2)
(4.3)
(0.1)
-95%
95%
Belowground Biomass
(0.3)
(0.8)
0.1
-123%
123%
Dead Wood
(1.0)
(1.6)
(0.4)
-57%
57%
Litter
(1.1)
(1.8)
(0.5)
-56%
56%
Non-federal Mineral Soils
(0.9)
(2.1)
0.3
-129%
129%
Federal Mineral Soils
(0.0)
(0.2)
0.1
-538%
538%
Settlements Converted to Forest Land
(32.5)
(39.0)
(26.0)
-20%
20%
Aboveground Biomass
(18.4)
(24.6)
(12.2)
-34%
34%
Belowground Biomass
(3.1)
(4.4)
(1.8)
-42%
42%
Dead Wood
(6.1)
(7.3)
(5.0)
-19%
19%
Litter
(4.8)
(5.7)
(3.9)
-19%
19%
Non-federal Mineral Soils
(0.1)
(0.1)
(0.0)
-44%
44%
Federal Mineral Soils
(0.0)
(0.0)
0.0
-157%
157%
Wetlands Converted to Forest Land
(9.0)
(9.1)
(8.8)
-2%
2%
Aboveground Biomass
(4.7)
(4.8)
(4.5)
-3%
3%
6-64 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Uncertainty Range Relative to Flux Rangea
2023 Flux
(MMT CO2 Eq.)
(%)
Estimate
Lower
Upper
Lower
Upper
Land Use/Carbon Pool
(MMT CO2 Eq.)
Bound
Bound
Bound
Bound
Belowground Biomass
(0.8)
(0.9)
(0.8)
-4%
4%
Dead Wood
(1.7)
(1.8)
(1.7)
-3%
3%
Litter
(1.7)
(1.8)
(1.7)
-4%
4%
Non-federal Mineral Soils
0.0
0.0
0.0
0%
0%
Federal Mineral Soils
0.0
0.0
0.0
0%
0%
Total: Aboveground Biomass
(57.6)
(68.3)
(46.9)
-19%
19%
Total: Belowground Biomass
(8.6)
(10.4)
(6.9)
-20%
20%
Total: Dead Wood
(18.0)
(19.8)
(16.2)
-10%
10%
Total: Litter
(18.7)
(20.4)
(17.2)
-9% 8%
Total: Mineral Soils
(0.9)
(1.3)
(0.5)
-42%
42%
Total: Lands Converted to Forest Lands
(103.8)
(114.9)
(92.7)
-11%
11%
+ Absolute value does not exceed 0.05 MMT C02 Eq.
a Range of flux estimate for 95 percent confidence interval.
Notes: Totals may not sum due to independent rounding. Parentheses indicate net uptake. It is not possible to separate
emissions from drained organic soils between forest land remaining forest land and land converted to forest land so estimates
for organic soils are included in Table 6-8 and Table 6-9 of the Forest Land Remaining Forest Land section of the Inventory.
QA/QC and Verification
See QA/QC and Verification sections under Forest Land Remaining Forest Land and for mineral soil
estimates, Cropland Remaining Cropland.
Recalculations Discussion
The approach for estimating carbon stock changes in land converted to forest land is consistent with the
methods used for forest land remaining forest Land and is described in Annex 3.14. The land converted
to forest land estimates in this Inventory are based on the land-use change information in the annual
NFI. All conversions are based on empirical estimates compiled using plot remeasurements from the
NFI, IPCC (2006) default biomass carbon stocks removed from croplands and grasslands in the year of
conversion on individual plots and the Tier 2 method for estimating mineral soil carbon stock changes
(Ogle et al. 2003, 2006; IPCC 2006). All annual NFI plots included in the public FIA database as of August
2024 were used in this Inventory. This is the fifth year that remeasurement data from the annual NFI
were available throughout the conterminous United States (with the exception of Wyoming) and coastal
southeast and southcentral Alaska to estimate land-use conversion. The availability of remeasurement
data from the annual NFI allowed for consistent plot-level estimation of carbon stocks and stock
changes for forest land remaining forest land and the land converted to forest land categories.
Overall, the land converted to forest land carbon stock changes increased by 3.5 percent in 2022 in the
current Inventory (Table 6-29) compared to the previous Inventory (1990 through 2022). While the overall
change is relatively small, there were large changes in the dead wood carbon pool and, to a lesser
extent, the aboveground and belowground biomass pools in this Inventory. These changes can be
attributed to the implementation of the National Scale Volume Biomass (NSVB) modeling framework
(Westfall et al. 2024) in the previous Inventory for estimating aboveground biomass carbon in live and
standing dead trees which resulted in changes across all land use conversion categories in the
Land Use, Land-Use Change, and Forestry 6-65
-------
aboveground and belowground biomass, dead wood, and litter pools, and new FIA data. While
estimates of aboveground live tree biomass, belowground live tree biomass, standing dead wood, and
litter pools were updated in the previous Inventory, understory aboveground biomass, understory
belowground biomass estimates and downed dead wood estimates were compiled in advance of the
implementation of the NSVB modeling framework in the previous Inventory, so changes to understory
aboveground biomass, understory belowground biomass, and downed dead wood estimates resulting
from those methodological improvements were not included in the previous Inventory. The NSVB
methods have been fully implemented in this Inventory for the land converted to forest land category,
resulting in relatively large increases in downed dead wood carbon stock changes which are part of the
dead wood carbon pool and understory aboveground and belowground carbon which are a part of the
aboveground and belowground carbon pools, respectively (Table 6-29). Please refer to other categories,
including settlement trees, where future improvements will incorporate the effects of this NSVB
approach.
Table 6-29: Recalculations of the Net Carbon Flux from Forest Carbon Pools in Land
Converted to Forest Land by Land Use Change Category (MMT C)
Conversion category
and Carbon pool (MMT C)
2022 Estimate,
Previous Inventory
2022 Estimate,
Current Inventory
2023 Estimate,
Current Inventory
Cropland Converted to Forest Land
(4.7)
(4.9)
(4.9)
Aboveground Biomass
(2.7)
(2.7)
(2.7)
Belowground Biomass
(0.5)
(0.5)
(0.5)
Dead Wood
(0.6)
(0.8)
(0.8)
Litter
(0.9)
(0.9)
(0.9)
Mineral Soil
(0.0)
(0.0)
(0.0)
Grassland Converted to Forest Land
(10.1)
(10.6)
(10.6)
Aboveground Biomass
(6.2)
(6.1)
(6.1)
Belowground Biomass
(0.7)
(0.7)
(0.7)
Dead Wood
(1.1)
(1.6)
(1.6)
Litter
(2.1)
(2.1)
(2.1)
Mineral Soil
0.1
0.1
0.1
Other Land Converted to Forest Land
(1.5)
(1.5)
(1.5)
Aboveground Biomass
(0.6)
(0.6)
(0.6)
Belowground Biomass
(0.1)
(0.1)
(0.1)
Dead Wood
(0.2)
(0.3)
(0.3)
Litter
(0.3)
(0.3)
(0.3)
Mineral Soil
(0.3)
(0.3)
(0.3)
Settlements Converted to Forest Land
(8.6)
(8.9)
(8.9)
Aboveground Biomass
(5.3)
(5.0)
(5.0)
Belowground Biomass
(0.9)
(0.8)
(0.8)
Dead Wood
(1.0)
(1.7)
(1.7)
Litter
(1.3)
(1.3)
(1.3)
Mineral soil
(0.0)
(0.0)
(0.0)
Wetlands Converted to Forest Land
(2.4)
(2.4)
(2.4)
Aboveground Biomass
(1.4)
(1.3)
(1.3)
Belowground Biomass (0.2) (0.2) (0.2)
6-66 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Conversion category
and Carbon pool (MMT C)
2022 Estimate,
Previous Inventory
2022 Estimate,
Current Inventory
2023 Estimate,
Current Inventory
Dead Wood
(0.3)
(0.5)
(0.5)
Litter
(0.5)
(0.5)
(0.5)
Mineral Soil
0.0
0.0
0.0
Total Aboveground Biomass Flux
(16.3)
(15.7)
(15.7)
Total Belowground Biomass Flux
(2.4)
(2.3)
(2.3)
Total Dead Wood Flux
(3.3)
(4.9)
(4.9)
Total Litter Flux
(5.1)
(5.1)
(5.1)
Total SOC (Mineral) Flux
(0.2)
(0.2)
(0.2)
Total Flux
(27.4)
(28.3)
(28.3)
Notes: Totals may not sum due to independent rounding. Parentheses indicate net sequestration.
Planned Improvements
There are many improvements necessary to improve the estimation of carbon stock changes associated
with land-use conversion to forest land over the entire time series.
• Soil carbon has historically been reported to a depth of 100 cm in the forest land remaining
forest land category (Domke et al. 2017) while other land-use categories (e.g., grasslands and
croplands) report soil carbon to a depth of 30 cm. To ensure greater consistency in the land
converted to forest land category where carbon stock transfers occur between land-use
categories, all mineral soil estimates in the land converted to forest land category in this
Inventory are based on methods from Ogle et al. (2003, 2006) and IPCC (2006). Methods have
been developed (Domke et al. 2017) to estimate soil carbon to depths of 20, 30, and 100 cm in
the forest land category using in situ measurements from the FIA program within the USDA
Forest Service and the International Soil Carbon Network. In future Inventories, a common
reporting depth will be defined for all land-use conversion categories and Domke et al. (2017)
will be used in the forest land remaining forest land and land converted to forest land categories
to ensure consistent reporting across all forest land.
• Due to the 5 to 10-year remeasurement periods within the FIA program and limited land-use
change information available over the entire time series, estimates presented in this section
may not reflect the entire 20-year conversion history.
• Since the sum of all land converted to forest land is used to adjust specific land-use
conversions into forest land for the state-level estimates in the NRI and NLCD, there is the
potential for differences in area estimates in states where specific land-use conversions into
forest land do not exist in the FIA data. These difference in area estimates may result in
differences between the summed estimates for mineral soil carbon stock changes across all
states and the estimates reported in Table 6-25 through Table 6-28. Work is underway to
integrate the dense time series of remotely sensed data into a new estimation system, which
will facilitate land conversion estimation over the entire time series.
Land Use, Land-Use Change, and Forestry 6-67
-------
6.4 Cropland Remaining Cropland
(Source Category 4B1)
Carbon in cropland ecosystems occurs in biomass, dead organic matter, and soils. However, carbon
storage in cropland biomass and dead organic matter is relatively ephemeral and does not need to be
reported according to the IPCC (2006), with the exception of carbon stored in perennial woody crop
biomass, such as citrus groves and apple orchards, in addition to the biomass, downed wood and dead
organic matter in agroforestry systems. This inventory includes aboveground biomass carbon stock
changes for perennial woody crops using the Tier 1 method (and total live biomass for conversions
between annual crops and woody perennial crops), but does not include biomass changes for
agroforestry. Dead organic matter is assumed to not change based on the Tier 1 method. Within soils,
carbon is found in organic and inorganic forms of carbon, but soil organic carbon is the main source and
sink for atmospheric C02. IPCC (2006) recommends quantifying changes in soil organic carbon stocks
due to agricultural land use and management activities for mineral and organic soils.34 This inventory
includes soil organic carbon stock changes for mineral and organic soils.
Well-drained mineral soils typically contain from 1 to 6 percent organic carbon by weight, whereas
mineral soils with high water tables for substantial periods of a year may contain significantly more
carbon (NRCS 1999). Conversion of mineral soils from their native state to agricultural land uses can
cause up to half of the soil organic carbon to be lost to the atmosphere due to enhanced microbial
decomposition. The rate and ultimate magnitude of carbon loss depends on subsequent management
practices, climate and soil type (Ogle et al. 2005). Agricultural practices, such as clearing, drainage,
tillage, planting, grazing, crop residue management, fertilization, application of biosolids (i.e., treated
sewage sludge) and flooding, can modify both organic matter inputs and decomposition, and thereby
result in a net carbon stock change (Paustian et al. 1997a; Lai 1998; Conant et al. 2001; Ogle et al. 2005;
Griscom et al. 2017; Ogle et al. 2019). Eventually, the soil can reach a new equilibrium that reflects a
balance between carbon inputs (e.g., decayed plant matter, roots, and organic amendments such as
manure and crop residues) and carbon loss through microbial decomposition of organic matter
(Paustian et al. 1997b).
Organic soils, also referred to as histosols, include all soils with more than 12 to 20 percent organic
carbon by weight, depending on clay content (NRCS 1999; Brady and Weil 1999). The organic layer of
these soils can be very deep (i.e., several meters), and form under inundated conditions that results in
minimal decomposition of plant residues. When organic soils are prepared for crop production, they are
drained and tilled, leading to aeration of the soil that accelerates both the decomposition rate and C02
emissions.35 Due to the depth and richness of the organic layers, carbon loss from drained organic soils
can continue over long periods of time, which varies depending on climate and composition (i.e.,
decomposability) of the organic matter (Armentano and Menges 1986). Due to deeper drainage and
more intensive management practices, the use of organic soils for annual crop production leads to
higher carbon loss rates than drainage of organic soils in grassland or forests (IPCC 2006).
34 Carbon dioxide emissions associated with liming and urea application are also estimated but are included in the Liming
and Urea Fertilization sections of the Agriculture chapter of the Inventory.
35 N2O emissions from drained organic soils are included in the Agricultural Soil Management section of the Agriculture
chapter of the Inventory.
6-68 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Cropland remaining cropland includes all cropland in an inventory year that has been cropland for a
continuous time period of at least 20 years. This determination is based on the United States
Department of Agriculture (USDA) National Resources Inventory (NRI) for non-federal lands (USDA-
NRCS 2020) and the National Land Cover Dataset for federal lands (Yanget al. 2018; Homer et al. 2007;
Fry et al. 2011; Homer et al. 2015). Cropland includes all land that is used to produce food and fiber,
forage that is harvested and used as feed (e.g., hay and silage), in addition to cropland that has been
enrolled in the Conservation Reserve Program (CRP)36 (i.e., considered set-aside cropland).
There are two discrepancies between the current land representation (see Section 6.1) and the area
data that have been used in the Inventory for cropland remaining cropland. First, croplands in Alaska are
not included in the Inventory, and second, some miscellaneous croplands that occur throughout the
United States are also not included in the Inventory due to limited understanding of greenhouse gas
emissions from these management systems (e.g., aquaculture). These differences lead to discrepancies
between the managed area in cropland remaining cropland and the cropland area included in the
Inventory analysis (Table 6-41). Improvements are underway to incorporate croplands in Alaska and
miscellaneous croplands as part of future Inventories (see Planned Improvements section).
Land use and land management of mineral soils are the largest contributor to total net carbon stock
change, especially in the early part of the time series (see Table 6-30 and Table 6-31). In 2023, mineral
soils are estimated to sequester 61.0 MMT C02 Eq. (16.6 MMT C). This level of carbon storage in mineral
soils represents a 56 percent increase since the initial inventory year of 1990. Carbon dioxide emissions
from organic soils are 30.4 MMT C02 Eq. (8.3 MMT C) in 2023, which is an 11 percent decrease in losses
of soil carbon compared to 1990. Biomass C stock changes in woody perennial crops is a smaller
portion of the overall change, estimated at 0.1 MMT C02 Eq. (less than 0.05 MMT C) in 2023. In total,
United States agricultural soils in cropland remaining cropland sequestered approximately 30.5 MMT
C02 Eq. (8.3 MMT C) in 2023.
Table 6-30: Net C02 Flux from Live Biomass and Soil Carbon Stock Changes in
Cropland Remaining Cropland (MMT C02 Eq.)
Soil Type
1990 2005 | 2019
2020
2021
2022
2023
Aboveground Live Biomass1
6.1 I
0.6 I
0.1
0.1
0.1
0.1
0.1
Mineral Soils
(39.2)
(61.8)
(48.5)
(38.2)
(62.2)
(62.0)
(61.0)
Organic Soils
34.2 |
30.2 |
29.1
29.4
30.2
30.3
30.4
Total Net Flux
1.0
(31.0)
(19.3)
(8.7)
(31.9)
(31.6)
(30.5)
1 The Tier 1 method for perennial woody crops only estimates the change in aboveground live biomass. The Tier 1 method for non-
woody crops includes total live biomass (aboveground and belowground combined), but is included in the estimates for
aboveground live biomass
Notes: Totals may not sum due to independent rounding. Parentheses indicate net sequestration.
36 The Conservation Reserve Program (CRP) is a land conservation program administered by the Farm Service Agency (FSA).
In exchange for a yearly rental payment, farmers enrolled in the program agree to remove environmentally sensitive land
from agricultural production and plant species that will improve environmental health and quality. Contracts for land
enrolled in CRP are 10 to 15 years in length. The long-term goal of the program is to re-establish valuable land cover to
help improve water quality, prevent soil erosion, and reduce loss of wildlife habitat.
Land Use, Land-Use Change, and Forestry 6-69
-------
Table 6-31: Net C02 Flux from Live Biomass and Soil Carbon Stock Changes in
Cropland Remaining Cropland (MMT C)
Soil Type
1990
2005
2019
2020
2021
2022
2023
Aboveground Live Biomass1
"1.7
0.2
+
+
+
+
+
Mineral Soils
(10.7)
(16.9)
(13.2)
(10.4)
(17.0)
(16.9)
(16.6)
Organic Soils
9.3 |
8.2
7.9
8.0
8.2
8.3
8.3
Total Net Flux
0.3
(8.5)
(5.3)
(2.4)
(8.7)
(8.6)
(8.3)
+ Does not exceed 0.05 MMT C.
1 The Tier 1 method for perennial woody crops only estimates the change in aboveground live biomass. The Tier 1 method for non-
woody crops includes total live biomass (aboveground and belowground combined), but is included in the estimates for
aboveground live biomass.
Notes: Totals may not sum due to independent rounding. Parentheses indicate net sequestration.
Soil organic carbon stocks increase in cropland remaining cropland largely due to conservation tillage
(i.e., reduced- and no-till practices), land set-aside from production in the Conservation Reserve
Program, annual crop production with hay or pasture in rotations, and manure amendments (Ogle et al.
2023). The mineral soil carbon stock changes between 1990 and 2023 range from 38.2 to 69.6 MMT C02
Eq. per year, with a mean of 56.1 MMT C02 Eq. Soil organic carbon losses from drainage of organic soils
are relatively stable across the time series with a mean emission of 30.2 MMT C02Eq. per year.
The spatial variability in the 2020 annual soil organic carbon stock changes37 are displayed in Figure 6-6
and Figure 6-7 for mineral and organic soils, respectively. Isolated areas with high rates of carbon
accumulation occur throughout the agricultural land base in the United States, but there are more
concentrated areas, such as the Maryland, Delaware, and Virginia where there have been relatively high
adoption rates of cover crop management. The regions with the highest rates of emissions from drainage
of organic soils occur in the Southeastern Coastal Region (particularly Florida and Louisiana), Northeast
and upper Midwest surrounding the Great Lakes, and isolated areas along the Pacific Coast (particularly
California), which coincides with the largest concentrations of organic soils in the United States that are
used for agricultural production.
37 Only national-scale emissions are estimated for 2021 to 2023 in this Inventory using the surrogate data method, and
therefore the fine-scale emission patterns in this map are based on Inventory data from 2020.
6-70 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Figure 6-6: Total Net Annual Soil Carbon Stock Changes for Mineral Soils under
Agricultural Management within States, 2020, Cropland Remaining Cropland
Notes: Only national-scale soil organic carbon stock changes are estimated for 2021 to 2023 in the current Inventory using a
surrogate data method, and therefore the fine-scale emission patterns in this map are based on Inventory data from 2020.
Negative values represent a net increase in soil organic carbon stocks, and positive values represent a net decrease in soil
organic carbon stocks. This figure was developed using a kriging method to develop a continuous surface from the NRI sample.
Land Use, Land-Use Change, and Forestry 6-71
-------
Figure 6-7: Total Net Annual Soil Carbon Stock Changes for Organic Soils under
Agricultural Management within States, 2020, Cropland Remaining Cropland
MT C02 ha"1 yr1
~ <10
~ 10 to 20
E 20 to 30
¦ 30 to 40
¦ >40
Note: Only national-scale soil organic carbon stock changes are estimated for 2021 to 2023 in the current Inventory using a
surrogate data method, and therefore the fine-scale emission patterns in this map are based on Inventory data from 2020. This
figure was developed using a kriging method to develop a continuous surface from the NRI sample.
Methodologyand Time-Series Consistency
The following section includes a description of the methodology used to estimate changes in carbon
stock changes for cropland remaining cropland, including (1) woody biomass for perennial crops and
conversions between annual crops and woody perennial crops, (2) agricultural land use and
management activities on mineral soils, and (3) agricultural land use and management activities on
organic soils.
Carbon stock changes on non-federal lands are estimated for cropland remaining cropland (as well as
agricultural land falling into the IPCC categories land converted to cropland, grassland remaining
grassland, and land converted to grassland) according to land-use histories recorded in the USDA NRI
survey through 2017 (USDA-NRCS 2020), and the cropping histories were extended through 2020 using
the USDA-NASS Crop Data Layer Product (CDL) (USDA-NASS 2021). The areas have been modified in
the original NRI survey through a process in which the Forest Inventory and Analysis (FIA) survey data
and the National Land Cover Dataset (Yanget al. 2018) are harmonized with the NRI data. This process
ensures that the land-use areas are consistent across all land- use categories (see Section 6.1 for more
information).
6-72 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
The NRI is a statistically-based sample and includes approximately 604,000 survey locations in
agricultural land for the conterminous United States and Hawaii. There are 364,333 survey locations
that are included in the Tier 3 method for soil carbon stock changes, and another 239,757 locations
included in the Tier 2 method for soil carbon stock changes. All locations that have perennial woody
crops are included in the Tier 1 method for this sub source. Each survey location is associated with a
weight that allows scaling of carbon stock changes from NRI survey locations to the entire country (i.e.,
each weight represents the amount of area that is expected to have the same land use/management
history as the sample point).
Land use and some management information (e.g., crop type, soil attributes, and irrigation) are
collected for each NRI point on a 5-year cycle beginning from 1982 through 1997. For cropland, data has
been collected for 4 out of 5 years during each survey cycle (i.e., 1979 through 1982,1984 through 1987,
1989 through 1992, and 1994 through 1997). In 1998, the NRI program began collecting annual data, and
the annual data are currently available through 2017 (USDA-NRCS 2020). For 2018 to 2020, the time
series is extended with the crop data provided in USDA-NASS CDL (USDA-NASS 2021), by overlaying NRI
survey locations on the CDL in a geographic information system and extracting the crop types to extend
the cropping histories. NRI survey locations are classified as cropland remaining cropland in a given
year between 1990 and 2020 if the land use has been cropland for a continuous time period of at least
20 years. The NRI survey locations are classified according to land use histories starting in 1979, and
consequently the classifications are based on less than 20 years from 1990 to 1998. This may have led
to an overestimation of cropland remaining cropland in the early part of the time series to the extent that
some areas are converted to cropland between 1971 and 1978.
Carbon dioxide emissions and removals38 due to changes in perennial woody crops, and conversions
between annual crops and perennial woody crops are estimated with the Tier 1 method (IPCC 2006,
2019). The emissions and removals associated with mineral soil organic carbon stocks are estimated
using a Tier 3 method for the majority of annual crops (Ogle et al. 2010,2023). A Tier 2 IPCC method is
used for the remaining crops not included in the Tier 3 method (see list of crops in the Mineral Soil
Carbon Stock Changes section below) (Ogle et al. 2003, 2006). In addition, a Tier 2 method is used for
very gravelly, cobbly, or shaley soils (i.e., classified as soils that have greater than 35 percent of soil
volume comprised of gravel, cobbles, or shale, regardless of crop). Emissions from organic soils are
estimated using a Tier 2 IPCC method. While a combination of Tier 2 and 3 methods are used to
estimate carbon stock changes across most of the time series, a data splicing method has been applied
to estimate stock changes in the last three to six years of the Inventory. Stock change estimates based
on data splicing will be recalculated in a future Inventory report using the Tier 1, 2 and 3 methods when
data become available.
Biomass Carbon Stock Changes
The IPCC Tier 1 approach (IPCC 2006, 2019) is used to estimate biomass carbon stock changes for
Cropland Remaining Croplands. Biomass carbon stock changes for Cropland Remaining Cropland are
estimated as (1) annual growth of perennial woody crops, (2) conversions between annual and woody
perennial croplands, and (3) management changes that affect biomass stocks (i.e., conversion between
annual, perennial hay crops, other crop systems with no biomass (e.g., bare summer fallow), and
perennial woody crop harvesting and replanting of different woody perennial types). Land-use area data
38 Removals occur through uptake of CCbinto crop and forage biomass that is later incorporated into soil carbon pools.
Land Use, Land-Use Change, and Forestry 6-73
-------
for estimating changes in cropland biomass C stock changes are based on the NRI through 2017 (USDA-
NRCS 2020) (Table 6-32). Data splicing methods are used for the remainder of the time series. For all
cases, the area from the NRI associated with each subcategory of Cropland Remaining Cropland was
multiplied by applicable factors from either 2006 IPCC or 2019 IPCC Refinements (Table 6-33). Perennial
woody crops are classified by type and age and tracked annually. Information on land use and
management from NRI begins in 1979, and all perennial woody cropland were assumed to be mature at
the beginning of the time series.
Table 6-32: Thousands of Hectares of Land for Total Live Biomass Associated with
Subcategory Land-Use Conversions within Croplands
1990 | 2005 | 2015
2016
2017
2018-2023
Annual Crop Conversions
822.0
248.8 I
252.3
585.3
478.2
*
Non-Woody Crops
814.5
236.5
226.6
555.1
452.5
~
Perennial Woody Crops
7.5
12.3
25.6
30.2
25.6
~
Non-Woody Crop Conversions
477.4
210.3
226.5
240.7
154.0
*
Annual Crops
475.9
202.7
221.3
233.6
152.6
~
Perennial Woody Crops
1.5
7.6
5.2
7.0
1.4
~
Perennial Woody Crop Conversions
34.3
51.5
25.5
41.7
60.2
*
Annual Crops
25.5
35.9
22.4
33.0
41.4
~
Non-Woody Crops
8.9
15.6 |
3.1
8.7
18.8
~
Total: Land Remaining Croplands
1,333.8
510.6 |
504.2
867.7
692.3
*
* NRI data have not been incorporated into the Inventory after 2017, designated with asterisks (*). Data splicing methods are
used to estimate the carbon stock changes for the remaining years in the time series.
Table 6-33: Carbon stock factors for total live biomass associated with Cropland
Remaining Cropland
Land Use
Climate
Subcategory/
Type
Maturity
Biomass C
Stock/Annual
C gain*
Source
Biomass Carbon Stock
t ha-1
Perennial Woody Crops
Temperate
Orchards
Mature
8.50 ±19%
IPCC 2019 (Table 5.3)
Vineyards
Mature
5.50 ± 18%
IPCC 2019 (Table 5.3)
Non-Woody Crops
-
Annual Crops
-
4.70 ± 75%
IPCC 2019 (Table 8.4)
Cold Temperate - Dry
Hay
-
3.07 ± 75%
IPCC 2006 (Table 6.4)
Cold Temperate - Moist
-
6.39 ± 75%
IPCC 2006 (Table 6.4)
Warm Temperate - Dry
-
2.87 ± 75%
IPCC 2006 (Table 6.4)
Warm Temperate - Moist
-
6.35 ± 75%
IPCC 2006 (Table 6.4)
Tropical - Dry
-
4.09 ± 75%
IPCC 2006 (Table 6.4)
Tropical - Moist & Wet
-
7.57 ± 75%
IPCC 2006 (Table 6.4)
Annual Carbon Gain
t ha1 yr1
Perennial Woody Crops
Temperate
Orchards
Immature
0.43 ± 46%
IPCC 2019 (Table 5.3)
Vineyards
Immature
0.28 ± 26%
IPCC 2019 (Table 5.3)
f Biomass carbon stock for Hay and Grasslands obtained by multiplying biomass values by 0.47 carbon fraction (IPCC 2019,
Table 5.8).
6-74 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
The aboveground biomass carbon accumulation rate for woody perennials is applied following
methodology described in the IPCC 2019 Refinement, Table 5.3, for the Temperate climate domain. A
nominal maturity cycle is assumed at 20 years (IPCC 2019, Table 5.3 updated). After reaching maturity,
biomass gain is assumed to be offset by biomass loss (mortality, self-thinning, etc.) and no biomass
carbon accrual or loss is accounted for until a conversion to another crop type or land use occurs. There
are no default root:shoot ratios for perennial woody crops available in the IPCC Guidelines, so
belowground biomass carbon is not estimated. Other non-woody crop systems are assumed to be in
steady state after the first year of conversion and no biomass changes are estimated unless there is a
change among crop types, i.e., perennial hay, annual crops, or systems with no biomass, such as bare
summer fallow. In these cases, biomass carbon stock changes are accounted for by assigning either
zero or default factors (Table 6-33) for the year that the management change occurred. For conversions
from perennial woody crops, the age of the stand is tracked and multiplied by the annual carbon gain
factor if the conversion occurred before reaching maturity. For conversions from perennial woody crops
that occur in mature stands, the maximum aboveground biomass carbon stock at harvest is applied
(IPCC 2019, Table 5.3).
In order to ensure time-series consistency, the same methods are applied from 1990 to 2017 so that
changes reflect anthropogenic activity and not methodological adjustments, and a data splicing method
is used to estimate biomass carbon loss for the remainder of the 2018 to 2023 time series (see Box 6-3).
Specifically, a linear regression model with moving-average (ARMA) errors (Brockwelland Davis 2016) is
used to impute the missing carbon stock changes using trends from 1990 to 2017. This method is a type
of linear extrapolation, which is a standard data splicing method for estimating emissions at the end of a
time series (IPCC 2006). The time series will be recalculated in a future Inventory with the methods
described previously for biomass carbon stock changes.
Soil Carbon Stock Changes for Mineral Soils
An IPCC Tier 3 model-based approach (Ogle et al. 2010) is applied to estimate organic carbon stock
changes for mineral soils on the majority of land that is used to produce annual crops and forage crops
that are harvested and used as feed (e.g., hay and silage) in the United States. These crops include
alfalfa hay, barley, corn, cotton, dry beans, grass hay, grass-clover hay, lentils, oats, onions, peanuts,
peas, potatoes, rice, sorghum, soybeans, sugar beets, sunflowers, tobacco, tomatoes, and wheat, but is
not applied to estimate organic carbon stock changes from other crops or rotations with other crops.
The model-based approach uses the DayCent ecosystem model (Parton et al. 1998; Del Grosso et al.
2001, 2011) to estimate soil organic carbon stock changes, soil nitrous oxide (N20) emissions from
agricultural soil management, and methane (CH4) emissions from rice cultivation. Carbon and nitrogen
dynamics are linked in plant-soil systems through the biogeochemical processes of microbial
decomposition and plant production (McGilland Cole 1981). Coupling the two source categories (i.e.,
agricultural soil carbon and N20) in a single inventory analysis ensures that there is a consistent
treatment of the processes and interactions between carbon and nitrogen cycling in soils.
The remaining crops on mineral soils are estimated using an IPCC Tier 2 method (Ogle et al. 2003),
including some vegetables, perennial/horticultural crops, and crops that are rotated with these crops.
The Tier 2 method is also used for very gravelly, cobbly, or shaley soils (greater than 35 percent by
volume), and soil organic carbon stock changes on federal croplands. Mineral soil organic carbon stocks
are estimated using a Tier 2 method for these areas because the DayCent model, which is used for the
Tier 3 method, has not been fully tested for estimating carbon stock changes associated with these
Land Use, Land-Use Change, and Forestry 6-75
-------
crops and rotations, as well as cobbly, gravelly, or shaley soils. In addition, there is insufficient
information to simulate croplands on federal lands using DayCent.
A surrogate data method is used to estimate soil organic carbon stock changes from 2021 to 2023 at the
national scale for land areas included in the Tier 2 and Tier 3 methods. Specifically, linear regression
models with autoregressive moving-average (ARMA) errors (Brockwell and Davis 2016) are used to
estimate the relationship between surrogate data and the 1990 to 2020 stock change data that are
derived using the Tier 2 and 3 methods. Surrogate data for these regression models include corn and
soybean yields from USDA-NASS statistics,39 and weather data from the PRISM Climate Group (PRISM
2022). See Box 6-3 for more information about the surrogate data method. Stock change estimates for
2021 to 2023 will be recalculated in future Inventories with an updated time series of activity data.
Box 6-3: Surrogate Data Method
Time series extension is needed because there are typically gaps at the end of the time series. This is
mainly because the NRI, which provides critical data for estimating greenhouse gas emissions and
removals, does not release new activity data every year.
A surrogate data method has been used to impute missing emissions at the end of the time series for
soil organic carbon stock changes in cropland remaining cropland, land converted to cropland,
grassland remaining grassland, and land converted to grassland. A linear regression model with
autoregressive moving-average (ARMA) errors (Brockwell and Davis 2016) is used to estimate the
relationship between the surrogate data and the modeled 1990 to 2020 emissions data that has been
compiled using the inventory methods described in this section. The model to extend the time series is
given by
Y = X(3 + e,
where Y is the response variable (e.g., soil organic carbon), X(B contains specific surrogate data
depending on the response variable, and s is the remaining unexplained error. Models with a variety of
surrogate data were tested, including commodity statistics, weather data, or other relevant information.
Parameters are estimated from the emissions data for 1990 to 2020 using standard statistical
techniques, and these estimates are used to predict the missing emissions data for 2021 to 2023.
A critical issue with the application of splicing methods is to adequately account for the additional
uncertainty introduced by predicting emissions rather than compiling the full inventory. Consequently,
uncertainty will increase for years with imputed estimates based on the splicing methods, compared to
those years in which the full inventory is compiled. This added uncertainty is quantified within the model
framework using a Monte Carlo approach. The approach requires estimating parameters for results in
each iteration of the Monte Carlo analysis for the full inventory (i.e., the surrogate data model is refit with
the emissions estimated in each Monte Carlo iteration from the full inventory analysis with data from
1990 to 2020), estimating emissions from each model and deriving confidence intervals combining
uncertainty across all iterations. This approach propagates uncertainties through the calculations from
the original inventory and the surrogate data method. Furthermore, the 95 percent confidence intervals
are estimated using the 3 sigma rules assuming a unimodal density (Pukelsheim 1994).
39 See https://qijickstats.nass.usda.gov/.
6-76 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Tier 3 Approach. Mineral soil organic carbon stocks and stock changes are estimated to a 30 cm depth
using the DayCent ecosystem model (Parton et al. 1998; Del Grosso et al. 2001, 2011), which simulates
cycling of carbon, nitrogen, and other nutrients in cropland, grassland, forest, and savanna ecosystems.
The DayCent model utilizes the soil carbon modeling framework developed in the Century model
(Parton et al. 1987,1988,1994; Metherell et al. 1993), but has been refined to simulate dynamics at a
daily time-step. Input data on land use and management are specified at a daily resolution and include
land-use type, crop/forage type, and management activities (e.g., planting, harvesting, fertilization,
manure amendments, tillage, irrigation, cover crops, and grazing; more information is provided below).
The model simulates net primary productivity (NPP) using the NASA-CASA production algorithm MODIS
Enhanced Vegetation Index (EVI) products, MOD13Q1 and MYD13Q1, for most croplands40 (Potter et al.
1993, 2007). The model simulates soil temperature and water dynamics, using daily weather data from a
4-kilometer gridded product developed by the PRISM Climate Group (2022), and soil attributes from the
Soil Survey Geographic Database (SSURGO) (Soil Survey Staff 2020). This method is more accurate than
the Tier 1 and 2 approaches provided by the IPCC (2006) because the simulation model treats changes
in land use and management as continuous over time as opposed to the simplified discrete changes
represented in the default method (see Box 6-4 for additional information).
Box 6-4: Tier 3 Approach for Soil Carbon Stocks Compared to Tier 1 or 2 Approaches
A Tier 3 model-based approach is used to estimate soil organic carbon stock changes for the majority of
agricultural land with mineral soils. This approach results in a more complete and accurate estimation
of soil organic carbon stock changes and entails several fundamental differences from the IPCC Tier 1 or
2 methods, as described below.
1. The IPCC Tier 1 and 2 methods are simplified approaches for estimating soil organic carbon stock
changes and classify land areas into discrete categories based on highly aggregated information
about climate (six regions), soil (seven types), and management (eleven management systems) in
the United States. In contrast, the Tier 3 model incorporates the same variables (i.e., climate, soils,
and management systems) with considerably more detail both temporally and spatially, and
captures multi-dimensional interactions through the more complex model structure.
2. The IPCC Tier 1 and 2 methods have a coarser spatial resolution in which data are aggregated to soil
types in climate regions, of which there about 30 combinations in the United States. In contrast, the
Tier 3 model simulates soil carbon dynamics at about 364,000 individual NRI survey locations in
crop fields and grazing lands.
The IPCC Tier 1 and 2 methods use a simplified approach for estimating changes in carbon stocks that
assumes a step-change from one equilibrium level of the carbon stock to another equilibrium level. In
contrast, the Tier 3 approach simulates a continuum of carbon stock changes that may reach a new
equilibrium over an extended period of time depending on the environmental conditions (i.e., a new
equilibrium often requires hundreds to thousands of years to reach). More specifically, the DayCent
model, which is used in the United States Inventory, simulates soil carbon dynamics (and C02
emissions and uptake) on a daily time step based on carbon emissions and removals from plant
40 NPP is estimated with the NASA-CASA algorithm for most of the cropland that is used to produce major commodity crops
in the central United States from 2000 to 2020. Other regions and years prior to 2000 are simulated with a method that
incorporates water, temperature and moisture stress on crop production (see Metherell et al. 1993), but does not
incorporate the additional information about crop condition provided with remote sensing data.
Land Use, Land-Use Change, and Forestry 6-77
-------
production and decomposition processes. These changes in soil organic carbon stocks are influenced
by multiple factors that affect primary production and decomposition, including changes in land use
and management, weather variability and secondary feedbacks between management activities,
climate, and soils.
Historical land-use patterns and irrigation histories are simulated with DayCent based on the 2017
USDA NRI survey (USDA-NRCS 2020). Additional sources of activity data are used to supplement the
activity data from the NRI. The USDA-NRCS Conservation Effects and Assessment Project (CEAP)
provides data on a variety of cropland management activities, and is used to inform the inventory
analysis about tillage practices, mineral fertilization, manure amendments, cover crop management, as
well as planting and harvest dates (USDA-NRCS 2022; USDA-NRCS 2018; USDA-NRCS 2012). CEAP data
are collected at a subset of NRI survey locations, and currently provide management information from
approximately 2002 to 2006 and 2013 to 2016. These data are combined with other datasets in an
imputation analysis. This imputation analysis is comprised of three steps: a) determine the trends in
management activity across the time series by combining information from several datasets (discussed
below); b) use Gradient Boosting (Friedman 2001) to determine the likely management practice at a
given NRI survey location; and c) assign management practices from the CEAP survey to the specific NRI
locations using a predictive mean matching method for certain variables that are adapted to reflect the
trending information (Little 1988; van Buuren 2012). Gradient boosting is a machine learning technique
used in regression and classification tasks, among others. It combines predictions from multiple weak
prediction models and outperforms many complicated machine learning algorithms. It makes the best
predictions at specific NRI survey locations or at state or region level models. The predictive mean
matching method identifies the most similar management activity recorded in the CEAP surveys that
match the prediction from the gradient boosting algorithm. The matching ensures that imputed
management activities are realistic for each NRI survey location, and not odd or physically unrealizable
results that could be generated by the gradient boosting. There are six complete imputations of the
management activity data using these methods.
To determine trends in mineral fertilization and manure amendments, CEAP data are combined with
information on fertilizer use and rates by crop type for different regions of the United States from the
USDA Economic Research Service. The data collection program was known as the Cropping Practices
Surveys through 1995 (USDA-ERS 1997), and is now part of data collection known as the Agricultural
Resource Management Surveys (ARMS) (USDA-ERS 2020). Additional data on fertilization practices are
compiled through other sources particularly the National Agricultural Statistics Service (USDA-NASS
1992,1999, 2004). To determine the trends in tillage management, CEAP data are combined with
Conservation Technology Information Center data between 1989 and 2004 (CTIC 2004) and OpTIS Data
Product41 for 2008 to 2020 (Hagen et al. 2020). The CTIC data are adjusted for long-term adoption of no-
till agriculture (Towery 2001). For cover crops, CEAP data are combined with information from USDA
Census of Agriculture (USDA-NASS 2012, 2017) and the OpTIS data (Hagen etal. 2020). It is assumed
that cover crop management was minimal prior to 1990 and the rates increased linearly over the decade
to the levels of cover crop management in the CEAP survey.
Uncertainty in the carbon stock estimates from DayCent associated with management activity includes
input uncertainty due to missing management data in the NRI survey, which is imputed from other
41 OpTIS data on tillage and cover crop practices provided by Regrow Agriculture, Inc.
6-78 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
sources as discussed above; model uncertainty due to incomplete specification of carbon and nitrogen
dynamics in the DayCent model algorithms and associated parameterization; and sampling uncertainty
associated with the statistical design of the NRI survey. Uncertainty is estimated with two variance
components (Ogle et al. 2010). The first variance component quantifies the uncertainty in management
activity data, model structure and parameterization. To assess this uncertainty, carbon and nitrogen
dynamics at each NRI survey location are simulated six times using the imputation product and other
model driver data. Uncertainty in parameterization and model algorithms are determined using a
structural uncertainty estimator derived from fitting a linear mixed-effect model (Ogle et al. 2007, 2010,
2023). The data are combined in a Monte Carlo stochastic simulation with 1,000 iterations for 1990
through 2020. For each iteration, there is a random selection of management data from the imputation
product (select one of the six imputations), and random selection of parameter values and random
effects for the linear mixed-effect model (i.e., structural uncertainty estimator). The second variance
component quantifies uncertainty in scaling from the NRI survey to the entire land base, and is
computed with the NRI replicate weights using a standard variance estimator for a two-stage sample
design (Sarndal et al. 1992). The two variance components are combined using simple error propagation
methods provided by the IPCC (2006), i.e., by taking the square root of the sum of the squares of the
standard deviations of the uncertain quantities. Carbon stocks and 95 percent confidence intervals are
estimated for each year between 1990 and 2020 using the DayCent model. Further elaboration on the
methodology and data used to estimate carbon stock changes from mineral soils are described in
Annex 3.13.
In order to ensure time-series consistency, the Tier 3 method is applied from 1990 to 2020 so that
changes reflect anthropogenic activity and not methodological adjustments. In addition, soil organic
carbon stock changes from 2021 to 2023 are approximated with a linear regression model with moving-
average (ARMA) errors (see Box 6-3) using stock change patterns from 1990 to 2020. This method is a
type of linear extrapolation, which is a standard data splicing method for stock changes at the end of a
time series (IPCC 2006). The time series of activity data will be updated in a future inventory, and stock
changes from 2021 to 2023 will be recalculated.
Tier 2 Approach. In the IPCC Tier 2 method, data on climate, soil types, land use, and land management
activity are used to classify land area and apply appropriate factors to estimate soil organic carbon
stock changes to a 30 cm depth (Ogle et al. 2003, 2006). The primary source of activity data for land use,
crop and irrigation histories is the 2017 NRI survey (USDA-NRCS 2020). Each NRI survey location is
classified by soil type, climate region, and management condition using data from other sources. Survey
locations on federal lands are included in the NRI, but land use and cropping history are not compiled
for these locations in the survey program (i.e., NRI is restricted to data collection on non-federal lands).
Therefore, land-use patterns for the NRI survey locations on federal lands are based on the National
Land Cover Database (NLCD) (Yang et al. 2018; Fry et al. 2011; Homer et al. 2007; Homer et al. 2015).
Additional management activities needed for the Tier 2 method are based on the imputation product
described for the Tier 3 approach, including tillage practices, mineral fertilization, and manure
amendments that are assigned to NRI survey locations. Activity data used exclusively in the Tier 2
method are wetland restoration for Conservation Reserve Program land from Euliss and Gleason (2002).
Climate zones in the United States are determined from the IPCC climate map (IPCC 2006), and then
assigned to NRI survey locations.
Reference carbon stocks are estimated using the National Soil Survey Characterization Database (NRCS
1997) with cultivated cropland as the reference condition, rather than native vegetation as used in IPCC
Land Use, Land-Use Change, and Forestry 6-79
-------
(2006). Soil measurements under agricultural management are much more common and easily
identified in the National Soil Survey Characterization Database (NRCS 1997) than are soils under a
native condition, and therefore cultivated cropland provides a more robust sample for estimating the
reference condition. Country-specific carbon stock change factors are derived from published literature
to determine the impact of management practices on soil organic carbon storage (Ogle et al. 2003,
2006). The factors represent changes in tillage, cropping rotations, intensification, and land-use change
between cultivated and uncultivated conditions. However, country-specific factors associated with
organic matter amendments are not estimated due to an insufficient number of studies in the United
States to analyze the impacts of this practice. Instead, factors from IPCC (2006) are used to estimate the
effect of those activities.
Uncertainty in soil carbon stock changes is estimated with two variance components (Ogle et al., 2010).
The first variance component quantifies the uncertainty in management activity data and carbon stock
change factors. To assess this uncertainty, changes in soil organic carbon stocks for mineral soils are
estimated 1,000 times for 1990 through 2020 using a Monte Carlo stochastic simulation approach and
probability distribution functions for the country-specific stock change factors, reference carbon stocks,
and land use activity data (Ogle et al. 2003; Ogle et al. 2006). The second variance component
quantifies uncertainty in scaling from the NRI survey to the entire land base, and is computed with the
NRI replicate weights using a standard variance estimator for a two-stage sample design (Sarndaletal.
1992). The two variance components are combined using simple error propagation methods provided by
the IPCC (2006), i.e., by taking the square root of the sum of the squares of the standard deviations of
the uncertain quantities. Further elaboration on the methodology and data used to estimate stock
changes from mineral soils are described in Annex 3.13.
In order to ensure time-series consistency, the Tier 2 method is applied from 1990 to 2020 so that
changes reflect anthropogenic activity and not methodological adjustments. In addition, soil organic
carbon stock changes from 2021 to 2023 are approximated with a linear regression model with moving-
average (ARMA) errors (see Box 6-3) using stock change patterns from 1990 to 2020. This method is a
type of linear extrapolation, which is a standard data splicing method for stock changes at the end of a
time series (IPCC 2006). The time series of activity data will be updated in a future inventory, and stock
changes from 2021 to 2023 will be recalculated.
Soil Carbon Stock Changes for Organic Soils
Annual carbon emissions from drained organic soils in cropland remaining cropland are estimated using
the Tier 2 method provided in IPCC (2006), with country-specific carbon loss rates (Ogle et al. 2003)
rather than default IPCC rates. As with mineral soils, uncertainty is estimated with two variance
components (Ogle et al., 2010). The first variance component quantifies the uncertainty in management
activity data and emission factors. A Monte Carlo stochastic simulation with 1,000 iterations is used to
quantify this uncertainty with probability distribution functions for the country-specific organic soil
emission factors and land use activity data. The second variance component quantifies uncertainty in
scaling from the NRI survey to the entire land base, and is computed with the NRI replicate weights
using a standard variance estimator for a two-stage sample design (Sarndal et al. 1992). The two
variance components are combined using simple error propagation methods provided by the IPCC
(2006), i.e., by taking the square root of the sum of the squares of the standard deviations of the
uncertain quantities. Further elaboration on the methodology and data used to estimate stock changes
from organic soils are described in Annex 3.13.
6-80 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
In order to ensure time-series consistency, the same Tier 2 method is applied from 1990 to 2020 so that
changes reflect anthropogenic activity and not methodological adjustments. In addition, soil organic
carbon stock changes from 2021 to 2023 are approximated with a linear regression model with moving-
average (ARMA) errors (see Box 6-3) using stock change patterns from 1990 to 2020. This method is a
type of linear extrapolation, which is a standard data splicing method for stock changes at the end of a
time series (IPCC 2006). The time series of activity data will be updated in a future inventory, and stock
changes from 2021 to 2023 will be recalculated.
Uncertainty
Uncertainty is quantified for changes in biomass and soil organic carbon stocks associated with
cropland remaining cropland. Uncertainty estimates are presented in Table 6-34 for each sub-source
(biomass carbon stocks, mineral and organic soil carbon stocks) and the methods that are used in the
Inventory analyses (i.e., Tier 2 and Tier 3). Uncertainty for the Tier 2 and 3 approaches is derived from two
variance components (Ogle et al. 2010). For the first component, a Monte Carlo approach is used to
address uncertainties in management activity data as well as model parameterization and structure or
emissions factors for the Tier 3 and Tier 2 methods, respectively (Ogle et al. 2010, 2023). The second
variance component is quantifying uncertainty in scaling from the NRI survey to the entire land base,
and is computed using a standard variance estimator for a two-stage sample design (Sarndal et al.
1992). The two variance components are combined using simple error propagation methods provided by
the IPCC (2006), i.e., by taking the square root of the sum of the squares of the standard deviations of
the uncertain quantities (see Annex 3.13 for further discussion). For biomass soil carbon stocks in 2021
to 2023, additional uncertainty is propagated through a Monte Carlo analysis that is associated with the
data splicing method (see Box 6-3). Soil organic carbon stock changes from the Tier 2 and 3 approaches
are combined using the simple error propagation method provided by the IPCC (2006). The combined
uncertainty is calculated by taking the square root of the sum of the squares of the standard deviations
of the uncertain quantities.
The combined uncertainty for soil organic carbon stocks in cropland remaining cropland ranges from
229 percent below to 229 percent above the 2023 stock change estimate of -30.5 MMT C02 Eq. The large
relative uncertainty around the 2023 stock change estimate is mostly due to variation in soil organic
carbon stock changes that is not explained by the surrogate data method, leading to high prediction
error.
Table 6-34: Approach 2 Quantitative Uncertainty Estimates for Soil and Biomass
Carbon Stock Changes occurring within Cropland Remaining Cropland (MMT C02 Eq.
and Percent)
2023 Flux
Estimate
Source (MMT CO2 Eq.)
Uncertainty Range Relative to Flux Estimate"
(MMT CO2 Eq.)
(%)
Lower Upper
Bound Bound
Lower Upper
Bound Bound
Biomass C Stocks: Cropland Remaining 0.1
Cropland, Tier 1 Inventory Methodology
Mineral Soil C Stocks: Cropland Remaining (57.8)
Cropland, Tier 3 Inventory Methodology
(17.3) 17.5
(122.5) 6.9
-17,828% +17,828%
-112% +112%
Land Use, Land-Use Change, and Forestry 6-81
-------
2023 Flux
Estimate
Source (MMTC02Eq.)
Uncertainty Range Relative to Flux Estimate"
(MMTCO2 Eq.)
(%)
Lower Upper
Bound Bound
Lower Upper
Bound Bound
Mineral Soil C Stocks: Cropland Remaining (3.2)
Cropland, Tier 2 Inventory Methodology
Organic Soil C Stocks: Cropland Remaining 30.4
Cropland, Tier 2 Inventory Methodology
(8.6) 2.1
10.7 50.1
-166% +166%
-65% +65%
Combined Uncertainty for Flux associated (30.5)
with Agricultural Soil Carbon Stock Change
in Cropland Remaining Cropland
(100.6) 39.5
-229% +229%
a Range of C stock change estimates is a 95 percent confidence interval.
Notes: Totals may not sum due to independent rounding. Parentheses indicate net sequestration.
Uncertainty is also associated with lack of reporting of agroforestry practices, such as shelterbelts,
riparian forests and intercropping with trees, which may have more significant changes over the
Inventory time series at least in some regions of the United States. This source of biomass carbon stock
changes is not reported because there are currently no datasets to evaluate the trends. Changes in litter
carbon stocks are also assumed to be negligible in croplands over annual time frames, although there
are certainly significant changes at sub-annual time scales across seasons. This trend may change in
the future, particularly if crop residue becomes a viable feedstock for bioenergy production.
QA/QC and Verification
Quality control measures included checking input data, model scripts, and results to ensure data are
properly handled throughout the inventory process consistent with the U.S. Inventory QA/QC plan,
which is in accordance with Volume 1, Chapter 6 of the 2006IPCC Guidelines (see Annex 8 for more
details). Inventory forms and text are reviewed and revised as needed to correct transcription errors. In
addition, results from the DayCent model are compared to field measurements and soil monitoring sites
associated with the NRI (Spencer et al. 2011), and a statistical relationship has been developed to
assess uncertainties in the predictive capability of the model (Ogle et al. 2007). The comparisons
include 69 long-term experiment sites and 145 NRI soil monitoring network sites, with 1406
observations across all of the sites (see Annex 3.13 for more information).
Recalculations Discussion
Recalculations are associated with estimating perennial woody biomass C stock changes and
conversions between crop types, including perennial hay, annual crops, perennial woody crops, or
systems with no biomass, such as bare summer fallow. The combined impact from these improvements
resulted in an average annual decrease in biomass and soil C stocks of 1.4 MMT C02 Eq., or 8.0 percent,
from 1990 to 2023 relative to the previous Inventory.
Planned Improvements
There are several planned improvements underway related to the plant production module in DayCent
which are expected to be completed by the next 1990 through 2024 Inventory:
6-82 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
• Improving crop parameters associated with temperature effects on plant production in DayCent
with additional model calibration.
• Modifying and incorporating senescence events following grain filling in crops, such as wheat,
based on recent model algorithm development.
• Testing and parameterization of the DayCent model to reduce uncertainty, particularly the
submodules that are used to approximate the cycling of nitrogen through the plant-soil system,
which will also have impacts on carbon cycling in the model simulations.
A number of other improvements are planned for Cropland Remaining Cropland, including:
• Conducting an analysis of carbon stock changes in Alaska for cropland. The improvement will
be conducted using the Tier 2 method for mineral and organic soils that is described earlier in
this section, and is expected to be completed for the next Inventory. The analysis will initially
focus on land-use change, which typically has a larger impact on soil organic carbon stock
changes than management practices, but will be further refined over time to incorporate
management data.
• Incorporating new land representation area data into the next Inventory. The current Inventory
for cropland remaining cropland is based on the land representation from the previous
Inventory. These two improvements will resolve most of the differences between the managed
land base for cropland remaining cropland and amount of area currently included in cropland
remaining cropland (see Table 6-35).
Additional improvements are considered longer-term and are expected over the next 2-3 years:
• Incorporating residue burning into the Tier 3 method. Simulating crop residue burning in the
DayCent model will be based on the amount of crop residues burned according to the data that
are used in the field burning of agricultural residues source category (see Section 5.7). This
improvement will more accurately represent the reduction in carbon inputs to the soil that are
associated with residue burning, reducing soil organic C stocks and soil nitrous oxide
emissions.
• Refining the application of biosolids to land uses. A review of available data on biosolids (i.e.,
treated sewage sludge) application will be undertaken to improve the distribution of biosolids
application on croplands, grasslands and settlements.
• Estimating biomass C stock for agroforestry systems. Methods combining survey data and
remote sensing imagery are under development to determine the extent of land with
agroforestry. In addition, a meta-analysis is being conducted to derive country-specific factors
for biomass C stock changes in agroforestry systems
Table 6-35: Comparison of Managed Land Area in Cropland Remaining Cropland and
Area in the Current Cropland Remaining Cropland Inventory (Thousand Hectares)
Year
Area (Thousand Hectares)
Managed Land
Inventory
Difference
1990
162,272
162,247
25
1991
161,839
161,814
25
1992
161,344
161,317
27
Land Use, Land-Use Change, and Forestry 6-83
-------
Year
Area (Thousand Hectares)
Managed Land
Inventory
Difference
1993
159,573
159,551
22
1994
157,888
157,864
24
1995
157,279
157,251
28
1996
156,641
156,613
28
1997
156,020
155,992
28
1998
152,338
152,309
29
1999
151,438
151,406
32
2000
151,262
151,231
31
2001
150,733
150,708
25
2002
150,425
150,400
25
2003
151,054
151,029
25
2004
150,782
150,761
20
2005
150,411
150,391
20
2006
149,903
149,882
21
2007
150,112
150,091
21
2008
149,713
149,692
22
2009
149,654
149,634
20
2010
149,221
149,196
25
2011
148,625
148,600
25
2012
148,296
148,271
25
2013
148,661
148,633
28
2014
149,142
149,115
27
2015
148,527
148,499
28
2016
148,437
148,410
26
2017
148,331
148,305
27
2018
149,721
149,694
27
2019
149,504
149,477
27
2020
149,816
149,796
20
2021
150,582
~
~
2022
151,267
~
~
2023
151,617
~
~
* Note: Activity data on land use have not been incorporated into the Inventory after 2020, designated with asterisks (*).
6-84 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
6.5 Land Converted to Cropland (Source
Category 4B2)
Land converted to cropland includes all current cropland in an inventory year that had been in another
land use(s) during the previous 20 years (IPCC 2006), and used to produce food or fiber, or forage that is
harvested and used as feed (e.g., hay and silage). For example, grassland or forest land converted to
cropland during the past 20 years would be reported in this category. Recently converted lands are
retained in this category for 20 years as recommended by IPCC (2006).
Land use change can lead to large losses of carbon to the atmosphere, particularly conversions from
forest land (Houghton et al. 1983; Houghton and Nassikas 2017). Moreover, conversion of forest to
another land use (i.e., deforestation) is one of the largest anthropogenic sources of emissions to the
atmosphere globally, although this source may be declining (Tubiello et al. 2015).
The 2006 IPCC Guidelines recommend quantifying changes in biomass, dead organic matter and soil
organic carbon stocks with land use change. All carbon stock changes are estimated and reported for
forest land converted to cropland, which includes the quantification of carbon stock changes for
aboveground and belowground biomass, dead wood, and litter. For other land use changes, the total live
biomass and soil carbon stock changes are estimated.
Grassland converted to cropland is the largest source of emissions from 1990 to 2001, while forest land
converted to cropland is the largest source of emissions from 2002 to 2023. This shift is largely due to
reduced losses of carbon from mineral soils after 2001. The high losses of carbon from forest land
converted to cropland is due to reductions in biomass and dead organic matter carbon following
conversion from forests (Table 6-36 and Table 6-37). The net change in total carbon stocks for 2023 led
to C02 emissions to the atmosphere of 35.635.6 MMT C02 Eq. (9.7 MMT C), including 11.3 MMT C02 Eq.
(3.1 MMTC) from aboveground biomass carbon losses, 2.0 MMT C02 Eq. (0.5 MMTC) from belowground
biomass carbon losses, 3.1 MMT C02 Eq. (0.8 MMT C) from dead wood carbon losses, 3.3 MMT C02 Eq.
(0.9 MMT C) from litter carbon losses, 13.1 MMT C02 Eq. (3.6 MMT C) from mineral soils and 2.8 MMT
C02 Eq. (0.8 MMT C) from drainage and cultivation of organic soils. The overall net loss of carbon has
declined by 26.7 percent from 1990 to 2023.
Land Use, Land-Use Change, and Forestry 6-85
-------
Table 6-36: Net C02 Flux from Soil, Dead Organic Matter and Biomass Carbon Stock
Changes in Land Converted to Cropland by Land-Use Change Category (MMT C02 Eq.)
1990
2005
2019
2020
2021
2022
2023
Grassland Converted to Cropland
30.5
18.3
13.2
10.7
16.2
16.4
16.9
Aboveground Live Biomass1
3.3
1.2
0.2
0.2
0.2
0.2
0.2
Belowground Live Biomass
+
+
+
+
+
+
+
Dead Wood
0.1
0.1
+
+
+
+
+
Litter
+
+
+
+
+
+
+
Mineral Soils
24.6
13.7
10.1
8.0
13.5
13.6
14.1
Organic Soils
2.4
3.3
2.7
2.4
2.4
2.4
2.5
Forest Land Converted to Cropland
19.4
19.4
19.9
19.9
19.9
19.9
19.9
Aboveground Live Biomass1
10.9
11.1
11.5
11.5
11.5
11.5
11.5
Belowground Live Biomass
1.9
1.9
2.0
2.0
2.0
2.0
2.0
Dead Wood
3.0
3.0
3.0
3.0
3.0
3.0
3.0
Litter
3.1
3.2
3.3
3.3
3.3
3.3
3.3
Mineral Soils
0.4
0.2
0.1
0.1
0.1
0.1
0.1
Organic Soils
0.1
+
+
+
+
+
+
Other Lands Converted to Cropland
(2.1)
(2.6)
(1.7)
(1.4)
(1.2)
(1.2)
(1.2)
Total Live Biomass1
(0.3)
{+)
(0.1)
(0.2)
(0.1)
(0.2)
(0.1)
Mineral Soils
(1.9)
(2.6)
(1.6)
(1.2)
(1.1)
(1.1)
(1.1)
Organic Soils
0.1
0.1
+
0.0
0.0
0.0
0.0
Settlements Converted to Cropland
0.0
(0.2)
(0.2)
(0.2)
(0.2)
(0.2)
(0.2)
Total Live Biomass1
0.1
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
Mineral Soils
(0.1)
(0.1)
(0.2)
(0.2)
(0.2)
(0.2)
(0.2)
Organic Soils
+
+
+
+
+
+
+
Wetlands Converted to Cropland
0.7
0.6
0.3
0.2
0.2
0.3
0.3
Total Live Biomass1
0.0
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
Mineral Soils
0.2
0.2
0.2
0.2
0.1
0.1
0.1
Organic Soils
0.5
0.5
0.2
0.2
0.2
0.2
0.2
Aboveground Live Biomass1
14.0
12.1
11.4
11.3
11.3
11.2
11.3
Belowground Live Biomass
1.9
1.9
2.0
2.0
2.0
2.0
2.0
Dead Wood
3.0
3.0
3.1
3.1
3.1
3.1
3.1
Litter
3.2
3.2
3.3
3.3
3.3
3.3
3.3
Total Mineral Soil Flux
23.2
11.3
8.6
6.9
12.5
12.6
13.1
Total Organic Soil Flux
3.2
3.9
3.0
2.6
2.6
2.7
2.8
Total Net Flux
48.5
35.5
31.4
29.2
34.9
35.0
35.6
+ Does not exceed 0.05 MMT C02 Eq.
1 Biomass C stock changes associated with conversions to croplands have been estimated as the total live biomass change and
reported as aboveground live biomass.
Notes: 0.0 indicates a true zero. Totals may not sum due to independent rounding. Parentheses indicate negative values or net
sequestration.
6-86 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Table 6-37: Net C02 Flux from Soil, Dead Organic Matter and Biomass Carbon Stock
Changes in Land Converted to Cropland (MMT C)
1990
2005
2019
2020
2021
2022
2023
Grassland Converted to Cropland
8.3
5.0
3.6
2.9
4.4
4.5
4.6
Aboveground Live Biomass1
0.9
0.3
0.1
0.1
+
+
+
Belowground Live Biomass
+
+
+
+
+
+
+
Dead Wood
+
+
+
+
+
+
+
Litter
+
+
+
+
+
+
+
Mineral Soils
6.7
3.7
2.8
2.2
3.7
3.7
3.8
Organic Soils
0.7
0.9
0.7
0.6
0.6
0.7
0.7
Forest Land Converted to Cropland
5.3
5.3
5.4
5.4
5.4
5.4
5.4
Aboveground Live Biomass1
3.0
3.0
3.1
3.1
3.1
3.1
3.1
Belowground Live Biomass
0.5
0.5
0.5
0.5
0.5
0.5
0.5
Dead Wood
0.8
0.8
0.8
0.8
0.8
0.8
0.8
Litter
0.9
0.9
0.9
0.9
0.9
0.9
0.9
Mineral Soils
0.1
+
+
+
+
+
+
Organic Soils
+
+
+
+
+
+
+
Other Lands Converted to Cropland
(0.6)
(0.7)
(0.5)
(0.4)
(0.3)
(0.3)
(0.3)
Total Live Biomass1
(0.1)
{+)
{+)
(0.1)
(+)
(0.1)
(+)
Mineral Soils
(0.5)
(0.7)
(0.4)
(0.3)
(0.3)
(0.3)
(0.3)
Organic Soils
+
+
+
0.0
0.0
0.0
0.0
Settlements Converted to Cropland
+
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
Total Live Biomass1
+
{+)
{+)
{+)
(+)
(+)
(+)
Mineral Soils
{+)
{+)
{+)
{+)
(+)
(+)
(+)
Organic Soils
+
+
+
+
+
+
+
Wetlands Converted to Cropland
0.2
0.2
0.1
0.1
0.1
0.1
0.1
Total Live Biomass1
+
{+)
{+)
{+)
(+)
(+)
(+)
Mineral Soils
0.1
0.1
+
+
+
+
+
Organic Soils
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Aboveground Live Biomass
3.8
3.3
3.1
3.1
3.1
3.1
3.1
Belowground Live Biomass
0.5
0.5
0.5
0.5
0.5
0.5
0.5
Dead Wood
0.8
0.8
0.8
0.8
0.8
0.8
0.8
Litter
0.9
0.9
0.9
0.9
0.9
0.9
0.9
Total Mineral Soil Flux
6.3
3.1
2.4
1.9
3.4
3.4
3.6
Total Organic Soil Flux
0.9
1.1
0.8
0.7
0.7
0.7
0.8
Total Net Flux
13.2
9.7
8.6
8.0
9.5
9.6
9.7
+ Does not exceed 0.05 MMT C.
1 Biomass C stock changes associated with conversions to croplands have been estimated as the total live biomass change and
reported as aboveground live biomass.
Notes: Totals may not sum due to independent rounding. Parentheses indicate negative values or net sequestration.
Land Use, Land-Use Change, and Forestry 6-87
-------
Methodology and Time-Series Consistency
The following section includes a description of the methodology used to estimate carbon stock changes
for land converted to cropland, including (1) loss of aboveground and belowground biomass, dead wood
and litter carbon with conversion to croplands from forest lands and woodlands designated in the
grassland, (2) loss of total live biomass with conversion from other land uses to cropland, as well as (2)
the impact from all land-use conversions to croplands on soil organic carbon stocks in mineral and
organic soils.
Biomass, Dead Wood and Litter Carbon Stock Changes
The IPCC Tier 1 approach is used to estimate biomass carbon stock changes to croplands from
grasslands, settlements, wetlands and other lands, according to land use histories recorded in the 2017
USDA NRI survey for non-federal lands (USDA-NRCS 2020). For federal lands, the land use history is
derived from land cover changes in the National Land Cover Dataset (Yang et al. 2018; Fry et al. 2011;
Homer et al. 2007, 2015). The areas have been modified through a process in which the Forest Inventory
and Analysis (FIA) survey data are harmonized with the NRI data (Nelson et al. 2020). This process
ensures that the land use areas are consistent across all land use categories (see Section 6.1 for more
information). Land use and some management information were originally collected for each NRI survey
location on a 5-year cycle beginning in 1982. In 1998, the NRI program began collecting annual data, and
the annual data have been incorporated from the NRI into the inventory analysis through 2017 (USDA-
NRCS 2020) (Table 6-38). Data splicing methods are used for the remainder of the time series. The
difference between the stocks is reported as the stock change under the assumption that the change
occurred in the year of the conversion. Biomass carbon stock changes from grasslands converted to
croplands include aboveground and belowground herbaceous biomass carbon. Biomass carbon stock
factors are assigned for each land use subcategory (e.g., annual or perennial cropland), perennial type,
and maturity class. Biomass carbon stocks in settlements, wetlands and other lands were assumed to
be zero (IPCC 2006). The total area of each land converted to croplands was multiplied by applicable
factors from IPCC (2006 and 2019) (Table 6-38 and Table 6-39). For non-woodland grassland biomass,
biomass values are disaggregated by climate zones (IPCC 2006). For conversion to perennial croplands,
factors vary by climate domain, perennial type, and maturity class as indicated in IPCC (2019).
Table 6-38: Thousands of Hectares of Land for Total Live Biomass Associated with
Land-Use Conversions to Cropland
1990
2005
2015
2016
2017
2018-2023
Grasslands Converted to Croplands
440.5
456.0
324.1
275.6
245.4
*
Annual Crops
313.2
407.3
310.0
259.1
218.0
~
Non-Woody Crops
124.0
37.7
10.4
11.5
23.8
~
Perennial Woody Crops
3.4
11.0
3.7
5.0
3.6
~
Other Lands Converted to Croplands
15.5
4.0
4.4
5.1
0.0
*
Annual Crops
9.8
2.3
3.9
5.1
0.0
~
Non-Woody Crops
5.7
1.5
0.5
-
-
~
Perennial Woody Crops
0.2
-
-
-
~
Settlements Converted to Croplands
0.1
4.8
5.7
5.2
18.0
*
Annual Crops
0.1
3.8
4.0
3.6
15.4
~
6-88 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
1990
2005
2015
2016
2017
2018-2023
Non-Woody Crops
0.8
1.7
1.6
2.4
~
Perennial Woody Crops
0.1
0.1
-
0.2
~
Wetlands Converted to Croplands
2.3
2.5
2.7
2.8
1.4
*
Annual Crops
1.5
2.3
2.7
2.6
1.3
~
Non-Woody Crops
0.8
0.2
-
0.2
0.1
~
Perennial Woody Crops
0.0 |
_ _ _ ~k
Total: Land Converted to Croplands
458.4
467.4 |
336.9
288.7
264.8
*
* NRI data have not been incorporated into the Inventory after 2017, designated with asterisks (*). Data splicing methods are
used to estimate the carbon stock changes for the remaining years in the time series.
Table 6-39: Carbon Stock Change Factors for Total Live Biomass Associated with Land-
Use Conversions to Cropland
Land Use
Climate
Subcategory/
Type
Maturity
Biomass C
Stock/
Annual
C gain*
Source
Biomass C Stock
t ha-1
Settlements,
Wetlands and
Other Lands
0
IPCC 2006
Grasslands
Cold Temperate - Dry
-
-
3.07 ± 75%
IPCC 2006
Table 6.4)
Cold Temperate - Moist
-
-
6.39 ± 75%
IPCC 2006
Table 6.4)
Warm Temperate - Dry
-
-
2.87 ± 75%
IPCC 2006
Table 6.4)
Warm Temperate - Moist
-
-
6.35 ± 75%
IPCC 2006
Table 6.4)
Tropical - Dry
-
-
4.09 ± 75%
IPCC 2006
Table 6.4)
Tropical - Moist & Wet
-
-
7.57 ± 75%
IPCC 2006
Table 6.4)
Non-Woody
Crops
-
Annual Crops
-
4.70 ± 75%
IPCC 2019
Table 8.4)
Cold Temperate - Dry
Hay
-
3.07 ± 75%
IPCC 2006
Table 6.4)
Cold Temperate - Moist
-
6.39 ± 75%
IPCC 2006
Table 6.4)
Warm Temperate - Dry
-
2.87 ± 75%
IPCC 2006
Table 6.4)
Warm Temperate - Moist
-
6.35 ± 75%
IPCC 2006
Table 6.4)
Tropical - Dry
-
4.09 ± 75%
IPCC 2006
Table 6.4)
Tropical - Moist & Wet
-
7.57 ± 75%
IPCC 2006
Table 6.4)
Annual C Gain
t ha"1 yr"1
Perennial Woody
Crops
Temperate
Orchards
Immature
0.43 ± 46%
IPCC 2019
Table 5.3)
Vineyards
Immature
0.28 ± 26%
IPCC 2019
Table 5.3)
f Biomass C Stock for Hay and Grasslands obtained by multiplying biomass values by 0.47 carbon fraction (IPCC 2019, Table 5.8).
A Tier 2 method is applied to estimate biomass, dead wood, and litter carbon stock changes for forest
land converted to cropland. Estimates are calculated in the same way as those in the forest land
remaining forest land category using data from the USDA Forest Service, Forest Inventory and Analysis
(FIA) program (USDA Forest Service 2024). However, there is no country-specific data for cropland
biomass, so only a default biomass estimate (IPCC 2006) for croplands was used to estimate carbon
stock changes (litter and dead wood carbon stocks were assumed to be zero since no reference carbon
density estimates exist for croplands). The difference between the stocks is reported as the stock
Land Use, Land-Use Change, and Forestry 6-89
-------
change under the assumption that the change occurred in the year of the conversion. Details for each of
the carbon attributes described below are available in Domke et al. (2022) and Westfall et al. (2024). If
FIA plots include data on individual trees, aboveground and belowground carbon density estimates are
based on Woodall et al. (2011) and Westfall et al. (2024). Aboveground and belowground biomass
estimates also include live understory which is a minor component of biomass defined as all biomass of
undergrowth plants in a forest, including woody shrubs and trees less than 2.54 cm dbh. For this
Inventory, it was assumed that 10 percent of total understory carbon mass is belowground (Smith et al.
2006). Estimates of carbon density are based on information in Birdsey (1996) and biomass estimates
from Jenkins et al. (2003).
For dead organic matter, if FIA plots include data on standing dead trees, standing dead tree carbon
density is estimated following the basic method applied to live trees (Woodall et al. 2011; Westfall et al.
2024) with additional modifications for woodland species to account for decay and structural loss
(Domke et al. 2011; Harmon et al. 2011). If FIA plots include data on downed dead wood, downed dead
wood carbon density is estimated based on measurements of a subset of FIA plots for downed dead
wood (Domke et al. 2013; Woodall and Monleon 2008). Downed dead wood is defined as pieces of dead
wood greater than 7.5 cm diameter, at transect intersection, that are not attached to live or standing
dead trees. This includes stumps and roots of harvested trees. To facilitate the downscaling of downed
dead wood carbon estimates from the state-wide population estimates to individual plots, downed
dead wood models specific to regions and forest types within each region are used. Litter carbon is the
pool of organic carbon (also known as duff, humus, and fine woody debris) above the mineral soil and
includes woody fragments with diameters of up to 7.5 cm. A subset of FIA plots are measured for litter
carbon. If FIA plots include litter material, a modeling approach using litter carbon measurements from
FIA plots is used to estimate litter carbon density (Domke et al. 2016). See Annex 3.14 for more
information about reference carbon density estimates for forest land and the compilation system used
to estimate carbon stock changes from forest land.
In order to ensure time-series consistency, the same methods are applied from 1990 to 2023 for the
forest lands and woodlands converted to croplands so that changes reflect anthropogenic activity and
not methodological adjustments. See Annex 3.14 for more information about reference carbon density
estimates for forest land and woodlands, and the compilation system used to estimate carbon stock
changes from forest land. For settlements, wetlands, other lands, and other grasslands (non-woodland
conversion) converted to croplands, the same methods are applied from 1990 to 2017, and a data
splicing method is used to estimate biomass carbon loss for the remainder of the 2018 to 2023 time
series (see Box 6-3 in Section 6.4). Specifically, a linear regression model with moving-average (ARMA)
errors (Brockwell and Davis 2016) is used to impute the missing C stock changes using trends from 1990
to 2018. This method is type of a linear extrapolation, which is a standard data splicing method for
estimating emissions at the end of a time series (IPCC 2006). The time series will be recalculated in a
future Inventory with the methods described previously for biomass carbon stock changes.
Soil Carbon Stock Changes
Soil organic stock changes are estimated for land converted to cropland according to land use histories
recorded in the 2017 USDA NRI survey for non-federal lands (USDA-NRCS 2020). Land use and some
management information (e.g., crop type, soil attributes, and irrigation) had been collected for each NRI
point on a five-year cycle beginning in 1982. In 1998, the NRI program began collecting annual data,
which are currently available through 2017 (USDA-NRCS 2020), and the time series for cropping
6-90 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
histories was extended through 2020 using the USDA-NASS Crop Data Layer Product (CDL) (USDA-NASS
2021) and National Land Cover Dataset (NLCD) (Yanget al. 2018; Fry et al. 2011; Homer et al. 2007,
2015). The areas have been modified in the original NRI survey through a process in which the Forest
Inventory and Analysis (FIA) survey data and the National Land Cover Dataset (NLCD; Yanget al. 2018)
are harmonized with the NRI data. This process ensures that the land use areas are consistent across all
land use categories (see Section 6.1 for more information).
NRI survey locations are classified as land converted to cropland in a given year between 1990 and 2020
if the land use is cropland but had been another use during the previous 20 years. NRI survey locations
are classified according to land use histories starting in 1979, and consequently the classifications are
based on less than 20 years from 1990 to 1998, which may have led to an underestimation of land
converted to cropland in the early part of the time series to the extent that some areas are converted to
cropland from 1971 to 1978. For federal lands, the land use history is derived from land cover changes in
the NLCD (Yang etal. 2018; Homer etal. 2007; Fryetal. 2011; Homer etal. 2015).
Soil Carbon Stock Changes for Mineral Soils
An IPCC Tier 3 model-based approach using the DayCent ecosystem model (Ogle et al. 2010, 2023) is
applied to estimate carbon stock changes from 1990 to 2020 for mineral soils on the majority of land
that is used to produce annual crops and forage crops that are harvested and used as feed (e.g., hay and
silage) in the United States. These crops include alfalfa hay, barley, corn, cotton, dry beans, grass hay,
grass-clover hay, lentils, oats, onions, peanuts, peas, potatoes, rice, sorghum, soybeans, sugar beets,
sunflowers, tobacco, tomatoes, and wheat. Soil organic carbon stock changes on the remaining mineral
soils are estimated with the IPCC Tier 2 method (Ogle et al. 2003, 2006), including land used to produce
some vegetables and perennial/horticultural crops and crops rotated with these crops; land on very
gravelly, cobbly, or shaley soils (greater than 35 percent by volume); and land converted from another
land use or federal ownership.42
For the years 2021 to 2023, a surrogate data method is used to estimate soil organic carbon stock
changes at the national scale for land areas included in the Tier 2 and Tier 3 methods. Specifically, linear
regression models with autoregressive moving-average (ARMA) errors (Brockwelland Davis 2016) are
used to estimate the relationship between surrogate data and the 1990 to 2020 stock change data from
the Tier 2 and 3 methods. Surrogate data for these regression models include corn and soybean yields
from USDA-NASS statistics,43 and weather data from the PRISM Climate Group (PRISM 2022). See Box
6-3 in the Methodology section of Cropland Remaining Cropland for more information about the
surrogate data method. Stock change estimates for 2021 to 2023 will be recalculated in future
Inventories when the time series of activity data are updated.
Tier 3 Approach. For the Tier 3 method, mineral soil organic carbon stocks and stock changes are
estimated using the DayCent ecosystem model (Parton et al. 1998; Del Grosso et al. 2001, 2011). The
DayCent model utilizes the soil carbon modeling framework developed in the Century model (Parton et
al. 1987,1988,1994; Metherellet al. 1993), but has been refined to simulate dynamics at a daily time-
step. National estimates are obtained by using the model to simulate historical land use change
patterns as recorded in the USDA NRI survey (USDA-NRCS 2020). Carbon stocks and 95 percent
42 Federal [and is not a [and use, but rather an ownership designation that is treated as grass[and for purposes of these
ca[cu[ations. The specific [and use on federa[ [ands is not identified in the NRI survey (USDA-NRCS 2018).
43 See https://quickstats.nass.usda.gov/.
Land Use, Land-Use Change, and Forestry 6-91
-------
confidence intervals are estimated for each year between 1990 and 2020. See the cropland remaining
cropland section and Annex 3.13 for additional discussion of the Tier 3 methodology for mineral soils.
In order to ensure time-series consistency, the same methods are applied from 1990 to 2020 so that
changes reflect anthropogenic activity and not methodological adjustments. In addition, soil organic
carbon stock changes from 2021 to 2023 are approximated using a linear regression model with moving-
average (ARMA) errors (Brockwell and Davis 2016) to impute the missing carbon stock changes based on
trends from 1990 to 2020. This method is type of a linear extrapolation, which is a standard data splicing
method for estimating emissions at the end of a time series (IPCC 2006). Stock change estimates for
2021 to 2023 will be recalculated in future Inventories with an updated time series of activity data (see
the Planned Improvements section in Cropland Remaining Cropland).
Tier 2 Approach. For the mineral soils not included in the Tier 3 analysis, soil organic carbon stock
changes are estimated using a Tier 2 Approach, as described in the Tier 2 Approach for mineral soils in
cropland remaining cropland. In order to ensure time-series consistency, the Tier 2 method is applied
from 1990 to 2020 so that changes reflect anthropogenic activity and not methodological adjustments.
In addition, soil organic carbon stock changes from 2021 to 2023 are approximated using a linear
regression model with moving-average (ARMA) errors (Brockwell and Davis 2016) to impute the missing
carbon stock changes based on trends from 1990 to 2020. This method is type of a linear extrapolation,
which is a standard data splicing method for estimating emissions at the end of a time series (IPCC
2006). As with the Tier 3 method, stock change estimates for 2021 to 2023 will be recalculated in future
Inventories with an updated time series of activity data.
Soil Carbon Stock Changes for Organic Soils
Annual carbon emissions from drained organic soils in land converted to cropland are estimated using
the Tier 2 method provided in IPCC (2006), with country-specific carbon loss rates (Ogle et al. 2003) as
described in the cropland remaining cropland section for organic soils. Further elaboration on the
methodology is also provided in Annex 3.13.
In order to ensure time-series consistency, the Tier 2 method is applied from 1990 to 2020 so that
changes reflect anthropogenic activity and not methodological adjustments. In addition, soil organic
carbon stock changes are approximated for the remainder of the time series with a linear regression
model with moving-average (ARMA) errors (see Box 6-3 of the Methodology section in Cropland
Remaining Cropland). This method is type of a linear extrapolation, which is a standard data splicing
method for estimating emissions at the end of a time series (IPCC 2006). Annual carbon emissions from
drained organic soils from 2021 to 2023 will be recalculated in future Inventories with an updated time
series of activity data.
Uncertainty
The uncertainty analyses for biomass, dead wood and litter carbon losses with forest land converted to
cropland and grassland converted to cropland for woodland conversions are conducted in the same
way as the uncertainty assessment for forest ecosystem carbon flux associated with forest land
remaining forest land. Sample and model-based errors are combined using simple error propagation
methods provided by the IPCC (2006) by taking the square root of the sum of the squares of the standard
deviations of the uncertain quantities. For additional details, see the Uncertainty Analysis in Annex 3.14.
6-92 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
The uncertainty analyses for soil organic carbon stock changes using the Tier 3 and Tier 2 methodologies
are quantified from two variance components (Ogle et al. 2010), as described in cropland remaining
cropland. For 2021 to 2023, there is additional uncertainty propagated through the Monte Carlo analysis
associated with the surrogate data method, which is also described in Cropland Remaining Cropland.
Uncertainty estimates are presented in Table 6-40 for each sub-source (i.e., biomass carbon stocks,
dead wood carbon stocks, litter carbon stocks, soil organic carbon stocks for mineral and organic soils)
and the method applied in the Inventory analysis (i.e., Tier 2 and Tier 3). Uncertainty estimates for the
total carbon stock changes for biomass, dead organic matter and soils are combined using the simple
error propagation methods provided by the IPCC (2006). The combined uncertainty for total carbon
stock changes in land converted to cropland ranged from 105 percent below to 105 percent above the
2023 stock change estimate of 35.6 MMT C02 Eq. The large relative uncertainty in the 2023 estimate is
mostly due high prediction error with the data splicing method.
Table 6-40: Approach 2 Quantitative Uncertainty Estimates for Soil, Dead Organic
Matter and Biomass Carbon Stock Changes occurring within Land Converted to
Cropland (MMT C02 Eq. and Percent)
Uncertainty Range Relative to Flux Estimate"
2023 Flux
(MMT CO2 Eq.)
(%)
Estimate
Lower
Upper
Lower
Upper
Source
(MMT CO2 Eq.)
Bound
Bound
Bound
Bound
Grassland Converted to Cropland
16.9
(16.9)
50.6
-200%
200%
Aboveground Live Biomass
0.2
(17.3)
17.6
-11309%
11309%
Belowground Live Biomass
0.0
0.0
0.0
-100%
67%
Dead Wood
0.0
0.0
0.1
-100%
121%
Litter
0.0
0.0
0.1
-100%
148%
Mineral Soil C Stocks: Tier 3
11.5
(17.1)
40.1
-249%
249%
Mineral Soil C Stocks: Tier 2
2.6
0.3
4.8
-87%
87%
Organic Soil C Stocks: Tier 2
2.5
(0.1)
5.1
-105%
105%
Forest Land Converted to Cropland
19.9
4.0
35.7
-80%
80%
Aboveground Live Biomass
11.5
(3.1)
26.1
-127%
127%
Belowground Live Biomass
2.0
(0.6)
4.5
-128%
127%
Dead Wood
3.0
(0.8)
6.9
-127%
127%
Litter
3.3
(0.9)
7.4
-127%
127%
Mineral Soil C Stocks: Tier 2
0.1
(0.0)
0.2
-134%
134%
Organic Soil C Stocks: Tier 2
0.0
(0.1)
0.1
-504%
504%
Other Lands Converted to Cropland
(1.2)
(2.8)
0.4
-135%
135%
Aboveground Live Biomass
(0.1)
(0.8)
0.6
-564%
564%
Mineral Soil C Stocks: Tier 2
(1.1)
(2.5)
0.4
-135%
135%
Organic Soil C Stocks: Tier 2
0.0
0.0
0.0
0%
0%
Settlements Converted to Cropland
(0.2)
(0.6)
0.1
-142%
142%
Aboveground Live Biomass
(0.1)
(0.4)
0.2
-306%
306%
Mineral Soil C Stocks: Tier 2
(0.2)
(0.3)
(0.0)
-77%
77%
Organic Soil C Stocks: Tier 2
0.0
(0.0)
0.1
-152%
152%
Land Use, Land-Use Change, and Forestry 6-93
-------
Uncertainty Range Relative to Flux Estimate"
2023 Flux
(MMT CO2 Eq.)
(%)
Estimate
Lower
Upper
Lower
Upper
Source
(MMT CO2 Eq.)
Bound
Bound
Bound
Bound
Wetlands Converted to Croplands
0.3
(0.6)
1.1
-320%
320%
Aboveground Live Biomass
(0.1)
(0.6)
0.3
-383%
383%
Mineral Soil C Stocks: Tier 2
0.1
(0.1)
0.4
-144%
144%
Organic Soil C Stocks: Tier 2
0.2
(0.5)
1.0
-303%
303%
Total: Land Converted to Cropland
35.6
(1.7)
72.9
-105%
105%
Aboveground Live Biomass
11.3
(11.5)
34.1
-202%
202%
Belowground Live Biomass
2.0
(0.5)
4.5
-126%
126%
DeadWood
3.1
(0.8)
6.9
-125%
125%
Litter
3.3
(0.8)
7.5
-125%
126%
Mineral Soil C Stocks: Tier 3
11.5
(17.1)
40.1
-249%
249%
Mineral Soil C Stocks: Tier 2
1.6
(1.1)
4.2
-171%
171%
Organic Soil C Stocks: Tier 2
2.8
0.1
5.5
-98%
98%
+ Does not exceed 0.05 MMT C02 Eq.
a Range of C stock change estimates is a 95 percent confidence interval.
Notes: Totals may not sum due to independent rounding. Parentheses indicate negative values or net sequestration.
Uncertainty is also associated with a lack of reporting on biomass, dead wood and litter carbon stock
changes for conversions to agroforestry systems in croplands. The influence of agroforestry is difficult to
address because there are currently no datasets to evaluate the trends in the area and associated
carbon stocks in agroforestry systems. The influence of land use change to agroforestry will be further
explored in a future Inventory.
QA/QC and Verification
See the QA/QC and Verification section in Cropland Remaining Cropland for information on QA/QC
steps.
Recalculations Discussion
Recalculations are associated with two revisions to the Inventory. First, updated FIA data from 1990 to
2023 on woody biomass, dead wood and litter carbon stocks were incorporated into forest land
converted to cropland estimates, including an update to the methodology to estimate downed
deadwood on forest land by implementing the National-Scale Volume and Biomass (NSVB) framework
to estimate carbon stock changes in this pool. See the Recalculations Discussion in Section 6.2 for
more information on this recalculation. Second, Tier 1 methodology to estimate biomass carbon stock
changes for cropland, settlements, wetlands, and other lands conversions to cropland (non-woodlands)
was newly implemented in this Inventory. As a result, land converted to cropland has an estimated
reduced loss by 0.5 MMT C02 Eq. on average over the time series. This represents a 1.3 percent average
decrease in carbon stock change losses for land converted to cropland compared to the previous
Inventory.
6-94 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Planned Improvements
A key improvement planned for the next Inventory includes the incorporation of new Land
Representation area data. The current Inventory for land converted to cropland is based on the land
representation from the previous Inventory. This improvement will resolve the majority of the
discrepancy between the managed land base for land converted to cropland and amount of area
currently included in land converted to cropland emissions and removals calculations (see Table 6-41).
This improvement is expected to be completed by the next Inventory. Additional planned improvements
are discussed in the Planned Improvements section of Cropland Remaining Cropland.
Table 6-41: Comparison of Managed Land Area in Land Converted to Cropland and the
Area in the current Land Converted to Cropland Inventory (Thousand Hectares)
Year
Area (Thousand Hectares)
Managed Land
Inventory
Difference
1990
12,222
12,225
-3
1991
12,556
12,557
-2
1992
12,850
12,853
-4
1993
14,087
14,088
-1
1994
15,261
15,262
-1
1995
15,432
15,436
-4
1996
15,734
15,737
-3
1997
15,914
15,916
-3
1998
17,258
17,261
-3
1999
17,654
17,658
-4
2000
17,505
17,510
-5
2001
17,433
17,436
-3
2002
17,303
17,305
-2
2003
16,055
16,057
-2
2004
15,127
15,128
-0
2005
15,215
15,215
-0
2006
15,140
15,141
-0
2007
14,726
14,726
-0
2008
14,242
14,242
0
2009
13,756
13,756
0
2010
13,883
13,884
-1
2011
14,204
14,205
-1
2012
14,445
14,446
-1
2013
13,986
13,989
-3
2014
13,461
13,463
-2
2015
13,557
13,559
-2
2016
13,516
13,518
-2
2017
13,592
13,594
-1
2018
11,672
11,673
-1
2019
11,188
11,190
-2
2020
10,286
10,290
-3
Land Use, Land-Use Change, and Forestry 6-95
-------
Year
Area (Thousand Hectares)
Managed Land
Inventory
Difference
2021
9,480
~
~
2022
8,749
~
~
2023
8,343
~
~
* Note: Activity data on land use has not been incorporated into the Inventory after 2020, designated with asterisks (*).
6.6 Grassland Remaining Grassland
(Source Category 4C1)
Carbon in grassland ecosystems occurs in biomass, dead organic matter, and soils. Soils are the largest
pool of carbon in grasslands, and have the greatest potential for longer-term storage or release of
carbon. Biomass and dead organic matter carbon pools are relatively ephemeral compared to the soil
carbon pool, with the exception of carbon stored in tree and shrub biomass that occurs in grasslands.
The 2006IPCC Guidelines recommend quantifying changes in biomass, dead organic matter and soil
organic carbon stocks with land use and management. Carbon stock changes for aboveground and
belowground biomass, dead wood and litter pools are reported for woodlands (i.e., a subcategory of
grasslands44), and may be extended to include agroforestry management associated with grasslands in
the future. For soil organic carbon, the 2006 IPCC Guidelines (IPCC 2006) recommend quantifying
changes due to (1) agricultural land use and management activities on mineral soils, and (2) agricultural
land use and management activities on organic soils.45
Grassland remaining grassland includes all grassland in an inventory year that had been grassland for a
continuous time period of at least 20 years (USDA-NRCS 2018). Grassland includes pasture and
rangeland that are primarily, but not exclusively used for livestock grazing. Rangelands are typically
extensive areas of native grassland that are not intensively managed, while pastures are typically
seeded grassland (possibly following tree removal) that may also have additional management, such as
irrigation or inter-seeding of legumes. Woodlands are also considered grassland and are areas of
continuous tree cover that do not meet the definition of forest land (see Section 6.1 for more information
about the criteria for forest land).
There is a discrepancy between the current land representation (see Section 6.1) and the area data that
have been used in the inventory for grassland remaining grassland. Specifically, grasslands in Alaska are
not included in the Inventory, and this land base is approximately 50 million hectares. This difference
leads to a discrepancy between the managed area in grassland remaining grassland in the land
representation and the grassland area included in the emissions and removals estimation for the
grassland remaining grassland land-use category (Table 6-45). Improvements are underway to
incorporate grasslands in Alaska as part of future Inventories (see Planned Improvements section).
For grassland remaining grassland, there has been considerable variation in carbon stocks between
1990 and 2023. These changes are driven by variability in weather patterns and associated interaction
44 Woodlands are considered grasslands in the U.S. land representation because they do not meet the definition of forest
land.
45 CO2 emissions associated with liming and urea fertilization are also estimated but included in the Agriculture chapter of
the report.
6-96 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
with land management activity. Moreover, changes are small on a per hectare rate basis across the time
series even in the years with a larger total change in stocks. The net change in total carbon stocks for
2023 led to net C02 emissions to the atmosphere of 22.0 MMT C02 Eq. (6.0 MMT C), including -1.3 MMT
C02 Eq. (-0.4 MMT C) from net gains of aboveground biomass C, -0.2 MMT C02 Eq. (-0.1 MMT C) from net
gains in belowground biomass carbon, 2.5 MMT C02 Eq. (0.7 MMT C) from net losses in dead wood
carbon, less than 0.05 MMT C02 Eq. (less than 0.05 MMT C) from net gains in litter C, 15.5 MMT C02 Eq.
(4.2 MMT C) from net losses in mineral soil organic carbon, and 5.5 MMT C02 Eq. (1.5 MMT C) from
losses of carbon due to drainage and cultivation of organic soils (Table 6-42 and Table 6-43). Losses of
carbon are 8.2 percent lower in 2023 compared to 1990, but as noted previously, stock changes are
highly variable from 1990 to 2023, with an average annual change of 19.6 MMT C02Eq. (5.3 MMT C).
Table 6-42: Net C02 Flux from Soil, Dead Organic Matter and Biomass Carbon Stock
Changes in Grassland Remaining Grassland (MMT C02 Eq.)
1990
2005
2019
2020
2021
2022
2023
Aboveground Live Biomass
(2.5)
(2.1)
(1.4)
(1.4)
(1.3)
(1.3)
(1.3)
Belowground Live Biomass
(0.4)
(0.3)
(0.2)
(0.2)
(0.2)
(0.2)
(0.2)
Dead Wood
2.7
2.6
2.6
2.5
2.5
2.5
2.5
Litter
(0.4)
(0.2)
+
+
+
+
+
Mineral Soils
18.6
18.6
22.0
9.3
3.8
6.6
15.5
Organic Soils
6.1
5.1
5.3
5.5
5.5
5.5
5.5
Total Net Flux
24.0
23.7
28.2
15.8
10.2
13.1
22.0
+ Does not exceed 0.05 MMT C02 Eq.
Notes: Totals may not sum due to independent rounding. Parentheses indicate negative values or net sequestration.
Table 6-43: Net C02 Flux from Soil, Dead Organic Matter and Biomass Carbon Stock
Changes in Grassland Remaining Grassland (MMT C)
1990
2005
2019
2020
2021
2022
2023
Aboveground Live Biomass
(0.7)
(0.6)
(0.4)
(0.4)
(0.4)
(0.4)
(0.4)
Belowground Live Biomass
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
Dead Wood
0.7
0.7
0.7
0.7
0.7
0.7
0.7
Litter
(0.1)
(0.1)
+
+
+
+
+
Mineral Soils
5.1
5.1
6.0
2.5
1.0
1.8
4.2
Organic Soils
1.7
1.4
1.4
1.5
1.5
1.5
1.5
Total Net Flux
6.5
6.5
7.7
4.3
2.8
3.6
6.0
+ Does not exceed 0.05 MMT C.
Notes: Totals may not sum due to independent rounding. Parentheses indicate negative values or net sequestration.
The spatial variability in soil organic carbon stock changes for 202046 is displayed in Figure 6-8 for
mineral soils and in Figure 6-9 for organic soils. Although relatively small on a per-hectare basis,
grassland soils gained carbon in isolated areas that mostly occurred in pastures of the upper Midwest
and eastern United States; losses occurred primarily in the northwestern region. For organic soils, the
regions with the highest rates of emissions coincide with the largest concentrations of organic soils that
occur in managed grassland, including the Southeastern Coastal Region (particularly Florida), areas
46 Only national-scale emissions are estimated for 2021 to 2023 in the current Inventory using the surrogate data method,
and therefore the fine-scale emission patterns in this map are based on land use data from 2020.
Land Use, Land-Use Change, and Forestry 6-97
-------
surrounding the Great Lakes in the upper Midwest and Northeast, and a few isolated areas along the
Pacific Coast.
Figure 6-8: Total Net Annual Soil Carbon Stock Changes for Mineral Soils under
Agricultural Management within States, 2020, Grassland Remaining Grassland
Notes: Only national-scale soil organic carbon stock changes are estimated for 2021 to 2023 in the current Inventory using a
surrogate data method, and therefore the fine-scale emission patterns in this map are based on inventory data from 2020.
Negative values represent a net increase in soil organic carbon stocks, and positive values represent a net decrease in soil
organic carbon stocks. This figure was developed using a kriging method to develop a continuous surface from the NRI sample.
6-98 inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Figure 6-9: Total Net Annual Soil Carbon Stock Changes for Organic Soils under
Agricultural Management within States, 2020, Grassland Remaining Grassland
L $
\
r
,,::v
\
—3
MT C02 ha"1 yr1
~ <10
~ 10 to 20
¦ 20 to 30
¦ 30 to 40
¦ > 40
t:-W ffcisRg
'A ) JL
A
Notes: Only national-scale soil organic carbon stock changes are estimated for 2021 to 2023 in the current Inventory using a
surrogate data method, and therefore the fine-scale emission patterns in this map are based on inventory data from 2020. This
figure was developed using a kriging method to develop a continuous surface from the NRI sample.
Methodology and Time-Series Consistency
The following section includes a description of the methodology used to estimate carbon stock changes
for grassland remaining grassland, including (1) aboveground and belowground biomass, dead wood
and litter carbon for woodlands, as well as (2) soil organic carbon stocks for mineral and organic soils.
Biomass, Dead Wood and Litter Carbon Stock Changes
Woodlands are lands that do not meet the definition of forest land or agroforestry (see Section 6.1), but
include woody tree vegetation with carbon storage in aboveground and belowground biomass, dead
wood and litter carbon (IPCC 2006) as described in the Forest Land Remaining Forest Land section.
Carbon stocks and net annual carbon stock change were determined according to the stock-difference
method for the conterminous United States, which involved applying carbon estimation factors to
annual forest inventories across time to obtain carbon stocks and then subtracting the values between
years to estimate the stock changes. The methods for estimating carbon stocks and stock changes for
woodlands in grassland remaining grassland are consistent with those in the forest land remaining
forest land section and are described in Annex 3.14. All annual National Forest Inventory (NFI) plots
available in the public FIA database (USDA Forest Service 2024) were used in the current Inventory.
Land Use, Land-Use Change, and Forestry 6-99
-------
While the NFI is an all-lands inventory, only those plots that meet the definition of forest land are
typically measured. However, in some cases, particularly in the Central Plains and Southwest United
States, woodlands have been measured as part of the survey. This analysis is limited to those plots and
is not considered a comprehensive assessment of trees outside of forest land that meet the definition of
grassland. The same methods are applied from 1990 to 2023 to ensure time-series consistency. This
methodology is consistent with IPCC guidance (2006).
Soil Carbon Stock Changes
The following section includes a brief description of the methodology used to estimate changes in soil
organic carbon stocks for grassland remaining grassland, including (1) agricultural land use and
management activities on mineral soils; and (2) agricultural land use and management activities on
organic soils. Further elaboration on the methodologies and data used to estimate stock changes from
mineral and organic soils is provided in the Cropland Remaining Cropland section and Annex 3.13.
Soil organic carbon stock changes are estimated for grassland remaining grassland on non-federal lands
according to land use histories recorded in the USDA National Resources Inventory (NRI) (USDA-NRCS
2020). Land use and some management information (e.g., grass type, soil attributes, and irrigation) were
originally collected for each NRI survey location on a five-year cycle beginning in 1982. In 1998, the NRI
program began collecting annual data, and the annual data are currently available through 2017 (USDA-
NRCS 2020). For 2018-2020, the time series is extended with the data provided in the National Land
Cover Dataset (NLCD) (Yang et al. 2018; Fry et al. 2011; Homer et al. 2007, 2015). The areas have been
modified in the original NRI survey through a process in which the Forest Inventory and Analysis (FIA)
survey data and the NLCD are harmonized with the NRI data. This process ensures that the land use
areas are consistent across all land use categories (see Section 6.1 for more information).
NRI survey locations are classified as grassland remaining grassland in a given year between 1990 and
2020 if the land use had been grassland for 20 years. NRI survey locations are classified according to
land use histories starting in 1979, and consequently the classifications are based on less than 20 years
from 1990 to 1998. This may have led to an overestimation of grassland remaining grassland in the early
part of the time series to the extent that some areas are converted to grassland between 1971 and 1978.
For federal lands, the land use history is derived from land cover changes in the NLCD (Yanget al. 2018;
Homer et al. 2007; Fry et al. 2011; Homer et al. 2015).
Soil Carbon Stock Changes for Mineral Soils
An IPCC Tier 3 model-based approach (Ogle et al. 2010) is applied to estimate carbon stock changes
from 1990 to 2020 for most mineral soils in grassland remaining grassland. The carbon stock changes
for the remaining soils are estimated with an IPCC Tier 2 method (Ogle et al. 2003), including gravelly,
cobbly, or shaley soils (greater than 35 percent by volume), as well as additional stock changes
associated with biosolids (i.e., treated sewage sludge) amendments and federal land.47
A surrogate data method is used to estimate soil organic carbon stock changes from 2021 to 2023 at the
national scale for land areas included in the Tier 2 and Tier 3 methods. Specifically, linear regression
models with autoregressive moving-average (ARMA) errors (Brockwell and Davis 2016) are used to
estimate the relationship between surrogate data and the 1990 to 2020 emissions data from the Tier 2
47 Federal [and is not a [and use, but rather an ownership designation that is treated as grass[and for purposes of these
ca[cu[ations. The specific [and use on federa[ [ands is not identified in the NRI survey (USDA-NRCS 2020).
6-100 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
and 3 methods. Surrogate data for these regression models are based on weather data from the PRISM
Climate Group (PRISM Climate Group 2022). See Box 6-3 in the Methodology section of Cropland
Remaining Cropland for more information about the surrogate data method.
Tier 3 Approach. Mineral soil organic carbon stocks and stock changes for grassland remaining
grassland are estimated using the DayCent ecosystem model (Parton et al. 1998; Del Grosso et al. 2001,
2011), as described in Cropland Remaining Cropland. The DayCent model utilizes the soil carbon
modeling framework developed in the Century model (Parton et al. 1987,1988,1994; Metherell et al.
1993), but has been refined to simulate dynamics at a daily time-step. Historical land-use patterns and
irrigation histories are simulated with DayCent based on the 2017 USDA NRI survey (USDA-NRCS 2020).
The amount of manure produced by each livestock type is calculated for managed and unmanaged
waste management systems based on methods described in Section 5.2 and Annex 3.12. Manure
nitrogen deposition from grazing animals (i.e., pasture/range/paddock (PRP) manure) is an input to the
DayCent model to estimate the influence of PRP manure on carbon stock changes for lands included in
the Tier 3 method. Carbon stocks and 95 percent confidence intervals are estimated for each year
between 1990 and 2020 using the NRI survey data. Further elaboration on the Tier 3 methodology and
data used to estimate carbon stock changes from mineral soils are described in Annex 3.13.
In order to ensure time-series consistency, the same methods are applied from 1990 to 2020 so that
changes reflect anthropogenic activity and not methodological adjustments. In addition, soil organic
carbon stock changes from 2021 to 2023 are approximated using a linear regression model with moving-
average (ARMA) errors, described in Box 6-4 of the Methodology section in Cropland Remaining
Cropland. This method is type of a linear extrapolation, which is a standard data splicing method for
estimating emissions at the end of a time series (IPCC 2006). Stock change estimates for 2021 to 2023
will be recalculated in future Inventories with an updated time series of activity data (see the Planned
Improvements section in Cropland Remaining Cropland).
Tier 2 Approach. The Tier 2 approach is based on the same methods described in the Tier 2 portion of
the Cropland Remaining Cropland section for mineral soils, with the exception of the manure nitrogen
deposition from grazing animals (i.e., PRP manure), and the land use and management data that are
used in the /nt/enfo/yforfederalgrasslands. First, the PRP nitrogen manure is included in the Tier 2
method that is not deposited on lands included in the Tier 3 method. Second, the NRI (USDA-NRCS
2020) provides land use and management histories for all non-federal lands, and is the basis for the Tier
2 analysis for these areas. However, NRI does not provide land use information on federal lands. The
land use data for federal lands is based on the NLCD (Yang et al. 2018; Fry et al. 2011; Homer et al. 2007;
Homer et al. 2015). In addition, the Bureau of Land Management (BLM) manages some of the federal
grasslands, and compiles information on grassland conditions through the BLM Rangeland Inventory
(BLM 2014). To estimate soil organic carbon stock changes from federal grasslands, rangeland
conditions in the BLM data are aligned with IPCC grassland management categories of nominal,
moderately degraded, and severely degraded in order to apply the appropriate emission factors. Further
elaboration on the Tier 2 methodology and data used to estimate carbon stock changes from mineral
soils are described in Annex 3.13.
In order to ensure time-series consistency, the Tier 2 method is applied from 1990 to 2020 so that
changes reflect anthropogenic activity and not methodological adjustments. In addition, soil organic
carbon stock changes are approximated for the remainder of the time series using a linear regression
model with moving-average (ARMA) errors (see Box 6-3 in the Methodology section in Cropland
Remaining Cropland). This method is type of a linear extrapolation, which is a standard data splicing
Land Use, Land-Use Change, and Forestry 6-101
-------
method for estimating emissions at the end of a time series (IPCC 2006). This method is type of a linear
extrapolation, which is a standard data splicing method for estimating emissions at the end of a time
series (IPCC 2006). As with the Tier 3 method, time series of activity data will be updated in a future
Inventory, and emissions from 2021 to 2023 will be recalculated.
Additional Mineral Carbon Stock Change Calculations
A Tier 2 method is used to adjust annual carbon stock change estimates for mineral soils between 1990
and 2023 to account for additional carbon stock changes associated with biosolids (i.e., treated sewage
sludge) amendments. Estimates of the amounts of biosolids nitrogen applied to agricultural land are
derived from national data on biosolids generation, disposition, and nitrogen content (see Section 7.2
for a detailed discussion of the methodology for estimating treated sewage sludge available for land
application). Although biosolids can be added to land managed for other land uses, it is assumed that
agricultural amendments only occur in grassland remaining grassland. Total biosolids generation data
for 1988,1996, and 1998, in dry mass units, are obtained from EPA (1999) and estimates for 2004 are
obtained from an independent national biosolids survey (NEBRA 2007). These values are linearly
interpolated to estimate values for the intervening years, and linearly extrapolated to estimate values for
years since 2004. Nitrogen application rates from Kellogg et al. (2000) are used to determine the amount
of area receiving biosolids amendments. The soil organic carbon storage rate is estimated at 0.38 metric
tons carbon per hectare per year for biosolids amendments to grassland as described above. The stock
change rate is based on country-specific factors and the IPCC default method (see Annex 3.13 for
further discussion).
Soil Carbon Stock Changes for Organic Soils
Annual carbon emissions from drained organic soils in grassland remaining grassland are estimated
using the Tier 2 method in IPCC (2006), which utilizes country-specific carbon loss rates (Ogle et al.
2003) rather than default IPCC rates. For more information, see the cropland remaining cropland section
for organic soils and Annex 3.13.
In order to ensure time-series consistency, the Tier 2 methods are applied from 1990 to 2020 so that
changes reflect anthropogenic activity and not methodological adjustments. In addition, soil organic
carbon stock changes for the remainder of the time series (i.e., 2021 to 2023) are approximated using a
linear regression model with moving-average (ARMA) errors (see Box 6-3 of the Methodology section in
cropland remaining cropland). This method is type of a linear extrapolation, which is a standard data
splicing method for estimating emissions at the end of a time series (IPCC 2006). This method is type of
a linear extrapolation, which is a standard data splicing method for estimating emissions at the end of a
time series (IPCC 2006). Estimates for 2021 to 2023 will be recalculated in future Inventories with an
updated time series of activity data.
Uncertainty
The uncertainty analysis for biomass, dead wood and litter carbon losses with woodlands is conducted
in the same way as the uncertainty assessment for forest ecosystem carbon flux associated with forest
land remaining forest land. Sample and model-based errors are combined using simple error
propagation methods provided by the IPCC (2006) by taking the square root of the sum of the squares of
the standard deviations of the uncertain quantities. For additional details, see the Uncertainty Analysis
in Annex3.14.
6-102 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
The uncertainty analysis for soil organic carbon stock changes using the Tier 3 and Tier 2 methodologies
is quantified from two variance components (Ogle et al. 2010), as described in Cropland Remaining
Cropland. For 2021 to 2023, there is additional uncertainty propagated through the Monte Carlo analysis
associated with the surrogate data method.
Uncertainty estimates are presented in Table 6-44 for each subcategory (i.e., soil organic carbon stocks
for mineral and organic soils) and the method applied in the Inventory analysis (i.e., Tier 2 and Tier 3).
Uncertainty estimates from the Tier 2 and 3 approaches are combined using the simple error
propagation methods provided by the IPCC (2006), i.e., by taking the square root of the sum of the
squares of the standard deviations of the uncertain quantities. The combined uncertainty for total
carbon stock changes in grassland remaining grassland ranges from more than 561 percent below and
above the 2023 stock change estimate of 22.0 MMT C02 Eq. The large relative uncertainty in the 2023
estimate is mostly due to variation in soil organic carbon stock changes that is not explained by the
surrogate data method, leading to high prediction error with the data splicing method.
Table 6-44: Approach 2 Quantitative Uncertainty Estimates for Carbon Stock Changes
Occurring Within Grassland Remaining Grassland (MMT C02 Eq. and Percent)
2023 Flux
Uncertainty Range Relative to Flux
Estimate3
Estimate
(MMT CO2 Eq.)
(%)
Source
(MMT CO2
Eq.)
Lower
Bound
Upper
Bound
Lower
Bound
Upper
Bound
Woodland Biomass:
Aboveground live biomass
(1.3)
(1.3)
(1.2)
-12%
10%
Belowground live biomass
(0.2)
(0.2)
(0.2)
-8%
8%
Dead wood
2.5
2.5
2.8
-13%
13%
Litter
+
+
+
-9%
9%
Mineral Soil C Stocks Grassland Remaining
Grassland, Tier 3 Methodology
16.4
16.4
139.8
-753%
753%
Mineral Soil C Stocks: Grassland Remaining
Grassland, Tier 2 Methodology
0.1
0.1
0.7
-491%
491%
Mineral Soil C Stocks: Grassland Remaining
Grassland, Tier 2 Methodology (Change in Soil C
due to Biosolids [i.e., Treated Sewage Sludge]
Amendments)
(1.0)
(1.0)
(0.5)
-50%
50%
Organic Soil C Stocks: Grassland Remaining
Grassland, Tier 2 Methodology
5.5
5.5
10.0
-82%
82%
Combined Uncertainty for Flux Associated with
Carbon Stock Changes Occurring in Grassland
Remaining Grassland
22.0
(101.5)
145.5
-561%
561%
+ Does not exceed 0.05 MMT C02 Eq.
a Range of C stock change estimates is a 95 percent confidence interval.
Notes: Totals may not sum due to independent rounding. Parentheses indicate negative values or net sequestration.
Uncertainty is also associated with a lack of reporting on biomass, dead wood and litter carbon stock
changes for agroforestry systems. Changes in biomass and dead organic matter carbon stocks are
assumed to be negligible in other grasslands, largely comprised of herbaceous biomass, although there
are significant changes at sub-annual time scales across seasons.
Land Use, Land-Use Change, and Forestry 6-103
-------
QA/QC and Verification
See the QA/QC and Verification section in Cropland Remaining Cropland.
Recalculations Discussion
Recalculations are associated with updated FIA data from 1990 to 2023 on biomass, dead wood and
litter carbon stocks in woodlands for grassland remaining grassland. These updates resulted in a slightly
lower average loss of 0.4 MMT C02 Eq. across the time series compared to the previous Inventory, which
is a 1.4 percent decrease on average over the time series.
Planned Improvements
The following are key planned improvements for Grassland Remaining Grassland aimed to resolving the
majority of the discrepancy between the managed land base for grassland remaining grassland and
amount of area currently included in grassland remaininggrassland emissions and removals
calculations (see Table 6-45). These improvements are planned for the next Inventory:
• Conducting an analysis of carbon stock changes for grasslands in Alaska.
• Incorporating the latest Land Representation area data. The current Inventory for grassland
remaininggrassland is based on the land representation from the previous Inventory.
Additionally, a review of available data on biosolids (i.e., treated sewage sludge) application will be
undertaken to improve the distribution of biosolids application on croplands, grasslands and
settlements. For information about other improvements, see the Planned Improvements section in
Cropland Remaining Cropland.
Table 6-45: Comparison of Managed Land Area in Grassland Remaining Grassland and
the Area in the current Grassland Remaining Grassland Inventory (Thousand Hectares)
Year
Area (Thousand Hectares)
Managed Land
Inventory
Difference
1990
330,060
279,705
50,355
1991
329,546
279,205
50,342
1992
329,090
278,755
50,335
1993
327,349
277,030
50,319
1994
325,724
275,418
50,306
1995
324,846
274,549
50,297
1996
323,995
273,701
50,293
1997
323,219
272,944
50,275
1998
321,267
271,010
50,257
1999
320,354
270,110
50,244
2000
319,360
269,131
50,229
2001
318,488
268,282
50,205
2002
317,865
267,883
49,982
2003
317,961
268,206
49,755
2004
317,759
268,232
49,527
6-104 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Year
Area (Thousand Hectares)
Managed Land
Inventory
Difference
2005
317,338
268,034
49,305
2006
316,831
267,748
49,084
2007
316,568
267,712
48,856
2008
316,485
267,861
48,624
2009
316,546
268,159
48,387
2010
316,142
267,984
48,158
2011
315,637
267,712
47,924
2012
315,337
267,586
47,750
2013
315,990
268,416
47,574
2014
316,668
269,271
47,397
2015
316,602
269,535
47,067
2016
316,757
269,602
47,155
2017
316,883
270,339
46,544
2018
319,784
273,168
46,616
2019
320,576
274,471
46,105
2020
321,736
275,079
46,657
2021
322,861
~
~
2022
323,760
~
~
2023
324,509
~
~
* Note: Activity data on land use has not been incorporated into the Inventory after 2020, designated with asterisks (*).
Non-C02 Emissions from Grassland Fires (Source
Category 4C1)
Fires are common in grasslands and are thought to have been a key feature shaping the evolution of the
grassland vegetation in North America (Daubenmire 1968; Anderson 2004). Fires can occur naturally
through lightning strikes but are also an important management practice to remove standing dead
vegetation and improve forage for grazing livestock. Woody and herbaceous biomass will be oxidized in a
fire, although in this section the current focus is primarily on herbaceous biomass.48 Biomass burning
emits a variety of trace gases including non-C02 greenhouse gases such as CH4 and N20, as well as CO
and NOx that can become greenhouse gases when they react with other gases in the atmosphere
(Andreae and Merlet 2001). IPCC (2006) recommends quantifying non-C02 greenhouse gas emissions
from all wildfires and prescribed burning occurring in managed grasslands.
Biomass burning in grasslands of the United States (including burning emissions in grassland remaining
grassland and land converted to grassland) is a relatively small source of emissions, but it has increased
by 228 percent since 1990. In 2023, CH4 and N20 emissions from biomass burning in grasslands were
0.4 MMT C02 Eq. (14 kt) and 0.3 MMT C02 Eq. (1 kt), respectively. Annual emissions from 1990 to 2023
48 A planned improvement is underway to incorporate woodland tree biomass into the Inventory tor non-CCb emissions
from grassland fires.
Land Use, Land-Use Change, and Forestry 6-105
-------
have averaged approximately 0.4 MMT C02 Eq. (14 kt) of CH4and 0.3 MMTC02 Eq. (1 kt) of N20 (see
Table 6-46 and Table 6-47).
Table 6-46: CH4 and N20 Emissions from Biomass Burning in Grassland (MMT C02 Eq.)
1990
2005 1
2019
2020
2021
2022
2023
CH4 0.1 1
0.4 I
0.2
0.6
0.5
0.3
0.4
n2o 0.1
0.4 |
0.2
0.5
0.4
0.3
0.3
Total Net Flux 0.2
0.8 |
0.3
1.1
0.9
0.6
0.7
Note: Totals may not sum due to independent rounding.
Table 6-47: CH4, N2Q, CO, and NOx Emissions from Biomass Burning in Grassland (kt)
1990 |
2005
2019
2020
2021
2022
2023
cm 4
15
6
20
18
12
14
n2o +
1
1
2
2
1
1
CO 122
430
170
575
509
346
399
z
0
^1
26
10
34
31
21
24
+ Does not exceed 0.5 kt.
Methodology and Time-Series Consistency
The following section includes a description of the methodology used to estimate non-C02 greenhouse
gas emissions from biomass burning in grassland, including (1) determination of the land base that is
classified as managed grassland; (2) assessment of managed grassland area that is burned each year,
and (3) estimation of emissions resulting from the fires. For this Inventory, the IPCC Tier 1 method is
applied to estimate non-C02 greenhouse gas emissions from biomass burning in grassland from 1990 to
2020 (IPCC 2006). A data splicing method is used to estimate the emissions from 2021 to 2023, which is
discussed later in this section.
The land area designated as managed grassland is based primarily on the USDA National Resources
Inventory (NRI) (Nusser and Goebel 1997; USDA-NRCS 2020). NRI has survey locations across the entire
United States, but does not classify land use on federally-owned areas, and so survey locations on
federal lands are designated as grassland using land cover data from the National Land Cover Dataset
(NLCD) (Fry et al. 2011; Homer et al. 2007; Homer et al. 2015) (see Section 6.1).
The area of biomass burning in grasslands (grassland remaining grassland and land converted to
grassland) is determined using 30-m burned area data from the Monitoring Trends in Burn Severity
(MTBS) program for 1990 through 2020 (MTBS 2023; Picotte, et al. 20 20).49 NRI survey locations on
grasslands are designated as burned in a year if there is a fire within 500 m of the survey point according
to the MTBS fire data. The area of biomass burning is estimated from the NRI spatial weights and
aggregated to the country (Table 6-48).
Table 6-48: Thousands of Grassland Hectares Burned Annually
Year
1990 |
2005
2019
2020
2021
2022
2023
Thousand Hectares
457 I
1,612 1
637
2,156
NE
NE
NE
NE (Not Estimated)
Notes: Burned area was not estimated (NE) for 2021 to 2023, but will be updated in a future Inventory. For 1990 to 2020, the total
area of grassland burned is multiplied by the IPCC default factor for grassland biomass (4.1 tonnes dry matter per ha) (IPCC
49 See https://www.mtbs.gov/.
6-106 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
2006) to estimate the amount of combusted biomass. A combustion factor of 1 is assumed in this Inventory, and the resulting
biomass estimate is multiplied by the IPCC default grassland emission factors for CH4<2.3g ChU per kg dry matter), N20 (0.21 g
N20 per kg dry matter), CO (65 g CO per kg dry matter) and NOx (3.9 g NOx per kg dry matter) (IPCC 2006).
In order to ensure time-series consistency, the same methods are applied from 1990 to 2020 so that
changes reflect anthropogenic activity and not methodological adjustments. In addition, emissions
from 2021 to 2023 are approximated using a linear regression model with moving-average (ARMA) errors
(Brockwell and Davis 2016), described in Box 6-4 of the Methodology section in Cropland Remaining
Cropland. This method is type of a linear extrapolation, which is a standard data splicing method for
estimating emissions at the end of a time series (IPCC 2006). The Tier 1 method described previously
will be applied to recalculate the 2021 to 2023 emissions in a future Inventory.
Uncertainty
Emissions are estimated using a linear regression model with ARMA errors from 2021 to 2023. The
model produces estimates for the upper and lower bounds of the emission estimate and the results are
summarized in Table 6-49. Methane emissions from biomass burning in grassland for 2023 are
estimated to be between approximately 0.0 and 0.9 MMT C02 Eq. at a 95 percent confidence level. This
indicates a range of 100 percent below and 120 percent above the 2023 emission estimate of 0.4 MMT
C02 Eq. Nitrous oxide emissions are estimated to be between approximately 0.0 and 0.8 MMTC02 Eq.,
or 100 percent below and 120 percent above the 2023 emission estimate of 0.3 MMT C02 Eq.
Table 6-49: Uncertainty Estimates for Non-C02 Greenhouse Gas Emissions from
Biomass Burning in Grassland (MMT C02 Eq. and Percent)
Uncertainty Range Relative to Emission Estimate"
2023 Emission
(MMT CO2 Eq.)
(%)
Estimate
Lower
Upper
Lower
Upper
Source
Gas
(MMT CO2 Eq.)
Bound
Bound
Bound
Bound
Grassland Burning
CH4
0.4
+
0.9
-100%
+ 120%
Grassland Burning
n2o
0.3
+
0.8
-100%
+ 120%
+ Does not exceed 0.05 MMT C02 Eq.
a Range of emission estimates predicted by linear regression time-series model for a 95 percent confidence interval.
Uncertainty is also associated with lack of reporting of emissions from biomass burning in grasslands of
Alaska. Grassland burning emissions could be relatively large in this region of the United States, and
therefore extending this analysis to include Alaska is a planned improvement for the Inventory. There is
also uncertainty due to lack of reporting on the combustion of woody biomass, and this is another
planned improvement.
QA/QC and Verification
Quality control measures included checking input data, model scripts, and results to ensure data are
properly handled throughout the inventory process consistent with the U.S. Inventory QA/QC plan,
which is in accordance with Volume 1, Chapter 6 of the 2006 IPCC Guidelines (see Annex 8 for more
details). Inventory forms and text are reviewed and revised as needed to correct transcription errors.
Land Use, Land-Use Change, and Forestry 6-107
-------
Recalculations Discussion
No recalculations were needed for this source category.
Planned Improvements
Two key planned improvements have been identified for this source category, including:
• Incorporating country-specific grassland biomass factors. In the current Inventory, biomass
factors are based on a global default for grasslands that is provided by the IPCC (2006). There is
considerable variation in grassland biomass, however, which would affect the amount of fuel
available for combustion in a fire.
• Extending the analysis to include Alaska. Alaska has an extensive area of grassland and includes
tundra vegetation, although some of the areas are not managed. There has been an increase in
fire frequency in boreal forest of the region (Chapin et al. 2008), and this may have led to an
increase in burning of neighboring grassland areas.
Other potential improvements include:
• Incorporating grassland fires into DayCent model simulations.
• Incorporating non-C02 greenhouse emissions from burning woodland tree biomass in
grasslands.
These improvements are expected to be completed over the next 2-3 years, and will reduce uncertainty
and produce more accurate estimates of non-C02 greenhouse gas emissions from grassland burning.
6.7 Land Converted to Grassland (Source
Category 4C2)
Land converted to grassland includes all current grassland in an inventory year that had been in another
land use(s) during the previous 20 years (IPCC 2006).50 For example, cropland or forest land converted to
grassland during the past 20 years would be reported in this category. Recently converted lands are
retained in this category for 20 years as recommended by IPCC (2006). Grassland includes pasture and
rangeland that are used primarily but not exclusively for livestock grazing. Rangelands are typically
extensive areas of native grassland that are not intensively managed, while pastures are typically
seeded grassland (possibly following tree removal) that may also have additional management, such as
irrigation or interseeding of legumes.
Land use change can lead to large losses of carbon to the atmosphere, particularly conversions from
forest land (Houghton et al. 1983; Houghton and Nassikas 2017). Moreover, conversion of forest to another
50 USDA NRI survey locations are classified according to land use histories starting in 1979, and consequently the
classifications are based on less than 20 years from 1990 to 2001 .This may have led to an underestimation of land
converted to grassland in the early part of the time series to the extent that some areas are converted to grassland
between 1971 and 1978.
6-108 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
land use (i.e., deforestation) is one of the largest anthropogenic sources of emissions to the atmosphere
globally, although this source maybe declining according to a recent assessment (Tubiello et a 1.2015).
IPCC (2006) recommends quantifying changes in biomass, dead organic matter, and soil organic carbon
stocks due to land use change. All soil organic carbon stock changes are estimated and reported for land
converted to grassland, but there is limited quantification of other pools in this Inventory. Losses of
aboveground and belowground biomass, dead wood and litter carbon from forest land converted to
grassland are reported, as well as gains and losses associated with conversions to woodlands51 from
other land uses, including croplands converted to grasslands, settlements converted to grasslands and
other lands converted to grasslands. Gains and losses in total live biomass and changes in soil organic
carbon stocks are also reported for all other land conversions to grasslands, but changes in dead organic
matter are not reported for these land use changes under the assumption that they are minor changes.
There is a discrepancy between the current land representation (see Section 6.1) and the area data that
have been used in the inventory for land converted to grassland. Specifically, grassland in Alaska is not
included in the Inventory, and this leads to a difference between the managed area in land converted to
grassland in the land representation and the grassland area included in the emissions and removal
calculations for land converted to grassland (Table 6-55). Improvements are underway to incorporate
grassland area in Alaska as part of future Inventories (see Planned Improvements section).
The largest carbon losses with land converted to grassland are associated with aboveground biomass,
belowground biomass, dead wood, and litter carbon losses from forest land converted to grassland (see
Table 6-50 and Table 6-51). These four pools led to net emissions in 2023 of 31.0,4.3, 5.2, and 8.1 MMT
C02 Eq. (8.5,1.2,1.4, and 2.2 MMT C), respectively. The losses associated with forest land converted to
grassland are partially offset by gains in mineral soil carbon stocks associated with other land converted
to grassland and due to cropland converted to grassland, which leads to less intensive management of
the soil. Drainage of organic soils for grassland management led to C02 emissions to the atmosphere of
1.4 MMT C02 Eq. (0.4 MMT C). The total net carbon stock change in 2023 for land converted to grassland
is estimated as a loss of 20.9 MMT C02 Eq. (5.7 MMT C) or a net source of emissions, which represents a
decrease in carbon stock loss by 41.3 percent compared to the initial inventory year of 1990.
Table 6-50: Net C02 Flux from Soil, Dead Organic Matter and Biomass Carbon Stock
Changes for Land Converted to Grassland (MMT C02 Eq.)
1990
2005 I
2019
2020
2021
2022
2023
Cropland Converted to Grassland
(11.5)
(18.1)
(14.5)
(13.4)
(17.6)
(16.5)
(16.4)
Aboveground Live Biomass1
(1.4)
(1.3)
(4.3)
(4.2)
(4.2)
(4.2)
(4.1)
Belowground Live Biomass
{+)
(+)
{+)
{+)
{+)
{+)
{+)
Dead Wood
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
Litter
(0.1)
(+)
{+)
{+)
{+)
{+)
{+)
Mineral Soils
(10.4)
(18.1)
(11.1)
(10.1)
(14.4)
(13.3)
(13.2)
Organic Soils
0.6
1.4
1.1
1.0
1.0
1.0
1.0
Forest Land Converted to Grassland
52.9
51.4
48.6
48.6
48.5
48.5
48.5
Aboveground Live Biomass1
34.3
33.0
31.0
31.0
31.0
31.0
31.0
Belowground Live Biomass
4.7
4.6
4.3
4.3
4.3
4.3
4.3
51 Woodlands are considered grasslands in the U.S. land representation because they do not meet the definition of forest
land.
Land Use, Land-Use Change, and Forestry 6-109
-------
1990
2005
2019
2020
2021
2022
2023
Dead Wood
5.4
5.3
5.2
5.2
5.2
5.2
5.2
Litter
8.6
8.4
8.1
8.1
8.1
8.1
8.1
Mineral Soils
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
Organic Soils
+
0.1
0.1
0.1
0.1
0.1
0.1
Other Lands Converted to Grassland
(4.8)
(10.3)
(12.2)
(10.0)
(9.9)
(9.9)
(10.1)
Aboveground Live Biomass1
(0.9)
(0.8)
(1.8)
(1.9)
(2.0)
(2.0)
(2.0)
Belowground Live Biomass
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Dead Wood
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Litter
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
Mineral Soils
(3.8)
(9.4)
(10.4)
(8.1)
(7.9)
(7.9)
(8.1)
Organic Soils
(+)
(+)
0.1
0.1
0.1
0.1
0.1
Settlements Converted to Grassland
(0.6)
(0.9)
(1.1)
(1.1)
(1.1)
(1.1)
(1.1)
Aboveground Live Biomass1
(0.2)
(0.3)
(0.4)
(0.5)
(0.5)
(0.5)
(0.5)
Belowground Live Biomass
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Dead Wood
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
Litter
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
Mineral Soils
(0.1)
(0.3)
(0.5)
(0.4)
(0.4)
(0.4)
(0.4)
Organic Soils
+
+
+
+
+
+
+
Wetlands Converted to Grassland
(0.4)
(0.1)
+
(+)
(+)
(+)
(+)
Aboveground Live Biomass1
(0.4)
(0.2)
(0.2)
(0.2)
(0.2)
(0.2)
(0.2)
Belowground Live Biomass
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Dead Wood
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Litter
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Mineral Soils
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Organic Soils
0.1
0.2
0.2
0.2
0.2
0.2
0.2
Aboveground Live Biomass1
31.4
30.4
24.3
24.3
24.2
24.2
24.2
Belowground Live Biomass
4.7
4.5
4.3
4.3
4.3
4.3
4.3
Dead Wood
5.1
5.1
5.0
5.0
5.0
5.0
5.0
Litter
8.3
8.2
7.9
7.9
7.8
7.8
7.8
Total Mineral Soil Flux
(14.6)
(27.9)
(22.1)
(18.7)
(22.8)
(21.7)
(21.8)
Total Organic Soil Flux
0.7
1.8
1.5
1.4
1.4
1.4
1.4
Total Net Flux
35.6
21.9
20.9
24.1
19.9
20.9
20.9
+ Does not exceed 0.05 MMT C02 Eq.
1 Biomass C stock changes associated with conversions to grasslands that are not woodlands have been estimated as the total
live biomass change and reported as aboveground live biomass. The Tier 1 method that has been used for the conversions to
non-woodland grasslands only estimates the total live biomass change. The exception is perennial woody crop conversions to
grasslands in which only the aboveground live biomass is estimated with the Tier 1 method.
Notes: Totals may not sum due to independent rounding. Parentheses indicate negative values or net sequestration.
Table 6-51: Net C02 Flux from Soil, Dead Organic Matter and Biomass Carbon Stock
Changes for Land Converted to Grassland (MMT C)
1990
2005 |
2019
2020
2021
2022
2023
Cropland Converted to Grassland
(3.1) I
(4.9)
(4.0)
(3.7)
(4.8)
(4.5)
(4.5)
Aboveground Live Biomass1
(0.4)
(0.4)
(1.2)
(1.2)
(1.1)
(1.1)
(1.1)
Belowground Live Biomass
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Dead Wood
<+>l
(+) | (+)
(+)
(+)
(+)
(+)
6-110 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
1990
2005
2019
2020
2021
2022
2023
Litter
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Mineral Soils
(2.8)
(4.9)
(3.0)
(2.8)
(3.9)
(3.6)
(3.6)
Organic Soils
0.2
0.4
0.3
0.3
0.3
0.3
0.3
Forest Land Converted to Grassland
14.4
14.0
13.3
13.2
13.2
13.2
13.2
Aboveground Live Biomass1
9.3
9.0
8.5
8.5
8.5
8.5
8.5
Belowground Live Biomass
1.3
1.3
1.2
1.2
1.2
1.2
1.2
Dead Wood
1.5
1.4
1.4
1.4
1.4
1.4
1.4
Litter
2.3
2.3
2.2
2.2
2.2
2.2
2.2
Mineral Soils
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Organic Soils
+
+
+
+
+
+
+
Other Lands Converted to Grassland
(1.3)
(2.8)
(3.3)
(2.7)
(2.7)
(2.7)
(2.8)
Aboveground Live Biomass1
(0.2)
(0.2)
(0.5)
(0.5)
(0.5)
(0.5)
(0.6)
Belowground Live Biomass
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Dead Wood
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Litter
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Mineral Soils
(1.0)
(2.6)
(2.8)
(2.2)
(2.2)
(2.1)
(2.2)
Organic Soils
+
+
+
+
+
+
+
Settlements Converted to Grassland
(0.2)
(0.2)
(0.3)
(0.3)
(0.3)
(0.3)
(0.3)
Aboveground Live Biomass1
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
Belowground Live Biomass
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Dead Wood
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Litter
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Mineral Soils
(+)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
Organic Soils
+
+
+
+
+
+
+
Wetlands Converted to Grassland
(0.1)
(+)
+
(+)
(+)
(+)
(+)
Aboveground Live Biomass1
(0.1)
(0.1)
(+)
(+)
(+)
(+)
(+)
Belowground Live Biomass
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Dead Wood
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Litter
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Mineral Soils
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Organic Soils
+
0.1
0.1
0.1
0.1
0.1
0.1
Aboveground Live Biomass1
8.6
8.3
6.6
6.6
6.6
6.6
6.6
Belowground Live Biomass
1.3
1.2
1.2
1.2
1.2
1.2
1.2
Dead Wood
1.4
1.4
1.4
1.4
1.4
1.4
1.4
Litter
2.3
2.2
2.1
2.1
2.1
2.1
2.1
Total Mineral Soil Flux
(4.0)
(7.6)
(6.0)
(5.1)
(6.2)
(5.9)
(5.9)
Total Organic Soil Flux
0.2
0.5
0.4
0.4
0.4
0.4
0.4
Total Net Flux
9.7
6.0
5.7
6.6
5.4
5.7
5.7
+ Does not exceed 0.05 MMT C.
1 Biomass C stock changes associated with conversions to grasslands that are not wood lands have been estimated as the total
live biomass change and reported as aboveground live biomass. The Tier 1 method that has been used for the conversions to
non-woodland grasslands only estimates the total live biomass change. The exception is perennial woody crop conversions to
grasslands in which only the aboveground live biomass is estimated with the Tier 1 method.
Notes: Totals may not sum due to independent rounding. Parentheses indicate negative values or net sequestration.
Land Use, Land-Use Change, and Forestry 6-111
-------
Methodology and Time-Series Consistency
For grassland biomass for other grassland conversions converted to grassland, including (1) loss of
aboveground and belowground biomass, dead wood and litter carbon with conversion to grasslands
from forest lands and woodlands designated in the grassland, (2) loss of total live biomass with
conversion to other land uses to grassland (i.e., non-woodlands), as well as (3) the impact from all land-
use conversions to grasslands on soil organic carbon stocks in mineral and organic soils.
Biomass, Dead Wood, and Litter Carbon Stock Changes
The IPCC Tier 1 approach is used to estimate biomass carbon stock changes to grasslands that are not
woodlands, according to land use histories recorded in the 2017 USDA NRI survey for non-federal lands
(USDA-NRCS 2020). For federal lands, the land use history is derived from land cover changes in the
National Land Cover Dataset (Yang et al. 2018; Fry et al. 2011; Homer et al. 2007, 2015). The areas have
been modified through a process in which the Forest Inventory and Analysis (FIA) survey data are
harmonized with the NRI data (Nelson et al. 2020). This process ensures that the land use areas are
consistent across all land use categories (see Section 6.1 for more information). Land use and some
management information were originally collected for each NRI survey location on a 5-year cycle
beginning in 1982. In 1998, the NRI program began collecting annual data, and the annual data have
been incorporated from the NRI into the Inventory analysis through 2017 (USDA-NRCS 2020) (Table
6-52). Data splicing methods are used for the remainder of the time series.
Table 6-52: Thousands of Hectares of Land for Total Live Biomass Associated with
Land-Use Conversions to Grasslands
1990 |
2005
2015
2016
2017
2018-2023
Croplands Converted to Grasslands
293.4
378.4
635.1
177.3
125.7
*
Annual Crops
288.3
357.9
630.8
176.2
125.7
~
Non-Woody Crops
0.6
1.6
1.7
1.0
-
~
Perennial Woody Crops
4.5
18.9
2.8
0.1
-
~
Other Lands Converted to Grasslands
47.1
40.4
58.7
12.8
89.7
*
Settlements Converted to Grasslands
26.8
10.0
5.9
5.5
2.7
*
Wetlands Converted to Grasslands
1.7 |
8.2 |
7.2
5.9
6.8
*
Total: Land Converted to Grasslands
369
437
707.2
201.5
225.0
*
* NRI data have not been incorporated into the Inventory after 2017, designated with asterisks (*). Data splicing methods are
used for the remainder of the time series.
indicates true zero.
The difference between the stocks is reported as the stock change under the assumption that the
change occurred in the year of the conversion. Biomass carbon losses from croplands converted to
grasslands include aboveground and belowground biomass carbon for non-woody crops, and
aboveground woody biomass from perennial croplands. Biomass carbon stock factors are assigned for
each land use subcategory (e.g., annual or perennial cropland), perennial type, and maturity class.
Biomass carbon stocks in settlements, wetlands and other lands were assumed to be zero (IPCC 2006).
The total area of each land converted to settlements was multiplied by applicable factors from IPCC
(2006 and 2019) (Table 6-53). Biomass values are disaggregated by climate zones for non-woodland
grasslands (IPCC 2006). For perennial croplands, factors vary by climate domain, perennial type, and
maturity class as indicated in IPCC (2019).
6-112 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Table 6-53: Carbon stock factors for total live biomass associated with land use
conversions to grassland (non-woodland).
Land Use
Climate
Subcategory/
Type
Maturity
Biomass C
Stock (t ha-1)*
Source
Settlements,
Wetlands and
Other Lands
0
IPCC 2006
Perennial Woody
Crops
Temperate
Orchards
Immature
0.43* ± 46%
IPCC 2019 (Table 5.3)
Vineyards
Immature
0.28* ± 26%
IPCC 2019
Table 5.3)
Orchards
Mature
8.50 ±19%
IPCC 2019
Table 5.3)
Vineyards
Mature
5.50 ± 18%
IPCC 2019
Table 5.3)
Non-Woody
Crops
-
Annual Crops
-
4.70 ± 75%
IPCC 2019
Table 8.4)
Cold Temperate - Dry
Hay
-
3.07 ± 75%
IPCC 2006
Table 6.4)
Cold Temperate - Moist
-
6.39 ± 75%
IPCC 2006
Table 6.4)
Warm Temperate - Dry
-
2.87 ± 75%
IPCC 2006
Table 6.4)
Warm Temperate -
Moist
-
6.35 ± 75%
IPCC 2006
Table 6.4)
Tropical - Dry
-
4.09 ± 75%
IPCC 2006
Table 6.4)
Tropical - Moist & Wet
-
7.57 ± 75%
IPCC 2006
Table 6.4)
Grasslands
Cold Temperate - Dry
-
-
3.07 ± 75%
IPCC 2006
Table 6.4)
Cold Temperate - Moist
-
-
6.39 ± 75%
IPCC 2006
Table 6.4)
Warm Temperate - Dry
-
-
2.87 ± 75%
IPCC 2006
Table 6.4)
Warm Temperate -
Moist
-
-
6.35 ± 75%
IPCC 2006
Table 6.4)
Tropical - Dry
-
-
4.09 ± 75%
IPCC 2006
Table 6.4)
Tropical - Moist & Wet
-
-
7.57 ± 75%
IPCC 2006
Table 6.4)
f Biomass C Stock for Hay and Grasslands obtained by multiplying biomass values by 0.47 carbon fraction (IPCC 2006, page
6.29).
* Biomass C stock from immature perennial woody crops converted to grasslands was obtained by multiplying annual
aboveground biomass C accumulation rate (IPCC 2019, Table 5.3) by the age of the stand.
A Tier 2 method is applied to estimate biomass, dead wood and litter carbon stock changes for forest
land converted to grassland. Estimates are calculated in the same way as those in the forest land
remaining forest land category using data from the USDA Forest Service, Forest Inventory and Analysis
(FIA) program (USDA Forest Service 2024) and in the Eastern U.S., IPCC (2006) defaults for biomass in
grasslands. There is limited data on grassland carbon stocks so only default biomass estimates (IPCC
2006) for grasslands were used to estimate carbon stock changes (litter and dead wood carbon stocks
were assumed to be zero since no reference carbon density estimates exist for croplands) in the eastern
United States. The difference between the stocks is reported as the stock change under the assumption
that the change occurred in the year of the conversion.
The amount of biomass carbon that is lost abruptly with forest land converted to grasslands is estimated
based on the amount of carbon before conversion and the amount of carbon following conversion
according to remeasurements in the FIA program. This approach is consistent with IPCC (2006) that
assumes there is an abrupt change during the first year, but does not necessarily capture the slower
change over the years following conversion until a new steady state is reached. It was determined that
using an IPCC Tier 1 approach that assumes all carbon is lost in the year of conversion for forest land
Land Use, Land-Use Change, and Forestry 6-113
-------
converted to grasslands in the West and Great Plains states does not accurately characterize the
transfer of carbon in woody biomass during abrupt or gradual land use change. To estimate this transfer
of carbon in woody biomass, state-specific carbon densities for woody biomass remaining on these
former forest lands following conversion to grasslands were developed and included in the estimation of
carbon stock changes from forest land converted to grasslands in the West and Great Plains states. A
review of the literature in grassland and rangeland ecosystems (Asner et al. 2003; Huang et al. 2009;
Tarhouni et al. 2016), as well as an analysis of FIA data, suggests that a conservative estimate of 50
percent of the woody biomass carbon density was lost during conversion from forest land to grasslands.
This estimate was used to develop state-specific carbon density estimates for biomass, dead wood, and
litter for grasslands in the West and Great Plains states, and these state-specific carbon densities were
applied in the compilation system to estimate the carbon losses associated with conversion from forest
land to grassland in the West and Great Plains states. Further, losses from forest land to what are often
characterized as woodlands are included in this category using FIA plot remeasurements and the
methods and models briefly described below and in detail in Domke et al. (2022) and Westfall et al.
(2024).
If FIA plots include data on individual trees, aboveground and belowground carbon density estimates are
based on Woodall et al. (2011) and Westfall et al. (2024). Aboveground and belowground biomass
estimates also include live understory which is a minor component of biomass defined as all biomass of
undergrowth plants in a forest, including woody shrubs and trees less than 2.54 cm dbh. For this
Inventory, it was assumed that 10 percent of total understory carbon mass is belowground (Smith et al.
2006). Estimates of carbon density are based on information in Birdsey (1996) and biomass estimates
from Jenkins et al. (2003). If FIA plots include data on standing dead trees, standing dead tree carbon
density is estimated following the basic method applied to live trees (Woodall et al. 2011, Westfall et al.
2024) with additional modifications to woodland species to account for decay and structural loss
(Domke et al. 2011; Harmon et al. 2011).
If FIA plots include data on downed dead wood, downed dead wood carbon density is estimated based
on measurements of a subset of FIA plots for downed dead wood (Domke et al. 2013; Woodall and
Monleon 2008). Downed dead wood is defined as pieces of dead wood greater than 7.5 cm diameter, at
transect intersection, that are not attached to live or standing dead trees. This includes stumps and
roots of harvested trees. To facilitate the downscaling of downed dead wood carbon estimates from the
state-wide population estimates to individual plots, downed dead wood models specific to regions and
forest types within each region are used. Litter carbon is the pool of organic carbon (also known as duff,
humus, and fine woody debris) above the mineral soil and includes woody fragments with diameters of
up to 7.5 cm. A subset of FIA plots are measured for litter carbon. If FIA plots include litter material, a
modeling approach using litter carbon measurements from FIA plots is used to estimate litter carbon
density (Domke et al. 2016). See Annex 3.14 for more information about reference carbon density
estimates for forest land.
In order to ensure time-series consistency, the same methods are applied from 1990 to 2023 for the
forest lands converted to grasslands and conversions to woodlands so that changes reflect
anthropogenic activity and not methodological adjustments. See Annex 3.14 for more information about
reference carbon density estimates for forest land and woodlands, and the compilation system used to
estimate carbon stock changes from forest land. For all other land use conversions to grasslands, the
same methods are applied from 1990 to 2017, and a data splicing method is used to estimate biomass
carbon loss for the remainder of the 2018 to 2023 time series (see Box 6-3 in Section 6.4 cropland
6-114 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
remaining cropland). Specifically, a linear regression model with moving-average (ARMA) errors
(Brockwell and Davis 2016) is used to impute the missing C stock changes using trends from 1990 to
2017. This method is type of a linear extrapolation, which is a standard data splicing method for
estimating emissions at the end of a time series (IPCC 2006). The time series will be recalculated in a
future Inventory with the methods described previously for biomass carbon stock changes.
Soil Carbon Stock Changes
Soil organic carbon stock changes are estimated for land converted to grassland according to land use
histories recorded in the 2017 USDA NRI survey for non-federal lands (USDA-NRCS 2020). Land use and
some management information (e.g., crop type, soil attributes, and irrigation) were originally collected
for each NRI survey location on a five-year cycle beginning in 1982. In 1998, the NRI Program began
collecting annual data, and the annual data are currently available through 2017 (USDA-NRCS 2020).
For 2018 through 2020, the time series is extended with the crop data provided in USDA-NASS CDL
(USDA-NASS 2021), while survey locations identified as grasslands are assumed to not change over this
time period. However, the areas have been modified in the original NRI survey through a process in
which the Forest Inventory and Analysis (FIA) survey data and the National Land Cover Dataset (NLCD;
Yangetal. 2018) are harmonized with the NRI data. This process ensures that the land use areas are
consistent across all land use categories (see Section 6.1 for more information).
NRI survey locations are classified as land converted to grassland in a given year between 1990 and
2020 if the land use is grassland but had been classified as another use during the previous 20 years.
NRI survey locations are classified according to land use histories starting in 1979, and consequently
the classifications are based on less than 20 years from 1990 to 1998. This may have led to an
underestimation of land converted to grassland in the early part of the time series to the extent that
some areas are converted to grassland between 1971 and 1978. For federal lands, the land use history is
derived from land cover changes in the NLCD (Yang et al. 2018; Homer et al. 2007; Fry et al. 2011;
Homer etal. 2015).
Soil Carbon Stock Changes for Mineral Soils
An IPCC Tier 3 model-based approach (Ogle et al. 2010) is applied to estimate carbon stock changes in
mineral soils for most of the area in land converted to grassland. Carbon stock changes on the
remaining area are estimated with an IPCC Tier 2 approach (Ogle et al. 2003), including prior cropland
used to produce vegetables, tobacco, and perennial/horticultural crops; land areas with very gravelly,
cobbly, or shaley soils (greater than 35 percent by volume); and land converted to grassland from
another land use other than cropland.
A surrogate data method is used to estimate soil organic carbon stock changes from 2021 to 2023 at the
national scale for land areas included in the Tier 2 and Tier 3 methods. Specifically, linear regression
models with autoregressive moving-average (ARMA) errors (Brockwell and Davis 2016) are used to
estimate the relationship between surrogate data and the 1990 to 2020 emissions data that are derived
using Tier 2 and 3 methods. Surrogate data for these regression models includes weather data from the
PRISM Climate Group (PRISM Climate Group 2022). See Box 6-3 in the Methodology section of cropland
remaining cropland for more information about the surrogate data method.
Tier 3 Approach. Mineral soil organic carbon stocks and stock changes are estimated using the
DayCent ecosystem model (Parton et al. 1998; Del Grosso et al. 2001, 2011). The DayCent model
utilizes the soil carbon modeling framework developed in the Century model (Parton et al. 1987,1988,
Land Use, Land-Use Change, and Forestry 6-115
-------
1994; Metherell et a 1.1993), but has been refined to simulate dynamics at a daily time-step. Historical
land use patterns and irrigation histories are simulated with DayCent based on the 2017 USDA NRI
survey (USDA-NRCS 2018). Carbon stocks and 95 percent confidence intervals are estimated for each
year between 1990 and 2020. See the cropland remaining cropland section and Annex 3.13 for
additional discussion of the Tier 3 methodology for mineral soils.
In order to ensure time-series consistency, the same methods are applied from 1990 to 2020 so that
changes reflect anthropogenic activity and not methodological adjustments. In addition, soil organic
carbon stock changes from 2021 to 2023 are approximated using a linear regression model with moving-
average (ARMA) errors (Brockwell and Davis 2016) to impute the missing carbon stock changes based on
trends from 1990 to 2020. This method is type of a linear extrapolation, which is a standard data splicing
method for estimating emissions at the end of a time series (IPCC 2006). Stock change estimates for
2021 to 2023 will be recalculated in future Inventories with an updated time series of activity data (see
the Planned Improvements section in cropland remaining cropland).
Tier 2 Approach. For the mineral soils not included in the Tier 3 analysis, soil organic carbon stock
changes are estimated using a Tier 2 Approach, as described in the Tier 2 Approach for mineral soils in
grassland remaining grassland and Annex 3.13. In order to ensure time-series consistency, the Tier 2
method is applied from 1990 to 2020 so that changes reflect anthropogenic activity and not
methodological adjustments. In addition, soil organic carbon stock changes from 2021 to 2023 are
approximated using a linear regression model with moving-average (ARMA) errors (Brockwell and Davis
2016) to impute the missing carbon stock changes based on trends from 1990 to 2020. This method is
type of a linear extrapolation, which is a standard data splicing method for estimating emissions at the
end of a time series (IPCC 2006). As with the Tier 3 method, stock change estimates for 2021 to 2023 will
be recalculated in future Inventories with an updated time series of activity data.
Soil Carbon Stock Changes for Organic Soils
Annual carbon emissions from drained organic soils in land converted to grassland are estimated using
the Tier 2 method provided in IPCC (2006), with country-specific carbon loss rates (Ogle et al. 2003) as
described in the cropland remaining cropland section. Further elaboration on the methodology is also
provided in Annex 3.13 for organic soils.
In order to ensure time-series consistency, the Tier 2 method is applied from 1990 to 2020 so that
changes reflect anthropogenic activity and not methodological adjustments. In addition, soil organic
carbon stock changes are approximated for the remainder of the time series with a linear regression
model with moving-average (ARMA) errors (see Box 6-3 of the Methodology section in Cropland
Remaining Cropland). This method is type of a linear extrapolation, which is a standard data splicing
method for estimating emissions at the end of a time series (IPCC 2006). Annual carbon emissions from
drained organic soils from 2021 to 2023 will be recalculated in future Inventories with an updated time
series of activity data.
Uncertainty
The uncertainty analyses for biomass, dead wood and litter carbon losses with forest land converted to
grassland and other land use conversions to woodlands are conducted in the same way as the
uncertainty assessment for forest ecosystem carbon flux in the forest land remaining forest land
category. Sample and model-based error are combined using simple error propagation methods
6-116 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
provided by the IPCC (2006), by taking the square root of the sum of the squares of the standard
deviations of the uncertain quantities. For additional details see the Uncertainty Analysis in Annex 3.14.
The uncertainty analyses for soil organic carbon stock changes using the Tier 3 and Tier 2 methodologies
are quantified from two variance components (Ogle et al. 2010), as described in cropland remaining
cropland. For 2021 to 2023, there is additional uncertainty propagated through the Monte Carlo analysis
associated with a data splicing method, which is also described in the Cropland Remaining Cropland
section.
Uncertainty estimates are presented in Table 6-54 for each sub-source (i.e., biomass carbon stocks,
mineral and organic carbon stocks in soils) and the method applied in the inventory analysis (i.e., Tier 2
and Tier 3). Uncertainty estimates from the Tier 2 and 3 approaches are combined using the simple error
propagation methods provided by the IPCC (2006), as discussed in the previous paragraph. The
combined uncertainty for total carbon stocks in land converted to grassland ranges from 206 percent
below to 206 percent above the 2023 stock change estimate of 20.9 MMT C02 Eq. The large relative
uncertainty around the 2023 stock change estimate is partly due to large uncertainties in biomass, dead
organic matter, and litter carbon estimates, in addition to variation in soil organic carbon stock changes
that is not explained by the data splicing method.
Table 6-54: Approach 2 Quantitative Uncertainty Estimates for Soil, Dead Organic
Matter and Biomass Carbon Stock Changes occurring within Land Converted to
Grassland (MMT C02 Eq. and Percent)
Source
2023 Flux
Estimate3
(MMTCO2 Eq.)
Uncertainty Range Relative to Flux Estimate3
(MMT CO2 Eq.)
(%)
Lower
Bound
Upper
Bound
Lower
Bound
Upper
Bound
Cropland Converted to Grassland
(16.4)
(41.0)
8.2
-150%
150%
Aboveground Live Biomass
(4.1)
(19.1)
10.8
-362%
362%
Belowground Live Biomass
{+)
{+)
+
-97%
100%
Dead Wood
(0.1)
(0.1)
+
-95%
100%
Litter
{+)
(0.1)
+
-95%
100%
Mineral Soil C Stocks: Tier 3
(11.5)
(30.9)
7.8
-168%
168%
Mineral Soil C Stocks: Tier 2
(1.6)
(3.8)
0.5
-132%
132%
Organic Soil C Stocks: Tier 2
1.0
(0.1)
2.2
-113%
113%
Forest Land Converted to Grassland
48.5
14.4
82.6
-70%
70%
Aboveground Live Biomass
31.0
(1.3)
63.3
-104%
104%
Belowground Live Biomass
4.3
(0.2)
8.9
-104%
105%
Dead Wood
5.2
(0.2)
10.5
-104%
104%
Litter
8.1
(0.3)
16.5
-104%
104%
Mineral Soil C Stocks: Tier 2
(0.1)
(0.2)
+
-140%
140%
Organic Soil C Stocks: Tier 2
0.1
{+)
0.2
-143%
143%
Other Lands Converted to Grassland
(10.1)
(18.8)
(1.5)
-85%
85%
Aboveground Live Biomass
(2.0)
(8.3)
4.3
-313%
313%
Belowground Live Biomass
{+)
+
+
-100%
100%
Dead Wood
{+)
(0.1)
+
-70%
100%
Litter
(0.1)
(0.1)
{+)
-59%
47%
Land Use, Land-Use Change, and Forestry 6-117
-------
Uncertainty Range Relative to Flux Estimate3
2023 Flux
(MMT CO2 Eq.)
(%)
Estimate3
Lower
Upper
Lower
Upper
Source
(MMTCO2 Eq.)
Bound
Bound
Bound
Bound
Mineral Soil C Stocks: Tier 2
(8.1)
(14.0)
(2.2)
-73%
73%
Organic Soil C Stocks: Tier 2
0.1
(0.1)
0.2
-212%
212%
Settlements Converted to Grassland
(1.1)
(2.2)
(0.0)
-97%
96%
Aboveground Live Biomass
(0.5)
(1.6)
0.5
-208%
208%
Belowground Live Biomass
{+)
{+)
+
-45%
100%
Dead Wood
(0.1)
(0.1)
{+)
-62%
46%
Litter
(0.1)
(0.1)
{+)
-55%
61%
Mineral Soil C Stocks: Tier 2
(0.4)
(0.6)
(0.2)
-54%
43%
Organic Soil C Stocks: Tier 2
+
{+)
+
-451%
451%
Wetlands Converted to Grasslands
(+)
(0.7)
0.7
-3,705%
3,705%
Aboveground Live Biomass
(0.2)
(0.8)
0.5
-411%
410%
Belowground Live Biomass
{+)
+
+
-100%
100%
Dead Wood
{+)
{+)
+
-46%
100%
Litter
{+)
{+)
+
-138%
100%
Mineral Soil C Stocks: Tier 2
{+)
(0.1)
+
-213%
213%
Organic Soil C Stocks: Tier 2
0.2
(0.1)
0.5
-147%
147%
Total: Land Converted to Grassland
20.9
(22.1)
63.8
-206%
206%
Aboveground Live Biomass
24.2
(12.0)
60.4
-150%
150%
Belowground Live Biomass
4.3
(0.2)
8.8
-106%
106%
Dead Wood
5.0
(0.4)
10.3
-108%
108%
Litter
7.8
(0.5)
16.2
-107%
107%
Mineral Soil C Stocks: Tier 3
(11.5)
(30.9)
7.8
-168%
168%
Mineral Soil C Stocks: Tier 2
(10.2)
(16.5)
(4.0)
-61%
61%
Organic Soil C Stocks: Tier 2
1.4
0.2
2.6
-88%
88%
+ Absolute value does not exceed 0.05 MMT C02 Eq.
a Range of C stock change estimates is a 95 percent confidence interval.
Notes: Totals may not sum due to independent rounding. Parentheses indicate negative values or net sequestration.
Uncertainty is also associated with a lack of reporting on biomass, dead wood and litter carbon stock
changes for conversions to agroforestry systems in grasslands. The influence of agroforestry is difficult
to address because there are currently no datasets to evaluate the trends in the area and associated
carbon stocks in agroforestry systems. The influence of land use change to agroforestry will be further
explored in a future Inventory.
QA/QC and Verification
See the QA/QC and Verification section in Cropland Remaining Cropland for information on QA/QC
steps. In addition, land use conversions had errors in identifying the C stocks for lands converted
between federal and non-federal ownership. A unit conversion error was identified in the uncertainty
analysis for perennial woody biomass. All errors were corrected.
6-118 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Recalculations Discussion
Recalculations are associated with two revisions to the Inventory. First, updated FIA data from 1990 to
2023 on woody biomass, dead wood and litter carbon stocks were incorporated into forest land
converted to cropland estimates, including an update to the methodology to estimate downed
deadwood on forest land by implementing the National-Scale Volume and Biomass (NSVB) framework
to estimate carbon stock changes in this pool. See the Recalculations Discussion in Section 6.2 for
more information on this recalculation. Second, Tier 1 methodology to estimate biomass carbon stock
changes for cropland, settlements, wetlands, and other lands conversions to grassland (non-
woodlands) was newly implemented in this Inventory. As a result, land converted to grassland has an
estimated decrease in losses of carbon stock changes, leading to a net change of 4.0 MMT C02 Eq. on
average over the time series, representing a 16 percent decrease on average compared to the previous
Inventory.
Planned Improvements
There are two key improvements planned for an upcoming Inventory to land converted to grassland
focused on resolving the majority of the discrepancy between the managed land base for land converted
to grassland and amount of area currently included in land converted to grassland emissions and
removals calculations (see Table 6-55):
• Conducting an analysis of carbon stock changes for grassland in Alaska. See Planned
Improvement chapter section in Grassland Remaining Grassland for more information.
• Incorporating new land representation area data. The current Inventory for land converted to
grassland is based on the land representation from the previous Inventory. This improvement is
expected to be completed by the next Inventory (i.e., 1990 through 2024).
In addition, other potential improvements include:
• The amount of biomass carbon that is lost abruptly or the slower changes that continue to occur
over a decade or longer with forest land converted to grasslands will be further refined in a
future Inventory. The current values are estimated based on the amount of carbon before
conversion and an estimated level of carbon left after conversion based on limited plot data
from the FIA and published literature for the Western United States and Great Plains Regions.
The amount of carbon left after conversion will be further investigated with additional data
collection, particularly in the Western United States and Great Plains, including tree biomass,
understory biomass, dead wood and litter carbon pools. This improvement is expected to be
completed over the next 2 to 3 years.
For information about other improvements, see the Planned Improvements section in Cropland
Remaining Cropland.
Table 6-55: Comparison of Managed Land Area in Land Converted to Grassland and
Area in the current Land Converted to Grassland Inventory (Thousand Hectares)
Area (Thousand Hectares)
Year
Managed Land
Inventory
Difference
1990
9,298
9,297
2
Land Use, Land-Use Change, and Forestry 6-119
-------
Year
Area (Thousand Hectares)
Managed Land
Inventory
Difference
1991
9,490
9,488
2
1992
9,703
9,706
-3
1993
11,613
11,615
-2
1994
13,371
13,368
4
1995
14,034
14,035
-1
1996
14,718
14,723
-5
1997
15,402
15,408
-5
1998
19,279
19,285
-6
1999
20,132
20,139
-7
2000
21,244
21,253
-8
2001
22,341
22,345
-4
2002
23,080
22,817
263
2003
22,978
22,445
533
2004
23,913
23,108
805
2005
24,084
23,009
1,075
2006
24,684
23,341
1,343
2007
24,683
23,072
1,611
2008
25,252
23,373
1,879
2009
25,412
23,260
2,152
2010
25,757
23,336
2,422
2011
26,164
23,471
2,693
2012
26,151
23,292
2,859
2013
25,137
22,116
3,020
2014
23,960
20,776
3,183
2015
24,080
20,730
3,351
2016
23,513
19,993
3,521
2017
22,792
19,270
3,522
2018
19,953
16,429
3,524
2019
19,525
16,008
3,517
2020
18,792
15,168
3,624
2021
17,500
~
~
2022
16,418
~
~
2023
15,525
~
~
Note: Activity data on land use have not been incorporated into the Inventory after 2020, designated with asterisks (*).
6.8 Wetlands Remaining Wetlands
(Source Category 4D1)
Wetlands remaining wetlands includes all wetlands in an inventory year that have been classified as a
wetland for the previous 20 years, and in this Inventory, the flux estimates include:
• peatlands,
6-120 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
• coastal wetlands, and
• flooded land.
Peatlands Remaining Peatlands
Emissions from Managed Peatlands
Managed peatlands are peatlands that have been cleared and drained for the production of peat. The
production cycle of a managed peatland has three phases: land conversion in preparation for peat
extraction (e.g., clearing surface biomass, draining), extraction (which results in the emissions reported
under peatlands remaining peatlands), and abandonment, restoration, rewetting, or conversion of the
land to another use.
Carbon dioxide emissions from the removal of biomass and the decay of drained peat constitute the
major greenhouse gas flux from managed peatlands. Managed peatlands may also emit CH4 and N20.
The natural production of CH4 is largely reduced but not entirely eliminated when peatlands are drained
in preparation for peat extraction (Strack et al. 2004 as cited in the 2006IPCC Guidelines). Drained land
surface and ditch networks contribute to the CH4 flux in peatlands managed for peat extraction.
Methane emissions were considered insignificant under the IPCC Tier 1 methodology (IPCC 2006), but
are included in the emissions estimates for peatlands remaining peatlands consistent with the 2073
Supplement to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories: Wetlands (I PCC
2013). Nitrous oxide emissions from managed peatlands depend on site fertility. In addition, abandoned
and restored peatlands continue to release greenhouse gas emissions. Although methodologies are
provided to estimate emissions and removals from rewetted organic soils (which includes
rewetted/restored peatlands) in IPCC (2013) guidelines, information on the areal extent of
rewetted/restored peatlands in the United States is currently unavailable. The current Inventory
estimates C02, CH4and N20 emissions from peatlands managed for peat extraction in accordance with
IPCC (2006 and 2013) guidelines.
CO2, N20, and CH4 Emissions from Peatlands Remaining Peatlands
IPCC (2013) recommends quantifying C02, N20, and CH4 emissions from lands undergoing active peat
extraction (i.e., peatlands remaining peatlands) as part of the estimate for emissions from managed
wetlands. Peatlands occur where plant biomass has sunk to the bottom of water bodies and water-
logged areas and exhausted the oxygen supply below the water surface during the course of decay. Due
to these anaerobic conditions, much of the plant matter does not decompose but instead forms layers
of peat over decades and centuries. In the United States, peat is extracted for horticulture and
landscaping growing media, and for a wide variety of industrial, personal care, and other products. It has
not been used for fuel in the United States for many decades. Peat is harvested from two types of peat
deposits in the United States: Sphagnum bogs in northern states (e.g., Minnesota) and wetlands in
states further south (e.g., Florida). The peat from Sphagnum bogs in northern states, which is nutrient-
poor, is generally corrected for acidity and mixed with fertilizer. Production from more southerly states is
relatively coarse (i.e., fibrous) but nutrient-rich.
IPCC (2006 and 2013) recommend considering both on-site and off-site emissions when estimating C02
emissions from peatlands remaining peatlands using the Tier 1 approach. Current IPCC methodologies
estimate only on-site N20 and CH4 emissions. This is because off-site N20 estimates are complicated
Land Use, Land-Use Change, and Forestry 6-121
-------
by the risk of double-counting emissions from nitrogen fertilizers added to horticultural peat where
subsequent runoff or leaching into waterbodies can result in indirect N20 emissions that are already
included within the agricultural soil management category
On-site emissions from managed peatlands occur as the land is drained and cleared of vegetation, and
the underlying peat is exposed to sun, weather and oxygen. As this occurs, some of the peat deposit is
lost and C02 is emitted from the oxidation of the peat. Since N20 emissions from saturated ecosystems
tend to be low unless there is an exogenous source of nitrogen, N20 emissions from drained peatlands
are dependent on nitrogen mineralization and therefore on soil fertility. Peatlands occurring on highly
fertile/nutrient-rich soils, mostly located in the southern peatlands in Florida, contain significant
amounts of organic nitrogen in inert/microbially inaccessible forms. Draining land in preparation for peat
extraction allows bacteria to convert the organic nitrogen into nitrates through nitrogen mineralization
which leach to the surface where they are reduced to N20 during nitrification. Nitrate availability also
contributes to the activity of methanogens and methanotrophs that result in CH4 emissions (Blodau
2002; Treat et al. 2007 as cited in IPCC 2013). Drainage ditches, which are constructed to drain the land
in preparation for peat extraction, also contribute to the flux of CH4 through in situ production and lateral
transfer of CH4 from the organic soil matrix (IPCC 2013).
Off-site C02 emissions from managed peatlands occur from waterborne dissolved organic carbon
losses and the horticultural and landscaping use of peat. Dissolved organic carbon from water drained
off peatlands reacts within aquatic ecosystems and is converted to C02, which is then emitted to the
atmosphere (Billet et al. 2004 as cited in IPCC 2013). During the horticultural and landscaping use of
peat, nutrient-poor (but fertilizer-enriched) peat tends to be used in bedding plants and in greenhouse
and plant nursery production, whereas nutrient-rich (but relatively coarse) peat is used directly in
landscaping, athletic fields, golf courses, and plant nurseries. Most (nearly 94 percent) of the C02
emissions from peat occur off-site, as the peat is processed and sold to firms which, in the United
States, use it predominantly for the aforementioned horticultural and landscaping purposes.
Total emissions from peatlands remaining peatlands are estimated to be 0.6 MMT C02 Eq. in 2023 (see
Table 6-56 and Table 6-57) comprising 0.6 MMT C02 Eq. (604 kt) of C02, 0.004 MMT C02 Eq. (0.13 kt) of
CH4 and 0.0004 MMT C02 Eq. (0.002 kt) of N20. Total emissions in 2023 are 2.7 percent greater than total
emissions in 2022.
Total emissions from peatlands remaining peatlands have fluctuated between 0.6 and 1.3 MMT C02 Eq.
across the time series with a decreasing trend from 1990 until 1993, followed by an increasing trend
until reaching peak emissions in 2000. After 2000, emissions generally decreased until 2006 and then
increased until 2009. The trend reversed in 2009 and total emissions have generally decreased between
2009 and 2021, however, total emissions from peatlands increased slightly in 2022 and 2023 compared
to 2021. Carbon dioxide emissions from peatlands remaining peatlands have fluctuated between 0.6
and 1.3 MMT C02 Eq. across the time series, and these emissions drive the trends in total emissions.
Methane and N20 emissions remained close to zero across the time series.
Table 6-56: Emissions from Peatlands Remaining Peatlands (MMT C02 Eq.)
Gas
1990
2005
2019
2020
2021
2022
2023
CO2
1.1
1.1
0.6
0.6
0.5
0.6
0.6
Off-site
1.0
1.0
0.6
0.5
0.5
0.5
0.6
On-site
0.1
0.1
+
+
+
+
+
6-122 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
CH4 (On-site)
+
+
+
+
+
+
+
N2O (On-site)
+ 1
+ I
+
+
+
+
+
Total
1.1 |
1.1 |
0.6
0.6
0.6
0.6
0.6
+ Does not exceed 0.05 MMT C02 Eq.
Note: Totals may not sum due to independent rounding.
Table 6-57: Emissions from Peatlands Remaining Peatlands (kt)
Gas
1990
2005
2019
2020
2021
2022
2023
CO2
1,055
1,101
613
590
547
588
604
Off-site
985
1,030
572
550
509
548
563
On-site
70
71
41
41
38
40
41
CH4 (On-site)
*
*
+
+
+
+
+
N2O (On-site)
*
*
+
+
+
+
+
+ Does not exceed 0.5 kt.
Note: Totals may not sum due to independent rounding.
Methodology and Time-Series Consistency
Off-Site C02 Emissions
Carbon dioxide emissions from domestic peat production were estimated using a Tier 1 methodology
consistent with IPCC (2006). Off-site C02 emissions from peatlands remaining peatlands were
calculated by apportioning the annual weight of peat produced in the United States (Table 6-58) into
peat extracted from nutrient-rich deposits and peat extracted from nutrient-poor deposits using annual
percentage-by-weight figures. These nutrient-rich and nutrient-poor production values were then
multiplied by the appropriate default C fraction conversion factor taken from IPCC (2006) in order to
obtain off-site emission estimates. For the conterminous 48 states, both annual percentages of peat
type by weight and domestic peat production data were sourced from estimates and industry statistics
provided in the Minerals Yearbook and Mineral Commodity Summaries from the U.S. Geological Survey
(USGS; USGS 1995 through 2018; USGS 2024a; USGS 2024b; USGS 2024c; USGS 2024d). Hawaii is
assumed to have no peat production due to its absence from these sources. To develop these data, the
USGS (U.S. Bureau of Mines prior to 1997) obtained production and use information by surveying
domestic peat producers. On average, about 75 percent of the peat operations respond to the survey;
USGS estimates data for non-respondents on the basis of prior-year production levels (Apodaca 2011).
The estimates for Alaska rely on reported peat production from the annual Alaska's Mineral Industry
reports (DGGS 1993 through 2015). Similar to the U.S. Geological Survey, the Alaska Department of
Natural Resources, Division of Geological & Geophysical Surveys (DGGS) solicits voluntary reporting of
peat production from producers for the Alaska's Mineral Industry report. However, the report does not
estimate production for the non-reporting producers, resulting in larger inter-annual variation in
reported peat production from Alaska depending on the number of producers who report in a given year
(Szumigala 2011). In addition, in both the conterminous United States and Alaska, large variations in
peat production can also result from variation in precipitation and the subsequent changes in moisture
conditions, since unusually wet years can hamper peat production. The methodology estimates
emissions from Alaska separately from the conterminous United States because Alaska previously
conducted its own mineral surveys and reported peat production by volume, rather than by weight (Table
6-59). However, volume production data were used to calculate off-site C02 emissions from Alaska
Land Use, Land-Use Change, and Forestry 6-123
-------
applying the same methodology but with volume-specific C fraction conversion factors from IPCC
(20 06).52 Peat production was not reported for 2015 in Alaska's Mineral Industry 2014 report (DGGS
2015), and reliable data are not available beyond 2012, so Alaska's peat production in 2013 through
2023 (reported in cubic yards) was assumed to be equal to the 2012 value.
Consistent with IPCC (2013) guidelines, off-site C02 emissions from dissolved organic carbon were
estimated based on the total area of peatlands managed for peat extraction, which is calculated from
production data using the methodology described in the On-Site C02 Emissions section below. Carbon
dioxide emissions from dissolved organic carbon were estimated by multiplying the area of managed
peatlands by the default emission factor for dissolved organic C provided in IPCC (2013).
The United States has largely imported peat from Canada for horticultural purposes; in 2023, imports of
Sphagnum moss (nutrient-poor) peat from Canada represented 96 percent of total U.S. peat imports
and 80 percent of U.S. domestic consumption (USGS 2024d). Most peat produced in the United States is
reed-sedge peat, generally from southern states, which is classified as nutrient-rich by IPCC (2006). To
be consistent with the Tier 1 method, only domestic peat production is accounted for when estimating
off-site emissions. Higher-tier calculations of C02 emissions from apparent consumption would involve
consideration of the percentages of peat types stockpiled (nutrient-rich versus nutrient-poor) as well as
the percentages of peat types imported and exported.
Table 6-58: Peat Production of Conterminous 48 States (kt)
Type of Deposit
1990
2005
2019
2020
2021
2022
2023
Nutrient-Rich
595.1
657.6
329.4
343.4
291.6
315.0
324.0
Nutrient-Poor
55.4
27.4
36.6
10.6
32.4
35.0
36.0
Total Production
692.0
685.0
366.0
354.0
324.0
350.0
360.0
Note: Totals may not sum due to independent rounding.
Sources: United States Geological Survey (USGS) (1991-2017) Minerals Yearbook: Peat (1994-2016); United States Geological
Survey (USGS) (2018-2021) Minerals Yearbook: Peat- Tables-only release (2018); United States Geological Survey (USGS)
(2024) Mineral Commodity Summaries: Peat (2024).
Table 6-59: Peat Production of Alaska (Thousand Cubic Meters)
1990
2005
2019
2020
2021
2022
2023
Total Production
49.7
47.8
93.1
93.1
93.1
93.1
93.1
Sources: Division of Geological & Geophysical Surveys (DGGS), Alaska Department of Natural Resources (1997-2015) Alaska's
Mineral Industry Report (1997-2014).
On-site C02 Emissions
IPCC (2006) suggests basing the calculation of on-site emission estimates on the area of peatlands
managed for peat extraction differentiated by the nutrient type of the deposit (rich versus poor).
Information on the area of land managed for peat extraction is currently not available for the United
States, but consistent with IPCC (2006), an average production rate for the industry was applied to
derive a land area estimate. In a mature industrialized peat industry, such as exists in the United States
and Canada, the vacuum method can extract up to 100 metric tons per hectare per year (Cleary et al.
52 Peat produced from Alaska was assumed to be nutrient poor; as is the case in Canada, "where deposits of high-quality
[but nutrient poor] Sphagnum moss are extensive" (USGS 2008).
6-124 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
2005 as cited in IPCC 20 06).53 The area of land managed for peat extraction in the conterminous United
States was estimated using both nutrient-rich and nutrient-poor production data and the assumption
that 100 metric tons of peat are extracted from a single hectare in a single year, see Table 6-60. The
annual land area estimates were then multiplied by the IPCC (2013) default emission factor in order to
calculate on-site C02 emission estimates.
Production data are not available by weight for Alaska. In order to calculate on-site emissions resulting
from peatlands remaining peatlands in Alaska, the production data by volume were converted to weight
using annual average bulk peat density values, and then converted to land area estimates using the
assumption that a single hectare yields 100 metric tons, see Table 6-61. The IPCC (2006) on-site
emissions equation also includes a term that accounts for emissions resultingfrom the change in
carbon stocks that occurs during the clearing of vegetation prior to peat extraction. Area data on land
undergoing conversion to peatlands for peat extraction is also unavailable for the United States.
However, USGS records show that the number of active operations in the United States has been
declining since 1990; therefore, it seems reasonable to assume that no new areas are being cleared of
vegetation for managed peat extraction. Other changes in carbon stocks in living biomass on managed
peatlands are also assumed to be zero under the Tier 1 methodology (IPCC 2006 and 2013).
Table 6-60: Peat Production Area of Conterminous 48 States (Hectares)
1990a |
2005 I
2019
2020
2021
2022
2023
Nutrient-Rich
5,951 I
6,576
3,294
3,434
2,916
3,150
3,240
Nutrient-Poor
554
274 |
366
106
324
350
360
Total Production
6,920
6,850 |
3,660
3,540
3,240
3,500
3,600
aA portion of the production in 1990 is of unknown nutrient type, resulting in a total production value greater than the sum of
nutrient-rich and nutrient-poor.
Note: Totals may not sum due to independent rounding.
Table 6-61: Peat Production Area of Alaska (Hectares)
1990
2005
2019
2020
2021
2022
2023
Nutrient-Rich
0
0
0
0
0
0
0
Nutrient-Poor
286
104
329
428
419
419
419
Total Production
286
104
329
428
419
419
419
Note: Totals may not sum due to independent rounding.
On-site N20 Emissions
IPCC (2006) indicates the calculation of on-site N20 emission estimates using Tier 1 methodology only
considers nutrient-rich peatlands managed for peat extraction. These area data are not available
directly for the United States, but the on-site C02 emissions methodology above details the calculation
of nutrient-rich area data from production data. In order to estimate N20 emissions, the land area
estimate of nutrient-rich peatlands remaining peatlands was multiplied by the appropriate default
emission factor taken from IPCC (2013). See the Planned Improvements section for additional
information on identified research activities to improve peatland land area estimates.
53 The vacuum method is one type of extraction that annually "mills" or breaks up the surface of the peat into particles,
which then dry during the summer months. The air-dried peat particles are then collected by vacuum harvesters and
transported from the area to stockpiles (IPCC 2006).
Land Use, Land-Use Change, and Forestry 6-125
-------
On-site CH4 Emissions
IPCC (2013) also suggests basing the calculation of on-site CH4 emission estimates on the total area of
peatlands managed for peat extraction. Area data is derived using the calculation from production data
described in the On-site C02 Emissions section above. In order to estimate CH4 emissions from drained
land surface, the land area estimate of peatlands remaining peatlands was multiplied by the emission
factor for direct CH4 emissions taken from IPCC (2013). In order to estimate CH4 emissions from
drainage ditches, the total area of peatland was multiplied by the default fraction of peatland area that
contains drainage ditches, and the appropriate emission factor taken from IPCC (2013). See Table 6-62
for the calculated area of ditches and drained land.
Table 6-62: Peat Production (Hectares)
1990
2005
2019
2020
2021
2022
2023
Conterminous 48 States
Area of Drained Land
6,574
6,508
3,477
3,363
3,078
3,325
3,420
Area of Ditches
346
343
183
177
162
175
180
Total Production
6,920
6,850
3,660
3,540
3,240
3,500
3,600
Alaska
Area of Drained Land
272
99
312
407
398
398
398
Area of Ditches
14
5
16
21
21
21
21
Total Production
286
104
329
428
419
419
419
Note: Totals may not sum due to independent rounding.
Methodological recalculations were applied to the entire time series to ensure time-series consistency
from 1990 through 2023. The same data sources were used throughout the time series, when available.
When data were unavailable or the available data were outliers, missing values were estimated based
on the past available data.
Uncertainty
A Monte Carlo (Approach 2) uncertainty analysis was applied to estimate the uncertainty of C02, CH4,
and N20 emissions from peatlands remaining peatlands for 2023, using the following assumptions:
• The uncertainty associated with peat production data was estimated to be ± 25 percent
(Apodaca 2008) and assumed to be normally distributed.
• The uncertainty associated with peat production data stems from the fact that the USGS
receives data from smaller peat producers but estimates production from some larger peat
distributors. The peat type production percentages were assumed to have the same uncertainty
values and distribution as the peat production data (i.e., ± 25 percent with a normal
distribution).
• The uncertainty associated with the reported production data for Alaska was assumed to be the
same as for the conterminous United States, or ± 25 percent with a normal distribution. It
should be noted that the DGGS estimates that around half of producers do not respond to their
survey with peat production data; therefore, the production numbers reported are likely to
underestimate Alaska peat production (Szumigala 2008).
6-126 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
• The uncertainty associated with the average bulk density values was estimated to be ± 25
percent with a normal distribution (Apodaca 2008).
• IPCC (2006 and 2013) gives uncertainty values for the emissions factors for the area of peat
deposits managed for peat extraction based on the range of underlying data used to determine
the emission factors. The uncertainty associated with the emission factors was assumed to be
triangularly distributed.
• The uncertainty values surrounding the C fractions were based on IPCC (2006) and the
uncertainty was assumed to be uniformly distributed.
• The uncertainty values associated with the fraction of peatland covered by ditches was
assumed to be ± 100 percent with a normal distribution based on the assumption that greater
than 10 percent coverage, the upper uncertainty bound, is not typical of drained organic soils
outside of The Netherlands (IPCC 2013).
The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 6-63. Carbon
dioxide emissions from peatlands remaining peatlands in 2023 were estimated to be between 0.5 and
0.7 MMT C02 Eq. at the 95 percent confidence level. This indicates a range of 15 percent below to 16
percent above the 2023 emission estimate of 0.6 MMT C02 Eq. Methane emissions from peatlands
remaining peatlands in 2023 were estimated to be between 0.002 and 0.007 MMT C02 Eq. This indicates
a range of 58 percent below to 80 percent above the 2023 emission estimate of 0.004 MMT C02 Eq.
Nitrous oxide emissions from peatlands remaining peatlands in 2023 were estimated to be between
0.0002 and 0.0006 MMT C02 Eq. at the 95 percent confidence level. This indicates a range of 53 percent
below to 54 percent above the 2023 emission estimate of 0.0004 MMT C02 Eq.
Table 6-63: Approach 2 Quantitative Uncertainty Estimates for C02, CH4, and N20
Emissions from Peatlands Remaining Peatlands (MMT C02 Eq. and Percent)
Uncertainty Range Relative to Emission Estimate"
2023 Emission
(MMT CO2 Eq.)
(%)
Estimate
Lower
Upper
Lower
Upper
Source
Gas
(MMT CO2 Eq.)
Bound
Bound
Bound
Bound
Peatlands Remaining Peatlands
C02
0.6
0.5
0.7
-15%
+ 16%
Peatlands Remaining Peatlands
CH4
+
+
+
-58%
+80%
Peatlands Remaining Peatlands
N2O
+
+
+
-53%
+54%
+ Does not exceed 0.05 MMT C02 Eq.
a Range of emission estimates predicted by Monte Carlo stochastic simulation for a 95 percent confidence interval.
QA/QC and Verification
A QA/QC analysis was performed to review input data and calculations, and no issues were identified. In
addition, the emission trends were analyzed to ensure they reflected activity data trends.
Recalculations Discussion
The conterminous United States peat production estimates for peatlands remaining peatlands were
updated using the Peat section of the Mineral Commodity Summaries 2024 and the updated data tables
from the Minerals Yearbook: Peat (2021) Tables-only release. The Mineral Commodity Summaries 2024
Land Use, Land-Use Change, and Forestry 6-127
-------
edition updated 2022 peat production data and provided peat type production estimates for 2023. The
Minerals Yearbook: Peat (2021) Tables-only release provided updated rich and poor-nutrient densities.
Updated data decreased previously estimated emissions for 2021 by 0.2 percent (due to C02 and CH4
changes) and increased 2022 by 2.4 percent versus estimated emissions for 2021 and 2022 in the
previous (i.e., 1990 through 2022) Inventory for peatlands remaining peatlands.
Although Alaska peat production data for 2015 through 2023 were unavailable, 2014 data are available
in the Alaska's Mineral Industry 2014 report. However, the reported values represented an apparent 98
percent decrease in production since 2012. Due to the uncertainty of the most recent data, 2013
through 2022 value were assumed to be equal to the 2012 value, seen in the Alaska's Mineral Industry
2013 report. If updated Alaska data are available for the next Inventory cycle, this will result in a
recalculation in the next (i.e., 1990 through 2024) Inventory report.
Planned Improvements
Edits to the trends and methodology sections are planned based on expert review comments.
EPA notes the following improvements may be implemented or investigated within the next two or three
Inventory cycles pending time and resources:
• The implied emission factors will be calculated and included in this chapter for future
Inventories. Currently, the N20 emissions calculation uses different land areas than the C02 and
CH4 emission calculations (see Methodology and Time Series Consistency in this chapter), so
estimating the implied emission factor per total land area is not appropriate. Another method of
estimating implied emission factors will be developed. The inclusion of implied emission factors
in this chapter will provide another method of QA/QC and verification for Inventory data.
EPA notes the following improvements will continue to be investigated as time and resources allow, but
there are no immediate plans to implement until data are available or identified:
• In order to further improve estimates of C02, N20, and CH4 emissions from peatlands remaining
peatlands, future efforts will investigate if improved data sources exist for determining the
quantity of peat harvested per hectare and the total area of land undergoing peat extraction. EPA
regularly monitors common data sources, such as USGS publications, for new sources of
national peat data.
• In an effort to harmonize data sets used between different wetlands land use categories, EPA is
monitoring current research efforts to create, modify and expand spatial land representation
datasets. As these data become available, EPA will investigate the potential to apply a Tier 2
approach to managed peatlands. While peatlands are not a key category, the goal of updating
the approach is to encourage consistency across the land use categories. While Tier 2
distinctions like peat extraction technology would not be addressed by these activity data
improvements, concepts like land use histories (e.g., previous vegetation cover), may be more
apparent and lend itself to an update. Promising research/advancements include:
¦ Potential use of the National Wetlands Inventory (NWI) to establish managed peatlands land
representation. These data are used in other categories (e.g., Section 6.8 Flooded Lands).
6-128 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
¦ The National Aeronautics and Space Administration (NASA)-funded Wetlands Intrinsic
Potential tool (WIP) incorporates remote sensing and machine learning and may be a
resource pending scalability to the national-level to fill in known gaps.
• As part of these new or expanded research areas, EPA is particularly interested in, and
monitoring for, identifying drained or degraded peatlands and differentiating managed and
unmanaged peatlands.
• EPA plans to identify a new source for Alaska peat production. The current source has not been
reliably updated since 2012 and Alaska Department of Natural Resources indicated future
publication of data has been discontinued.
Coastal Wetlands Remaining Coastal Wetlands
Consistent with ecological definitions of wetlands,54 the United States has historically included under
the category of wetlands those coastal shallow water areas of estuaries and bays that lie within the
extent of the Land Representation. Guidance on quantifying greenhouse gas emissions and removals on
coastal wetlands is provided in the 2013 Supplement to the 2006IPCC Guidelines for National
Greenhouse Gas Inventories: Wetlands (Wetlands Supplement), which recognizes the particular
importance of vascular plants in sequestering C02 from the atmosphere within biomass, dead organic
material (DOM; including litter and dead wood stocks) and soils. Thus, the Wetlands Supplement
provides specific guidance on quantifying emissions and removals on organic and mineral soils that are
covered or saturated for part of the year by tidal fresh, brackish or saline water and are vegetated by
vascular plants and may extend seaward to the maximum depth of vascular plant vegetation. The United
States calculates emissions and removals based upon the stock change method for soil carbon (C) and
the gain-loss method for biomass and DOM. Presently, this Inventory does not calculate the lateralflux
of carbon to or from any land use. Lateral transfer of organic carbon to coastal wetlands and to marine
sediments within U.S. waters is the subject of ongoing scientific investigation; there is currently no IPCC
methodological guidance for lateral fluxes of carbon.
The United States recognizes both vegetated wetlands and unvegetated open water as coastal
wetlands. Per guidance provided by the Wetlands Supplement, sequestration of carbon into biomass,
DOM and soil carbon pools is recognized only in vegetated coastal wetlands and does not occur in
unvegetated open water coastal wetlands. The United States takes the additional step of recognizing
that carbon stock losses occur when vegetated coastal wetlands are converted to Unvegetated open
water coastal wetlands.
This Inventory includes all privately- and publicly-owned coastal wetlands (i.e., mangroves and tidal
marsh) along the oceanic shores of the conterminous United States, including the District of Columbia.,
but does not include coastal wetlands remaining coastal wetlands in Alaska, Hawaii, or any of the
United States Territories. Seagrasses are not currently included within the Inventory due to insufficient
data on distribution, change through time and carbon stocks or carbon stock changes as a result of
anthropogenic influence (see Planned Improvements).
Under the coastal wetlands remaining coastal wetlands category, the following emissions and removals
are quantified in this chapter:
54 See https://water.usgs.gov/nwsum/W8P9495/definitions.html: accessed August ?0?4.
Land Use, Land-Use Change, and Forestry 6-129
-------
• Carbon stock changes and CH4 emissions on vegetated coastal wetlands remaining vegetated
coastal wetlands,
• Carbon stock changes on vegetated coastal wetlands converted to unvegetated open water
coastal wetlands,
• Carbon stock changes on unvegetated open water coastal wetlands converted to vegetated
coastal wetlands, and
• Nitrous oxide emissions from aquaculture in coastal wetlands.
Vegetated coastal wetlands hold carbon in all five carbon pools (i.e., aboveground biomass,
belowground biomass, dead organic matter [DOM; dead wood and litter], and soil), though typically soil
carbon and, to a lesser extent, aboveground and belowground biomass are the dominant pools,
depending on wetland type (i.e., forested vs. marsh). Vegetated coastal wetlands are net accumulators
of carbon over centuries to millennia as soils accumulate carbon under anaerobic soil conditions and
carbon accumulates in plant biomass. Large emissions from soil carbon and biomass stocks occur
when vegetated coastal wetlands are converted to unvegetated open water coastal wetlands (e.g., when
vegetated coastal wetlands are lost due to subsidence, channel cutting through vegetated coastal
wetlands), but are still recognized as coastal wetlands in this Inventory. These carbon stock losses
resulting from conversion to unvegetated open water coastal wetlands can cause the release of
decades to centuries of accumulated soil carbon, as well as the standing stock of biomass carbon.
Conversion of unvegetated open water coastal wetlands to vegetated coastalwetlands, either through
restoration efforts or naturally, initiates the building of carbon stocks within soils and biomass. In
applying the Wetlands Supplement methodologies for estimating CH4 emissions, coastalwetlands in
salinity conditions greater than 18 parts per thousand have little to no CH4 emissions compared to those
experiencing lower salinity brackish and freshwater conditions. Therefore, conversion of vegetated
coastalwetlands to or from unvegetated open water coastalwetlands are conservatively assumed to
not result in a change in salinity condition and are assumed to have no impact on CH4 emissions. The
Wetlands Supplement provides methodologies to estimate N20 emissions from coastalwetlands that
occur due to aquaculture. The N20 emissions from aquaculture result from the nitrogen derived from
consumption of the applied food stock that is then excreted as nitrogen load available for conversion to
N20. While N20 emissions can also occur due to anthropogenic nitrogen loading from the watershed
and atmospheric deposition, these emissions are not reported here to avoid double-counting of indirect
N20 emissions with the agricultural soils management, forest land and settlements categories.
The Wetlands Supplement provides methodologies for estimating carbon stock changes and CH4
emissions from mangroves, tidal marshes and seagrasses. Depending upon their height and area,
carbon stock changes from mangroves may be reported under the forest land category or under coastal
wetlands. If mangrove stature is 5 m or greater or if there is evidence that trees can obtain that height,
mangroves are reported under the forest land category because they meet the definition of forest land.
Mangrove forests that are less than 5 m are reported under coastalwetlands because they meet the
definition of wetlands. All other non-drained, intact coastal marshes are reported under coastal
wetlands.
Because of human activities and level of regulatory oversight, all coastalwetlands within the
conterminous United States are included within the managed land area described in Section 6.1, and as
such, estimates of carbon stock changes, emissions of CH4, and emissions of N20 from aquaculture
from all coastalwetlands are included in this Inventory. At the present stage of inventory development,
6-130 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
coastal wetlands are not explicitly shown in the land representation analysis while work continues to
harmonize data from NOAA's Coastal Change Analysis Program (C-CAP)55 with NRI, FIAand NLDC data
used to compile the land representation (see Section 6.1). However, a check was undertaken to confirm
that coastal wetlands recognized by C-CAP represented a subset ofwetlands recognized by the NRI for
marine coastal states.
The greenhouse gas fluxes for all four wetland categories described above are summarized in Table
6-64. Coastal wetlands remaining coastal wetlands are generally a net carbon sink, with the fluxes
rangingfrom -5.6to -6.7 MMT C02 Eq. across the majority of the time series; however, between 2006 and
2010, they were a net source of emissions (rangingfrom 3.2 to 53.5 MMT C02 Eq.), resulting from a large
loss of vegetated coastal wetlands to open water due to hurricanes (Table 6-64). Recognizing removals
of C02to soil of 12.5 MMT C02 Eq. and CH4 emissions of 4.3 MMT C02 Eq. in 2023, vegetated coastal
wetlands remaining vegetated coastal wetlands are a net sink of 8.2 MMT C02 Eq. Loss of coastal
wetlands, primarily in the Mississippi Delta as a result of hurricane impacts and sediment diversion and
other human impacts, recognized as vegetated coastal wetlands converted to unvegetated coastal
wetlands, drive an emission of 1.5 MMT C02 Eq. since 2011, primarily from soils. Building of new
wetlands from open water, recognized as unvegetated coastal wetlands converted to vegetated coastal,
results each year in removal of 0.1 MMT C02 Eq. Aquaculture is a minor industry in the United States,
resulting in an emission of N20 across the time series of between 0.1 to 0.2 MMT C02 Eq. In total,
coastal wetlands are a net sink of 6.7 MMT C02 Eq. in 2023.
Table 6-64: Emissions and Removals from Coastal Wetlands Remaining Coastal
Wetlands (MMT C02 Eq.)
Land Use/Carbon Pool
1990
2005
2019
2020
2021
2022
2023
Vegetated Coastal Wetlands Remaining
Vegetated Coastal Wetlands
(8.4)
(8.4)
(8.3)
(8.3)
(8.2)
(8.2)
(8.2)
Biomass C Flux
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Soil C Flux
(12.5)
(12.6)
(12.5)
(12.5)
(12.5)
(12.5)
(12.5)
Net CH4 Flux
4.2
4.2
4.3
4.3
4.3
4.3
4.3
Vegetated Coastal Wetlands Converted to
Unvegetated Open Water Coastal Wetlands
1.8
2.6
1.5
1.5
1.5
1.5
1.5
Biomass C Flux
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Dead Organic Matter C Flux
+ I
+
+
+
+
+
+
Soil C Flux
1.7
2.5
1.5
1.5
1.5
1.5
1.5
Unvegetated Open Water Coastal Wetlands
Converted to Vegetated Coastal Wetlands
(+)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
Biomass C Flux
(+)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
Dead Organic Matter C Flux
(+)
<+)
+
+
+
+
+
Soil C Flux
(+)
{+)
{+)
{+)
{+)
{+)
{+)
Net N2O Flux from Aquaculture in Coastal
Wetlands
0.1
0.2
0.1
0.1
0.1
0.1
0.1
Total Biomass C Flux
+
+
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
Total Dead Organic Matter C Flux
(+>
(+)
+
+
+
+
+
Total Soil C Flux
(10.8)
(10.1)
(11.0)
(11.0)
(11.0)
(11.1)
(11.1)
Total CH4 Flux
4.2 |
4.2
4.3
4.3
4.3
4.3
4.3
55 See https://coast.noaa.gov/digitalcoast/tools/lca.html: accessed September ?0?4.
Land Use, Land-Use Change, and Forestry 6-131
-------
Total N2O Flux
0.1 |
0.2 |
0.1
0.1
0.1
0.1
0.1
Total Flux
(6.5) |
(5.7) |
(6.7)
(6.7)
(6.7)
(6.7)
(6.7)
+ Absolute value does not exceed 0.05 MMT C02 Eq.
Note: Totals may not sum due to independent rounding. Parentheses indicate net sequestration.
Emissions and Removals from Vegetated Coastal
Wetlands Remaining Vegetated Coastal Wetlands
The conterminous United States currently has 2.98 million hectares of intertidal vegetated coastal
wetlands remaining vegetated coastal wetlands comprised of tidally influenced palustrine emergent
marsh (664,294 ha), palustrine scrub shrub (133,798 ha), estuarine emergent marsh (1,891,738 ha),
estuarine scrub shrub (95,782 ha) and estuarine forested wetlands (195,779 ha). Mangroves fall under
both estuarine forest and estuarine scrub shrub categories depending upon height. Dwarf mangroves,
found in subtropical Gulf Coast states, do not attain the height status to be recognized as forest land,
and are therefore always classified within vegetated coastal wetlands. Vegetated coastal wetlands
remaining vegetated coastal wetlands are found in cold temperate (53,965 ha), warm temperate
(896,878 ha), subtropical (1,966,960 ha) and Mediterranean (62,988 ha) climate zones.
Soils are the largest carbon pool in vegetated coastal wetlands remaining vegetated coastal wetlands,
reflecting long-term removal of atmospheric C02 by vegetation and transfer into the soil pool in the form
of both autochthonous and allochthonous decaying organic matter. Soil carbon emissions are not
assumed to occur in coastal wetlands that remain vegetated. This Inventory includes changes in carbon
stocks in both biomass and soils. Changes in DOM carbon stocks are not included. Methane emissions
from decomposition of organic matter in anaerobic conditions are present at salinity less than half that
of sea water. Mineral and organic soils are not differentiated in terms of carbon stock changes or CH4
emissions.
Table 6-65 through Table 6-67 summarize nationally aggregated biomass and soil carbon stock changes
and CH4emissions on vegetated coastal wetlands remaining vegetated coastal wetlands. Intact
vegetated coastal wetlands remaining vegetated coastal wetlands hold a total biomass carbon stock of
35.98 MMT C. Removals from biomass carbon stocks in 2023 were 0.05 MMT C02 Eq. (0.01 MMT C),
which has increased over the time series (Table 6-65 and Table 6-66). Carbon dioxide emissions from
biomass in vegetated coastal wetlands remaining vegetated coastal wetlands between 2002 and 2011,
with very low sequestration between 2002 and 2006 and emissions of 0.21 MMT C02 Eq. between 2007
and 2011, are not inherently typical and are a result of coastal wetland loss overtime. Most of the
coastal wetland loss has occurred in palustrine and estuarine emergent wetlands. Vegetated coastal
wetlands maintain a large carbon stock within the top 1 meter of soil (estimated to be 804 MMT C) to
which carbon accumulated at a rate of 12.6 MMT C02 Eq. (3.4 MMT C) in 2023, a value that has
remained relatively constant across the time-series. For vegetated coastal wetlands remaining
vegetated coastal wetlands, methane emissions of 4.3 of MMT C02 Eq. (154 kt CH4) in 2023 (Table 6-67)
offset carbon removals resulting in a net removal of 8.2 MMT C02 Eq. in 2023; this rate has been
relatively consistent across the time-series. Dead organic matter stock changes are not calculated in
vegetated coastal wetlands remaining vegetated coastal wetlands since this stock is considered to be in
a steady state when using Tier 1 methods (IPCC 2014). Due to federal regulatory protection, loss of
vegetated coastal wetlands through human activities slowed considerably in the 1970s and the current
annual rates of carbon stock change and CH4 emissions are relatively constant over time.
6-132 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Table 6-65: Net C02 Flux from Carbon Stock Changes in Vegetated Coastal Wetlands
Remaining Vegetated Coastal Wetlands (MMT C02 Eq.)
Year
1990 |
2005
2019
2020
2021
2022
2023
Biomass Flux
<+)
(+)
(+)
(+)
(+)
(+)
(+)
Soil Flux
(12.5)
(12.6)
(12.5)
(12.5)
(12.5)
(12.5)
(12.5)
Total C Stock Change
(12.6) |
(12.6)
(12.5)
(12.5)
(12.5)
(12.5)
(12.6)
+ Absolute value does not exceed 0.05 MMT C02 Eq.
Note: Parentheses indicate net sequestration. Totals may not sum due to independent rounding.
Table 6-66: Net C02 Flux from Carbon Stock Changes in Vegetated Coastal Wetlands
Remaining Vegetated Coastal Wetlands (MMT C)
Year
1990 | 2005 |
2019
2020
2021
2022
2023
Biomass Flux
(+) I
+ I
(+)
(+)
(+)
(+)
(+)
Soil Flux
(3.4)
(3.4)
(3.4)
(3.4)
(3.4)
(3.4)
(3.4)
Total C Stock Change
(3.4) |
(3.4)
(3.4)
(3.4)
(3.4)
(3.4)
(3.4)
+ Absolute value does not exceed 0.05 MMT C.
Note: Parentheses indicate net sequestration. Totals may not sum due to independent rounding.
Table 6-67: CH4 Emissions from Vegetated Coastal Wetlands Remaining Vegetated
Coastal Wetlands (MMT C02 Eq. and kt CH4)
Year
1990
2005
2019
2020
2021
2022
2023
Methane Emissions (MMT CO2 Eq.)
4.2
4.2
4.3
4.3
4.3
4.3
4.3
Methane Emissions (kt CH4)
149
151
153
154
154
154
155
Methodology and Time-Series Consistency
The following section includes a description of the methodology used to estimate changes in biomass
carbon stocks, soil carbon stocks and emissions of CH4 for vegetated coastal wetlands remaining
vegetated coastal wetlands. Dead organic matter is not calculated for vegetated coastal wetlands
remaining vegetated coastal wetlands since it is assumed to be in steady state (IPCC 2014).
Methodological recalculations were applied to the entire time series to ensure time-series consistency
from 1990 through 2023.
Biomass Carbon Stock Changes
Above- and belowground biomass carbon stocks for palustrine (freshwater) and estuarine (saline)
marshes are estimated for vegetated coastal wetlands remaining vegetated coastal wetlands on land
below the elevation of high tides (taken to be mean high water spring tide elevation) and as far seawards
as the extent of intertidal vascular plants according to the national LiDAR dataset, the national network
of tide gauges and land use histories recorded in the 1996, 2001, 2006, 2010, and 2016 NOAAC-CAP
surveys (NOAA OCM 2020). C-CAP areas are calculated at the state/territory level and summed
according to climate zone to national values. Federal and non-federal lands are represented. Trends in
land cover change are extrapolated to 1990 and 2023 from these datasets. Based upon NOAA C-CAP,
coastal wetlands are subdivided into palustrine and estuarine classes and further subdivided into
emergent marsh, scrub shrub and forest classes (Table 6-68). Biomass is not sensitive to soil organic
Land Use, Land-Use Change, and Forestry 6-133
-------
matter content but is differentiated based on climate zone. Aboveground biomass carbon stocks for
non-forested wetlands data are derived from a national assessment combining field plot data and
aboveground biomass mapping by remote sensing (Byrd et al. 2017; Byrd et al. 2018; Byrd et al. 2020).
The aboveground biomass carbon stock for subtropical estuarine forested wetlands (dwarf mangroves
that are not classified as forests due to their stature) is derived from a meta-analysis by Lu and
Megonigal (2017). Root to shoot ratios from the Wetlands Supplement (Table 6-70; IPCC2014) were
used to account for belowground biomass, which were multiplied by the aboveground carbon stock.
Above- and belowground values were summed to obtain total biomass carbon stocks. Biomass carbon
stock changes per year for wetlands remaining wetlands were determined by calculating the difference
in area between that year and the previous year to calculate gain/loss of area for each climate type,
which was multiplied by the mean biomass for that climate type.
Table 6-68: Area of Vegetated Coastal Wetlands Remaining Vegetated Coastal
Wetlands, Vegetated Coastal Wetlands Converted to Unvegetated Open Water
Coastal Wetlands, and Unvegetated Open Water Coastal Wetlands Converted to
Vegetated Coastal Wetlands (hectares)
Year
1990
2005
2019
2020
2021
2022
2023
Vegetated Coastal
Wetlands Remaining
Vegetated Coastal
Wetlands
2,975,477
2,985,783
2,975,789
2,977,055
2,978,322
2,979,588
2,981,231
Vegetated Coastal
Wetlands Converted to
Unvegetated Open Water
Coastal Wetlands
1,720
2,515
1,488
1,488
1,488
1,488
1,488
Unvegetated Open Water
Coastal Wetlands
Converted to Vegetated
Coastal Wetlands
952
1,769
2,406
2,406
2,406
2,406
2,406
Table 6-69: Aboveground Biomass Carbon Stocks for Vegetated Coastal Wetlands (t C
ha1)
Wetland Type
Climate Zone
Cold Temperate
Warm Temperate
Subtropical
Mediterranean
Palustrine Scrub/Shrub Wetland
3.25
3.17
2.24
4.69
Palustrine Emergent Wetland
3.25
3.17
2.24
4.69
Estuarine Forested Wetland
N/A
N/A
17.83
N/A
Estuarine Scrub/Shrub Wetland
3.05
3.05
2.43
3.44
Estuarine Emergent Wetland
3.05
3.10
2.43
3.44
Source: All data from Byrd et al. (2017, 2018 and 2020) except for subtropical estuarine forested wetlands, which is from Lu and
Megonigal (2017); N/A means there are currently no estuarine forested wetlands that are less than 5 meters tall; these forested
wetlands meet the definition of forest land and are included in the Forest Land section.
Table 6-70: Root to Shoot Ratios for Vegetated Coastal Wetlands
Climate Zone
Wetland Type
Cold Temperate
Warm Temperate Subtropical
Mediterranean
Palustrine Scrub/Shrub Wetland
1.15
1.15 3.65
3.63
6-134 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Wetland Type
Climate Zone
Cold Temperate
Warm Temperate
Subtropical
Mediterranean
Palustrine Emergent Wetland
1.15
1.15
3.65
3.63
Estuarine Forested Wetland
N/A
N/A
0.96
N/A
Estuarine Scrub/Shrub Wetland
2.11
2.11
3.65
3.63
Estuarine Emergent Wetland
2.11
2.11
3.65
3.63
Source: All values from IPCC (2014); N/A means there are currently no estuarine forested wetlands that are less than 5 meters
tall; these forested wetlands meet the definition of forest land and are included in the Forest Land section.
Soil Carbon Stock Changes
Soil carbon stock changes are estimated for vegetated coastal wetlands remaining vegetated coastal
wetlands for both mineral and organic soils. Soil carbon stock changes, stratified by climate zones and
wetland classes, are derived from a synthesis of peer-reviewed literature (Table 6-71; Lynch 1989; Orson
et al. 1990; Kearny & Stevenson 1991; Thom et al. 1992; Roman et al. 1997; Craft et al. 1998; Orson et al.
1998; Merrill 1999; Weis et al. 2001; Hussein et al. 2004; Church et al. 2006; Koster et al. 2007; Drexler
et al. 2009; Boyd 2012; Callaway et al. 2012a&b; Bianchi et al. 2013; Drexler et al. 2013; Watson and
Byrne 2013; Breithaupt et al. 2014; Crooks et al. 2014; Weston et al. 2014; Smith et al. 2015; Villa &
Mitsch 2015; Boyd and Sommerfield 2016; Marchio et al. 2016; Noe et al. 2016; Arriola and Cable 2017;
Boyd et al. 2017; Gerlach et al. 2017; Giblin and Forbrich 2018; Krauss et al. 2018; Abbott et al. 2019;
Drexler et al. 2019; Poppe and Rybczyk 2019; Ensign et al. 2020; Kemp et al. 2020; Lagomasino et al.
2020; Luk et al. 2020; McTigue et al. 2020; Peck et al. 2020; Vaughn et al. 2020; Weston et al. 2020;
Arias-Ortiz et al. 2021; Baustian et al. 2021; Allen et al. 2022; Miller et al. 2022).
Tier 2 estimates of soil carbon removals associated with annual soil carbon accumulation on managed
vegetated coastal wetlands remaining vegetated coastal wetlands were developed with country-specific
soil carbon removal factors multiplied by activity data of land area for vegetated coastal wetlands
remaining vegetated coastal wetlands. The methodology follows Eq. 4.7, Chapter4 of the Wetlands
Supplement, and is applied to the area of vegetated coastal wetlands remaining vegetated coastal
wetlands on an annual basis. To estimate soil carbon stock changes, no differentiation is made between
organic and mineral soils since currently, no statistical evidence supports disaggregation (Holmquist et
al. 2018).
Table 6-71: Annual Soil Carbon Accumulation Rates for Vegetated Coastal Wetlands (t
C ha1 yr1)
Climate Zone Cold Temperate Warm Temperate Subtropical Mediterranean
Palustrine Scrub/Shrub Wetland 1.010 1.544 0.450 0.845
Palustrine Emergent Wetland 1.010 1.544 0.454 0.845
Estuarine Forested Wetland N/A N/A 0.821 N/A
Estuarine Scrub/Shrub Wetland 1.254 1.039 0.821 0.845
Estuarine Emergent Wetland 1.254 1.039 1.587 0.845
Source: All data from CCRCN (2023)56; N/A means there are no estuarine forested wetlands outside of subtropical regions.
56 Coastal Carbon Network (2023). Database: Coastal Carbon Library (Version 1.0.0). Smithsonian Environmental
Research Center. Dataset. https://doi.org/10.95573/serc.91565671. Accessed September 2024.
Land Use, Land-Use Change, and Forestry 6-135
-------
Soil Methane Emissions
Tier 1 estimates of CH4 emissions for vegetated coastal wetlands remaining vegetated coastal wetlands
are derived from the same wetland map used in the analysis of wetland soil C fluxes, produced from C-
CAP, LiDAR and tidal data, in combination with default CH4 emission factors provided in Table 4.14 of
the Wetlands Supplement. The methodology follows Equation 4.9, Chapter 4 of the Wetlands
Supplement; Tier 1 emissions factors are multiplied by the area of freshwater (palustrine) coastal
wetlands. The CH4 fluxes applied are determined based on salinity; only palustrine wetlands are
assumed to emit CH4. Estuarine coastal wetlands in the C-CAP classification include wetlands with
salinity less than 18 ppt, a threshold at which methanogenesis begins to occur (Poffenbarger et a I.
2011), but the dataset currently does not differentiate estuarine wetlands based on their salinities and,
as a result, CH4 emissions from estuarine wetlands are not included at this time.
Uncertainty
Underlying uncertainties in the estimates of soil and biomass carbon stock changes and CH4 emissions
include uncertainties associated with Tier 2 literature values of soil carbon stocks, biomass carbon
stocks and CH4 flux, assumptions that underlie the methodological approaches applied and
uncertainties linked to interpretation of remote sensing data. Uncertainty specific to vegetated coastal
wetlands remaining vegetated coastal wetlands include differentiation of palustrine and estuarine
community classes, which determines the soil carbon stock and CH4 flux applied. Uncertainties for soil
and biomass carbon stock data for all subcategories are not available and thus assumptions were
applied using expert judgment about the most appropriate assignment of a carbon stock to a
disaggregation of a community class. Because mean soil and biomass carbon stocks for each available
community class are in a fairly narrow range, the same overall uncertainty was assigned to each,
respectively (i.e., applying approach for asymmetrical errors, the largest uncertainty for any soil carbon
stock value should be applied in the calculation of error propagation; IPCC 2000). Uncertainty for root to
shoot ratios, which are used for quantifying belowground biomass, are derived from the 2013 Wetlands
Supplement. Uncertainties for CH4 flux are the Tier 1 default values reported in the 2013 IPCC Wetlands
Supplement. Overall uncertainty of the NOAA C-CAP remote sensing product is 15 percent. This is in the
range of remote sensing methods (±10 to 15 percent; IPCC 2003). However, there is significant
uncertainty in salinity ranges for tidal and non-tidal estuarine wetlands and activity data used to apply
CH4 flux emission factors (delineation of an 18 ppt boundary) that will need significant improvement to
reduce uncertainties. Details on the emission/removal trends and methodologies through time are
described in more detail in the introduction and the Methodology section. The combined uncertainty
was calculated using the IPCC Approach 1 method of summing the squared uncertainty for each
individual source (C-CAP, soil, biomass and CH4) and taking the square root of that total.
Uncertainty estimates are presented in Table 6-72 for each subcategory (i.e., soil carbon, biomass
carbon and CH4 emissions). The combined uncertainty across all subcategory is 37.0 percent below and
above the estimate of-6.4 MMT C02 Eq, which is primarily driven by the uncertainty in the CH4 estimates
because there is high variability in CH4 emissions when the salinity is less than 18 ppt. In 2023, the total
flux was -8.2 MMT C02 Eq., with lower and upper estimates of -11.3 and -5.2 MMT C02 Eq.
6-136 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Table 6-72: IPCC Approach 1 Quantitative Uncertainty Estimates for Carbon Stock
Changes and CH4 Emissions occurring within Vegetated Coastal Wetlands Remaining
Vegetated Coastal Wetlands in 2023 (MMT C02 Eq. and Percent)
Uncertainty Range Relative to Estimate
(MMT CO2 Eq.) (%)
Source/Sink
Gas
2023 Estimate
(MMT CO2 Eq.)
Lower
Bound
Upper
Bound
Lower
Bound
Upper
Bound
Biomass C Stock Change
CO2
(0.05)
(0.06)
(0.03)
-24.1%
+24.1%
Soil C Stock Change
CO2
(12.5)
(14.7)
(10.3)
-17.7
+ 17.7%
CH4 emissions
cm
4.3
3.0
5.6
-29.9%
+29.9%
Total Flux
(8.2)
(11.3)
(5.2)
-36.5%
+36.5%
Notes: Parentheses indicate net sequestration. Totals may not sum due to independent rounding.
QA/QC and Verification
NOAA provided the National LiDAR Dataset, tide data, and C-CAP land cover and land cover change
mapping, all of which are subject to agency internal QA/QC assessment consistent with the general QC
checks outlined in the Inventory QA/QC Plan. Acceptance of final datasets into archive and
dissemination are contingent upon the product compilation being compliant with mandatory QA/QC
requirements (McCombs et al. 2016). QA/QC and verification of soil carbon stock datasets have been
provided by the Smithsonian Environmental Research Center and coastal wetland inventory team leads
who reviewed summary tables against reviewed sources. Biomass carbon stocks are derived from peer-
review literature and reviewed by the U.S. Geological Survey prior to publishing, by the peer-review
process during publishing, and by the coastal wetland inventory team leads before inclusion in this
Inventory. A team of two evaluated and verified there were no computational errors within the
calculation worksheets. Soil and biomass carbon stock change data are based upon peer-reviewed
literature and CH4 emission factors derived from the Wetlands Supplement.
Recalculations Discussion
No recalculations were performed for the current Inventory.
Planned Improvements
Multiple planned improvements are currently being investigated for vegetated coastal wetlands,
including:
• Harmonizing all spatial datasets used to calculate activity data is underway. Once completed, a
better representation of forested tidal wetlands, palustrine tidal wetlands, and forest land near
the tidal boundary will be obtained.
• Examining the feasibility of incorporating seagrass soil and biomass carbon stocks into the
vegetated coastal wetlands remaining vegetated coastal wetlands estimates. Seagrass
incorporation is being done on a state-by-state basis dependent upon data availability and
emissions factor data (see Box 6-5).
Land Use, Land-Use Change, and Forestry 6-137
-------
• Investigating and then quantifying the distribution, area, and emissions resulting from
impounded waters (i.e., coastal wetlands where tidal connection to the ocean has been
restricted or eliminated completely).
Box 6-5: State-Level Case Studies for the Estimation of GHG Removals in Seagrasses
North Carolina and Maryland are the first states to include seagrasses within their state-level inventory.
North Carolina has the largest extent of seagrass coverage along the U.S. Atlantic coast, measuring
approximately 86,412 acres in 2021. Seagrass mapping efforts occurred in 2007, 2013, and 2020 using a
field-validated aerial image classification. The Tier 1 soil carbon accumulation rate was used and
currently, biomass is not included due to lack of local data. The analysis shows that these high salinity
seagrass habitats provided a net carbon sink to the state, although greenhouse gas removals decreased
over time due to loss in seagrass coverage. Overall, seagrass beds in 2021 sequestered approximately
0.055 MMTC02 Eq. (55.14 kt C02 Eq.) in the soils alone.
In Maryland, the state greenhouse gas inventory comprises blue carbon stocks and fluxes from
estuarine wetlands and seagrasses. Maryland currently has long-term monitoring of submerged aquatic
vegetation (SAV) extent and density through annual surveying, and the rate of carbon sequestration and
methane emission was a regional average for coastal wetlands. This study at state-level calculation
offers an opportunity to maintain consistency in quantification efforts across spatial scales and allows
positioning SAV in its role as a carbon sink, in addition to its benefits in water quality and habitat
conservation.
These two case studies demonstrate the importance of refining emission factor data and harmonizing
the inclusion of this ecosystem in the land representation analysis (reconciling the National Ocean and
Atmospheric Administration [NOAA] Coastal Change Analysis Program [C-CAP] data with the National
Resource Inventory, Forest Inventory Analysis, and the National Land Cover Database).
Emissions from Vegetated Coastal Wetlands Converted to
Unvegetated Open Water Coastal Wetlands
Vegetated coastal wetlands converted to unvegetated open water coastal wetlands is a source of
emissions from soil, biomass, and DOM carbon stocks. An estimated 1,488 ha of vegetated coastal
wetlands were converted to unvegetated open water coastal wetlands in 2023, which largely occurred
within estuarine and palustrine emergent wetlands. Prior to 2006, annual conversion to unvegetated
open water coastal wetlands was higher than current rates: 1,720 between 1990 and 2000 and 2,515 ha
between 2001 and 2005. The Mississippi Delta represents more than 40 percent of the total coastal
wetland of the United States, and over 90 percent of the area of vegetated coastal wetlands converted to
unvegetated open water coastal wetlands. The drivers of coastal wetlands loss include legacy human
impacts on sediment supply through rerouting river flow, direct impacts of channel cutting on hydrology,
salinity and sediment delivery, and accelerated subsidence from aquifer extraction. Each of these
drivers directly contributes to wetland erosion and subsidence, while also reducing the resilience of the
wetland to build with sea-level rise or recover from hurricane disturbance. Over recent decades, the rate
of Mississippi Delta wetland loss has slowed, though episodic mobilization of sediment occurs during
hurricane events (Couvillion et al. 2011; Couvillion et al. 2016). The land cover analysis between the
6-138 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
2006 and 2011 C-CAP surveys coincides with two such events, hurricanes Katrina and Rita (both making
landfall in the late summer of 2005), that occurred between these C-CAP survey dates. The subsequent
2016 C-CAP survey determined that erosion rates had slowed.
Shallow nearshore open water within the U.S. land representation is recognized as falling under the
coastal wetlands category within this Inventory. While high resolution mapping of coastal wetlands
provides data to support IPCC Approach 2 methods for tracking land cover change, the depth in the soil
profile to which sediment is lost is less clear. This Inventory adopts the Tier 1 methodological guidance
from the Wetlands Supplement for estimating emissions following the methodology for excavation (see
Methodology section, below) when vegetated coastal wetlands are converted to unvegetated open
water coastal wetlands, assuming a 1 m depth of disturbed soil. This 1 m depth of disturbance is
consistent with estimates of wetland carbon loss provided in the literature and the Wetlands
Supplement (Crooks et al. 2009; Couvillion et al. 2011; Delaune and White 2012; IPCC 2014). The same
assumption on depth of soils impacted by erosion has been applied here. It is a reasonable Tier 1
assumption, based on experience, but estimates of emissions are sensitive to the depth to which the
assumed disturbances have occurred (Holmquist et al. 2018). ATier 1 assumption is also adopted in
that all mobilized carbon is immediately returned to the atmosphere (as assumed for terrestrial land-
use categories), rather than redeposited in long-term carbon storage. The science is currently under
evaluation to adopt more refined emissions factors for mobilized coastal wetland carbon based upon
the geomorphic setting of the depositional environment.
In 2023, there were 1,488 ha of vegetated coastal wetlands converted to unvegetated open water
coastal wetlands (Table 6-68) across all wetland types and climates, which resulted in 1.5 MMT C02 Eq.
(0.4 MMT C) and 0.06 MMT C02 Eq. (0.02 MMT C) lost through soil and biomass, respectively, with
minimal DOM C stock loss (Table 6-73, and Table 6-74). Across the time-series, the area of vegetated
coastal wetlands converted to unvegetated open water coastal wetlands was greatest between the 2006
to 2011 C-CAP time-series (11,373 ha) and has decreased since then to current levels (Table 6-68).
Table 6-73: Net C02 Flux from Carbon Stock Changes in Vegetated Coastal Wetlands
Converted to Unvegetated Open Water Coastal Wetlands (MMT C02 Eq.)
Year
1990
2005
2019
2020
2021
2022
2023
Biomass Flux
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Dead Organic Matter Flux
+
+
+
+
+
+
+
Soil Flux
1.7
2.5
1.5
1.5
1.5
1.5
1.5
Total C Stock Change
1.8
2.6
1.5
1.5
1.5
1.5
1.5
+ Absolute value does not exceed 0.05 MMT C02 Eq.
Note: Totals may not sum due to independent rounding.
Table 6-74: Net C02 Flux from Carbon Stock Changes in Vegetated Coastal Wetlands
Converted to Unvegetated Open Water Coastal Wetlands (MMT C)
Year
1990
2005
2019
2020
2021
2022
2023
Biomass Flux
+
+
+
+
+
+
+
Dead Organic Matter Flux
+
+
+
+
+
+
+
Soil Flux
0.5
0.7
0.4
0.4
0.4
0.4
0.4
Total C Stock Change
0.5
0.7
0.4
0.4
0.4
0.4
0.4
+ Absolute value does not exceed 0.05 MMT C.
Note: Totals may not sum due to independent rounding.
Land Use, Land-Use Change, and Forestry 6-139
-------
Methodology and Time-Series Consistency
The following section includes a brief description of the methodology used to estimate changes in soil,
biomass and DOM carbon stocks for vegetated coastal wetlands converted to unvegetated open water
coastal wetlands.
Methodological recalculations were applied to the entire time series to ensure time-series consistency
from 1990 through 2023.
Biomass Carbon Stock Changes
Biomass carbon stock changes for palustrine and estuarine marshes are estimated for vegetated
coastal wetlands converted to unvegetated open water coastal wetlands on lands below the elevation
of high tides (taken to be mean high water spring tide elevation) within the U.S. land representation
according to the national LiDAR dataset, the national network of tide gauges and land use histories
recorded in the 1996, 2001, 2006, 2010, and 2016 NOAA C-CAP surveys. C-CAP areas are calculated at
the state/territory level and summed according to climate zone to national values. Publicly-owned and
privately-owned lands are represented. Trends in land cover change are extrapolated to 1990 and 2023
from these datasets. The C-CAP database provides peer reviewed country-specific mapping to support
IPCC Approach 3 quantification of coastal wetland distribution, including conversion to and from open
water. Biomass carbon stocks are not sensitive to soil organic content but are differentiated based on
climate zone. Non-forested aboveground biomass carbon stock data are derived from a national
assessment combining field plot data and aboveground biomass mapping by remote sensing (Byrd et al.
2017; Byrd et al. 2018; Byrd et al. 2020). The aboveground biomass carbon stock for estuarine forested
wetlands (dwarf mangroves that are not classified as forests due to their stature) is derived from a meta-
analysis by Lu and Megonigal (2017;57 Table 6-69). Aboveground biomass carbon stock data for all
subcategories are not available and thus assumptions were applied using expert judgment about the
most appropriate assignment of a carbon stock to a disaggregation of a community class. Root to shoot
ratios from the Wetlands Supplement were used to account for belowground biomass, which were
multiplied by the aboveground carbon stock (Table 6-70; IPCC 2014). Above- and belowground values
were summed to obtain total biomass carbon stocks. Conversion to open water results in emissions of
all biomass carbon stocks during the year of conversion; therefore, emissions are calculated by
multiplying the C-CAP derived area of vegetated coastal wetlands lost that year in each climate zone by
its mean biomass.
Dead Organic Matter
Dead organic matter (DOM) carbon stocks, which include litter and dead wood stocks for subtropical
estuarine forested wetlands, are an emission from vegetated coastal wetlands converted to
unvegetated open water coastal wetlands across all years in the time series. Data on DOM carbon
stocks are not currently available for either palustrine or estuarine scrub/shrub wetlands for any climate
zone. Data for estuarine forested wetlands in other climate zones are not included since there is no
estimated loss of these forests to unvegetated open water coastal wetlands across any year based on C-
CAP data. For subtropical estuarine forested wetlands, Tier 1 estimates of mangrove DOM were used
(IPCC 2014). Trends in land cover change are derived from the NOAA C-CAP dataset and extrapolated to
cover the entire 1990 through 2023 time series. Conversion to open water results in emissions of all
57 See https://github.com/8mithsonian/Coastal-Wetland-NGGI-Data-Public: accessed September ?0?3.
6-140 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
DOM carbon stocks during the year of conversion; therefore, emissions are calculated by multiplying the
C-CAP derived area of vegetated coastal wetlands lost that year by its Tier 1 DOM carbon stock.
Soil Carbon Stock Changes
Soil carbon stock changes are estimated for vegetated coastal wetlands converted to unvegetated open
water coastal wetlands. Soil carbon stocks for all coastal states were updated in 2018 based upon
analysis of an assembled dataset of 1,959 cores from across the conterminous United States
(Holmquist et al. 2018). Based on carbon stock samples from all coastal climate zones in the
coterminous U.S., using a constant 2701C ha1 (the mean stock of the cores) resulted in higher
performance in predicting carbon stocks than models that accounted for soil type, climate zone,
salinity, and vegetation type. This analysis demonstrated that it was not justified to stratify carbon stocks
based upon mineral or organic soil classification, climate zone, or wetland classes; therefore, a single
soil carbon stock of 2701C ha1 was applied to all wetland classes as the stock present before LULC
change. Following the Tier 1 approach for estimating C02 emissions with extraction provided within the
Wetlands Supplement, soil carbon loss with conversion of vegetated coastal wetlands to unvegetated
open water coastal wetlands is assumed to affect soil carbon stock to one-meter depth (Holmquist et
al. 2018) with all emissions occurring in the year of wetland conversion, and multiplied by activity data
of vegetated coastal wetland area converted to unvegetated open water wetlands. The methodology
follows Eq. 4.6 in the Wetlands Supplement.
Soil Methane Emissions
ATier 1 assumption has been applied that salinity conditions are unchanged and hence CH4 emissions
are assumed to be zero with conversion of vegetated coastal wetlands to unvegetated open water
coastal wetlands.
Uncertainty
Underlying uncertainties in estimates of soil and biomass carbon stock changes are associated with
country-specific (Tier 2) literature values of these stocks, while the uncertainties with the Tier 1
estimates are associated with subtropical estuarine forested wetland DOM stocks. Assumptions that
underlie the methodological approaches applied and uncertainties linked to interpretation of remote
sensing data are also included in this uncertainty assessment. The IPCC default assumption of 1 m of
soil erosion with anthropogenic activities was adopted to provide standardization in U.S. tidal carbon
accounting (Holmquist et al. 2018). This depth of potentially erodible tidalwetland soil has not been
comprehensively addressed since most soil cores analyzed were shallow (e.g., less than 50 cm) and do
not necessarily reflect the depth to non-wetland soil or bedrock (Holmquist et al. 2018). Uncertainty
specific to coastal wetlands include differentiation of palustrine and estuarine community classes,
which determines the soil carbon stock applied. Because mean soil and biomass carbon stocks for
each available community class are in a fairly narrow range, the same overall uncertainty was assigned
to each (i.e., applying approach for asymmetrical errors, the largest uncertainty for any soil carbon stock
value should be applied in the calculation of error propagation; IPCC 2000). For aboveground biomass
carbon stocks, the mean standard error was very low and largely influenced by the uncertainty
associated with the estimated map area (Byrd et al. 2018). Uncertainty for root to shoot ratios, which are
used for quantifying belowground biomass, are derived from the Wetlands Supplement. Uncertainty for
subtropical estuarine forested wetland DOM stocks was derived from those listed for the Tier 1
Land Use, Land-Use Change, and Forestry 6-141
-------
estimates (IPCC 2014). Overall uncertainty of the NOAA C-CAP remote sensing product is 15 percent.
This is in the range of remote sensing methods (+/-10 to 15 percent; IPCC 2003). The combined
uncertainty was calculated by summing the squared uncertainty for each individual source (C-CAP, soil,
biomass, and DOM) and taking the square root of that total.
Uncertainty estimates are presented in Table 6-75 for each subcategory (i.e., soil carbon, biomass
carbon, and DOM emissions). The combined uncertainty across all subcategory is 32.0 percent above
and below the estimate of 1.5 MMT C02 Eq, which is driven by the uncertainty in the soil carbon
estimates. In 2023, the total carbon flux was 1.5 MMT C02 Eq., with lower and upper estimates of 1.0
and 2.0 MMT C02 Eq.
Table 6-75: Approach 1 Quantitative Uncertainty Estimates for C02 Flux Occurring
within Vegetated Coastal Wetlands Converted to Unvegetated Open Water Coastal
Wetlands in 2023 (MMT C02 Eq. and Percent)
Uncertainty Range Relative to Flux Estimate
2023 Flux
(MMT CO2 Eq.)
(%)
Estimate
Lower
Upper
Lower
Upper
Source
(MMTCO2 Eq.)
Bound
Bound
Bound
Bound
Biomass C Stock
0.06
0.05
0.08
-24.1%
+24.1%
Dead Organic Matter C Stock
0.0005
0.000
0.001
-25.8%
+25.8%
Soil C Stock
1.5
1.3
1.7
-15.0%
+ 15.0%
Total Flux
1.5
1.0
2.0
-32.0%
+32.0%
Note: Totals may not sum due to independent rounding.
QA/QC and Verification
General QA/QC procedures were applied to activity data, documentation, and emission calculations
consistent with the U.S. Inventory QA/QC Plan, which is in accordance with Vol. 1 Chapter 6 of the 2006
IPCC Guidelines (see Annex 8 for more details). Data provided by NOAA (i.e., National LiDAR Dataset,
NOS Tide Data, and C-CAP land cover and land cover change mapping) undergo internal agency QA/QC
procedures. Acceptance of final datasets into archive and dissemination are contingent upon assurance
that the data product is compliant with mandatory NOAA QA/QC requirements (McCombs et al. 2016).
QA/QC and Verification of the soil carbon stock dataset have been provided by the Smithsonian
Environmental Research Center and by the Coastal Wetlands project team leads who reviewed the
estimates against primary scientific literature. Biomass carbon stocks are derived from peer-review
literature and reviewed by the U.S. Geological Survey prior to publishing, by the peer-review process
during publishing, and by the coastal wetland inventory team leads before inclusion in the Inventory. For
subtropical estuarine forested wetlands, Tier 1 estimates of mangrove DOM were used (IPCC 2014).
Land cover estimates were assessed to ensure that the total land area did not change over the time
series in which the inventory was developed, and were verified by a second QA team. A team of two
evaluated and verified there were no computational errors within the calculation worksheets.
Recalculations Discussion
No recalculations were performed for the current Inventory.
6-142 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Planned Improvements
The following are improvements being investigated for future vegetated coastal wetlands converted to
unvegetated open water coastal wetlands inventories:
• Updating the depth of soil carbon affected byconversion of vegetated coastal wetlands
converted to unvegetated coastal wetlands from the IPCC default assumption of 1 m of soil
erosion when mapping and modeling advancements to quantitatively improve accuracy and
precision. This involves conducting a review of literature publications. Until the time where
these more detailed and spatially distributed data are available, the IPCC default assumption
that the top 1 m of soil is disturbed by anthropogenic activity will be applied. This is a long-term
improvement.
• Conducting a longer-term assessment and researching more highly refined rates of wetlands
loss across the Mississippi Delta (e.g., Couvillion et al. 2016, Blum et al. 2023, Creamer et al.
2024). The Mississippi Delta is the largest extent of coastal wetlands in the United States. Higher
resolution imagery analysis would improve quantification of conversation to open water, which
occurs not only at the edge of the marsh but also within the interior. Improved mapping could
provide a more refined regional Approach 2-3 land representation to support the national-scale
assessment provided by C-CAP.
• Reviewing an approach for calculating the fraction of remobilized coastal wetland soil carbon
returned to the atmosphere as C02.
• Investigating higher resolution mapping approaches to quantify conversion of coastal wetlands.
This research being conducted by the USGS is underway. Such approaches may form the basis
for a full Approach 3 land representation assessment in future years. C-CAP data harmonization
with the National Land Cover Dataset (NLCD) will be incorporated into a future iteration of the
Inventory.
Stock Changes from Unvegetated Open Water Coastal
Wetlands Converted to Vegetated Coastal Wetlands
Open water within the U.S. land base, as described in Section 6.1, is recognized as coastal wetlands
within this Inventory. The appearance of vegetated tidal wetlands on lands previously recognized as
open water reflects either the building of new vegetated marsh through sediment accumulation or the
transition from other lands uses through an intermediary open water stage as flooding intolerant plants
are displaced and then replaced by wetland plants. Biomass, DOM and soil carbon accumulation on
unvegetated open water coastal wetlands converted to vegetated coastal wetlands begins with
vegetation establishment.
Within the United States, conversion of unvegetated open water coastal wetlands to vegetated coastal
wetlands is predominantly due to engineered activities, which include active restoration of wetlands
(e.g., wetlands restoration in San Francisco Bay), dam removals or other means to reconnect sediment
supply to the nearshore (e.g., Atchafalaya Delta, Louisiana, Couvillion et al. 2011). Wetland restoration
projects have been ongoing in the United States since the 1970s. Early projects were small, a few
hectares in size. By the 1990s, restoration projects, each hundreds of hectares in size, were becoming
common in major estuaries. In several coastal areas e.g., San Francisco Bay, Puget Sound, Mississippi
Land Use, Land-Use Change, and Forestry 6-143
-------
Delta and south Florida, restoration activities are in planning and implementation phases, each with the
goal of recovering tens of thousands of hectares of wetlands.
In 2023, 2,406 ha of unvegetated open water coastal wetlands were converted to vegetated coastal
wetlands across all wetland types and climates, which has steadily increased over the time-series
(Table 6-67). This resulted in 0.008 MMTC02 Eq. (0.002 MMTC) and 0.1 MMTC02 Eq. (0.03 MMTC)
sequestered in soil and biomass, respectively (Table 6-76 and Table 6-77). The soil carbon stock has
increased during the Inventory time-series, likely due to increasing vegetated coastal wetland
restoration over time. While DOM carbon stock increases are present, they are minimal in the early part
of the time series and zero in the later because there are no conversions from unvegetated open water
coastal wetlands to subtropical estuarine forested wetlands between 2011 and 2016 (and by proxy
through 2023), and that is the only coastal wetland type where DOM data is currently available.
Throughout the time-series, the amount of open water coastal wetlands converted to vegetated coastal
wetlands has increased overtime, reflecting the increase in engineered restoration activities mentioned
above.
Table 6-76: C02 Flux from Carbon Stock Changes from Unvegetated Open Water
Coastal Wetlands Converted to Vegetated Coastal Wetlands (MMT C02 Eq.)
Year
1990
2005 | 2019
2020
2021
2022
2023
Biomass C Flux
(+)
(0.1) I
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
Dead Organic Matter C Flux
(+)
(+)
0.0
0.0
0.0
0.0
0.0
Soil C Flux
(+)
(+)
{+)
{+)
{+)
{+)
{+)
Total C Stock Change
(+)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
+ Absolute value does not exceed 0.05 MMT C02 Eq.
Notes: Parentheses indicate net sequestration. Totals may not sum due to independent rounding.
Table 6-77: C02 Flux from Carbon Stock Changes from Unvegetated Open Water
Coastal Wetlands Converted to Vegetated Coastal Wetlands (MMT C)
Year
1990 | 2005 | 2019
2020
2021
2022
2023
Biomass C Flux
(0.01) I
(0.02) I
(0.03)
(0.03)
(0.03)
(0.03)
(0.03)
Dead Organic Matter C Flux
(+)
(+)
0
0
0
0
0
Soil C Flux
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Total C Stock Change
(0.01) I
(0.02) |
(0.03)
(0.03)
(0.03)
(0.03)
(0.03)
+ Absolute value does not exceed 0.005 MMT C.
Notes: Parentheses indicate net sequestration. Totals may not sum due to independent rounding.
Methodology and Time-Series Consistency
The following section includes a brief description of the methodology used to estimate changes in soil,
biomass and DOM carbon stocks, and CH4 emissions for unvegetated open water coastal wetlands
converted to vegetated coastal wetlands.
Methodological recalculations were applied to the entire time series to ensure time-series consistency
from 1990 through 2023.
6-144 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Biomass Carbon Stock Changes
Quantification of regional coastal wetland biomass carbon stock changes for palustrine and estuarine
marsh vegetation are presented for unvegetated open water coastal wetlands converted to vegetated
coastal wetlands on lands below the elevation of high tides (taken to be mean high water spring tide
elevation) according to the national LiDAR dataset, the national network of tide gauges and land use
histories recorded in the 1996, 2001, 2005, 2011, and 2016 NOAA C-CAP surveys. C-CAP areas are
calculated at the state/territory level and summed according to climate zone to national values. Privately-
owned and publicly-owned lands are represented. Trends in land cover change are extrapolated to 1990 and
2023 from these datasets (Table 6-66). C-CAP provides peer reviewed high resolution level mapping of
coastal wetland distribution, including conversion to and from open water. Biomass carbon stock is not
sensitive to soil organic content but differentiated based on climate zone. Data for non-forested
wetlands are derived from a national assessment combining field plot data and aboveground biomass
mapping by remote sensing (Table 6-69; Byrd et al. 2017; Byrd et al. 2018; Byrd et al. 2020). The
aboveground biomass carbon stock for subtropical estuarine forested wetlands (dwarf mangroves that
are not classified as forests due to their stature) is derived from a meta-analysis by Lu and Megonigal
(20 1 758). Aboveground biomass carbon stock data for all subcategories are not available and thus
assumptions were applied using expert judgment about the most appropriate assignment of a carbon
stock to a disaggregation of a community class. Root to shoot ratios from the Wetlands Supplement
were used to account for belowground biomass, which were multiplied by the aboveground carbon
stock (Table 6-70; IPCC 2014). Above- and belowground values were summed to obtain total biomass
carbon stocks.
Conversion of open water to vegetated coastal wetlands results in the establishment of a standing
biomass carbon stock; therefore, stock changes that occur are calculated by multiplying the C-CAP
derived area gained that year in each climate zone by its mean biomass. While the process of
revegetation of unvegetated open water wetlands can take many years to occur, it is assumed in the
calculations that the total biomass is reached in the year of conversion.
Dead Organic Matter
Dead organic matter (DOM) carbon stocks, which include litter and dead wood stocks, are included for
subtropical estuarine forested wetlands for vegetated coastal wetlands converted to unvegetated open
water coastal wetlands across all years. Tier 1 default or country-specific data on DOM are not currently
available for either palustrine or estuarine scrub/shrub wetlands for any climate zone. Data for estuarine
forested wetlands in other climate zones are not included since there is no estimated loss of these
forests to unvegetated open water coastal wetlands across any year based on C-CAP data. Tier 1
estimates of subtropical estuarine forested wetland DOM were used (IPCC 2014). Trends in land cover
change are derived from the NOAA C-CAP dataset and extrapolated to cover the entire 1990 through
2023 time series. Dead organic matter removals are calculated by multiplying the C-CAP derived area
gained that year by its Tier 1 DOM C stock. Similar to biomass carbon stock gains, gains in DOM can take
many years to occur, but for this analysis, the total DOM stock is assumed to accumulate during the first
year of conversion.
58 See https://doi.Org/10.?5573/serc.?1565671: accessed September ?0?4.
Land Use, Land-Use Change, and Forestry 6-145
-------
Soil Carbon Stock Change
Soil carbon stock changes are estimated for unvegetated open water coastal wetlands converted to
vegetated coastal wetlands. Country-specific soil carbon removal factors associated with soil carbon
accretion, stratified by climate zones and wetland classes, are derived from a synthesis of peer-
reviewed literature and updated this year based upon refined review of the dataset (Lynch 1989; Orson
et al. 1990; Kearny & Stevenson 1991; Thom et al. 1992; Roman et al. 1997; Craft et al. 1998; Orson et al.
1998; Merrill 1999; Weis et al. 2001; Hussein et al. 2004; Church et al. 2006; Koster et al. 2007; Drexler
et al. 2009; Boyd 2012; Callaway et al. 2012 a & b; Bianchi et al. 2013; Drexler et al. 2013; Watson and
Byrne 2013; Crooks et al. 2014; Breithaupt et al. 2014; Weston et al. 2014; Smith et al. 2015; Villa &
Mitsch 2015; Boyd and Sommerfield 2016; Marchio et al. 2016; Noe et al. 2016; Arriola and Cable 2017;
Boyd et al. 2017; Gerlach et al. 2017; Giblin and Forbrich 2018; Krauss et al. 2018; Abbott et al. 2019;
Drexler et al. 2019; Poppe and Rybczyk 2019; Ensign et al. 2020; Kemp et al. 2020; Lagomasino et al.
2020; Luk et al. 2020; McTigue et al. 2020; Peck et al. 2020; Vaughn et al. 2020; Weston et al. 2020;
Arias-Ortiz et al. 2021; Baustian et al. 2021; Allen et al. 2022; Miller et al. 2022). Soil carbon stock
changes are stratified based upon wetland class (Estuarine, Palustrine) and subclass (Emergent Marsh,
Scrub Shrub). For soil carbon stock change, no differentiation is made for soil type (i.e., mineral,
organic). Soil carbon removal factors were developed from literature references that provided soil
carbon removal factors disaggregated by climate region and vegetation type by salinity range (estuarine
or palustrine) as identified using NOAAC-CAP as described above (see Table 6-71 for values).
Tier 2 level estimates of carbon stock changes associated with annual soil carbon accumulation in
vegetated coastal wetlands were developed using country-specific soil carbon removal factors
multiplied by activity data on unvegetated coastal wetlands converted to vegetated coastal wetlands.
The methodology follows Eq. 4.7, Chapter 4 of the Wetlands Supplement, and is applied to the area of
unvegetated coastal wetlands converted to vegetated coastal wetlands on an annual basis.
Soil Methane Emissions
ATier 1 assumption has been applied that salinity conditions are unchanged and hence CH4 emissions
are assumed to be zero with conversion of vegetated open water coastal wetlands to vegetated coastal
wetlands.
Uncertainty
Underlying uncertainties in estimates of soil and biomass carbon stock changes include uncertainties
associated with country-specific (Tier 2) literature values of these carbon stocks, assumptions that
underlie the methodological approaches applied and uncertainties linked to interpretation of remote
sensing data. Uncertainty specific to coastal wetlands include differentiation of palustrine and
estuarine community classes that determines the soil carbon stock applied. Because mean soil and
biomass carbon stocks for each available community class are in a fairly narrow range, the same overall
uncertainty was applied to each, respectively (i.e., applying approach for asymmetrical errors, the
largest uncertainty for any soil carbon stock value should be applied in the calculation of error
propagation; IPCC 2000). For aboveground biomass carbon stocks, the mean standard error was very
low and largely influenced by error in estimated map area (Byrd et al. 2018). Uncertainty for root to shoot
ratios, which are used for quantifying belowground biomass (Table 6-70), are derived from the Wetlands
Supplement. Uncertainty for subtropical estuarine forested wetland DOM stocks were derived from
those listed for the Tier 1 estimates (IPCC 2014). Overall uncertainty of the NOAAC-CAP remote sensing
6-146 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
product is 15 percent. This is in the range of remote sensing methods (±10 to 15 percent; IPCC 2003).
The combined uncertainty was calculated by summing the squared uncertainty for each individual
source (C-CAP, soil, biomass, and DOM) and taking the square root of that total.
Uncertainty estimates are presented in Table 6-78 for each subcategory (i.e., soil carbon, biomass
carbon and DOM emissions). The combined uncertainty across all subcategories is 33.43 percent above
and below the estimate of-0.1 MMT C02 Eq. In 2023, the total carbon flux was -0.1 MMT C02 Eq., with
lower and upper estimates of -0.1 and -0.08 MMT C02 Eq.
Table 6-78: Approach 1 Quantitative Uncertainty Estimates for Carbon Stock Changes
Occurring within Unvegetated Open Water Coastal Wetlands Converted to Vegetated
Coastal Wetlands in 2023 (MMT C02 Eq. and Percent)
Uncertainty Range Relative to Flux Estimate
2023 Flux
(MMT CO2 Eq.)
(%)
Estimate
Lower
Upper
Lower
Upper
Source
(MMT CO2 Eq.)
Bound
Bound
Bound
Bound
Biomass C Stock Flux
(0.1)
(0.12)
(0.08)
-20.0%
+20.0%
Dead Organic Matter C Stock Flux
0
0
0
-25.8%
+25.8%
Soil C Stock Flux
(0.008)
(0.009)
(0.006)
-17.7%
+ 17.7%
Total Flux
(0.1)
(0.14)
(0.01)
-33.3%
+33.3%
Notes: Parentheses indicate net sequestration. Totals may not sum due to independent rounding.
QA/QC and Verification
NOAA provided data (i.e., National LiDAR Dataset, NOS Tide Data, and C-CAP land cover and land cover
change mapping), which undergo internal agency QA/QC assessment procedures. Acceptance of final
datasets into the archive for dissemination are contingent upon assurance that the product is compliant
with mandatory NOAA QA/QC requirements (McCombs et al. 2016). QA/QC and Verification of soil
carbon stock dataset has been provided by the Smithsonian Environmental Research Center and
Coastal Wetlands project team leads who reviewed the summary tables against primary scientific
literature. Aboveground biomass carbon reference stocks are derived from an analysis by the Blue
Carbon Monitoring project and reviewed by U.S. Geological Survey prior to publishing, the peer-review
process during publishing, and the coastal wetland inventory team leads before inclusion in the
Inventory. Root to shoot ratios and DOM data are derived from peer-reviewed literature and undergo
review as per IPCC methodology. Land cover estimates were assessed to ensure that the total land area
did not change over the time series in which the inventory was developed and verified by a second QA
team. A team of two evaluated and verified there were no computational errors within calculation
worksheets. Two biogeochemists at the USGS, also members of the NASA Carbon Monitoring System
Science Team, corroborated the simplifying assumption thatwhere salinities are unchanged CH4
emissions are constant with conversion of unvegetated open water coastal wetlands to vegetated
coastal wetlands.
Recalculations Discussion
No recalculations were performed for the current Inventory.
Land Use, Land-Use Change, and Forestry 6-147
-------
Planned Improvements
The following are planned improvements being investigated for a future unvegetated open water coastal
wetlands converted to vegetated coastal wetlands Inventory:
• Investigating higher resolution mapping approaches to quantify conversion of coastal wetlands.
The USGS is current undertaking this effort. Such approaches may form the basis for a full
Approach 3 land representation assessment in future years. This is considered a medium- to
long-term improvement.
• Harmonizing C-CAP data with the National Land Cover Dataset (NLCD). This is considered a
medium-term improvement.
N20 Emissions from Aquaculture in Coastal Wetlands
Shrimp and fish cultivation in coastal areas increases nitrogen loads resulting in direct emissions of
N20. Nitrous oxide is generated and emitted as a byproduct of the conversion of ammonia (contained in
fish urea) to nitrate through nitrification and nitrate to N2 gas through denitrification (Hu et al. 2012).
Nitrous oxide emissions can be readily estimated from data on fish production (IPCC 2014).
Aquaculture production in the United States has fluctuated slightly from year to year, with resulting N20
emissions between 0.1 and 0.2 MMTC02 Eq. between 1990 and 2023 (Table 6-79). Aquaculture
production data were updated through 2019; data through 2023 are not yet available and in this analysis
are held constant with 2019 emissions of 0.2 MMT C02 Eq. (0.5 Kt N20).
Table 6-79: N20 Emissions from Aquaculture in Coastal Wetlands (MMT C02 Eq. and kt
N20)
Year
1990 |
2005
2019
2020
2021
2022
2023
Emissions (MMT CO2 Eq.)
0.1
0.2
0.1
0.1
0.1
0.1
0.1
Emissions (kt N2O)
0.4 |
0.6
0.5
0.5
0.5
0.5
0.5
Methodology and Time-Series Consistency
The methodology to estimate N20 emissions from aquaculture in coastal wetlands follows the Tier 1
guidance in the Wetlands Supplement by applying country-specific fisheries production data and the
IPCC Tier 1 default emission factor.
NOAA Fisheries document the status of U.S. marine fisheries in the report of Fisheries of the United
States (National Marine Fisheries Service 2022), from which activity data for this analysis is derived.59
The fisheries report has been produced in various forms for more than 100 years, primarily at the
national level, on U.S. recreational catch and commercial fisheries landings and values. In addition,
data are reported on U.S. aquaculture production, the U.S. seafood processing industry, imports and
exports of fish-related products, and domestic supply and per capita consumption of fisheries
products. Within the aquaculture chapter, the mass of production for catfish, striped bass, tilapia, trout,
crawfish, salmon and shrimp are reported. While some of these fisheries are produced on land and
some in open water cages within coastal wetlands, all have data on the quantity of food stock produced,
59 See https://www.fisheries.noaa.gov/resoijrce/docijment/fisheries-united-states-90??: accessed October ?0?4.
6-148 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
which is the activity data that is applied to the IPCC Tier 1 default emissions factor to estimate
emissions of N20 from aquaculture from 1990 to 2023. It is not apparent from the data as to the amount
of aquaculture occurring above the extent of high tides on river floodplains. While some aquaculture
occurs on coastal lowland floodplains, this is likely a minor component of tidal aquaculture production
because of the need for a regular source of water for pond flushing. The estimation of N20 emissions
from aquaculture is not sensitive to salinity using IPCC approaches, and as such, the location of
aquaculture ponds within the boundaries of coastal wetlands does not influence the calculations.
Values from the 2022 report were extrapolated to 2023 due to lack of data from 2023.
Other open water shellfisheries for which no food stock is provided, and thus no additional N inputs, are
not applicable for estimating N20 emissions (e.g., clams, mussels, and oysters) and have not been
included in the analysis. The IPCC Tier 1 default emissions factor of 0.00169 kg N20-N per kg of
fish/shellfish produced is applied to the activity data to calculate total N20 emissions.
Methodological recalculations were applied to the entire time series to ensure time-series consistency
from 1990 through 2023.
Uncertainty
Uncertainty estimates are based upon the Tier 1 default 95 percent confidence interval provided in Table
4.15, chapter 4 of the Wetlands Supplement for N20 emissions and on expert judgment of the NOAA
Fisheries of the United States fisheries production data. Given the overestimate of fisheries production
from coastal wetland areas due to the inclusion offish production in non-coastal wetland areas, this is a
reasonable initial first approximation for an uncertainty range.
Uncertainty estimates for N20 emissions from aquaculture production are presented in Table 6-80 for
N20 emissions. The combined uncertainty is 116 percent above and below the estimate of 0.13 MMT
C02 Eq. In 2023, the total flux was 0.13 MMT C02 Eq., with lower and upper estimates of 0.00 and 0.29
MMTC02 Eq.
Table 6-80: Approach 1 Quantitative Uncertainty Estimates for N20 Emissions from
Aquaculture Production in Coastal Wetlands in 2023 (MMT C02 Eq. and Percent)
2023
Emissions
Estimate
Source (MMT CO2 Eq.)
Uncertainty Range Relative to Emissions Estimate"
(MMT CO2 Eq.)
(%)
Lower Upper
Bound Bound
Lower Upper
Bound Bound
Combined Uncertainty for N2O
Emissions for Aquaculture Production in 0.13
Coastal Wetlands
0.00 0.29
-116% +116%
a Range of emission estimates predicted by Monte Carlo stochastic simulation for a 95 percent confidence interval.
QA/QC and Verification
NOAA provided internal QA/QC review of reported fisheries data. The coastal wetlands inventory team
consulted with the coordinating lead authors of the coastal wetlands chapter of the Wetlands
Supplement to assess which fisheries production data to include in estimating emissions from
aquaculture. It was concluded that N20 emissions estimates should be applied to any fish production to
Land Use, Land-Use Change, and Forestry 6-149
-------
which food supplement is supplied be they pond or coastal open water and that salinity conditions were
not a determining factor in production of N20 emissions.
Recalculations Discussion
No recalculations were performed for the current Inventory.
Planned Improvements
No planned improvements have been identified for N20 emissions from aquaculture production in
coastal wetlands.
Flooded Land Remaining Flooded Land
Flooded lands are defined as water bodies where human activities have 1) caused changes in the
amount of surface area covered by water, typically through water level regulation (e.g., constructing a
dam), 2) waterbodies where human activities have changed the hydrology of existing natural
waterbodies there by altering water residence times and/or sedimentation rates, in turn causing
changes to the natural emission of greenhouse gases, and 3) waterbodies that have been created by
excavation, such as canals, ditches and ponds (IPCC 2019). Flooded lands include waterbodies with
seasonally variable degrees of inundation, but these waterbodies would be expected to retain some
inundated area throughout the year under normal conditions.
Flooded lands are broadly classified as "reservoirs" or "other constructed waterbodies" (IPCC 2019).
Reservoirs are defined as flooded land greater than 8 ha. Other constructed waterbodies include
canals/ditches and ponds. Ponds are defined as flooded land that is not a canal/ditch with surface area
<8 ha. IPCC guidance (IPCC 2019) provides default emission factors for reservoirs and other
constructed waterbodies.
Land that has been flooded for greater than 20 years is defined as flooded land remaining flooded land.
Land flooded for 20 years or less is defined as land converted to flooded land. The distinction is based
on literature reports that CH4 and C02 emissions are relatively high immediately following flooding but
decline to a steady background level approximately 20 years after flooding (Abril et al. 2005; Barros et al.
2011; Teodoru et al. 2012). Emissions of CH4 are estimated for flooded land remaining flooded land, but
C02 emissions are not included as they are primarily the result of decomposition of organic matter
entering the waterbody from the catchment or contained in inundated soils and are captured in Chapter
6, Land Use, Land-Use Change, and Forestry.
Nitrous oxide emissions from flooded lands are largely related to input of organic or inorganic nitrogen
from the watershed. These inputs from runoff/leaching/deposition are largely driven by anthropogenic
activities such as land-use change, wastewater disposal or fertilizer application in the watershed or
application of fertilizer or feed in aquaculture. These emissions are not included here to avoid double-
counting of N20 emissions which are captured in other source categories, such as indirect N20
emissions from managed soils (Section 5.4, Agricultural Soil Management) and wastewater
management (Section 7.2, Wastewater Treatment and Discharge).
6-150 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Emissions from Flooded Land Remaining Flooded Land-
Reservoirs
Reservoirs are designed to store water for a wide range of purposes including hydropower, flood control,
drinking water, and irrigation. In 2023, the United States ana Puerto Rico contained 9,6 million ha of
reservoir surface area in the flooded land remaining flooded land category. These reservoirs are
distributed across all six of the aggregated climate zones used to define flooded land emission factors
(Figure 6-10) (IPCC2019).
Figure 6-10: U.S. Reservoirs (black polygons) in the Flooded Land Remaining Flooded
Land Category in 2023
500 mi
100 mi
Climate Zone
| boreal
I] cool temperate
| tropical dry/montane
| tropical moist/wet
H warm temperate dry
J warm temperate moist
Alaska
1000 mi
Hawaii
o
500 mi
Note: Colors represent climate zone used to derive IPCC default emission factors, Alaska map scale is 1000 miles, Hawaii and
contiguous United States map scale is 500 miles, Puerto Rico map scale is 100 miles.
Methane is produced in reservoirs through the microbial breakdown of organic matter. Per unit area, CH -,
emission rates tend to scale positively with temperature and system productivity (i.e., abundance of
algae), but negatively with system size (i.e., depth, surface area). Methane produced in reservoirs can
also be emitted from the reservoir when CH4-rich water passes through the dam. This exported CH4can
be released to the atmosphere as the water passes through hydropower turbines or the downstream
river channel. Methane emitted to the atmosphere via this pathway is referred to as "downstream
emissions."
Table 6-81 and Table 6-82 below summarize nationally aggregated CH4 emissions from reservoirs. The
increase in CH4 emissions through the time series is attributable to reservoirs matriculating from the
land converted to flooded land category into the flooded land remaining flooded land category.
Land Use, Land-Use Change, and Forestry 6-151
-------
Table 6-81: CH4 Emissions from Flooded Land Remaining Flooded Land—Reservoirs
(MMT C02 Eq.)
Source
1990
2005
2019
2020
2021
2022
2023
Reservoirs
Surface Emission
24.2
25.6
25.8
25.8
25.8
25.8
25.8
Downstream Emission
2.2
2.3
2.3
2.3
2.3
2.3
2.3
Total
26.4
27.9
27.9
28.1
28.2
28.2
28.2
Note: Totals may not sum to due independent rounding.
Table 6-82: CH4 Emissions from Flooded Land Remaining Flooded Land—Reservoirs
(kt CH4)
Source
1990
2005
2019
2020
2021
2022
2023
Reservoirs
Surface Emission
866
913
922
992
992
992
993
Downstream Emission
78
82
83
83
83
83
83
Total
944
995
1,005
1,005
1,005
1,005
1,006
Note: Totals may not sum to due independent rounding.
Methane emissions from reservoirs in Texas, Florida, and Louisiana (Figure 6-11, Table 6-83) compose
31.1 percent of national CH4 emissions from reservoirs in 2023. Emissions from these states are
particularly high due to 1) the large expanse of reservoirs in these states (Table 6-86) and 2) the high CH4
emission factor for the tropical dry/montane and topical moist climate zones which encompass a
majority of the flooded land area in these states (Figure 6-10, Table 6-84).
Methane emissions from reservoirs in flooded land remaining flooded land increased 6.6 percent from
1990 to 2023 due to the matriculation of reservoirs in land converted to flooded land to flooded land
remaining flooded land.
6-152 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Figure 6-11: Total CH4 Emissions (Downstream + Surface) from Reservoirs in Flooded
Land Remaining Flooded Land in 2023 (kt CH4)
500 mi
100 mi
500 mi
Note: Alaska map scale is 1000 miles, Hawaii and contiguous United States map scale is 500 miles, Puerto Rico map scale
is 100 miles.
Table 6-83: Surface and Downstream CH4 Emissions from Reservoirs in Flooded Land
Remaining Flooded Land in 2023 (kt CH4)
State
Surface
Downstream
Total
Alabama
22
2
24
Alaska
1
+
1
Arizona
14
1
15
Arkansas
25
2
27
California
36
3
40
Colorado
6
1
7
Connecticut
3
+
3
Delaware
4
+
4
District of Columbia
1
+
1
Florida
99
9
108
Georgia
33
3
36
Hawaii
1
+
1
Idaho
11
1
12
Illinois
16
1
17
Indiana
5
+
5
Iowa
6
1
7
Kansas
10
1
10
Kentucky
14
1
15
Land Use, Land-Use Change, and Forestry 6-153
-------
State
Surface
Downstream
Total
Maine 14 1 16
Maryland
12
1
13
Massachusetts
5
+
5
Michigan
9
1
10
Minnesota
20
2
22
Mississippi
19
2
21
Missouri
16
1
17
Montana
15
1
17
Nebraska
6
1
7
Nevada
17
2
18
New Hampshire
3
+
4
New Jersey
9
1
10
New Mexico
6
1
7
New York
17
1
18
North Carolina
32
3
35
North Dakota
14
1
15
Ohio
7
1
7
Oklahoma
25
2
28
Oregon
15
1
16
Pennsylvania
7
1
8
Puerto Rico
+
+
+
Rhode Island
1
+
1
South Carolina
36
3
40
South Dakota
12
1
13
Tennessee
19
2
21
Texas
131
12
143
Utah
21
2
23
Vermont
5
+
5
Virginia
24
2
27
Washington
22
2
23
West Virginia
3
+
3
Wisconsin
11
1
11
Wyoming
7
1
8
+ Indicates values less than 0.5 kt.
Note: Totals may not sum due to independent rounding.
Methodology and Time-Series Consistency
Estimates of CH4 emission for reservoirs in flooded land remaining flooded land follow the Tier 1
methodology in the 2019 Refinement to the 2006IPCC Guidelines (IPCC 2019). Methane emissions from
the surface of these flooded lands are calculated as the product of flooded land surface area and a
climate-specific emission factor (Table 6-84). Downstream emissions are calculated as nine percent of
the surface emission (Tier 1 default). Total CH4 emissions from reservoirs are calculated as the sum of
surface and downstream emissions. National emissions are calculated as the sum of state emissions.
6-154 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
The IPCC default surface emission factors used in the Tier 1 methodology are derived from model-
predicted (G-res model, Prairie et al. 2017) emission rates for all reservoirs in the Global Reservoir and
Dam (GRanD) database (Lehner et al. 2011). Predicted emission rates were aggregated by the 11 IPCC
climate zones (IPCC 2019, Table 7A.2) which were collapsed into six climate zones using a regression
tree approach. All six aggregated climate zones are present in the United States.
Table 6-84: IPCC (2019) Default CH4 Emission Factors for Surface Emission from
Reservoirs in Flooded Land Remaining Flooded Land
Climate Surface emission factor (MT Cm ha"1 y"1)
Boreal 0.0136
Cool Temperate
0.0540
Warm Temperate Dry
0.1509
Warm Temperate Moist
0.0803
Tropical Dry/Montane
0.2837
Tropical Moist/Wet
0.1411
Notes: Downstream ChU emissions are calculated as 9 percent of surface emissions. Downstream emissions are not calculated
for C02.
Area Estimates
U.S. reservoirs were identified from the "NHDWaterbody" layer in the National Hydrography Dataset Plus
V2 (NHD),60 the National Inventory of Dams (NID),61 the Hydropower Infrastructure—Lakes, Reservoirs,
and Rivers (HILARRI) database,62 the National Wetlands Inventory (NWI),63 the Navigable Waterways
(NW) network,64 and the EPA's Safe Drinking Water Information System (SDWIS).65The NHD only covers
the conterminous U.S., whereas the NID, NW and NWI also include Alaska, Hawaii, and Puerto Rico.
Waterbodies in the NHDWaterbody layer that were greater than or equal to 8 ha in surface area, not
identified as canal/ditch in NHD, and met any of the following criteria were considered reservoirs: 1) the
waterbody was classified as "Subtype = Reservoir" in the "NHDWaterbody" layer, 2) the GNIS name in
the "NHDWaterbody" layer contained the text "Reservoir", 3) the "NHDWaterbody" or "NHDArea" layer
was located in close proximity (up to 200 m) to a dam in the NID, 4) the "NHDWaterbody" GNIS name
was similar to a nearby NID feature (up to a 1000 m radius), 5) the "NHDWaterbody" intersected a public
drinking water intake, 6) the "NHDWaterbody" intersected an operational hydropower dam defined by
the HILARRI dataset.
EPA assumes that all features included in the NW network are subject to water-level management to
maintain minimum water depths required for navigation and are therefore managed flooded lands.
Navigable Waterway features greater than 8 ha in surface area are defined as reservoirs.
NWI features were considered "managed" if they had a Special Modifier value indicating the presence of
management activities (Figure 6-12). To be included in the flooded lands inventory, the managed flooded
land had to be wet or saturated for at least one season per year (see "Water Regime" in Figure 6-12). NWI
60 See https://www.usgs.gov/core-science-systems/ngp/national-hydrography.
61 See https://nid.sec.usace.army.mil.
62 See https://www.osti.gov/biblio/1960141.
63 See https://www.fws.gov/program/national-wetlands-inventorv/data-download.
64 See https://geodata.bts.gov/datasets/usdot::navigable-waterwav-network-lines/explore.
65 See https://www.epa.gov/enviro/sdwis-overview. Not publicly available due to security concerns
Land Use, Land-Use Change, and Forestry 6-155
-------
features that met these criteria, were greater than 8 ha in surface area, and were not a canal/ditch (see
emissions from land converted to flooded land - other constructed waterbodies) were defined as
reservoirs.
Any NWI or NHD feature that intersected a drinking water intake point from SDWIS was assumed to be
"managed ."The rational being that a waterbody used as a source for public drinking water is typically
managed in some capacity - by flow and/or volume control,
Surface areas for identified flooded lands were taken from the NHD, NWI orNW. If features from the
NHD, NWI, or NW datasets overlapped, duplicated areas were erased. The first step was to take the final
NWI flooded lands features and use it to identify overlapping NHD features. If the NHD feature had its
center in a NWI feature, it was removed from analysis. Next, remaining NHD features were erased from
any remaining overlapping NWI features. Final selections of NHD and NWI features were used to erase
any overlapping NW waterbodies.
Reservoir age was determined by assuming the waterbody was created the same year as a nearby (up to
100 m) NID feature. If no nearby NID feature was identified, it was assumed the waterbody was greater
than 20-years old throughout the time series.
Figure 6-12: Selected Features from NWI that Meet Flooded Lands Criteria
MODIFIERS
In order to more adequately describe the wetland and deepwater habitats, one each of the water regime, water chemistry, soil, or
special modifiers may be applied at the class or lower level in the hierarchy
Water Regime
Special Modifiers
Water Chemistry
Soil
Nontidal
A Temporarily Flooded
Saltwater Tidal
|L Subtidal
Freshwater Tidal
Q Regularly Flooded-Fresh Tidal
R Seasonally Flooded-Fresh Tidal
S Temporarily Flooded- Fresh Tidal
T Semipermanently Flooded-Fresh Tidal
V Permanently Flooded-Fresh Tidal
b Beaver
Halinity/Salinity pH Modifiers for
Fresh Water
1 Hyperhaline / Hypersaline a Acid
2 Euhaline / Eusaline t Circumneutral
3 Mixohaline / M ixosaline (Brackish) i Alkaline
4 Polyhaline
5 Mesohaline
6 Oligohaline
0 Fresh
g Organic
n Mineral
B Seasonally Saturated
C Seasonally Flooded
D Continuously Saturated
M Irreqularlv Exposed
N Regularly Flooded
P Irregularly Flooded
d Partly Drained/Ditched
f Farmed
m Managed
h Diked/impounded
r Artificial Substrate
s Spoil
x Excavated |
E Seasonally Flooded /
Saturated
F Semipermanently Flooded
G Intermittently Exposed
H Permanently Flooded
J Intermittently Flooded
K Artificially Flooded |
] Must also meet one selected special modifier (red box) to be included in the flooded lands inventory
I ~l Included in the flooded lands inventory if it meets water regime qualifier (gold box)
Source (modified): https://www.fws.gov/sites/default/files/documents/wetlands-and-deepwater-map-code-diagram.pdf
IPCC (2019) allows for the exclusion of managed waterbodies from the Inventory If the water surface
area or residence time was not substantially changed by the construction of the dam. The guidance
does not quantify what constitutes a "substantial" change, but here EPA excludes the U.S. Great Lakes
from the Inventoiy based on expert judgment that neither the surface area nor water residence time was
substantially altered by their associated dams.
Reservoirs were disaggregated by state (using boundaries from the 2016 U.S. Census Bureau66) and
climate zone. Downstream and surface emissions for cross-state reservoirs were allocated to states
based on the surface area that the reservoir occupied in each state. Only the U.S. portion of reservoirs
that cross country borders were included in the Inventory.
The surface area of reservoirs in flooded land remaining flooded land increased by approximately 6.4
percent from 1990 to 2023 (Table 6-85) due to reservoirs matriculating into flooded land remaining
flooded land when they reached 20 years of age.
68 See https://www.censLis.gov/geographie,s/mapping-filas/time-series/geo/carto-bouridary-file.html.
6-156 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Table 6-85: National Totals of Reservoir Surface Area in Flooded Land Remaining
Flooded Land (millions of ha)
Surface Area (millions of ha)
1990
2005
2019
2020
2021
2022
2023
Reservoir
9.1
9.5
9.6
9.6
9.6
9.6
9.6
Table 6-86: State Breakdown of Reservoir Surface Area in Flooded Land Remaining
Flooded Land (millions of ha)
State
1990
2005
2019
2020
2021
2022
2023
Alabama
0.22
0.22
0.22
0.22
0.22
0.22
0.22
Alaska
0.02
0.02
0.02
0.02
0.02
0.02
0.02
Arizona
0.06
0.06
0.06
0.06
0.06
0.06
0.06
Arkansas
0.25
0.28
0.28
0.28
0.28
0.28
0.28
California
0.34
0.35
0.35
0.35
0.35
0.35
0.35
Colorado
0.07
0.08
0.08
0.08
0.08
0.08
0.08
Connecticut
0.03
0.03
0.04
0.04
0.04
0.04
0.04
Delaware
0.05
0.05
0.05
0.05
0.05
0.05
0.05
District of Columbia
0.01
0.01
0.01
0.01
0.01
0.01
0.01
Florida
0.69
0.70
0.70
0.70
0.70
0.70
0.70
Georgia
0.27
0.28
0.28
0.28
0.28
0.28
0.28
Hawaii
+
+
+
+
+
+
+
Idaho
0.16
0.18
0.18
0.18
0.18
0.18
0.18
Illinois
0.15
0.16
0.20
0.20
0.20
0.20
0.20
Indiana
0.05
0.06
0.06
0.06
0.06
0.06
0.06
Iowa
0.08
0.09
0.09
0.09
0.09
0.09
0.09
Kansas
0.09
0.10
0.10
0.10
0.10
0.10
0.10
Kentucky
0.16
0.17
0.17
0.17
0.17
0.17
0.17
Louisiana
0.39
0.40
0.40
0.40
0.40
0.40
0.40
Maine
0.25
0.26
0.27
0.27
0.27
0.27
0.27
Maryland
0.15
0.15
0.15
0.15
0.15
0.15
0.15
Massachusetts
0.07
0.07
0.07
0.07
0.07
0.07
0.07
Michigan
0.15
0.16
0.17
0.17
0.17
0.17
0.17
Minnesota
0.36
0.36
0.37
0.37
0.37
0.37
0.37
Mississippi
0.17
0.18
0.18
0.18
0.18
0.18
0.18
Missouri
0.18
0.20
0.20
0.20
0.20
0.20
0.20
Montana
0.27
0.28
0.28
0.28
0.28
0.28
0.28
Nebraska
0.06
0.07
0.07
0.07
0.07
0.07
0.07
Nevada
0.08
0.09
0.09
0.09
0.09
0.09
0.09
New Hampshire
0.06
0.06
0.06
0.06
0.06
0.06
0.06
New Jersey
0.12
0.12
0.12
0.12
0.12
0.12
0.12
New Mexico
0.05
0.05
0.05
0.05
0.05
0.05
0.05
New York
0.28
0.28
0.29
0.29
0.29
0.29
0.29
North Carolina
0.39
0.40
0.40
0.40
0.40
0.40
0.40
North Dakota
0.25
0.25
0.25
0.25
0.25
0.25
0.25
Ohio
0.08
0.09
0.09
0.09
0.09
0.09
0.09
Oklahoma
0.26
0.29
0.29
0.29
0.29
0.29
0.29
Oregon
0.20
0.21
0.21
0.21
0.21
0.21
0.21
Land Use, Land-Use Change, and Forestry 6-157
-------
State
1990
2005
2019
2020
2021
2022
2023
Pennsylvania
0.10
0.10
0.10
0.10
0.10
0.10
0.10
Puerto Rico
+
+
+
+
+
+
+
Rhode Island
0.02
0.02
0.02
0.02
0.02
0.02
0.02
South Carolina
0.30
0.31
0.32
0.32
0.32
0.32
0.32
South Dakota
0.22
0.22
0.23
0.23
0.23
0.23
0.23
Tennessee
0.18
0.23
0.23
0.23
0.23
0.23
0.23
Texas
0.60
0.67
0.67
0.67
0.67
0.67
0.67
Utah
0.17
0.17
0.17
0.17
0.17
0.17
0.17
Vermont
0.09
0.09
0.09
0.09
0.09
0.09
0.09
Virginia
0.30
0.30
0.30
0.30
0.30
0.30
0.30
Washington
0.22
0.22
0.22
0.22
0.22
0.22
0.22
West Virginia
0.04
0.04
0.04
0.04
0.04
0.04
0.04
Wisconsin
0.19
0.19
0.19
0.19
0.19
0.19
0.19
Wyoming
0.11
0.13
0.13
0.13
0.13
0.13
0.13
Total
9.06
9.51
9.63
9.63
9.63
9.63
9.64
+ Indicates values less than 0.005 million ha.
Note: Totals may not sum due to independent rounding.
Uncertainty
Uncertainty in estimates of CH4 emissions from reservoirs in flooded land remaining flooded land (Table
6-87) are developed using Monte Carlo simulations (IPCC Approach 2) and include uncertainty in the
default emission factors and land areas. Each iteration of the simulation draws surface and downstream
emission factors from a statistical distribution based on the mean and variance in the 2019 Refinement
to the 2006 IPCC Guidelines (IPCC 2019). The CH4 emission factors for surface and downstream
emissions are modeled using normal and lognormal distributions, respectively Uncertainties in the
spatial data include 1) uncertainty in area estimates from the NHD, NWI, and NW, and 2) uncertainty in
the location of dams in the NID. Overall uncertainties in these spatial datasets are unknown, but
uncertainty for remote sensing products is assumed to be ± 10 -15 percent based on IPCC guidance
(IPCC 2003). An uncertainty range of ± 15 percent for the reservoir area estimates is assumed and is
based on expert judgment. Each iteration of the simulation draws a surface area for each waterbody
from a uniform distribution bounded by ± 15 percent of the estimated surface area.
Table 6-87: Approach 2 Quantitative Uncertainty Estimates for CH4 Emissions from
Reservoirs in Flooded Land Remaining Flooded Land
Uncertainty Range Relative to Emission Estimate"
2023 Emission
(MMTCO2 Eq.)
(%)
Estimate
Lower
Upper
Lower
Upper
Source
Gas
(MMT CO2 Eq.)
Bound
Bound
Bound
Bound
Reservoir
Surface
CH4
25.8
25.4
26.2
-1.7%
+ 1.3%
Downstream
ch4
2.3
2.2
2.7
-4.3%
+ 15.7%
Total
ch4
28.2
27.9
28.7
-0.8%
+2.0%
a Range of emission estimates predicted by Monte Carlo stochastic simulation for a 95 percent confidence interval.
6-158 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
QA/QC and Verification
The National Hydrography Data (NHD) is managed by the USGS in collaboration with many other federal,
state, and local entities. Extensive QA/QC procedures are incorporated into the curation of the NHD. The
National Inventory of Dams (NID) is maintained by the U.S. Army Corps of Engineers (USACE) in
collaboration with the Federal Emergency Management Agency (FEMA) and state regulatory offices.
USACE resolves duplicative and conflicting data from 68 data sources, which helps obtain the more
complete, accurate, and updated NID. The Navigable Waterways (NW) dataset is part of the U.S.
Department of Transportation (USDOT)/Bureau of Transportation Statistics (BTS) National Transportation
Atlas Database (NTAD). The NW is a comprehensive network database of the nation's navigable
waterways updated on a continuing basis. U.S. Fish and Wildlife Service is the principal agency in
charge of wetland mapping including the National Wetlands Inventory (NWI). Quality and consistency of
the Wetlands Layer is supported by federal wetlands mapping and classification standards, which were
developed under the oversight of the Federal Geographic Data Committee (FGDC) with input by
the FGDC Wetlands Subcommittee. This dataset is part of the FGDC Water-Inland Theme, which is co-
chaired by the FWS and the U.S. Geological Survey. The EPA's Safe Drinking Water Information System
(SDWIS) tracks information on drinking water contamination levels as required by the 1974 Safe Drinking
Water Act and its 1986 and 1996 amendments.
General QA/QC procedures were applied to activity data, documentation, and emission calculations
consistent with the U.S. Inventory QA/QC plan, which is in accordance with Vol. 1 Chapter 6 of the 2006
IPCC Guidelines (see Annex 8 for more details). All calculations were executed independently in Excel
and R. Ten percent of state and national totals were randomly selected for comparison between the two
approaches to ensure there were no computational errors.
Recalculations Discussion
The National Inventory of Dams (NID) data are updated regularly. The version of NID used for the current
Inventory contains 610 new dams relative to the previous (1990 through 2022) Inventory data. Similarly,
the National Wetlands Inventory (NWI) is periodically updated. The NWI version used for this Inventory
incorporated recent New Mexico, Ohio, Maryland, Michigan, and West Virginia feature updates
compared to the previous Inventory.
The net effect of these recalculations was an average annual decrease in CH4 emission estimates from
reservoirs of 2.26 MMT C02 Eq., or 7.5 percent, over the time series from 1990 to 2022 compared to the
previous Inventory.
Planned Improvements
The following are ongoing planned improvements which will be incorporated into a future Inventory:
• Developing country-specific emission factors for U.S. reservoirs based on a recently completed
EPA survey of greenhouse gas emissions from 108 reservoirs in the conterminous United
States.67 This improvement is expected for the next Inventory (1990 through 2024, publishing in
2026).
67 See https://www.epa.gov/air-research/research-emissions-us-reservoirs.
Land Use, Land-Use Change, and Forestry 6-159
-------
• Distinguishing inland wetlands from reservoirs and other types of flooded lands. See the
Planned Improvement chapter section of 6.1 Representation of the U.S. Land Base for additional
information. This is a long-term improvement but efforts are underway.
Emissions from Flooded Land Remaining Flooded Land-
Other Constructed Waterbodies
The IPCC (IPCC 2019) provides emission factors for several types of "other constructed waterbodies"
including freshwater ponds and canals/ditches. IPCC (2019) describes ponds as waterbodies that are
"...constructed by excavation and/or construction of walls to hold water in the landscape for a range of
uses, including agricultural water storage, access to water for livestock, recreation, and aquaculture."
Furthermore, the IPCC "Decision tree for types of Flooded Land" (IPCC 2019, Fig. 7.2) defines a size
threshold of 8 ha to distinguish reservoirs from "other constructed waterbodies." For this Inventory,
ponds are defined as managed flooded land that are 1) less than 8 ha in surface area, and 2) not
categorized as canals/ditches. IPCC (2019) further distinguishes saline versus brackish ponds, with the
former supporting lower CH4 emissions than the latter. Activity data on pond salinity are not uniformly
available for the conterminous United States and all ponds in the inventory are assumed to be
freshwater. Ponds often receive high organic matter and nutrient loadings, may have low oxygen levels,
and are often sites of substantial CH4 emissions from anaerobic sediments.
Canals and ditches (terms are used interchangeably) are linear water features constructed to transport
water (i.e., stormwater drainage, aqueduct), to irrigate or drain land, to connect two or more bodies of
water, or to serve as a waterway for watercraft. The geometry and construction of canals and ditches
varies widely and includes narrow earthen channels (<1 m wide) and concrete lined aqueducts in excess
of 50 m wide. Canals and ditches are typically angular, linear, follow man-made features such as roads,
powerlines, and/or agricultural parcels, and much less sinuous than natural riverine systems. Canals
and ditches can be extensive in many agricultural, forest and settlement areas, and may also be
significant sources of emissions in some circumstances.
Methane emissions from freshwater ponds in flooded land remaining flooded land increased by
approximately 1.4 percent from 1990 to 2023. Methane emissions from canals and ditches have
remained constant throughout the time series because age data are not available for canals and
ditches, thus they are assumed to be greater than 20-years old in 1990 and are included in flooded land
remaining flooded land throughout the time series. Overall, CH4 emissions from other constructed
waterbodies have remained fairly constant since 1990 (Table 6-88 and Table 6-89).
Table 6-88: CH4 Emissions from Other Constructed Waterbodies in Flooded Land
Remaining Flooded Land (MMT C02 Eq.)
Source
1990
2005
2019
2020
2021
2022
2023
Other Constructed Waterbodies
Canals and Ditches
7.2
7.2
7.2
7.2
7.2
7.2
7.2
Freshwater Ponds
10.3
10.4
10.4
10.4
10.4
10.4
10.4
Total
17.5
17.6
17.6
17.6
17.6
17.6
17.6
Note: Totals may not sum due to independent rounding.
6-160 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Table 6-89: CH4 Emissions from Other Constructed Waterbodies in Flooded Land
Remaining Flooded Land (kt CH4)
Source
1990
2005
2019
2020
2021
2022
2023
Other Constructed Waterbodies
Canals and Ditches
258.4
258.4
258.4
258.4
258.4
258.4
258.4
Freshwater Ponds
366.8
371.0
371.7
371.7
371.8
371.8
371.9
Total
625.2
629.4
630.1
630.2
630.2
630.2
630.3
Note: Totals may not sum due to independent rounding.
Florida has the greatest methane emissions from canals and ditches in the United States (Figure 6-13,
Table 6-90), with the other Gulf Coast states of Texas and Louisiana making significant contributions.
The Caloosahatchee River and Lake Okeechobee drainage areas in Florida were brought into federal
management by the Central and Southern Florida Project of 1948. This plexus of drainage canals and the
ongoing construction of new flood control ditches represent some of the densest network of ditches and
canals in the country. California has the second greatest methane emissions from canals and ditches.
Canals and ditches in California primarily serve to convey water from the mountains to urban and
agricultural areas. Texas, Florida, and Georgia have the greatest methane emissions from freshwater
ponds, although states throughout the eastern United States make significant contributions to the
national total. These patterns of emissions are in accordance with the distribution of other constructed
waterbodies in the United States.
Table 6-90: CH4 Emissions from Other Constructed Waterbodies in Flooded Land
Remaining Flooded Land in 2023 (kt CH4)
State Canals and Ditches Freshwater Ponds Total
Alabama + 10.4 10.8
Alaska + + +
Arizona 1.8 0.6 2.3
Arkansas 6.3 8.1 14.4
California 21.6 6.4 28.0
Colorado 6.9 3.2 10.1
Connecticut + 1.7 1.8
Delaware 1.0 0.8 1.8
District of Columbia + + +
Florida 50.4 27.0 77.4
Georgia 2.6 20.6 23.2
Hawaii + + 0.7
Idaho 7.5 1.2 8.7
Illinois 8.1 10.4 18.5
Indiana 8.7 9.6 18.3
Iowa 10.1 8.9 19.0
Kansas 1.8 14.8 16.6
Kentucky 0.6 7.5 8.1
Louisiana 12.2 4.4 16.6
Maine 0.6 3.3 3.9
Maryland + 3.7 4.0
Land Use, Land-Use Change, and Forestry 6-161
-------
State
Canals and Ditches
Freshwater Ponds
Total
Massachusetts 1.3 2.0 3.3
Michigan
5.6
8.6
14.2
Minnesota
14.8
10.1
24.8
Mississippi
5.1
12.6
17.7
Missouri
6.6
19.8
26.4
Montana
6.4
8.8
15.2
Nebraska
7.7
7.3
14.9
Nevada
1.8
+
2.1
New Hampshire
+
1.0
1.1
New Jersey
2.1
2.5
4.5
New Mexico
+
2.1
2.4
New York
3.7
7.8
11.4
North Carolina
6.2
11.2
17.4
North Dakota
4.3
19.3
23.6
Ohio
1.8
9.3
11.1
Oklahoma
0.9
19.0
19.9
Oregon
5.3
2.3
7.6
Pennsylvania
+
4.0
4.2
Puerto Rico
+
+
+
Rhode Island
+
+
+
South Carolina
3.4
9.6
13.0
South Dakota
4.7
14.8
19.5
Tennessee
2.3
6.3
8.6
Texas
17.5
30.1
47.6
Utah
3.8
1.3
5.1
Vermont
+
0.7
1.1
Virginia
1.3
7.0
8.3
Washington
3.3
1.5
4.8
West Virginia
+
2.0
2.0
Wisconsin
0.7
3.6
4.4
Wyoming
5.0
3.6
8.6
Total
258.4
371.9
630.3
+ Indicates values less than 0.5 kt.
Note: Totals may not sum due to independent rounding.
6-162 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Figure 6-13: 2023 CH4 Emissions from A) Ditches and Canals and B) Freshwater Ponds
in Flooded Land Remaining Flooded Land (kt CH4)
A. CH4 Emissions from Ditches and Canals
B. CH4 Emissions from Freshwater Ponds
kt CH4 y
¦
30
-
-
20
_
15
—
10
—
5
—
0
Note: Alaska map scale is 1000 miles, Hawaii arid contiguous United States map scale is 500 miles, Puerto Rico map scale is
100 miles.
Land Use, Land-Use Change, and Forestry 6-163
-------
Methodology and Time-Series Consistency
Estimates of CH4 emissions for other constructed waterbodies in flooded land remaining flooded Land
follow the Tier 1 methodology in IPCC (2019). All calculations are performed at the state level and
summed to obtain national estimates. Based on IPCC guidance, methane emissions from the surface of
these flooded lands are calculated as the product of flooded land surface area and an emission factor
(Table 6-91). Although literature data on greenhouse gas emissions from canals and ditches is relatively
sparse, they have the highest default emission factor of all flooded land types (Table 6-91). Default
emission factors for freshwater ponds are on the higher end of those for reservoirs. There are insufficient
data to support climate-specific emission factors for ponds or canals and ditches. Downstream
emissions are not inventoried for other constructed waterbodies because 1) many of these systems are
not associated with dams (e.g., excavated ponds and ditches), and 2) there are insufficient data to
derive downstream emission factors for other constructed waterbodies that are associated with dams
(IPCC 2019).
Table 6-91: IPCC (2019) Default CH4 Emission Factors for Surface Emissions from
Other Constructed Waterbodies in Flooded Land Remaining Flooded Land
Other Constructed Waterbody
Surface emission factor (MT CH4 ha-1 y1)
Freshwater ponds
0.183
Canals and ditches
0.416
Area Estimates
Other constructed waterbodies were identified from the "NHDWaterbody" layer in the National
Hydrography Dataset Plus V2 (NHD),68 the National Inventory of Dams (NID),69 the National Wetlands
Inventory (NWI),70 the Navigable Waterways (NW) network,71 the Hydropower Infrastructure-Lakes,
Reservoirs and Rivers (HILARRI) database,72 and the EPA's Safe Drinking Water Information System
(SDWIS).73 The NHD only covers the conterminous United States, whereas the NID, NW and NWI also
include Alaska, Hawaii, District of Columbia, and Puerto Rico. The following paragraphs present the
criteria used to identify other constructed waterbodies in the NHD, NW, and NWI.
Waterbodies in the "NHDWaterbody" layer that were greater than 20-years old, less than 8 ha in surface
area, not identified as canal/ditch in NHD, and met any of the following criteria were considered
freshwater ponds in flooded land remaining flooded land: 1) the waterbody was classified "Reservoir" in
the "NHDWaterbody" layer, 2) the waterbody name in the "NHDWaterbody" layer included "Reservoir",
3) the waterbody in the "NHDWaterbody" layer was located in close proximity (up to 200 m) to a dam in
the NID, 4) the "NHDWaterbody" GNIS name was similar to a nearby NID feature (up to a 1000 m radius),
5) the waterbody intersected a drinking water intake, or 6) the "NHDWaterbody" intersected an
operational hydropower dam defined by the HILARRI dataset.
68 See https://www.usgs.gov/core-science-systems/ngp/national-hydrography.
69 See https://nid.sec.usace.army.mil.
70 See https://www.fws.gov/program/national-wetlands-inventorv/data-download.
71 See https://geodata.bts.gov/datasets/usdot::navigab[e-waterwav-network-[ines/exp[ore.
72 See https://www.osti.gov/biblio/1960141.
73 See https://www.epa.gov/enviro/sdwis-overview. Not publicly available due to security concerns.
6-164 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
EPA assumes that all features included in the NW are subject to water-level management to maintain
minimum water depths required for navigation and are therefore managed flooded lands. NW features
that were less than 8 ha in surface area and not identified as canals/ditch (see below) were considered
freshwater ponds.
NWI features were considered "managed" if they had a special modifier value indicating the presence of
management activities (Figure 6-12). To be included in the flooded lands inventory, the managed flooded
land had to be wet or saturated for at least one season per year (see "Water Regime" in Figure 6-12). NWI
features that met these criteria, were less than 8 ha in surface area, and were not a canal/ditch (see
below) were defined as freshwater ponds.
Any NWI or NHD feature that intersected a drinking water intake point from SDWIS was assumed to be
"managed "The rational being that a waterbody used as a source for public drinking water is typically
managed in some capacity - by flow and/or volume control.
Canals and ditches, a subset of other constructed waterbodies, were identified in the NWI by their
attribute codes.74 The following NWI attribute codes were interpreted as canals/ditches due to their
angular, linear morphology: R5UBFx, R2UBFx, R2UBHx, R4SBCx, PEMICx.
Surface areas for other constructed waterbodies were taken from NHD, NWI or the NW. If features from
the NHD, NWI, or the NWdatasets overlapped, these areas were erased. The first step was to take the
final NWI flooded lands features and use it to identify overlapping NHD features. If the NHD feature had
its center in a NWI feature, it was removed from analysis. Next, remaining NHD features were erased
from any remaining overlapping NWI features. Final selections of NHD and NWI features were used to
erase any overlapping NW waterbodies.
The age of other constructed waterbody features was determined by assuming the waterbody was
created the same year as a nearby (up to 200 m) NID feature. If no nearby NID feature was identified, it
was assumed the waterbody was greater than 20-years old throughout the time series. No canal/ditch
features were associated with a nearby dam, therefore all canal/ditch features were assumed to be
greater than 20-years old throughout the time series.
For the year 2023, this Inventory contains 2,031,716 ha of freshwater ponds and 621,220 ha of canals
and ditches in flooded land remaining flooded land. The surface area of freshwater ponds increased by
27,305 (1.3%) from 1990 to 2023 due to flooded lands matriculating from land converted to flooded land
to flooded land remaining flooded land. All canals and ditches were assumed to be greater than 20-
years old throughout the time series, thus the surface area of these flooded lands is constant
throughout the time series.
Table 6-92: National Surface Area Totals in Flooded Land Remaining Flooded Land -
Other Constructed Waterbodies (hectares)
1990 2005
2019
2020
2021
2022
2023
Canals and ditches
621,220 I
621,220 I
621,220
621,220
621,220
621,220
621,220
Freshwater ponds
2,004,411
| 2,027,241
| 2,031,084
2,031,395
2,031,606
2,031,716
2,031,987
Total
2,625,631
2,648,461
2,652,304
2,652,615
2,652,827
2,652,936
2,653,208
Note: Totals may not sum due to independent rounding.
74 See https://www.fws.gov/program/national-wetlands-inventory/classification-codes.
Land Use, Land-Use Change, and Forestry 6-165
-------
Canals arid ditches in the conterminous United States are most abundant in the Gulf Coast states and
California (Figure 6-14, Table 6-93). Florida contains 19.5 percent of all U.S. canal and ditch surface
area, most of which were constructed in the early 1900s for drainage, flood protection, and water
storage purposes. Texas has the greatest surface area of freshwater ponds, equivalent to 8,1 percent of
all freshwater pond surface area in the United States, closely followed by Florida.
Figure 6-14: 2023 Surface Area of A) Ditches and Canals and B) Freshwater Ponds in
Flooded Land Remaining Flooded Land (hectares)
A. Area of Ditches and Canals
hectares
B 100000
50000
o
Alaska
Hawaii
1000 mi
¦¦ft
500 mi
100 mi
500 mi
B. Area of Freshwater Ponds
Note: Alaska map scale is 1000 miles, Hawaii and contiguous United States map scale is 500 miles, Puerto Rico map scale is
100 miles.
6-166 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Table 6-93: State Totals of Surface Area in Flooded Land Remaining Flooded Land—
Canals and Ditches (hectares)
State
1990
2005
2019
2020
2021
2022
2023
Alabama
1,125
1,125
1,125
1,125
1,125
1,125
1,125
Alaska
217
217
217
217
217
217
217
Arizona
4,221
4,221
4,221
4,221
4,221
4,221
4,221
Arkansas
15,155
15,155
15,155
15,155
15,155
15,155
15,155
California
51,834
51,834
51,834
51,834
51,834
51,834
51,834
Colorado
16,694
16,694
16,694
16,694
16,694
16,694
16,694
Connecticut
249
249
249
249
249
249
249
Delaware
2,405
2,405
2,405
2,405
2,405
2,405
2,405
District of Columbia
4
4
4
4
4
4
4
Florida
121,192
121,192
121,192
121,192
121,192
121,192
121,192
Georgia
6,175
6,175
6,175
6,175
6,175
6,175
6,175
Hawaii
1,170
1,170
1,170
1,170
1,170
1,170
1,170
Idaho
18,080
18,080
18,080
18,080
18,080
18,080
18,080
Illinois
19,394
19,394
19,394
19,394
19,394
19,394
19,394
Indiana
21,026
21,026
21,026
21,026
21,026
21,026
21,026
Iowa
24,174
24,174
24,174
24,174
24,174
24,174
24,174
Kansas
4,397
4,397
4,397
4,397
4,397
4,397
4,397
Kentucky
1,506
1,506
1,506
1,506
1,506
1,506
1,506
Louisiana
29,310
29,310
29,310
29,310
29,310
29,310
29,310
Maine
1,422
1,422
1,422
1,422
1,422
1,422
1,422
Maryland
641
641
641
641
641
641
641
Massachusetts
3,239
3,239
3,239
3,239
3,239
3,239
3,239
Michigan
13,361
13,361
13,361
13,361
13,361
13,361
13,361
Minnesota
35,480
35,480
35,480
35,480
35,480
35,480
35,480
Mississippi
12,196
12,196
12,196
12,196
12,196
12,196
12,196
Missouri
15,804
15,804
15,804
15,804
15,804
15,804
15,804
Montana
15,453
15,453
15,453
15,453
15,453
15,453
15,453
Nebraska
18,429
18,429
18,429
18,429
18,429
18,429
18,429
Nevada
4,324
4,324
4,324
4,324
4,324
4,324
4,324
New Hampshire
462
462
462
462
462
462
462
New Jersey
4,936
4,936
4,936
4,936
4,936
4,936
4,936
New Mexico
750
750
750
750
750
750
750
New York
8,809
8,809
8,809
8,809
8,809
8,809
8,809
North Carolina
14,873
14,873
14,873
14,873
14,873
14,873
14,873
North Dakota
10,230
10,230
10,230
10,230
10,230
10,230
10,230
Ohio
4,282
4,282
4,282
4,282
4,282
4,282
4,282
Oklahoma
2,068
2,068
2,068
2,068
2,068
2,068
2,068
Oregon
12,753
12,753
12,753
12,753
12,753
12,753
12,753
Pennsylvania
393
393
393
393
393
393
393
Puerto Rico
656
656
656
656
656
656
656
Rhode Island
6
6
6
6
6
6
6
South Carolina
8,064
8,064
8,064
8,064
8,064
8,064
8,064
Land Use, Land-Use Change, and Forestry 6-167
-------
State
1990
2005
2019
2020
2021
2022
2023
South Dakota
11,402
11,402
11,402
11,402
11,402
11,402
11,402
Tennessee
5,494
5,494
5,494
5,494
5,494
5,494
5,494
Texas
41,969
41,969
41,969
41,969
41,969
41,969
41,969
Utah
9,196
9,196
9,196
9,196
9,196
9,196
9,196
Vermont
1,120
1,120
1,120
1,120
1,120
1,120
1,120
Virginia
3,138
3,138
3,138
3,138
3,138
3,138
3,138
Washington
8,010
8,010
8,010
8,010
8,010
8,010
8,010
West Virginia
40
40
40
40
40
40
40
Wisconsin
1,779
1,779
1,779
1,779
1,779
1,779
1,779
Wyoming
12,110
12,110
12,110
12,110
12,110
12,110
12,110
Total
621,220
621,220
621,220
621,220
621,220
621,220
621,220
Note: Totals may not sum due to independent rounding.
Table 6-94: State Totals of Surface Area in Flooded Land Remaining Flooded Land—
Freshwater Ponds (hectares)
State
1990
2005
2019
2020
2021
2022
2023
Alabama
56,255
56,619
56,643
56,643
56,643
56,643
56,643
Alaska
2,192
2,199
2,199
2,199
2,199
2,199
2,199
Arizona
3,016
3,056
3,066
3,069
3,069
3,069
3,069
Arkansas
43,706
44,041
44,043
44,043
44,043
44,043
44,043
California
34,981
35,155
35,227
35,230
35,230
35,236
35,242
Colorado
17,161
17,380
17,412
17,412
17,412
17,412
17,415
Connecticut
9,464
9,534
9,539
9,539
9,539
9,539
9,539
Delaware
4,099
4,102
4,102
4,102
4,102
4,102
4,102
District of Columbia
11
11
11
11
11
11
11
Florida
147,166
147,252
147,283
147,283
147,288
147,290
147,290
Georgia
110,800
112,443
112,521
112,521
112,521
112,521
112,521
Hawaii
921
929
931
931
931
931
931
Idaho
6,293
6,394
6,394
6,394
6,395
6,395
6,395
Illinois
56,236
56,752
56,834
56,844
56,845
56,849
56,850
Indiana
51,757
52,209
52,286
52,286
52,286
52,295
52,302
Iowa
45,414
47,366
48,558
48,676
48,770
48,800
48,809
Kansas
78,258
80,482
80,576
80,581
80,603
80,604
80,618
Kentucky
40,608
40,967
41,001
41,001
41,001
41,001
41,001
Louisiana
24,017
24,137
24,147
24,153
24,153
24,153
24,154
Maine
18,046
18,070
18,079
18,079
18,079
18,079
18,079
Maryland
20,045
20,214
20,272
20,274
20,276
20,277
20,277
Massachusetts
10,733
10,776
10,820
10,826
10,827
10,831
10,833
Michigan
46,821
46,960
47,004
47,011
47,011
47,011
47,011
Minnesota
54,666
54,905
54,964
54,982
54,991
54,999
54,999
Mississippi
68,315
68,638
68,753
68,765
68,771
68,779
68,789
Missouri
104,956
108,267
108,395
108,399
108,406
108,411
108,417
Montana
47,596
47,942
47,962
47,963
47,963
47,963
48,049
Nebraska
38,290
39,492
39,701
39,720
39,729
39,733
39,742
6-168 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
State
1990
2005
2019
2020
2021
2022
2023
Nevada
1,833
1,836
1,872
1,875
1,880
1,883
1,887
New Hampshire
5,084
5,184
5,228
5,228
5,229
5,229
5,229
New Jersey
13,577
13,603
13,617
13,617
13,617
13,617
13,617
New Mexico
11,202
11,226
11,251
11,251
11,254
11,258
11,260
New York
42,126
42,381
42,494
42,494
42,499
42,499
42,510
North Carolina
60,787
61,231
61,312
61,315
61,317
61,318
61,318
North Dakota
105,261
105,382
105,492
105,502
105,510
105,513
105,524
Ohio
50,221
50,562
50,757
50,787
50,799
50,801
50,818
Oklahoma
101,972
103,867
103,944
103,960
103,966
103,974
103,977
Oregon
12,490
12,643
12,685
12,690
12,690
12,690
12,690
Pennsylvania
21,686
21,917
21,955
21,955
21,956
21,956
21,956
Puerto Rico
406
406
406
406
406
406
406
Rhode Island
2,198
2,206
2,213
2,213
2,213
2,213
2,213
South Carolina
51,627
52,222
52,444
52,455
52,455
52,455
52,455
South Dakota
80,332
80,600
80,678
80,679
80,687
80,687
80,702
Tennessee
33,954
34,346
34,384
34,390
34,390
34,391
34,391
Texas
161,800
164,486
164,579
164,580
164,581
164,581
164,588
Utah
6,898
6,983
6,989
6,994
6,994
6,994
7,004
Vermont
3,509
3,576
3,587
3,587
3,587
3,587
3,587
Virginia
38,292
38,350
38,354
38,354
38,354
38,354
38,354
Washington
7,943
8,071
8,110
8,113
8,115
8,116
8,117
West Virginia
10,738
10,853
10,887
10,887
10,887
10,887
10,887
Wisconsin
19,591
19,738
19,747
19,747
19,747
19,747
19,747
Wyoming
19,059
19,280
19,375
19,377
19,379
19,383
19,421
Total
2,004,411
2,027,241
2,031,084
2,031,395
2,031,606
2,031,716
2,031,987
Note: Totals may not sum due to independent rounding.
Uncertainty
Uncertainty in estimates of CH4 emissions from other constructed waterbodies (ponds, canals/ditches)
in flooded land remaining flooded land (Table 6-95) are estimated using IPCC Approach 2 and include
uncertainty in the default emission factors and the flooded land area inventory. Uncertainty in default
emission factors is provided in the 2019 Refinement to the 2006 IPCC Guidelines (IPCC 2019).
Uncertainties in the spatial data include 1) uncertainty in area estimates from the NHD, NWI, and NW,
and 2) uncertainty in the location of dams in the NID. Overall uncertainties in these spatial datasets are
unknown, but uncertainty for remote sensing products is assumed to be ± 10 to 15 percent based on
IPCC guidance (IPCC 2003). An uncertainty range of ± 15 percent for the flooded land area estimates is
assumed and is based on expert judgment.
Land Use, Land-Use Change, and Forestry 6-169
-------
Table 6-95: Approach 2 Quantitative Uncertainty Estimates for CH4 Emissions from
Other Constructed Waterbodies in Flooded Land Remaining Flooded Land
Uncertainty Range Relative to Emission Estimate"
2023 Emission
(MMT CO2 Eq.)
(%)
Estimate
Lower
Upper
Lower
Upper
Source
Gas
(MMT CO2 Eq.)
Bound
Bound
Bound
Bound
Canals and ditches
CH4
7.2
6.8
7.6
-6.0%
+5.5%
Freshwater pond
ch4
10.4
10.4
10.4
-0.04%
+0.04%
Total
ch4
17.6
17.2
17.9
-1.4%
+1.2%
a Range of emission estimates predicted by Monte Carlo stochastic simulation for a 95 percent confidence interval.
QA/QC and Verification
The National Hydrography Data (NHD) is managed by the USGS in collaboration many other federal,
state, and local entities. Extensive QA/QC procedures are incorporated into the curation of the NHD. The
National Inventory of Dams (NID) is maintained by the U.S. Army Corps of Engineers (USACE) in
collaboration with the Federal Emergency Management Agency (FEMA) and state regulatory offices.
USACE resolves duplicative and conflicting data from 68 data sources, which helps obtain the more
complete, accurate, and updated NID.75 The Navigable Waterways (NW) dataset is part of the U.S.
Department of Transportation (USDOT)/Bureau of Transportation Statistics (BTS) National Transportation
Atlas Database (NTAD). The NW is a comprehensive network database of the nation's navigable
waterways updated on a continuing basis. U.S. Fish and Wildlife Service is the principal agency in
charge of wetland mapping including the National Wetlands Inventory (NWI). Quality and consistency of
the Wetlands Layer is supported by federal wetlands mapping and classification standards, which were
developed under the oversight of the Federal Geographic Data Committee (FGDC) with input by
the FGDC Wetlands Subcommittee. This dataset is part of the FGDC Water-Inland Theme, which is co-
chaired by the FWS and the U.S. Geological Survey. The EPA's Safe Drinking Water Information System
(SDWIS) tracks information on drinking water contamination levels as required by the 1974 Safe Drinking
Water Act and its 1986 and 1996 amendments.
General QA/QC procedures were applied to activity data, documentation, and emission calculations
consistent with the U.S. Inventory QA/QC plan, which is in accordance with Vol. 1 Chapter 6 of 2006
IPCC Guidelines (see Annex 8 for more details). All calculations were executed independently in Excel
and R. Ten percent of state and national totals were randomly selected for comparison between the two
approaches to ensure there were no computational errors.
Recalculations Discussion
The National Inventory of Dams (NID) data are updated regularly. The version of NID used for the current
Inventory contains 610 new dams relative to the previous (1990 through 2022) Inventory data. Similarly,
the National Wetlands Inventory (NWI) is periodically updated. The NWI version used for this Inventory
incorporated recent New Mexico, Ohio, Maryland, Michigan, and West Virginia feature updates
compared to the previous Inventory.
75 See https://www.epa.gov/national-aqijatic-resoijrce-sijrveys/national-lakes-assessment-9017-qijality-assurance-
project-plan.
6-170 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
The net effect of these recalculations was an average annual increase in CH4 emission estimates from
other constructed waterbodies of 3.85 MMT C02 Eq., or 28 percent, over the time series from 1990 to
2022 compared to the previous Inventory.
Planned Improvements
The following are ongoing planned improvements which will be incorporated into a future Inventory:
• Conducting a literature review to determine if sufficient data are available to derive a country-
specific emission factorforthe next Inventory (i.e., publishing in 2026). Default emission factors
for canals/ditches were derived from a global dataset that include few measurements from U.S.
systems.
• Conducting a survey, led by EPA, of GHG emissions from U.S. ponds from 2024 through 2026 to
assess the accuracy of IPCC default emission factors. The IPCC default emission factors for
ponds were derived from a global dataset that include few measurements from U.S. systems.
The results of this survey will determine if default or country specific EFs will be used for future
inventories. This is an ongoing improvement but will take multiple years to fully implement.
• Distinguishing inland wetlands from ponds and other types of flooded lands. See the Planned
Improvement chapter section of 6.1 Representation of the U.S. Land Base for additional
information. This is a long-term improvement but efforts are underway.
6.9 Land Converted to Wetlands (Source
Category 4D2)
Emissions and Removals from Land Converted to
Vegetated Coastal Wetlands
Conversion to coastal wetlands resulted in a biomass carbon stock loss of 0.12 MMT C02 Eq. (0.03 MMT
C) in 2023 (Table 6-96 and Table 6-97). Loss of forest biomass through conversion of forest lands to
vegetated coastal wetlands is the primary driver behind biomass carbon stock change being a source
rather than a sink across the time series. Conversion of cropland, grassland, settlement and other lands
result in a net increase in biomass stocks. Conversion of lands to vegetated coastal wetlands resulted in
a DOM loss of 0.03 MMTC02 Eq. (0.008 MMTC) in 2023 (Table 6-96 and Table 6-97), which is driven by
the loss of DOM when forest land is converted to vegetated coastal wetlands. This is likely an
overestimate of loss because wetlands inherently preserve dead organic material. Conversion of
cropland, grassland, settlement and other land results in a net increase in DOM. Across all time periods,
soil carbon accumulation resulting from lands converted to vegetated coastal wetlands is a carbon sink
and has ranged between -0.25 and -0.13 MMTC02 Eq. (-0.04 and -0.07 MMTC; Table 6-96 and Table
6-97). Conversion of lands to coastal wetlands resulted in CH4 emissions of 0.16 MMT C02 Eq. (5.8 kt
CH4) in 2023 (Table 6-98). Methane emissions due to the conversion of lands to vegetated coastal
wetlands are largely the result of forest land converting to palustrine emergent and scrub shrub coastal
Land Use, Land-Use Change, and Forestry 6-171
-------
wetlands in warm temperate climates. Methane emissions were the highest between 1990 and 2002
(0.28 MMT C02 Eq., 10.0 kt CH4) and have continually decreased to current levels. This decrease was
driven by a reduction in the rate of conversion of forest land to palustrine scrub-shrubs and emergent
wetlands.
Table 6-96: Net C02 Flux from Carbon Stock Changes in Land Converted to Vegetated
Coastal Wetlands (MMT C02 Eq.)
Land Use/Carbon Pool
1990
2005
2019
2020
2021
2022
2023
Cropland Converted to Vegetated Coastal Wetlands
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Biomass C Stock
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Soil C Stock
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Forest Land Converted to Vegetated Coastal Wetlands
0.49
0.50
0.01
0.02
0.03
0.04
0.05
Biomass C Stock
0.62
0.62
0.13
0.13
0.13
0.13
0.13
Dead Organic Matter C Flux
0.11
0.12
0.03
0.03
0.03
0.03
0.03
Soil C Stock
(0.24)
(0.24)
(0.15)
(0.14)
(0.13)
(0.12)
(0.11)
Grassland Converted to Vegetated Coastal Wetlands
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Biomass C Stock
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Soil C Stock
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Other Land Converted to Vegetated Coastal Wetlands
(0.03)
(0.03)
(0.02)
(0.02)
(0.02)
(0.02)
(0.02)
Biomass C Stock
(0.01)
(0.02)
(0.01)
(0.01)
(0.01)
(0.01)
(0.01)
Soil C Stock
(0.02)
(0.02)
(0.02)
(0.02)
(0.02)
(0.02)
(0.02)
Settlements Converted to Vegetated Coastal Wetlands
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Biomass C Stock
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Soil C Stock
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Total Biomass Flux
0.60
0.60
0.12
0.12
0.12
0.12
0.12
Total Dead Organic Matter Flux
0.11
0.12
0.03
0.03
0.03
0.03
0.03
Total Soil C Flux
(0.25)
(0.25)
(0.17)
(0.16)
(0.15)
(0.14)
(0.13)
Total Flux
0.46
0.47
(0.02)
(0.01)
(0.00)
0.01
0.02
+ Absolute value does not exceed 0.005 MMT C02 Eq.
Notes: Totals may not sum due to independent rounding. Parentheses indicate net sequestration.
Table 6-97: Net C02 Flux from Carbon Stock Changes in Land Converted to Vegetated
Coastal Wetlands (MMT C)
Land Use/Carbon Pool
1990
2005
2019
2020
2021
2022
2023
Cropland Converted to Vegetated Coastal Wetlands
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Biomass C Stock
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Soil C Stock
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Forest Land Converted to Vegetated Coastal Wetlands
0.13
0.14
+
0.1
0.01
0.01
0.01
Biomass C Stock
0.17
0.17
0.04
0.04
0.04
0.04
0.04
Dead Organic Matter C Flux
0.03
0.03
0.01
0.01
0.01
0.01
0.01
Soil C Stock
(0.06)
(0.06)
(0.04)
(0.04)
(0.04)
(0.03)
(0.03)
Grassland Converted to Vegetated Coastal Wetlands
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Biomass C Stock
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Soil C Stock
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Other Land Converted to Vegetated Coastal Wetlands
(0.01)
(0.01)
(0.01)
(0.01)
(0.01)
(0.01)
(0.01)
6-172 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Land Use/Carbon Pool
1990
2005
2019
2020
2021
2022
2023
Biomass C Stock
(+)
(0.01)
{+)
{+)
{+)
{+)
{+)
Soil C Stock
(+)
(0.01)
{+)
{+)
{+)
{+)
{+)
Settlements Converted to Vegetated Coastal Wetlands
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Biomass C Stock
(+)
{+)
{+)
{+)
{+)
{+)
{+)
Soil C Stock
(+)
<+)
{+)
{+)
{+)
{+)
{+)
Total Biomass Flux
0.16
I 0.16
0.03
0.03
0.03
0.03
0.03
Total Dead Organic Matter Flux
0.03
0.03
0.01
0.01
0.01
0.01
0.01
Total Soil C Flux
(0.07)
(0.07)
(0.05)
(0.04
(0.04)
(0.04)
(0.03)
Total Flux
0.13
0.13
(+)
(+)
+
+
+
+ Absolute value does not exceed 0.005 MMT C.
Notes: Totals may not sum due to independent rounding. Parentheses indicate net sequestration.
Table 6-98: CH4 Emissions from Land Converted to Vegetated Coastal Wetlands (MMT
C02 Eq. and kt CH4)
Land Use/Carbon Pool
1990 |
2005
2019
2020
2021
2022
2023
Cropland Converted to Vegetated Coastal Wetlands
ChU Emissions (MMT CO2 Eq.)
+
+
+
+
+
+
+
CH4 Emissions (kt CH4)
+
0.01
0.04
0.05
0.05
0.05
0.05
Forest Land Converted to Vegetated Coastal Wetlands
CH4 Emissions (MMT CO2 Eq.)
0.28
0.27
0.18
0.17
0.16
0.15
0.14
CH4 Emissions (kt CH4)
9.88
9.74
6.48
6.10
5.76
5.41
5.07
Grassland Converted to Vegetated Coastal Wetlands
CH4 Emissions (MMT CO2 Eq.)
+
+
+
+
+
+
+
CH4 Emissions (kt CH4)
0.01
0.01
0.07
0.08
0.08
0.09
0.09
Other Land Converted to Vegetated Coastal Wetlands
CH4 Emissions (MMT CO2 Eq.)
+
1 +
0.01
0.01
0.01
0.02
0.02
CH4 Emissions (kt CH4)
0.08
0.14
0.47
0.50
0.52
0.54
0.56
Settlements Converted to Vegetated Coastal Wetlands
CH4 Emissions (MMT CO2 Eq.)
+
+
+
+
+
+
+
CH4 Emissions (kt CH4)
0.01
1 +
I +
+
+
+
+
Total CH4 Emissions (MMT CO2 Eq.)
0.28 |
0.28 |
0.20
0.190
0.18
0.17
0.16
Total CH4 Emissions (kt CH4)
9.98 |
9.91
7.06
6.73
6.41
6.09
5.78
+ Absolute value does not exceed 0.005 MMT C02 Eq. or 0.005 kt ChU.
Note: Totals may not sum due to independent rounding.
Methodology and Time-Series Consistency
The following section provides a description of the methodology used to estimate changes in biomass,
dead organic matter and soil carbon stocks and CH4emissions for land converted to vegetated coastal
wetlands. Methodological recalculations were applied to the entire time series to ensure time-series
consistency from 1990 through 2023.
Biomass Carbon Stock Changes
Biomass carbon stocks for land converted to vegetated coastal wetlands are estimated for palustrine
and estuarine marshes for land below the elevation of high tides (taken to be mean high water spring
Land Use, Land-Use Change, and Forestry 6-173
-------
tide elevation) and as far seawards as the extent of intertidal vascular plants within the U.S. land
representation according to the national LiDAR dataset, the national network of tide gauges and land
use histories recorded in the 1996, 2001, 2005, 2011, and 2016 NOAA C-CAP surveys (NOAA OCM
2020). Both federal and non-federal lands are represented. Delineating vegetated coastal wetlands from
ephemerally flooded upland grasslands represents a particular challenge in remote sensing. Moreover,
at the boundary between wetlands and uplands, which may be gradual on low lying coastlines, the
presence of wetlands may be ephemeral depending upon weather and climate cycles and as such,
impacts on the emissions and removals will vary over these timeframes. Trends in land cover change
are extrapolated to 1990 and 2023 from these datasets using the C-CAP change data closest in date to a
given year. Biomass is not sensitive to soil organic content. Aboveground biomass carbon stocks for
non-forested coastal wetlands are derived from a national assessment combining field plot data and
aboveground biomass mapping by remote sensing (Byrd et al. 2017; Byrd et al. 2018; Byrd et al. 2020).
Aboveground biomass carbon removal data for all subcategories are not available and thus
assumptions were applied using expert judgment about the most appropriate assignment to a
disaggregation of a community class. The aboveground biomass carbon stock for estuarine forested
wetlands (dwarf mangroves that are not classified as forests due to their stature) is derived from a meta-
analysis by Lu and Megonigal (201776). Root to shoot ratios from the Wetlands Supplement were used to
account for belowground biomass, which were multiplied by the aboveground carbon stock (IPCC 2014)
and summed with aboveground biomass to obtain total biomass carbon stocks. Aboveground biomass
carbon stocks for forest land, cropland, and grassland that are lost with the conversion to vegetated
coastal wetlands were derived fromTierl default values (IPCC 2006; IPCC 2019). Biomass carbon stock
changes are calculated by subtracting the biomass carbon stock values of each land use category (i.e.,
forest land, cropland, and grassland) from those of vegetated coastal wetlands in each climate zone and
multiplying that value by the corresponding C-CAP derived area gained that year in each climate zone.
The difference between the stocks is reported as the stock change under the assumption that the
change occurred in the year of the conversion. The total coastal wetland biomass carbon stock change
is accounted for during the year of conversion; therefore, no interannual changes are calculated during
the remaining years it is in the category.
Dead Organic Matter
Dead organic matter (DOM) carbon stocks, which include litter and dead wood stocks, are accounted
for in subtropical estuarine forested wetlands for lands converted to vegetated coastal wetlands across
all years. Tier 1 estimates of mangrove DOM carbon stocks were used for subtropical estuarine forested
wetlands (IPCC 2014). Neither Tier 1 or 2 data on DOM are currently available for either palustrine or
estuarine scrub/shrub wetlands for any climate zone or estuarine forested wetlands in climates other
than subtropical climates. Tier 1 DOM C stocks for forest land converted to vegetated coastal wetlands
were derived from IPCC (2019) to account for the loss of DOM that occurs with conversion. Changes in
DOM are assumed to be negligible for other land use conversions (i.e., other than forest land) to coastal
wetlands based on the Tier 1 method in IPCC (2006). Trends in land cover change are derived from the
NOAA C-CAP dataset and extrapolated to cover the entire 1990 through 2023 time series. Dead organic
matter removals are calculated by multiplying the C-CAP derived area gained that year by the difference
between Tier 1 DOM carbon stocks for vegetated coastal wetlands and forest land. The difference
between the stocks is reported as the stock change under the assumption that the change occurred in
76 See https://github.com/8mithsonian/Coastal-Wetland-NGGI-Data-Public: accessed October 2023.
6-174 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
the year of the conversion. The coastal wetland DOM stock is assumed to be in steady state once
established in the year of conversion; therefore, no interannual changes are calculated.
Soil Carbon Stock Changes
Soil carbon removals are estimated for land converted to vegetated coastal wetlands across all years.
Soil carbon stock changes, stratified by climate zones and wetland classes, are derived from a synthesis
of peer-reviewed literature77 (Lynch 1989; Orson et al. 1990; Kearny & Stevenson 1991; Thom et al. 1992;
Roman et al. 1997; Craft et al. 1998; Orson et al. 1998; Merrill 1999; Weis et al. 2001; Hussein et al.
2004; Church et al. 2006; Koster et al. 2007; Drexler et al. 2009; Boyd 2012; Callaway et al. 2012 a & b;
Bianchi et al. 2013; Drexler et al. 2013; Watson and Byrne 2013; Breithaupt et al. 2014; Crooks et al.
2014; Weston et al. 2014; Smith et al. 2015; Villa & Mitsch 2015; Boyd and Sommerfield 2016; Marchio
et al. 2016; Noe et al. 2016; Arriola and Cable 2017; Boyd et al. 2017; Gerlach et al. 2017; Giblin and
Forbrich 2018; Krauss et al. 2018; Abbott et al. 2019; Drexler et al. 2019; Poppe and Rybczyk 2019;
Ensign et al. 2020; Kemp et al. 2020; Lagomasino et al. 2020; Luk et al. 2020; McTigue et al. 2020; Peck
et al. 2020; Vaughn et al. 2020; Weston et al. 2020; Arias-Ortiz et al. 2021; Baustian et al. 2021; Allen et
al. 2022; Miller et al. 2022). To estimate soil carbon stock changes, no differentiation is made for soil
type (i.e., mineral, organic). Soil C removal data for all subcategories are not available and thus
assumptions were applied using expert judgment about the most appropriate assignment to a
disaggregation of a community class.
As per IPCC (2014) guidance, land converted to vegetated coastal wetlands is assumed to remain in this
category for up to 20 years before transitioning to vegetated coastal wetlands remaining vegetated
coastal wetlands. Tier 2 level estimates of soil carbon stock changes associated with annual soil carbon
accumulation from land converted to vegetated coastal wetlands were developed using country-
specific soil carbon removal factors multiplied by activity data of land area for land converted to
vegetated coastal wetlands for a given year in addition to the previous 19-year cumulative area.
Guidance from the Wetlands Supplement allows for the rate of soil carbon accumulation to be
instantaneously equivalent to that in natural settings and that soil carbon accumulation is initiated
when natural vegetation becomes established; this is assumed to occur in the first year of conversion.
No loss of soil carbon as a result of land conversion to coastal wetlands is assumed to occur. Since the
C-CAP coastal wetland area dataset begins in 1996, the area converted prior to 1996 is assumed to be
the same as in 1996. Similarly, the coastal wetland area data for 2017 through 2023 is assumed to be
the same as in 2016. The methodology follows Eq. 4.7, Chapter 4 of the IPCC Wetlands Supplement
(IPCC 2014) and is applied to the area of land converted to vegetated coastal wetlands on an annual
basis.
Soil Methane Emissions
Tier 1 estimates of CH4 emissions for land converted to vegetated coastal wetlands are derived from the
same wetland map used in the analysis of wetland soil carbon fluxes for palustrine wetlands, and are
produced from C-CAP, LiDAR and tidal data, in combination with default CH4 emission factors provided
in Table 4.14 of the IPCC Wetlands Supplement. The methodology follows Eq. 4.9, Chapter 4 of the IPCC
Wetlands Supplement and a global warming potential of 28 is used (IPCC 2013). Because land
converted to vegetated coastal wetlands is held in this category for up to 20 years before transitioning to
77 Coastal Carbon Network (2023). Database: Coastal Carbon Library (Version 1.0.0). Smithsonian Environmental
Research Center. Dataset. https://doi.org/10.25573/serc.21565671. Accessed October 2023.
Land Use, Land-Use Change, and Forestry 6-175
-------
vegetated coastal wetlands remaining to vegetated coastal wetlands, CH4 emissions in a given year
represent the cumulative area held in this category for that year and the prior 19 years.
Uncertainty
Underlying uncertainties in estimates of soil carbon removal factors, biomass change, DOM, and CH4
emissions include error in uncertainties associated with Tier 2 literature values of soil carbon removal
estimates, biomass stocks, DOM, and IPCC default CH4 emission factors, uncertainties linked to
interpretation of remote sensing data, as well as assumptions that underlie the methodological
approaches applied.
Uncertainty specific to coastal wetlands include differentiation of palustrine and estuarine community
classes, which determines what flux is applied. Because mean soil and biomass carbon removal for
each available community class are in a fairly narrow range, the same overall uncertainty was assigned
to each, respectively (i.e., applying approach for asymmetrical errors, the largest uncertainty for any soil
carbon stock value should be applied in the calculation of error propagation; IPCC 2000). Uncertainties
for CH4 flux are the Tier 1 default values reported in the Wetlands Supplement. Overall uncertainty of the
NOAA C-CAP remote sensing product is 15 percent. This is in the range of remote sensing methods (±10
to 15 percent; IPCC 2003). However, there is significant uncertainty in salinity ranges for tidal and non-
tidal estuarine wetlands and activity data used to estimate the CH4 flux (e.g., delineation of an 18 ppt
boundary), which will need significant improvement to reduce uncertainties. The combined uncertainty
was calculated by summing the squared uncertainty for each individual source (C-CAP, soil, biomass,
and DOM) and taking the square root of that total.
Uncertainty estimates are presented in Table 6-99 for each carbon pool and the CH4 emissions. In 2023,
the total flux was 0.19 MMT C02 Eq., 42.2 percent above and below the estimate with lower and upper
estimates of 0.11 and 0.26 MMTC02 Eq.
Table 6-99: Approach 1 Quantitative Uncertainty Estimates for Carbon Stock Changes
occurring within Land Converted to Vegetated Coastal Wetlands in 2023 (MMT C02 Eq.
and Percent)
Uncertainty Range Relative to Estimate"
(MMT CO2 Eq.)
(%)
2023 Estimate
Lower
Upper
Lower
Upper
Source
(MMT CO2 Eq.)
Bound
Bound
Bound
Bound
Biomass C Stock Flux
0.12
0.1
0.15
-20.0%
+20.0%
Dead Organic Matter Flux
0.03
0.02
0.03
-25.8%
+25.8%
Soil C Stock Flux
(0.13)
(0.2)
(0.1)
-17.7%
+ 17.7%
Methane Emissions
0.16
0.11
0.21
-29.9%
+29.9%
Total Uncertainty
0.19
0.11
0.26
-42.2%
+42.2%
a Range of flux estimates based on error propagation at 95 percent confidence interval.
Note: Totals may not sum due to independent rounding. Parentheses indicate net sequestration.
QA/QC and Verification
NOAA provided National LiDAR Dataset, tide data, and C-CAP land cover and land cover change
mapping, all of which are subject to agency internal mandatory QA/QC assessment (McCombs et al.
6-176 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
2016). QA/QC and verification of soil carbon stock dataset has been provided by the Smithsonian
Environmental Research Center and coastal wetland inventory team leads. Biomass carbon stocks are
derived from peer-review literature, reviewed by U.S. Geological Survey prior to publishing, by the peer-
review process during publishing, and by the coastal wetland inventory team leads prior to inclusion in
the Inventory and from IPCC reports. As a QC step, a check was undertaken confirming that coastal
wetlands recognized by C-CAP represent a subset of wetlands recognized by the NRI for marine coastal
states. A team of two evaluated and verified there were no computational errors within the calculation
worksheets. Soil carbon stock, emissions/removals data are based upon peer-reviewed literature and
CH4 emission factors are derived from the Wetlands Supplement.
Recalculations Discussion
No recalculations were performed for the current Inventory.
Planned Improvements
Currently, the only coastal wetland conversion that is reported in the Inventory is lands converted to
vegetated coastal wetlands. The next Inventory (i.e., publishing in 2026) is expected to include carbon
stock change data for lands converted to unvegetated open water coastal wetlands.
Land Converted to Flooded Land
Flooded lands are defined as water bodies where human activities have 1) caused changes in the
amount of surface area covered by water, typically through water level regulation (e.g., constructing a
dam), 2) waterbodies where human activities have changed the hydrology of existing natural
waterbodies thereby altering water residence times and/or sedimentation rates, in turn causing changes
to the natural production of greenhouse gases, and 3) waterbodies that have been created by
excavation, such as canals, ditches and ponds (IPCC 2019). Flooded lands include waterbodies with
seasonally variable degrees of inundation but would be expected to retain some inundated area
throughout the year under normal conditions.
Flooded lands are broadly classified as "reservoirs" or "other constructed waterbodies" (IPCC 2019).
Reservoirs are defined as flooded land greater than 8 ha and includes the seasonally flooded land on the
perimeter of permanently flooded land (i.e., inundation areas). IPCC guidance (IPCC 2019) provides
default emission factors for reservoirs and several types of "other constructed waterbodies" including
freshwater ponds and canals/ditches.
Land that has been flooded for 20 years or greater is defined as flooded land remaining flooded land and
land flooded for less than 20 years is defined as land converted to flooded land. The distinction is based
on literature reports that C02 and CH4 emissions are high immediately following flooding as labile
organic matter is rapidly degraded but decline to a steady background level approximately 20 years after
flooding (Abril et al. 2005, Barros et al. 2011, Teodoru et al. 2012). Both C02 and CH4 emissions are
estimated for land converted to flooded land.
Nitrous oxide emissions from flooded lands are largely related to inputs of organic or inorganic nitrogen
from the watershed. These inputs from runoff/leaching/deposition are largely driven by anthropogenic
activities such as land-use change, wastewater disposal or fertilizer application in the watershed or
application of fertilizer or feed in aquaculture. These emissions are not included here to avoid double-
Land Use, Land-Use Change, and Forestry 6-177
-------
counting N20 emissions which are captured in other source categories, such as indirect N20 emissions
from managed soils (Section 5.4, Agricultural Soil Management) and wastewater management (Section
7.2, Wastewater Treatment and Discharge).
Reservoirs are designed to store water for a wide range of purposes including hydropower, flood control,
drinking water, and irrigation. Lakes converted into reservoirs without substantial changes in water
surface area or water residence times are not considered to be managed flooded land (see Area
Estimates below) (IPCC 2019).
In 2023, the United States and Puerto Rico contained 66,850 ha of reservoir surface area in land
converted to flooded land (see Methodology and Time-Series Consistency below for calculation details)
distributed across all six of the aggregated climate zones used to define flooded land emission factors
(Figure 6-15) (IPCC 2019).
Emissions from Land Converted to Flooded Land-
Reservoirs
Figure 6-15: U.S. Reservoirs (black polygons) in the Land Converted to Flooded Land
Category in 2023
Climate Zone
| boreal
| cool temperate
| tropical dry/montane
| tropical moist/wet
| warm temperate dry
HI warm temperate moist
Alaska
Hawaii
500 mi
100 mi
500 mi
Note: Colors represent climate zone used to derive IPCC default emission factors. Reservoirs (indicated by black polygons) are
sparsely distributed across United States, but can be seen in MN, IL, and IN in this image. Alaska map scale is 1000 miles, Hawaii
and contiguous United States map scale is 500 miles, Puerto Rico map scale is 100 miles.
Methane and C02 are produced in reservoirs through the natural breakdown of organic matter. Per unit
area emission rates tend to scale positively with temperature and system productivity (i.e., abundance
of algae). Greenhouse gases produced in reservoirs can be emitted directly from the water surface and
inundation areas or as greenhouse gas-enriched water passes through the dam and the downstream
river. Sufficient information exists to estimate downstream CH4 emissions using Tier 1 IPCC guidance
(IPCC 2019), but no guidance is provided for downstream C02 emissions. Table 6-100 and Table 6-101
below summarize nationally aggregated CH4 and C02 emissions from reservoirs in land converted to
6-178 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
flooded land. The decrease in C02 and CH4 emissions through the time series is attributable to
reservoirs matriculating from the land converted to flooded land category into the flooded land
remaining flooded land category. Emissions have been stable since 2005, reflecting the low rate of new
flooded land creation over the past 18 years.
Table 6-100: CH4 Emissions from Land Converted to Flooded Land—Reservoirs (MMT
C02 Eq.)
Source
1990
2005
2019
2020
2021
2022
2023
Reservoirs
Surface Emissions
2.4
0.4
0.2
0.2
0.2
0.2
0.2
Downstream Emissions
0.2
+
+
+
+
+
+
Total
2.6
0.5
0.2
0.2
0.2
0.2
0.2
+lndicates values less than 0.05 MMT C02
Note: Totals may not sum due to independent rounding.
Table 6-101: CH4 Emissions from Land Converted to Flooded Land—Reservoirs (kt
CH4)
Source
1990
2005
2019
2020
2021
2022
2023
Reservoirs
Surface Emissions
84
15
7
7
6
6
6
Downstream Emissions
8
1
1
1
1
1
1
Total
92
16
8
8
7
7
7
Note: Totals may not sum due to independent rounding.
Table 6-102: C02 Emissions from Land Converted to Flooded Land—Reservoirs (MMT
C02)
Source
1990 2005 | 2019
2020
2021
2022
2023
Reservoir
00
d
CO
d
CNJ
CO
0.3
0.3
0.3
0.3
Table 6-103: C02 Emissions from Land Converted to Flooded Land—Reservoirs (MMT
C)
Source
1990 | 2005 | 2019
2020
2021
2022
2023
Reservoir
0.9 | 0.2 | 0.1
0.1
0.1
0.1
0.1
Minnesota was the largest source of CH4and C02 emissions from reservoirs in land converted to
flooded land, constituting 77.1 percent of total emissions from this category (Figure 6-16 and Table
6-104). These emissions are attributed to twelve reservoirs created after 2001 which impounded 54,338
ha of water, 95 percent of which is located in Mille Lacs Lake.
North Dakota is the second largest source of C02 and CH4 from reservoirs in land converted to flooded
land. Over ninety-eight percent of land converted to flooded land reservoir surface area in North Dakota
is attributed to Devils Lake. Both Mille Lacs and Devils Lakes are natural waterbodies provisioned with
dams for water level management. The dams have not substantially increased the surface area of these
reservoirs, but serve to reduce flooding and/or maintain minimum water levels for recreation, thereby
affecting water residence time.
Land Use, Land-Use Change, and Forestry 6-179
-------
Figure 6-16: 2023 A) CH4 and B) C02 Emissions from U.S. Reservoirs in Land Converted
to Flooded Land
A. CH4 Emissions from Reservoirs B. C02 Emissions from Reservoirs
Note: Alaska map scale is 1000 miles, Hawaii and contiguous United States map scale is 500 miles, Puerto Rico map scale is
100 miles.
Table 6-104: Methane and C02 Emissions from Reservoirs in Land Converted to
Flooded Land in 2023 (kt CH4; kt C02)
State
CH4
co2a
Surface
Downstream
Total
Surface
Alabama
0.0
0.0
0.0
0.0
Alaska
0.0
0.0
0.0
0.0
Arizona
0.0
0.0
0.0
0.0
Arkansas
0.0
0.0
0.0
0.0
California
+
+
+
+
Colorado
+
+
+
1.4
Connecticut
+
+
+
+
Delaware
0.0
0.0
0.0
0.0
District of Columbia
0.0
0.0
0.0
0.0
Florida
+
+
+
13.4
Georgia
+
+
+
0.8
Hawaii
0.0
0.0
0.0
0.0
Idaho
+
+
+
1.5
Illinois
+
+
+
4.1
Indiana
0.0
0.0
0.0
0.0
Iowa
+
+
+
1.9
Kansas
+
+
+
+
Kentucky
0.0
0.0
0.0
0.0
Louisiana
+
+
+
+
Maine
+
+
+
+
Maryland
+
+
+
+
Massachusetts
+
+
+
4.3
Michigan
+
+
+
+
Minnesota
4.6
+
5.0
203.2
Mississippi
+
+
+
+
6-180 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
State
CH4
co2a
Surface
Downstream
Total
Surface
Missouri
0.0
0.0
0.0
0.0
Montana
+
+
+
+
Nebraska
+
+
+
0.9
Nevada
+
+
+
+
New Hampshire
0.0
0.0
0.0
0.0
New Jersey
0.0
0.0
0.0
0.0
New Mexico
+
+
+
0.7
New York
+
+
+
+
North Carolina
+
+
+
2.1
North Dakota
0.5
+
0.6
22.5
Ohio
+
+
+
0.8
Oklahoma
0.0
0.0
0.0
0.0
Oregon
+
+
+
+
Pennsylvania
+
+
+
0.5
Puerto Rico
0.0
0.0
0.0
0.0
Rhode Island
0.0
0.0
0.0
0.0
South Carolina
0.0
0.0
0.0
0.0
South Dakota
+
+
+
+
Tennessee
+
+
+
+
Texas
+
+
+
+
Utah
+
+
+
0.8
Vermont
0.0
0.0
0.0
0.0
Virginia
0.0
0.0
0.0
0.0
Washington
+
+
+
+
West Virginia
+
+
+
+
Wisconsin
+
+
+
+
Wyoming
+
+
+
0.5
+ Indicates values greater than zero and less than 0.5 kt.
aC02: Only surface C02 emissions are included in the Inventory.
Methodology and Time-Series Consistency
Estimates of CH4 and C02 emissions for reservoirs in land converted to flooded land follow the Tier 1
methodology in the IPCC guidance (IPCC 2019). All calculations are performed at the state level and
summed to obtain national estimates. Emissions from the surface of these flooded lands are calculated
as the product of flooded land surface area and a climate-specific emission factor (Table 6-105).
Downstream CH4 emissions are calculated as 9 percent of the surface CH4 emission (Tier 1 default). The
IPCC guidance (IPCC 2019) does not address downstream C02 emissions, presumably because there
are insufficient data in the literature to estimate this emission pathway.
The IPCC default surface emission factors are derived from model-predicted (G-res model, Prairie et al.
2017) emission rates for all reservoirs in the Global Reservoir and Dam (GRanD) database (Lehner et al.
2011). Predicted emission rates were aggregated by the 11 IPCC climate zones (IPCC 2019, Table 7A.2)
which were collapsed into six climate zones using a regression tree approach. All six aggregated climate
zones are present in the United States.
Land Use, Land-Use Change, and Forestry 6-181
-------
Table 6-105: IPCC (2019) Default CH4and C02 Emission Factors for Surface Emissions
from Reservoirs in Land Converted to Flooded Land
Climate
Surface emission factor
MTCHtha"1
1 y1 MT CO2 ha1 y1
Boreal
0.0277
3.45
Cool Temperate
0.0847
3.74
Warm Temperate Dry
0.1956
6.23
Warm Temperate Moist
0.1275
5.35
Tropical Dry/Montane
0.3923
10.82
Tropical Moist/Wet
0.2516
10.16
Notes: Downstream ChU emissions are calculated as 9 percent of surface emissions. Downstream emissions are not calculated
for C02.
Area Estimates
U.S. reservoirs were identified from the "NHDWaterbody" layer in the National Hydrography Dataset Plus
V2 (NHD),78 the National Inventory of Dams (NID),79 operational hydropower dams in the Hydropower
Infrastructure-Lakes, Reservoirs and Rivers (HILARRI) database,80 the National Wetlands Inventory
(NWI),81 the Navigable Waterways (NW) network,82 and the EPA's Safe Drinking Water Information
System (SDWIS).83The NHD only covers the conterminous United States, whereas the NID, NW and NWI
include Alaska, Hawaii, and Puerto Rico. The following paragraphs present the criteria used to identify
reservoirs in the NHD, NW, and NWI.
Waterbodies in the "NHDWaterbody" layer that were less than or equal to 20-years old, greater than or
equal to 8 ha in surface area, not identified as canal/ditch in NHD, and met any of the following criteria
were considered reservoirs in land converted to flooded land: 1) the waterbody was classified
"Reservoir" in the "NHDWaterbody" layer, 2) the waterbody name in the "NHDWaterbody" layer
included "Reservoir", 3) the waterbody in the "NHDWaterbody" layer was located in close proximity (up
to 200 m) to a dam in the NID, 4) the "NHDWaterbody" GNIS name was similar to nearby NID feature (up
to a 1000 m radius).
EPA assumes that all features included in the NW are subject to water-level management to maintain
minimum water depths required for navigation and are therefore managed flooded lands. NW features
greater than 8 ha in surface area are defined as reservoirs.
NWI features were considered "managed" if they had a special modifier value indicating the presence of
management activities (Figure 6-19). To be included in the flooded lands inventory, the managed flooded
land had to be wet or saturated for at least one season per year (see 'Water Regime' in Figure 6-19). NWI
features that met these criteria, were greater than 8 ha in surface area, and were not a canal/ditch (see
emissions from land converted to flooded land-other constructed waterbodies) were defined as
reservoirs.
78 See https://www.usgs.gov/core-science-systems/ngp/national-hydrography.
79 See https://nid.sec.usace.army.mil.
80 See https://www.osti.gov/biblio/1960141.
81 See https://www.fws.gov/program/national-wetlands-inventorv/data-download.
82 See https://geodata.bts.gov/datasets/usdot::navigable-waterwav-network-lines/explore.
83 See https://www.epa.gov/enviro/sdwis-overview. Not publicly available due to security concerns.
6-182 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Any NWI or NHD feature that intersected a drinking water intake point from SDWIS was assumed to be
"managed "The rational being that a waterbody used as a source for public drinking water is typically
managed in some capacity - by flow and/or volume control.
Surface areas for identified flooded lands were taken from NHD, NWI or the NW. if features from the
NHD, NWI, or the NW datasets overlapped, duplicate areas were erased. The first step was to take the
final NWI flooded lands features and use it to identify overlapping NHD features. If the NHD feature had
its center in a NWI feature, it was removed from analysis. Next, remaining NHD features were erased
from any remaining overlapping NWI features. Final selections of NHD and NWI features were used to
erase any overlapping NW waterbodies.
Reservoir age was determined by assuming they were created the same year as a nearby (up to 200 m)
NID feature. If no nearby NID feature was identified, it was assumed the feature was greater than 20-
years old throughout the time series. Only reservoirs less than or equal to 20-years old are included in
land converted to flooded land.
Figure 6-17: Selected Features from NWI that meet Flooded Lands Criteria
MODIFIERS
In order to more adequately describe the wetland and deepwater habitats, one each of the water regime, water chemistry, soil, or
special modifiers may be applied at the class or lower level in the hierarchy
Water Regime
Special Modifiers
Water Chemistry
Soil
Nontidal
A Temporarily Flooded
B Seasonally Saturated
Saltwater Tidal
|L Subtidal
Freshwater Tidal
Q Regularly Flooded-Fresh Tidal
R Seasonally Flooded-Fresh Tidal
b Beaver
Halinity/Salinity pH Modifiers for
Fresh Water
1 Hyperhaline / Hypersaline a Acid
2 Euhaline / Eusaline t Circumneutral
3 Mixohaline / M ixosaline (Brackish) i Alkaline
4 Polyhaline
5 Mesohaline
6 Oligohaline
0 Fresh
g Organic
n Mineral
M Irreqularly Exposed
d Partly Drained/Ditched
f Farmed
m Managed
h Diked/impounded
r Artificial Substrate
s SDOil
x Excavated |
C Seasonally Flooded
N Regularly Flooded
P Irregularly Flooded
S Temporarily Flooded- Fresh Tidal
D Continuously Saturated
E Seasonally Flooded /
Saturated
F Semipermanently Flooded
G Intermittently Exposed
H Permanently Flooded
J Intermittently Flooded
K Artificially Flooded |
T Semipermanently Flooded-Fresh Tidal
V Permanently Flooded-Fresh Tidal
Must also meet one selected special modifier (red box) to be included in the flooded lands inventory
I "I Included in the flooded lands inventory if it meets water regime qualifier (gold box)
Source (modified): https://www.1ws.gov/sites/default/files/documents/wetlands-and-deepwater-map-code-diagram.pdf
IPCC (2019) allows for the exclusion of managed waterbodies from the Inventory if the water surface
area or residence time was not substantially changed by the construction of the dam. The guidance
does not quantify what constitutes a "substantial" change, but here EPA excludes the U.S. Great Lakes
from the Inventory based on expert judgment that neither the surface area nor water residence time was
substantially altered by their associated dams.
Reservoirs were disaggregated by state (using boundaries from the 2016 U.S. Census Bureau84) and
climate zone. Downstream and surface emissions for cross-state reservoirs were allocated to states
based on the surface area that the reservoir occupied in each state. Only the U.S. portion of reservoirs
that cross country borders were included in the Inventory.
The surface area of reservoirs in land converted to flooded land decreased by nearly 88 percent from
1990 to 2023 (Table 6-106). This is due to reservoirs that were less than 20-years old at the beginning of
time series entering the flooded land remaining flooded land category when they exceeded 20 years of
age. The rate at which flooded land has aged out of the land converted to flooded land category has
w See https://www.censLis.gov/geographies/mapping-files/time-series/geo/carto-boundary-file.html.
Land Use, Land-Use Change, and Forestry 6-183
-------
outpaced the rate of new dam construction. New dam construction has slowed considerably during the
time series with only nine new dams constructed in 20 23,85 versus 552 in 1990 (Figure 6-18).
Table 6-106: National Totals of Reservoir Surface Area in Land Converted to Flooded
Land (thousands of ha)
Surface Area (thousands of ha)
1990 |
2005
2019
2020
2021
2022
2023
Reservoir
550 |
125 |
73
71
69
68
67
Figure 6-18: Number of Dams Built per Year from 1990 through 2023
O O O O
Table 6-107: State Breakdown of Reservoir Surface Area in Land Converted to Flooded
Land (thousands of ha)
State
1990
2005
2019
2020
2021
2022
2023
Alabama
8.6
0.1
0.0
0.0
0.0
0.0
0.0
Alaska
0.6
0.0
0.0
0.0
0.0
0.0
0.0
Arizona
0.1
0.1
0.0
0.0
0.0
0.0
0.0
Arkansas
33.5
2.9
0.0
0.0
0.0
0.0
0.0
California
15.7
1.8
0.1
0.1
0.1
0.1
0.1
Colorado
7.4
1.3
0.4
0.4
0.4
0.4
0.4
Connecticut
1.2
1.1
0.0
0.0
0.0
0.0
0.0
Delaware
0.0
0.0
0.0
0.0
0.0
0.0
0.0
District of Columbia
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Florida
10.1
3.2
1.5
1.5
1.3
1.3
1.3
Georgia
15.6
0.0
0.1
0.1
0.1
0.1
0.1
85 See https://nid.sec.usace.army.mil.
6-184 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
State
1990
2005
2019
2020
2021
2022
2023
Hawaii
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Idaho
17.8
1.0
0.4
0.4
0.4
0.4
0.4
Illinois
49.3
39.2
1.3
1.3
0.8
0.8
0.8
Indiana
9.8
0.2
0.0
0.0
0.0
0.0
0.0
Iowa
9.5
2.2
0.6
0.5
0.5
0.4
0.4
Kansas
9.7
0.4
0.2
0.1
0.1
0.0
0.0
Kentucky
3.8
0.1
0.0
0.0
0.0
0.0
0.0
Louisiana
8.8
3.0
0.9
0.9
0.0
0.0
0.0
Maine
11.4
4.5
0.0
0.0
0.0
0.0
0.0
Maryland
0.6
0.0
0.0
0.0
0.0
0.0
0.0
Massachusetts
1.4
0.2
1.1
1.0
1.0
1.0
1.0
Michigan
11.9
1.0
0.1
0.1
0.1
0.1
0.1
Minnesota
8.1
5.1
54.0
53.8
53.7
53.5
54.3
Mississippi
5.2
2.5
0.1
0.1
0.1
0.1
0.1
Missouri
16.3
0.1
0.0
0.0
0.0
0.0
0.0
Montana
14.3
3.9
2.1
2.1
2.1
2.1
0.0
Nebraska
5.4
1.3
0.3
0.2
0.2
0.2
0.2
Nevada
1.3
1.1
0.1
0.1
0.1
0.1
0.1
New Hampshire
0.4
0.0
0.0
0.0
0.0
0.0
0.0
New Jersey
0.6
0.5
0.0
0.0
0.0
0.0
0.0
New Mexico
3.6
1.5
0.2
0.2
0.2
0.2
0.1
New York
13.9
12.0
0.1
0.1
0.1
0.1
0.1
North Carolina
11.6
0.3
0.4
0.4
0.4
0.4
0.4
North Dakota
1.9
3.5
6.4
6.4
6.3
6.0
6.0
Ohio
6.6
0.5
0.2
0.2
0.2
0.1
0.1
Oklahoma
34.0
6.8
0.0
0.0
0.0
0.0
0.0
Oregon
9.7
0.3
0.2
0.2
0.1
0.1
0.0
Pennsylvania
6.9
0.3
0.1
0.1
0.1
0.1
0.1
Puerto Rico
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Rhode Island
0.1
0.0
0.0
0.0
0.0
0.0
0.0
South Carolina
17.7
9.5
0.0
0.0
0.0
0.0
0.0
South Dakota
0.6
3.9
0.8
0.0
0.0
0.0
0.0
Tennessee
58.5
0.0
0.1
0.1
0.1
0.1
0.1
Texas
72.3
0.5
0.0
0.0
0.0
0.0
0.0
Utah
1.8
0.1
0.2
0.2
0.2
0.2
0.2
Vermont
0.2
0.1
0.0
0.0
0.0
0.0
0.0
Virginia
6.6
0.2
0.0
0.0
0.0
0.0
0.0
Washington
5.6
1.1
0.0
0.0
0.0
0.0
0.0
West Virginia
3.4
1.6
0.2
0.2
0.2
0.1
0.1
Wisconsin
1.9
0.3
0.1
0.1
0.1
0.1
0.1
Wyoming
14.6
5.2
0.2
0.2
0.2
0.2
0.1
Total
549.9
124.7
72.6
71.2
69.3
68.4
66.9
Note: Totals may not sum due to independent rounding
Land Use, Land-Use Change, and Forestry 6-185
-------
Uncertainty
Uncertainty in estimates of CH4 and C02 emissions from reservoirs on land converted to flooded land
were developed using IPCC Approach 2 and include uncertainty in the default emission factors and the
flooded land area inventory (Table 6-108). Uncertainty in emission factors is provided in the 2019
Refinement to the 2006 IPCC Guidelines (IPCC 2019). Uncertainties in the spatial data include 1)
uncertainty in area estimates from the NHD, NWI, and NW, and 2) uncertainty in the location of dams in
the NID and drinking water intakes in SDWIS. Overall uncertainties in these spatial datasets are
unknown, but uncertainty for remote sensing products is assumed to be ± 10 to 15 percent based on
IPCC guidance (IPCC 2003). An uncertainty range of ± 15 percent for the flooded land area estimates is
assumed and is based on expert judgment.
Table 6-108: Approach 2 Quantitative Uncertainty Estimates for CH4 and C02
Emissions from Reservoirs in Land Converted to Flooded Land
Source
Gas
2023 Emission
Estimate
(MMT CO2 Eq.)
Uncertainty Range Relative to Emission Estimate"
(MMT CO2 Eq.)
(%)
Lower Bound
Upper Bound
Lower Bound
Upper Bound
Reservoir
Surface
CH4
0.18
0.15
0.19
-12.7%
+ 12.9%
Surface
CO2
0.26
0.23
0.30
-13.3%
+ 13.6%
Downstream
cm
+
+
0.05
-63.4%
+239.6%
Total
0.44
0.38
0.51
-13.1%
+12.7%
+ Indicates values less than 0.05 MMT C02 Eq.
a Range of emission estimates predicted by Monte Carlo stochastic simulation for a 95 percent confidence interval.
QA/QC and Verification
The National Hydrography Data (NHD) is managed by the USGS in collaboration many other federal,
state, and local entities. Extensive QA/QC procedures are incorporated into the curation of the NHD. The
National Inventory of Dams (NID) is maintained by the U.S. Army Corps of Engineers (USACE) in
collaboration with the Federal Emergency Management Agency (FEMA) and state regulatory offices.
USACE resolves duplicative and conflicting data from 68 data sources, which helps obtain the more
complete, accurate, and updated NID. The Navigable Waterways (NW) dataset is part of the U.S.
Department of Transportation (USDOT)/Bureau of Transportation Statistics (BTS) National Transportation
Atlas Database (NTAD). The NW is a comprehensive network database of the nation's navigable
waterways updated on a continuing basis. U.S. Fish and Wildlife Service is the principal agency in
charge of wetland mapping including the National Wetlands Inventory (NWI). Quality and consistency of
the Wetlands Layer is supported by federal wetlands mapping and classification standards, which were
developed under the oversight of the Federal Geographic Data Committee (FGDC) with input by
the FGDC Wetlands Subcommittee. This dataset is part of the FGDC Water-Inland Theme, which is co-
chaired by the FWS and the U.S. Geological Survey. The EPA's Safe Drinking Water Information System
(SDWIS) tracks information on drinking water contamination levels as required by the 1974 Safe Drinking
Water Act and its 1986 and 1996 amendments.
General QA/QC procedures were applied to activity data, documentation, and emission calculations
consistent with the U.S. Inventory QA/QC plan, which is in accordance with Vol. 1 Chapter 6 of the 2006
6-186 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
IPCC Guidelines (see Annex 8 for more details). All calculations were executed independently in Excel
and R. Ten percent of state and national totals were randomly selected for comparison between the two
approaches to ensure there were no computational errors.
Recalculations Discussion
The National Inventory of Dams (NID) data are updated regularly. The version of NID used for the current
Inventory contains 610 new dams relative to the previous (1990 through 2022) Inventory data. Similarly,
the National Wetlands Inventory (NWI) is periodically updated. The NWI version used for the current
Inventory incorporated recent New Mexico, Ohio, Maryland, Michigan, and West Virginia feature updates
compared to the previous Inventory.
The net effect of these recalculations was an average annual change in CH4 and C02 emission estimates
from reservoirs of -0.1 MMT C02 Eq., or-3.2 percent, over the time series from 1990 to 2022 compared
to the previous Inventory.
Planned Improvements
The following are ongoing planned improvements which will be incorporated into a future Inventory:
• Developing country-specific emission factors for U.S. reservoirs based on a recently completed
EPA survey of greenhouse gas emissions from 108 reservoirs in the conterminous United
States.86 This improvement is expected for the next Inventory (1990 through 2024, publishing in
2026).
• Distinguishing inland wetlands from reservoirs and other types of flooded lands. See the
Planned Improvement chapter section of 6.1 Representation of the U.S. Land Base for additional
information. This is a long-term improvement but efforts are underway.
The data will be used to develop country-specific emission factors for U.S. reservoirs to be used in the
next Inventory (e.g. publishing in 2026). Efforts are also underway to distinguish inland wetlands from
reservoirs and other types of flooded lands.
Emissions from Land Converted to Flooded Land-Other
Constructed Waterbodies
Freshwater ponds are the only type of flooded lands within the "other constructed waterbodies"
subcategory of land converted to flooded land that are included in this Inventory (see Methodology for
details) because age data are not available for canals and ditches. All canals and ditches are assumed
to be greater than 20-years old throughout the time series and are included in flooded land remaining
flooded land.
IPCC (2019) describes ponds as waterbodies that are "...constructed by excavation and/or construction
of walls to hold water in the landscape for a range of uses, including agricultural water storage, access
to water for livestock, recreation, and aquaculture."The IPCC "Decision tree for types of Flooded Land"
(IPCC 2019, Fig. 7.2) elaborates on this description by defining waterbodies less than 8 ha as a subset of
"other constructed waterbodies." For this Inventory, ponds are defined as managed flooded land not
86 See https://www.epa.gov/air-research/research-emissions-us-reservoirs.
Land Use, Land-Use Change, and Forestry 6-187
-------
identified as "canal/ditch" (see Methods below) with surface area less than 8 ha. IPCC (2019) further
distinguishes saline versus brackish ponds, with the former supporting lower CH4 emission rates than
the latter. Activity data on pond salinity is not uniformly available for the United States and all ponds in
land converted to flooded land are assumed to be freshwater. Ponds often receive high organic matter
and nutrient loadings, may have low oxygen levels, and are sites of substantial CH4 and C02 emissions
from anaerobic sediments.
Methane and C02 emissions from freshwater ponds decreased 95.6 and 96.2 percent, respectively,
from 1990 to 2023 due to flooded land matriculating from land converted to flooded land to flooded land
remaining flooded land. In 2023, states in the Great Plains region generally had the greatest C02 and CH4
emissions from freshwater ponds in land converted to flooded land (Table 6-109 through Table 6-113,
Note: Totals may not sum due to independent rounding.
Figure 6-19). Mississippi had the second greatest emissions of all states, partly due to the relatively high
C02 emission factor for the tropical moist/wet climate zone (Figure 6-15, Table 6-114).
Table 6-109: CH4 Emissions from Other Constructed Waterbodies in Land Converted
to Flooded Land (MMT C02 Eq.)
Source
1990 | 2005 | 2019
2020
2021
2022
2023
Freshwater Ponds
0.1 | +| +
+
+
+
+
+ Indicates values less than 0.05 MMT C02Eq.
Table 6-110: CH4 Emissions from Other Constructed Waterbodies in Land Converted
to Flooded Land (kt CH4)
Source
1990 |
2005 | 2019
2020
2021
2022
2023
Freshwater Ponds
4
1
+
+
+
+
+
+ Indicates values less than 0.5 kt.
Table 6-111: C02 Emissions from Other Constructed Waterbodies in Land Converted
to Flooded Land (MMT C02 Eq.)
Source
1990
2005
2019
2020
2021
2022
2023
Freshwater Ponds
0.1
+
+
+
+
+
+
+ Indicates values less than 0.05 MMT C02 Eq.
Table 6-112: C02 Emissions from Other Constructed Waterbodies in Land Converted
to Flooded Land (MMT C)
Source
1990 |
2005 |
2019
2020
2021
2022
2023
Freshwater Ponds
0.04 |
0.01 I
+
+
+
+
+
+ Indicates values less than 0.005 MMT C.
Table 6-113: CH4 and C02 Emissions from Other Constructed Waterbodies in Land
Converted to Flooded Land in 2023 (MT C02 Eq.)
Freshwater Ponds
State
CH4
CO2
Total
Alabama
1
1
1
6-188 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Freshwater Ponds
State
CH4
CO2
Total
Alaska
0
0
0
Arizona
0
0
0
Arkansas
2
2
5
California
38
44
82
Colorado
274
213
487
Connecticut
0
0
1
Delaware
0
0
1
District of Columbia
0
0
0
Florida
15
30
45
Georgia
260
470
730
Hawaii
0
0
0
Idaho
4
5
9
Illinois
73
63
136
Indiana
29
30
59
Iowa
290
269
559
Kansas
369
387
756
Kentucky
13
13
26
Louisiana
56
112
168
Maine
1
1
2
Maryland
48
50
98
Massachusetts
307
283
590
Michigan
48
35
84
Minnesota
172
125
297
Mississippi
354
540
894
Missouri
177
185
363
Montana
69
50
119
Nebraska
406
367
772
Nevada
83
65
148
New Hampshire
90
66
156
New Jersey
0
0
0
New Mexico
58
59
117
New York
46
34
80
North Carolina
141
147
288
North Dakota
226
165
390
Ohio
183
175
358
Oklahoma
319
352
671
Oregon
84
66
150
Pennsylvania
20
19
39
Puerto Rico
0
0
0
Rhode Island
0
0
0
South Carolina
47
49
95
South Dakota
462
337
800
Tennessee
13
14
27
Texas
94
163
257
Land Use, Land-Use Change, and Forestry 6-189
-------
State
Freshwater Ponds
CH4
co2
Total
Utah
51
38
89
Vermont
16
11
27
Virginia
11
12
23
Washington
90
87
177
West Virginia
143
149
292
Wisconsin
106
77
183
Wyoming
173
126
299
Total
5,463
5,488
10,951
Note: Totals may not sum due to independent rounding.
Figure 6-19: 2023 A) GH4 and B) C02 Emissions from Other Constructed Waterbodies
(Freshwater Ponds) in Land Converted to Flooded Land (MT C02 Eq.)
A. CH4 Emissions from Freshwater Ponds B. C02 Emissions from Freshwater Ponds
Note: Alaska map scale is 1000 miles, Hawaii and contiguous United States map scale is 500 miles, Puerto Rico map scale is
100 miles.
Methodology and Time-Series Consistency
Estimates of CH4 and C02 emissions for other constructed waterbodies in land converted to flooded
land follow the Tier 1 methodology in IPCC (2019). All calculations are performed at the state level and
summed to obtain national estimates. Greenhouse gas emissions from the surface of these flooded
lands are calculated as the product of flooded land surface area and an emission factor (Table 6-114).
Due to a lack of empirical data on C02 emissions from recently created ponds, IPCC (2019) states "For
all types of ponds created by damming, the methodology described above to estimate CC2 emissions
from land converted to reservoirs may be used." This Inventory uses IPCC default C02 emission factors
for land converted to reservoirs when estimating C02 emissions from land converted to freshwater
ponds. IPCC guidance also states that "there is insufficient information available to derive separate CH4
emission factors for recently constructed ponds..." and allows for the use of IPCC default CH4 emission
factors for land remaining flooded land. Downstream emissions are not inventoried for other
constructed waterbodies because 1) many of these systems are not associated with dams (e.g.,
excavated ponds and ditches), and 2) there are insufficient data to derive downstream emission factors
for other constructed waterbodies that are associated with dams (IPCC 2019).
6-190 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Table 6-114: IPCC Default Methane and C02 Emission Factors for Other Constructed
Waterbodies in Land Converted to Flooded Land
Other Constructed Waterbody
Climate Zone
Emission Factor
MT CH4 ha-1 y1
MT C02 ha1 y1
Freshwater ponds
Boreal
0.183
3.45
Freshwater ponds
Cool Temperate
0.183
3.74
Freshwater ponds
Warm Temperate Dry
0.183
6.23
Freshwater ponds
Warm Temperate Moist
0.183
5.35
Freshwater ponds
Tropical Dry/Montane
0.183
10.82
Freshwater ponds
Tropical Moist/Wet
0.183
10.16
Note: downstream emissions are not estimated for freshwater ponds
Area Estimates
Other constructed waterbodies were identified from the "NHDWaterbody" layer in the National
Hydrography Dataset Plus V2 (NHD),87 the National Inventory of Dams (N ID),88 the National Wetlands
Inventory (NWI),89 the Navigable Waterways (NW) network,90 the Hydropower Infrastructure—Lakes,
Reservoirs, and Rivers (HILARRI) database,91 and the EPA's Safe Drinking Water Information System
(SDWIS)92. The NHD only covers the conterminous United States, whereas the NID, NW and NWI also
include Alaska, Hawaii, and Puerto Rico.
Waterbodies in the "NHDWaterbody" layer that were less than or equal to 20-years old, less than 8 ha in
surface area, not identified as canal/ditch in NHD, and met any of the following criteria were considered
freshwater ponds in land converted to flooded land: 1) the waterbody was classified "Reservoir" in the
"NHDWaterbody" layer, 2) the GNIS name in the "NHDWaterbody" layer included the text "Reservoir", 3)
the waterbody in the "NHDWaterbody" layer was located in close proximity (up to 200 m) to a dam in the
NID, 4) the "NHDWaterbody" GNIS name was similar to nearby NID feature (up to a 1000 m radius).
EPA assumes that all features included in the NW are subject to water-level management to maintain
minimum water depths required for navigation and are therefore managed flooded lands. NW features
that were less than 8 ha in surface area and not identified as canals/ditch (see below) were considered
freshwater ponds.
NWI features were considered "managed" if they had a special modifier value indicating the presence of
management activities (Figure 6-19). To be included in the flooded lands inventory, the managed flooded
land had to be wet or saturated for at least one season per year (see 'Water Regime' in Figure 6-19). NWI
features that met these criteria, were less than 8 ha in surface area, and were not a canal/ditch were
defined as freshwater ponds.
Any NWI or NHD feature that intersected a drinking water intake point from SDWIS was assumed to be
"managed". The rational being that a waterbody used as a source for public drinking water is typically
managed in some capacity - by flow and/or volume control.
87 See https://www.usgs.gov/core-science-systems/ngp/national-hydrography.
88 See https://nid.sec.usace.army.mil.
89 See https://www.fws.gov/program/national-wetlands-inventorv/data-download.
90 See https://geodata.bts.gov/datasets/usdot::navigab[e-waterwav-network-[ines/exp[ore.
91 See https://www.osti.gov/biblio/1960141.
92 See https://www.epa.gov/enviro/sdwis-overview. Not publicly available due to security concerns.
Land Use, Land-Use Change, and Forestry 6-191
-------
Surface areas for other constructed waterbodies were taken from NHD, NWI or the NW. If features from
the NHD, NWI, or the NWdatasets overlapped, duplicate areas were erased. The first step was to take
the final NWI flooded lands features and use it to identify overlapping NHD features. If the NHD feature
had its center in a NWI feature, it was removed from analysis. Next, remaining NHD features were erased
from any remaining overlapping NWI features. Final selections of NHD and NWI features were used to
erase any overlapping NW waterbodies.
The age of other constructed waterbody features was determined by assuming the waterbody was
created the same year as a nearby (up to 200 m) NID feature. If no nearby NID feature was identified, it
was assumed the waterbody was greater than 20-years old throughout the time series. No canal/ditch
features were associated with a nearby dam, therefore all canal/ditch features were assumed to be
greater than 20-years old through the time series.
For the year 2023, this Inventory contains 1,066 ha of freshwater ponds in land converted to flooded
land. The surface area of freshwater ponds decreased by 95.6 percent from 1990 to 2023 due to flooded
lands aging out of land converted to flooded land more quickly than new flooded lands entered the
category. Freshwater ponds in year 2023 of the Inventory are most abundant in Nebraska, South Dakota,
and Kansas (Figure 6-20).
Table 6-115: National Surface Area Totals of Other Constructed Waterbodies in Land
Converted to Flooded Land (hectares)
Other Constructed Waterbodies
1990 |
2005
2019
2020
2021
2022
2023
Freshwater Ponds
24,363 I
5,037 I
1,939
1,645
1,435
1,333
1,066
Figure 6-20: Surface Area of Other Constructed Waterbodies in Land Converted to
Flooded Land (hectares) in 2023
hectares
1100
75
50
Note: Alaska map scale is 1000 miles, Hawaii and contiguous United States map scale is 500 miles, Puerto Rico map scale is
100 miles.
Alaska
1000 mi
Hawaii
500 mi
100 mi
500 mi
6-192 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Table 6-116: State Surface Area Totals of Other Constructed Waterbodies in Land
Converted to Flooded Land (hectares)
State
1990
2005
2019
2020
2021
2022
2023
Alabama
381
24
1
1
1
0
0
Alaska
6
0
0
0
0
0
0
Arizona
41
13
4
0
0
0
0
Arkansas
336
2
0
0
0
0
0
California
226
90
23
20
20
14
7
Colorado
225
49
53
56
56
56
53
Connecticut
74
5
0
0
0
0
0
Delaware
4
0
0
0
0
0
0
District of Columbia
0
0
0
0
0
0
0
Florida
100
41
11
11
5
3
3
Georgia
1,690
86
51
51
51
51
51
Hawaii
7
2
0
0
0
0
0
Idaho
101
1
1
1
1
1
1
Illinois
539
102
30
20
18
15
14
Indiana
469
98
22
22
22
12
6
Iowa
2,328
1,460
303
189
96
66
57
Kansas
2,252
162
104
99
78
82
72
Kentucky
375
35
3
3
3
3
3
Louisiana
130
25
18
12
12
12
11
Maine
29
9
0
0
0
0
0
Maryland
214
64
15
13
11
10
9
Massachusetts
57
68
73
66
65
61
60
Michigan
149
51
17
9
9
9
9
Minnesota
275
96
67
50
42
34
34
Mississippi
348
165
104
92
86
78
69
Missouri
3,363
169
57
52
45
41
35
Montana
359
106
100
99
99
99
13
Nebraska
1,272
274
118
100
91
88
79
Nevada
17
51
31
28
23
21
16
New Hampshire
140
45
19
18
18
18
18
New Jersey
35
14
0
0
0
0
0
New Mexico
24
34
20
20
17
13
11
New York
304
130
25
25
20
20
9
North Carolina
482
90
33
30
28
28
28
North Dakota
149
160
76
66
58
55
44
Ohio
411
265
96
67
55
53
36
Oklahoma
1,923
144
95
79
73
64
62
Oregon
179
64
21
17
16
16
16
Pennsylvania
247
39
4
4
4
4
4
Puerto Rico
0
0
0
0
0
0
0
Rhode Island
9
7
0
0
0
0
0
South Carolina
756
234
21
9
9
9
9
Land Use, Land-Use Change, and Forestry 6-193
-------
State
1990
2005
2019
2020
2021
2022
2023
South Dakota
285
114
107
113
105
105
90
Tennessee
406
46
10
3
3
3
3
Texas
2,738
107
27
26
25
25
18
Utah
85
21
24
20
20
20
10
Vermont
72
11
3
3
3
3
3
Virginia
58
4
2
2
2
2
2
Washington
146
46
23
20
20
19
18
West Virginia
121
47
28
28
28
28
28
Wisconsin
149
19
21
21
21
21
21
Wyoming
277
150
79
78
75
72
34
Total
24,363
5,037
1,939
1,645
1,435
1,333
1,066
Note: Totals may not sum due to independent rounding
Uncertainty
Uncertainty in estimates of C02 and CH4emissions from land converted to flooded land-other
constructed water bodies include uncertainty in the default emission factors and the flooded land area
inventory. Uncertainty in emission factors is provided in the 2019 Refinement to the 2006IPCC
Guidelines (IPCC 2019). Uncertainties in the spatial data include 1) uncertainty in area estimates from
the NHD and NW, and 2) uncertainty in the location of dams in the NID and drinking water intakes in
SDWIS. Overall uncertainties in the NHD, NWI, NID, and NW are unknown, but uncertainty for remote
sensing products is ±10 to 15 percent (IPCC 2003). EPA assumes an uncertainty of ± 15 percent for the
flooded land area inventory based on expert judgment. These uncertainties do not include the
underestimate of pond surface area discussed above.
Table 6-117: Approach 2 Quantitative Uncertainty Estimates for CH4 and C02
Emissions from Other Constructed Waterbodies in Land Converted to Flooded Land
2023 Emission
Estimate
Uncertainty Range Relative to Emission Estimate"
(kt CO2 Eq.)
(%)
Source
Gas
(kt CO2 Eq.)
Lower Bound
Upper Bound
Lower Bound
Upper Bound
Freshwater ponds
CH4
5.44
5.35
5.51
-1.6%
+ 1.3%
Freshwater ponds
CO2
5.46
5.38
5.53
-1.4%
+ 1.4%
Total
10.90
10.74
11.03
-1.4%
+1.2%
a Range of emission estimates predicted by Monte Carlo stochastic simulation for a 95 percent confidence interval.
Note: Totals may not sum due to independent rounding.
QA/QC and Verification
The National Hydrography Data (NHD) is managed by the USGS with collaboration from many other
federal, state, and local entities. Extensive QA/QC procedures are incorporated into the curation of the
NHD. The National Inventory of Dams (NID) is maintained by the U.S. Army Corps of Engineers (USACE)
in collaboration with the Federal Emergency Management Agency (FEMA) and state regulatory offices.
USACE resolves duplicative and conflicting data from 68 data sources, which helps obtain the more
complete, accurate, and updated NID. The Navigable Waterways (NW) dataset is part of the U.S.
Department of Transportation (USDOT)/Bureau of Transportation Statistics (BTS) National Transportation
6-194 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Atlas Database (NTAD). The NW is a comprehensive network database of the nation's navigable
waterways updated on a continuing basis. U.S. Fish and Wildlife Service is the principal agency in
charge ofwetland mapping including the National Wetlands Inventory. Quality and consistency of the
Wetlands Layer is supported by federal wetlands mapping and classification standards, which were
developed under the oversight of the Federal Geographic Data Committee (FGDC) with input by
the FGDC Wetlands Subcommittee. This dataset is part of the FGDC Water-Inland Theme, which is co-
chaired by the FWS and the U.S. Geological Survey. The EPA's Safe Drinking Water Information System
(SDWIS) tracks information on drinking water contamination levels as required by the 1974 Safe Drinking
Water Act and its 1986 and 1996 amendments.
General QA/QC procedures were applied to activity data, documentation, and emission calculations
consistent with the U.S. Inventory QA/QC plan, which is in accordance with Vol. 1 Chapter 6 of the 2006
IPCC Guidelines (see Annex 8 for more details). All calculations were executed independently in Excel
and R. Ten percent of state and national totals were randomly selected for comparison between the two
approaches to ensure there were no computational errors.
Recalculations Discussion
The National Inventory of Dams (NID) data are updated regularly. The version of NID used for the current
Inventory contains 610 new dams relative to the version used for the previous (1990 through 2022)
Inventory. Similarly, the National Wetlands Inventory (NWI) is periodically updated, he NWI version used
for the current Inventory incorporated recent New Mexico, Ohio, Maryland, Michigan, and West Virginia
feature updates compared to the previous Inventory.
The net effect of these recalculations was an average annual change in CH4 and C02 emissions from
other constructed waterbodies of -0.004 MMT C02 Eq., or-4.3 percent, over the time series from 1990 to
2022 compared to the previous Inventory.
Planned Improvements
The following are ongoing planned improvements which will be incorporated into a future Inventory:
• Conducting a survey, led by EPA, of GHG emissions from U.S. ponds from 2024 through 2026 to
assess the accuracy of IPCC default emission factors. The IPCC default emission factors for
ponds were derived from a global dataset that include few measurements from U.S. systems.
The results of this survey will determine if default or country specific EFs will be used for future
inventories. This is an ongoing improvement but will take multiple years to fully implement.
• Distinguishing inland wetlands from ponds and other types of flooded lands. See the Planned
Improvement chapter section of 6.1 Representation of the U.S. Land Base for additional
information. This is a long-term improvement but efforts are underway.
Land Use, Land-Use Change, and Forestry 6-195
-------
6.10 Settlements Remaining Settlements
(Source Category 4E1)
Soil Carbon Stock Changes (Source Category 4E1)
Soil organic C stock changes for settlements remaining settlements occur in both mineral and organic
soils. However, the United States does not estimate changes in soil organic C stocks for mineral soils in
settlements remaining settlements. This approach is consistent with the assumption of the Tier 1
method in the 2006IPCC Guidelines (IPCC 2006) that inputs equal outputs, and therefore the soil
organic C stocks do not change in this land use category. This assumption may be re-evaluated in the
future if funding and resources are available to conduct an analysis of soil organic C stock changes for
mineral soils in settlements remaining settlements.
Drainage of organic soils is common when wetland areas have been developed for settlements. Organic
soils, also referred to as Histosols, include all soils with more than 12 to 20 percent organic carbon by
weight, depending on clay content (NRCS 1999; Brady and Weil 1999). The organic layer of these soils
can be very deep (i.e., several meters), and form under inundated conditions that results in minimal
decomposition of plant residues. Drainage of organic soils leads to aeration of the soil that accelerates
decomposition rate and C02 emissions.93 Due to the depth and richness of the organic layers, carbon
loss from drained organic soils can continue over long periods of time, which varies depending on
climate and composition (i.e., decomposability) of the organic matter (Armentano and Menges 1986).
Settlements remaining settlements includes all areas that have been settlements for a continuous time
period of at least 20 years according to the 2017 United States Department of Agriculture (USDA)
National Resources Inventory (NRI) (USDA-NRCS 2020)94 or according to the National Land Cover
Dataset (NLCD) for federal lands (Yang et al. 2018; Fry et al. 2011; Homer et al. 2007, 2015). There are
discrepancies between the current land representation (see Section 6.1) and the area data that have
been used in the Inventory for settlements remaining settlements. Specifically, Alaska and the small
amount of settlements on federal lands are not included in this Inventory even though these areas are
part of the U.S. managed land base. There is a planned improvement to include C02 emissions from
drainage of organic soils in settlements of Alaska and federal lands as part of a future Inventory (see
Planned Improvements section).
C02 emissions from drained organic soils in settlements are 16.4 MMT C02 Eq. (4.5 MMT C) in 2023 (see
Table 6-118 and Table 6-118). Although the flux is relatively small, the amount has increased by 66
percent since 1990 due to an increase in area of drained organic soils in settlements.
93 N2O emissions from drained organic soils are included in the N2O Emissions from Settlement Soils section.
94 NRI survey locations are classified accordingto land use histories starting in 1979, and consequently the classifications
are based on less than 20 years from 1990 to 1998. This may have led to an overestimation of settlements remaining
settlements in the early part of the time series to the extent that some areas are converted to settlements between 1971
and 1978.
6-196 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Table 6-118: Net C02 Flux from Soil C Stock Changes in Settlements Remaining
Settlements (MMT C02 Eq.)
Soil Type 1990
2005
2019 2020 2021 2022
2023
Organic Soils 9.9
10.1
14.6 15.1 15.6 16.0
16.4
Table 6-119: Net C02 Flux from Soil C Stock Changes in Settlements Remaining
Settlements (MMT C)
Soil Type 1990
2005
2019 2020 2021 2022
2023
Organic Soils 2.7
2.7
4.0 4.1 4.2 4.4
4.5
Methodology and Time-Series Consistency
An IPCC Tier 2 method is used to estimate soil organic C stock changes for organic soils in settlements
remaining settlements (IPCC 2006). Organic soils in settlements remaining settlements are assumed to
be losing C at a rate similar to croplands due to deep drainage, and therefore emission rates are based
on country-specific values for cropland (Ogle et al. 2003).
The land area designated as settlements is based primarily on the 2017 NRI (USDA-NRCS 2020) with
additional information from the NLCD to the extend the time series through 2020 (Yanget al. 2018). Soils
are classified as organic using data from the Soil Survey Geographic Database (SSURGO) (Soil Survey
Staff 2020). The areas have been modified through a process in which the Forest Inventory and Analysis
(FIA) survey data are harmonized with the NRI data (Nelson et al. 2020). This process ensures that the
land use areas are consistent across all land use categories (see Section 6.1 for more information). All
settlements occurring on organic soil are assumed to be drained for the purposes of approximating
greenhouse gas emissions. The area of drained organic soils is estimated from the NRI spatial weights
and aggregated to the country (Table 6-120). The area of land on organic soils in settlements remaining
settlements has increased from 216 thousand hectares in 1990 to over 327 thousand hectares in 2020.
Table 6-120: Thousands of Hectares of Drained Organic Soils in Settlements
Remaining Settlements
1990 |
2005
2019
2020
2021
2022
2023
Area (Thousand Hectares)
216 |
219 |
317
327
~
~
~
NRI data have not been incorporated into the Inventory after 2020, designated with asterisks (*).
To estimate C02 emissions from drained organic soils across the time series from 1990 to 2020, the area
of organic soils by climate (i.e., cool temperate, warm temperate, subtropical) in settlements remaining
settlements is multiplied by the appropriate country-specific emission factors for cropland remaining
cropland under the assumption that there is deep drainage of the soils. The emission factors are 11.2 MT
C per ha in cool temperate regions, 14.0 MT C per ha in warm temperate regions, and 14.3 MT C per ha in
subtropical regions (see Annex 3.13 for more information).
In order to ensure time-series consistency, the same methods are applied from 1990 to 2020, and a data
splicing method is used to approximate emissions for the remainder of the 2021 to 2023 time series (see
Box 6-3 in Section 6.4 cropland remaining cropland). The extrapolation is based on a linear regression
model with moving-average (ARMA) errors (Brockwelland Davis 2016) using the 1990 to 2020 emissions
data. This method is type of a linear extrapolation, which is a standard data splicing method for
Land Use, Land-Use Change, and Forestry 6-197
-------
estimating emissions at the end of a time series (IPCC 2006). The Tier 2 method described previously
will be applied in future inventories to recalculate the estimates beyond 2020 as new activity data are
integrated into the analysis.
Uncertainty
The total uncertainty was quantified with two variance components (Ogle et al. 2010) that are combined
using simple error propagation methods provided by the IPCC (2006), i.e., by taking the square root of
the sum of the squares of the standard deviations of the uncertain quantities. The first variance
component is associated with uncertainty in the emission factor, and the second variance component is
associated with scaling of the data from the NRI survey to the entire area of drained organic soils in
settlements remaining settlements, and is computed using a standard variance estimator for a two-
stage sample design (Sarndal et al. 1992). There is also additional uncertainty associated with the fit of
the linear regression model for the data splicing methods that was incorporated into the analysis for the
latter part of the time series. Soil carbon losses from drained organic soils in settlements remaining
settlements for 2023 are estimated to be between 6.7 and 26.2 MMT C02 Eq. at a 95 percent confidence
level (Table 6-121). This indicates a range of 59 percent below and 59 percent above the 2023 emission
estimate of 16.4 MMTC02 Eq.
Table 6-121: Uncertainty Estimates for C02 Emissions from Drained Organic Soils in
Settlements Remaining Settlements (MMT C02 Eq. and Percent)
Uncertainty Range Relative to Emission Estimate"
2023 Emission
(MMT CO2 Eq.)
(%)
Estimate
Lower
Upper
Lower
Upper
Source
Gas
(MMT CO2 Eq.)
Bound
Bound
Bound
Bound
Organic Soils
C02
16.4
6.7
26.2
-59%
+59%
a Range of emission estimates is a 95 percent confidence interval.
QA/QC and Verification
Quality control measures included checking input data, model scripts, and results to ensure data are
properly handled throughout the inventory process consistent with the U.S. Inventory QA/QC plan,
which is in accordance with Volume 1, Chapter 6 of the 2006 IPCC Guidelines (see Annex 8 for more
details). Inventory forms and text are reviewed and revised as needed to correct transcription errors. No
errors were found in this Inventory.
Recalculations Discussion
A recalculation was completed for the data splicing values that were estimated for 2021 and 2022. As a
result of this change, C02-equivalent emissions changed annually with an average annual increase of
0.4 MMT C02 Eq., or 2 percent, over the time series from 2021 to 2022 compared to the previous
Inventory.
6-198 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Planned Improvements
Key planned improvements that will resolve most of the differences between the managed land base for
settlements remaining settlements and amount of area currently included in this Inventory as
settlements remaining settlements (see Table 6-122) include:
• Estimating C02 emissions from drainage of organic soils in settlements of Alaska and federal
lands.
• Incorporating new Land Representation area data into the next Inventory. The current Inventory
for settlements remaining settlements is based on the land representation from the previous
Inventory.
Table 6-122: Area of Managed Land in Settlements Remaining Settlements that is not
included in the current Inventory (Thousand Hectares)
Year
Area (Thousand Hectares)
SRS Managed Land Area (Section 6.1)
SRS Area Included in Inventory
Difference
1990
30,540
30,366
173
1991
30,537
30,364
173
1992
30,534
30,361
173
1993
30,462
30,288
174
1994
30,374
30,203
171
1995
30,312
30,141
171
1996
30,252
30,081
171
1997
30,183
30,011
171
1998
30,117
29,945
171
1999
30,063
29,891
171
2000
30,005
29,834
172
2001
29,952
29,781
171
2002
29,945
29,774
171
2003
30,468
30,298
171
2004
30,962
30,791
171
2005
31,421
31,250
171
2006
31,929
31,758
171
2007
32,386
32,215
171
2008
33,005
32,833
171
2009
33,581
33,410
171
2010
34,156
33,984
171
2011
34,720
34,549
171
2012
35,291
35,120
172
2013
36,215
36,042
172
2014
37,149
36,977
172
2015
38,017
37,844
172
2016
38,927
38,756
171
2017
39,851
39,679
172
2018
40,749
40,574
175
Land Use, Land-Use Change, and Forestry 6-199
-------
Year
Area (Thousand Hectares)
SRS Managed Land Area (Section 6.1)
SRS Area Included in Inventory
Difference
2019
41,595
41,420
176
2020
42,446
42,270
176
2021
43,171
~
~
2022
43,731
~
~
NRI data have not been incorporated into the Inventory after 2020, designated with asterisks (*).
Changes in Carbon Stocks in Settlement Trees (Source
Category 4E1)
Settlements are land uses where human populations and activities are concentrated. In these areas, the
anthropogenic impacts on tree growth, stocking and mortality are particularly pronounced (Nowak
2012) in comparison to forest lands where non-anthropogenic forces can have more significant impacts.
Estimates included in this section include net C02 and carbon flux from trees on settlements remaining
settlements and land converted to settlements as it is not possible to report on these separately at this
time.
Trees in settlement areas of the United States are estimated to account for an average annual net
sequestration of 118.7 MMT C02 Eq. (32.4 MMT C) over the period from 1990 through 2023. Net carbon
sequestration from settlement trees in 2023 is estimated to be 139.0 MMT C02 Eq. (37.9 MMT C) (Table
6-123). Dominant factors affecting carbon flux trends for settlement trees are changes in the amount of
settlement area (increasing sequestration due to more land and trees) and net changes in tree cover
(e.g., tree losses versus tree gains through planting and natural regeneration), with percent tree cover
trending downward recently. In addition, changes in species composition, tree sizes and tree densities
affect base carbon flux estimates. Annual sequestration increased by 44.0 percent between 1990 and
2023 due to increases in settlement area and changes in total tree cover.
Trees in settlements often grow faster than forest trees because of their relatively open structure (Nowak
and Crane 2002). Because tree density in settlements is typically much lower than in forested areas, the
carbon storage per hectare of land is in fact smaller for settlement areas than for forest areas. Also,
percent tree cover in settlement areas is less than in forests and this urban tree cover varies significantly
across the United States (e.g., Nowak and Greenfield 2018a). To quantify the carbon stored in
settlement trees, the methodology used here requires analysis per unit area of tree cover, rather than
per unit of total land area (as is done for forest lands).
Table 6-123: Net Flux from Trees in Settlements Remaining Settlements (MMT C02 Eq.
and MMT C)a
Year
1990
2005 |
2019
2020
2021
2022
2023
MMTCO2 Eq.
(96.5)
(117.0)
(135.4)
(136.6)
(137.6)
(138.4)
(139.0)
MMT C
(26.3)
(31.9) |
(36.9)
(37.3)
(37.5)
(37.7)
(37.9)
a These estimates include net C02 and C flux from trees on settlements remaining settlements and land converted to
settlements as it is not possible to report on these separately at this time.
Note: Parentheses indicate net sequestration.
6-200 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Methodology and Time-Series Consistency
To estimate net carbon sequestration in settlement areas, three types of data are required for each
state:
1. Settlement area
2. Percent tree cover in settlement areas
3. Carbon sequestration density per unit of tree cover
Settlement Area
Settlement area is defined in Section 6.1 as a land-use category representing developed areas. The data
used to estimate settlement area within Section 6.1 comes from the latest NRI as updated through
2017, with the extension of the time series through 2023 based on assuming the settlement area is the
same as 2017. The NRI data is also harmonized with the USDA Forest Service (USFS) Forest Inventory
and Analysis (FIA) dataset, which is available through 2023, and the 2021 NLCD dataset. This process of
combiningthe datasets extends the time series to ensure that there is a complete and consistent
representation of land use data for all source categories in the LULUCF sector. Annual estimates of the
net C02 flux (Table 6-123) were developed based on estimates of annual settlement area and tree cover
derived from NLCD developed lands. Developed land, which was used to estimate tree cover in
settlement areas, is about six percent higher than the area categorized as settlements in the
representation of the U.S. land base developed for this report.
Percent Tree Cover in Settlement Areas
Percent tree cover in settlement area by state is needed to convert settlement land area to settlement
tree cover area. Converting to tree cover area is essential as tree cover, and thus carbon estimates, can
vary widely among states in settlement areas due to variations in the amount of tree cover area (e.g.,
Nowak and Greenfield 2018a). However, since the specific geography of settlement area is unknown
because they are based on NRI sampling methods, NLCD developed land was used to estimate the
percent tree cover to be used in settlement areas. The NLCD developed land cover classes 21-24
(developed, open space (21), low intensity (22), medium intensity (23), and high intensity (24)) were used
to estimate percent tree cover in settlement area by state (U.S. Department of Interior 2018; MRLC
2013).
a) "Developed, Open Space - areas with a mixture of some constructed materials, but mostly
vegetation in the form of lawn grasses. Impervious surfaces account for less than 20 percent of
total cover. These areas most commonly include large-lot single-family housing units, parks,
golf courses, and vegetation planted in developed settings for recreation, erosion control, or
aesthetic purposes." Plots designated as either park, recreation, cemetery, open space,
institutional or vacant land were classified as "Developed, Open Space".
b) "Developed, Low Intensity-areas with a mixture of constructed materials and vegetation.
Impervious surfaces account for 20 to 49 percent of total cover. These areas most commonly
include single-family housing units." Plots designated as single family or low-density residential
land were classified as "Developed, Low Intensity".
Land Use, Land-Use Change, and Forestry 6-201
-------
c) "Developed, Medium Intensity-areas with a mixture of constructed materials and vegetation.
Impervious surfaces account for 50 to 79 percent of the total cover. These areas most
commonly include single-family housing units." Plots designated as medium density residential,
other urban or mixed urban were classified as "Developed, Medium Intensity".
d) "Developed High Intensity-highly developed areas where people reside or work in high
numbers. Examples include apartment complexes, row houses and commercial/industrial.
Impervious surfaces account for 80 to 100 percent of the total cover." Plots designated as either
commercial, industrial, high density residential, downtown, multi-family residential, shopping,
transportation or utility were classified as "Developed, High Intensity".
As NLCD is known to underestimate tree cover (Nowak and Greenfield 2010), photo-interpretation of
tree cover within NLCD developed lands was conducted for the years of c. 2016 and 2020 using 1,000
random points to determine an average adjustment factor for NLCD tree cover estimates in developed
land and determine recent tree cover changes. This photo-interpretation of change followed methods
detailed in Nowak and Greenfield (2018b). Percent tree cover (%TC) in settlement areas by state was
estimated as:
%TC in state = state NLCD %TC x national photo-interpreted %TC / national NLCD %TC
Percent tree cover in settlement areas by year was set as follows:
• 1990 to 2011: used 2011 NLCD tree cover adjusted with 2011 photo-interpreted values
• 2012 to 2015: used 2011 NLCD tree cover adjusted with photo-interpreted values, which were
interpolated from values between 2011 and 2016
• 2016 to 2023: used 2016 NLCD tree cover adjusted with 2020 photo-interpreted values
Carbon Sequestration Density per Unit of Tree Cover
Methods for quantifying settlement tree biomass, carbon sequestration, and carbon emissions from
tree mortality and decomposition were taken directly from Nowak etal. (2013), Nowak and Crane
(2002), and Nowak (1994). In general, net carbon sequestration estimates followed three steps, each of
which is explained further in the paragraphs below. First, field data from cities and urban areas within
entire states were used to estimate carbon in tree biomass from field data on measured tree
dimensions. Second, estimates of annual tree growth and biomass increment were generated from
published literature and adjusted for tree condition, crown competition, and growing season to generate
estimates of gross carbon sequestration in settlement trees for all 50 states and the District of
Columbia. Third, estimates of carbon emissions due to mortality and decomposition were subtracted
from gross carbon sequestration estimates to obtain estimates of net carbon sequestration. Carbon
storage, gross and net sequestration estimates were standardized per unit tree cover based on tree
cover in the study area.
Settlement tree carbon estimates are based on published literature (Nowak et al. 2013; Nowak and
Crane 2002; Nowak 1994) as well as newer data from the i-Tree database95 and U.S. Forest Service
urban forest inventory data (e.g., Nowak et al. 2016, 2017) (Table 6-124). These data are based on
collected field measurements in several U.S. cities between 1989 and 2017. Carbon storage and
sequestration in these cities were estimated using the U.S. Forest Service's i-Tree Eco model (Nowak et
95 See http://www.itreetools.org.
6-202 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
al. 2008). This computer model uses standardized field data from randomly located plots, along with
local hourly air pollution and meteorological data, to quantify urban forest structure, monetary values of
the urban forest, and environmental effects, including total carbon stored and annual carbon
sequestration (Nowak et a I. 2013).
In each city, a random sample of plots were measured to assess tree stem diameter, tree height, crown
height and crown width, tree location, species, and canopy condition. The data for each tree were used
to estimate total dry-weight biomass using allometric models, a root-to-shoot ratio to convert
aboveground biomass estimates to whole tree biomass, and wood moisture content. Total dry weight
biomass was converted to carbon by dividing by two (50 percent carbon content). An adjustment factor
of 0.8 was used for open grown trees to account for settlement trees having less aboveground biomass
for a given stem diameter than predicted by allometric models based on forest trees (Nowak 1994).
Carbon storage estimates for deciduous trees include only carbon stored in wood. Estimated carbon
storage was divided by tree cover in the area to estimate carbon storage per square meter of tree cover.
Table 6-124: Carbon Storage (kg C/m2 tree cover), Gross and Net Sequestration (kg
C/m2 tree cover/year) and Tree Cover (percent) among Sampled U.S. Cities (see Nowak
et al. 2013)
Sequestration
Tree
City
Storage
SE
Gross
SE
Net
SE
Ratio3
Cover
SE
Adrian, Ml
12.17
1.88
0.34
0.04
0.13
0.07
0.36
22.1
2.3
Albuquerque, NM
5.61
0.97
0.24
0.03
0.20
0.03
0.82
13.3
1.5
Arlington, TX
6.37
0.73
0.29
0.03
0.26
0.03
0.91
22.5
0.3
Atlanta, GA
6.63
0.54
0.23
0.02
0.18
0.03
0.76
53.9
1.6
Austin, TX
3.57
0.25
0.17
0.01
0.13
0.01
0.73
30.8
1.1
Baltimore, MD
10.30
1.24
0.33
0.04
0.20
0.04
0.59
28.5
1.0
Boise, ID
7.33
2.16
0.26
0.04
0.16
0.06
0.64
7.8
0.2
Boston, MA
7.02
0.96
0.23
0.03
0.17
0.02
0.73
28.9
1.5
Camden, NJ
11.04
6.78
0.32
0.20
0.03
0.10
0.11
16.3
9.9
Casper, WY
6.97
1.50
0.22
0.04
0.12
0.04
0.54
8.9
1.0
Chester, PA
8.83
1.20
0.39
0.04
0.25
0.05
0.64
20.5
1.7
Chicago (region), IL
9.38
0.59
0.38
0.02
0.26
0.02
0.70
15.5
0.3
Chicago, IL
6.03
0.64
0.21
0.02
0.15
0.02
0.70
18.0
1.2
Corvallis, OR
10.68
1.80
0.22
0.03
0.20
0.03
0.91
32.6
4.1
El Paso, TX
3.93
0.86
0.32
0.05
0.23
0.05
0.72
5.9
1.0
Freehold, NJ
11.50
1.78
0.31
0.05
0.20
0.05
0.64
31.2
3.3
Gainesville, FL
6.33
0.99
0.22
0.03
0.16
0.03
0.73
50.6
3.1
Golden, CO
5.88
1.33
0.23
0.05
0.18
0.04
0.79
11.4
1.5
Grand Rapids, Ml
9.36
1.36
0.30
0.04
0.20
0.05
0.65
23.8
2.0
Hartford, CT
10.89
1.62
0.33
0.05
0.19
0.05
0.57
26.2
2.0
Houston, TX
4.55
0.48
0.31
0.03
0.25
0.03
0.83
18.4
1.0
lndianab
8.80
2.68
0.29
0.08
0.27
0.07
0.92
20.1
3.2
Jersey City, NJ
4.37
0.88
0.18
0.03
0.13
0.04
0.72
11.5
1.7
Kansasb
7.42
1.30
0.28
0.05
0.22
0.04
0.78
14.0
1.6
Kansas City (region), MO/KS
7.79
0.85
0.39
0.04
0.26
0.04
0.67
20.2
1.7
Lake Forest Park, WA
12.76
2.63
0.49
0.07
0.42
0.07
0.87
42.4
0.8
Land Use, Land-Use Change, and Forestry 6-203
-------
Sequestration
Tree
City
Storage
SE
Gross
SE
Net
SE
Ratio3
Cover
SE
Las Cruces, NM
3.01
0.95
0.31
0.14
0.26
0.14
0.86
2.9
1.0
Lincoln, NE
10.64
1.74
0.41
0.06
0.35
0.06
0.86
14.4
1.6
Los Angeles, CA
4.59
0.51
0.18
0.02
0.11
0.02
0.61
20.6
1.3
Milwaukee, Wl
7.26
1.18
0.26
0.03
0.18
0.03
0.68
21.6
1.6
Minneapolis, MN
4.41
0.74
0.16
0.02
0.08
0.05
0.52
34.1
1.6
Moorestown, NJ
9.95
0.93
0.32
0.03
0.24
0.03
0.75
28.0
1.6
Morgantown, WV
9.52
1.16
0.30
0.04
0.23
0.03
0.78
39.6
2.2
Nebraskab
6.67
1.86
0.27
0.07
0.23
0.06
0.84
15.0
3.6
New York, NY
6.32
0.75
0.33
0.03
0.25
0.03
0.76
20.9
1.3
North Dakotab
7.78
2.47
0.28
0.08
0.13
0.08
0.48
2.7
0.6
Oakland, CA
5.24
0.19
NA
NA
NA
NA
NA
21.0
0.2
Oconomowoc, Wl
10.34
4.53
0.25
0.10
0.16
0.06
0.65
25.0
7.9
Omaha, NE
14.14
2.29
0.51
0.08
0.40
0.07
0.78
14.8
1.6
Philadelphia, PA
8.65
1.46
0.33
0.05
0.29
0.05
0.86
20.8
1.8
Phoenix, AZ
3.42
0.50
0.38
0.04
0.35
0.04
0.94
9.9
1.2
Roanoke, VA
9.20
1.33
0.40
0.06
0.27
0.05
0.67
31.7
3.3
Sacramento, CA
7.82
1.57
0.38
0.06
0.33
0.06
0.87
13.2
1.7
San Francisco, CA
9.18
2.25
0.24
0.05
0.22
0.05
0.92
16.0
2.6
Scranton, PA
9.24
1.28
0.40
0.05
0.30
0.04
0.74
22.0
1.9
Seattle, WA
9.59
0.98
0.67
0.06
0.55
0.05
0.82
27.1
0.4
South Dakotab
3.14
0.66
0.13
0.03
0.11
0.02
0.87
16.5
2.2
Syracuse, NY
9.48
1.08
0.30
0.03
0.22
0.04
0.72
26.9
1.3
Tennesseeb
6.47
0.50
0.34
0.02
0.30
0.02
0.89
37.7
0.8
Washington, DC
8.52
1.04
0.26
0.03
0.21
0.03
0.79
35.0
2.0
Woodbridge, NJ
8.19
0.82
0.29
0.03
0.21
0.03
0.73
29.5
1.7
SE (Standard Error)
NA (Not Available)
a Ratio of net to gross sequestration.
bStatewide assessment of urban areas.
To determine gross sequestration rates, tree growth rates need to be estimated. Base growth rates were
standardized for open-grown trees in areas with 153 days of frost-free length based on measured data
on tree growth (Nowak et al. 2013). These growth rates were adjusted to local tree conditions based on
length of frost-free season, crown competition (as crown competition increased, growth rates
decreased), and tree condition (as tree condition decreased, growth rates decreased). Annual growth
rates were applied to each sampled tree to estimate gross annual sequestration-that is, the difference
in carbon storage estimates between year 1 and year (x + 1) represents the gross amount of carbon
sequestered. These annual gross carbon sequestration rates for each tree were then scaled up to city
estimates using tree population information. Total carbon sequestration was divided by total tree cover
to estimate a gross carbon sequestration density (kg C/m2 of tree cover/year). The area of assessment
for each city or state was defined by its political boundaries; parks and other forested urban areas were
thus included in sequestration estimates.
Where gross carbon sequestration accounts for all carbon sequestered, net carbon sequestration for
settlement trees considers carbon emissions associated with tree death and removals. The third step in
the methodology estimates net carbon emissions from settlement trees based on estimates of annual
6-204 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
mortality, tree condition, and assumptions about whether dead trees were removed from the site.
Estimates of annual mortality rates by diameter class and condition class were obtained from a study of
street-tree mortality (Nowak 1986). Different decomposition rates were applied to dead trees left
standing compared with those removed from the site. For removed trees, different rates were applied to
the removed/aboveground biomass in contrast to the belowground biomass (Nowak et al. 2002). The
estimated annual gross carbon emission rates for each plot were then scaled up to city estimates using
tree population information.
The full methodology development is described in the underlying literature, and key details and
assumptions were made as follows. The allometric models applied to the field data for the Nowak
methodology for each tree were taken from the scientific literature (see Nowak 1994, Nowak et al. 2002),
but if no allometric model could be found for the particular species, the average result for the genus or
botanical relative was used. The adjustment (0.8) to account for less live tree biomass in open-grown
urban trees was based on information in Nowak (1994). Measured tree growth rates for street (Frelich
1992; Fleming 1988; Nowak 1994), park (deVries 1987), and forest (Smith and Shifley 1984) trees were
standardized to an average length of growing season (153 frost free days) and adjusted for site
competition and tree condition. Standardized growth rates of trees of the same species or genus were
then compared to determine the average difference between standardized street tree growth and
standardized park and forest growth rates. Crown light exposure (CLE) measurements (number of sides
and/or top of tree exposed to sunlight) were used to represent forest, park, and open (street) tree growth
conditions. Local tree base growth rates were then calculated as the average standardized growth rate
for open-grown trees multiplied by the number of frost-free days divided by 153. Growth rates were then
adjusted for CLE. The CLE-adjusted growth rate was then adjusted based on tree condition to determine
the final growth rate. Assumptions for which dead trees would be removed versus left standing were
developed specific to each land use and were based on expert judgment of the authors. Decomposition
rates were based on literature estimates (Nowak et al. 2013).
Estimates of gross and net sequestration rates for each of the 50 states and the District of Columbia
(Table 6-125) were compiled in units of carbon sequestration per unit area of tree canopy cover. These
rates were used in conjunction with estimates of state settlement area and developed land percent tree
cover data to calculate each state's annual net carbon sequestration by urban trees. This method was
described in Nowak et al. (2013) and has been modified here to incorporate developed land percent tree
cover data.
Net annual carbon sequestration estimates were obtained for all 50 states and the District of Columbia
by multiplying the gross annual emission estimates by 0.73, the average ratio for net/gross sequestration
(Table 6-125). However, state specific ratios were used where available.
State Carbon Sequestration Estimates
The gross and net annual carbon sequestration values for each state were multiplied by each state's
settlement area of tree cover, which was the product of the state's settlement area and the state's tree
cover percentage based on NLCD developed land. The model used to calculate the total carbon
sequestration amounts for each state, can be written as follows:
Land Use, Land-Use Change, and Forestry 6-205
-------
Equation 6-1: Net State Annual Carbon Sequestration
Net state annual C sequestration (t C/yr) = Gross state sequestration rate (t C/ha/yr) x
Net to Gross state sequestration ratio x state settlement Area (hectares) x
% state tree cover in settlement area
The results for all 50 states and the District of Columbia are given in Table 6-125. This approach is
consistent with the default IPCC Gain-Loss methodology in IPCC (2006), although sufficient field data
are not yet available to separately determine interannual gains and losses in carbon stocks in the living
biomass of settlement trees. Instead, the methodology applied here uses estimates of net carbon
sequestration based on modeled estimates of decomposition, as given by Nowak et al. (2013).
Table 6-125: Estimated Annual Carbon Sequestration, Tree Cover, and Annual Carbon
Sequestration per Area of Tree Cover for settlement areas in the United States by State
and the District of Columbia (2023)
State
Gross Annual
Sequestration
(Metric Tons
C/Year)
Net Annual
Sequestration
(Metric Tons
C/Year)
Tree
Cover
(Percent)
Gross Annual
Sequestration
per Area of Tree
Cover
(kg C/m2/Year)
Net Annual
Sequestration
per Area of Tree
Cover
(kgC/m2/Year)
Net: Gross
Annual
Sequestration
Ratio
Alabama
2,298,740
1,675,034
53.0
0.376
0.274
0.73
Alaska
149,774
109,137
47.0
0.169
0.123
0.73
Arizona
165,296
120,447
4.5
0.388
0.283
0.73
Arkansas
1,323,132
964,133
48.5
0.362
0.264
0.73
California
2,015,159
1,468,395
16.8
0.426
0.311
0.73
Colorado
142,537
103,863
7.9
0.216
0.157
0.73
Connecticut
651,984
475,084
58.2
0.262
0.191
0.73
Delawa re
102,084
74,386
24.2
0.366
0.267
0.73
DC
12,904
9,403
24.8
0.366
0.267
0.73
Florida
4,645,833
3,385,300
39.9
0.520
0.379
0.73
Georgia
3,922,791
2,858,437
55.8
0.387
0.282
0.73
Hawaii
280,228
204,195
41.3
0.637
0.464
0.73
Idaho
59,753
43,541
7.3
0.201
0.146
0.73
Illinois
669,850
488,103
15.4
0.310
0.226
0.73
Indiana
479,906
443,749
16.9
0.274
0.254
0.92
Iowa
177,759
129,528
8.5
0.263
0.191
0.73
Kansas
287,941
224,067
10.7
0.310
0.241
0.78
Kentucky
988,545
720,327
36.4
0.313
0.228
0.73
Louisiana
1,587,645
1,156,877
46.6
0.435
0.317
0.73
Maine
451,851
329,253
55.0
0.242
0.176
0.73
Maryland
859,287
626,140
39.7
0.353
0.257
0.73
Massachusetts
1,098,217
800,243
56.7
0.278
0.203
0.73
Michigan
1,414,096
1,030,415
34.4
0.241
0.175
0.73
Minnesota
325,888
237,466
13.0
0.251
0.183
0.73
Mississippi
1,646,019
1,199,412
56.8
0.377
0.275
0.73
Missouri
881,832
642,569
22.9
0.313
0.228
0.73
Montana
45,446
33,116
4.8
0.201
0.147
0.73
6-206 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Gross Annual
Net Annual
Gross Annual
Net Annual
Sequestration
Sequestration
Net: Gross
Sequestration
Sequestration
Tree
per Area of Tree
per Area of Tree
Annual
(Metric Tons
(Metric Tons
Cover
Cover
Cover
Sequestration
State
C/Year)
C/Year)
(Percent)
(kg C/m2/Year)
(kgC/m2/Year)
Ratio
Nebraska
97,669
82,419
7.3
0.261
0.220
0.84
Nevada
36,193
26,373
4.8
0.226
0.165
0.73
New Hampshire
395,478
288,174
58.8
0.238
0.174
0.73
New Jersey
964,812
703,034
40.4
0.321
0.234
0.73
New Mexico
189,022
137,735
10.1
0.288
0.210
0.73
New York
1,610,390
1,173,450
39.6
0.263
0.192
0.73
North Carolina
3,482,601
2,537,682
53.7
0.341
0.249
0.73
North Dakota
18,729
8,899
1.7
0.244
0.116
0.48
Ohio
1,277,128
930,611
28.0
0.271
0.198
0.73
Oklahoma
717,179
522,590
21.9
0.364
0.265
0.73
Oregon
673,763
490,954
39.5
0.265
0.193
0.73
Pennsylvania
1,906,477
1,389,202
39.8
0.267
0.195
0.73
Rhode Island
128,574
93,688
49.5
0.283
0.206
0.73
South Carolina
2,074,371
1,511,541
53.3
0.370
0.269
0.73
South Dakota
29,314
25,420
2.8
0.258
0.224
0.87
Tennessee
1,686,697
1,508,105
40.7
0.332
0.297
0.89
Texas
4,412,562
3,215,321
28.2
0.403
0.294
0.73
Utah
119,655
87,190
11.6
0.235
0.172
0.73
Vermont
188,284
137,198
50.1
0.234
0.170
0.73
Virginia
2,125,574
1,548,851
52.4
0.321
0.234
0.73
Washington
1,144,137
833,703
37.2
0.282
0.206
0.73
West Virginia
777,972
566,889
63.5
0.264
0.192
0.73
Wisconsin
713,396
519,833
25.6
0.246
0.180
0.73
Wyoming
29,501
21,496
4.7
0.199
0.145
0.73
Total
51,483,975
37,912,981
Note: Totals may not sum due to independent rounding.
Uncertainty
Uncertainty associated with changes in carbon stocks in settlement trees includes the uncertainty
associated with settlement area, percent tree cover in developed land and how well it represents
percent tree cover in settlement areas, and estimates of gross and net carbon sequestration for each of
the 50 states and the District of Columbia. A ten percent uncertainty was associated with settlement
area estimates based on expert judgment. Uncertainty associated with estimates of percent settlement
tree coverage for each of the 50 states was based on standard error associated with the photo-
interpretation of national tree cover in developed lands. Uncertainty associated with estimates of gross
and net carbon sequestration for each of the 50 states and the District of Columbia was based on
standard error estimates for each of the state-level sequestration estimates (Table 6-126). These
estimates are based on field data collected in each of the 50 states and the District of Columbia, and
uncertainty in these estimates increases as they are scaled up to the national level.
Land Use, Land-Use Change, and Forestry 6-207
-------
Additional uncertainty is associated with the biomass models, conversion factors, and decomposition
assumptions used to calculate carbon sequestration and emission estimates (Nowak et al. 2002). These
results also exclude changes in soil carbon stocks, and there is likely some overlap between the
settlement tree carbon estimates and the forest tree carbon estimates (e.g., Nowak et al. 2013). Due to
data limitations, settlement soil flux is not quantified as part of this analysis, while reconciliation of
settlement tree and forest tree estimates will be addressed through the land-representation effort
described in the Planned Improvements section of this chapter.
A Monte Carlo (Approach 2) uncertainty analysis was applied to estimate the overall uncertainty of the
sequestration estimate in 2023. The results of this quantitative uncertainty analysis are summarized in
Table 6-126. The change in carbon stocks in settlement trees in 2023 was estimated to be between -
208.1 and -67.0 MMT C02 Eq. at a 95 percent confidence level. This analysis indicates a range of 50
percent more sequestration to 52 percent less sequestration than the 2023 flux estimate of -139.0 MMT
C02 Eq.
Table 6-126: Approach 2 Quantitative Uncertainty Estimates for Net C02 Flux from
Changes in Carbon Stocks in Settlement Trees (MMT C02 Eq. and Percent)
2023 Flux
Estimate
Source Gas (MMTC02Eq.)
Uncertainty Range Relative to Flux Estimate"
(MMT CO2 Eq.)
(%)
Lower Upper
Bound Bound
Lower Upper
Bound Bound
Changes in C Stocks in Settlement Trees CO2 (139.0)
(208.1) (67.0)
-50% +52%
a Range of C stock change estimates predicted by Monte Carlo stochastic simulation with a 95 percent confidence interval.
Note: Parentheses indicate negative values or net sequestration.
QA/QC and Verification
Tier 1 and Tier 2 QA/QC activities were conducted consistent with the U.S. QA/QC plan. Source-specific
quality control measures for settlement trees included checking input data, documentation, and
calculations to ensure data were properly handled through the inventory process. Errors that were found
during this process were corrected as necessary.
Recalculations Discussion
The compilation methods remained the same in the latest Inventory relative to the previous Inventory.
New data from the NLCD resulted in a small decrease in the settlement area for 2022, leading to no
substantial change in the net carbon sequestration (Table 6-127).
Table 6-127: Recalculations of the Settlement Tree Categories
Category
2022 Estimate,
Previous Inventory
2022 Estimate,
Current Inventory
2023 Estimate,
Current Inventory
Settlement Area (km2)
471,851
471,617
473,700
Settlement Tree Coverage (km2)
152,442
152,349
153,047
Net C Flux (MMT C)
(37.8)
(37.7)
(37.9)
Net CO2 Flux MMT CO2 Eq.
(138.5)
(138.4)
(139.0)
Note: Totals may not sum due to independent rounding.
6-208 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Planned Improvements
The following are planned improvements for future Settlement Tree inventories:
A consistent representation of the managed land base in the United States is discussed in Section 6.1,
and describes reconciliation in the overlap between settlement trees and the forest land categories.
Estimates for settlement trees are based on tree cover in settlement areas. Work is needed to clarify
how much of this settlement area tree cover may also be accounted for in "forest" area assessments as
some of these forests maybe adjacent to settlement areas. For example, "forest" as defined by the FIA
program fall within urban areas. Nowaket al. (2013) estimates that 1.5 percent of forest plots measured
by the FIA program fall within land designated as Census urban, suggesting that approximately 1.5
percent of the carbon reported in the forest source category might also be counted in the urban areas.
The potential overlap with settlement areas is unknown at this time but research is underway to develop
spatially explicit and spatially continuous land representation products which will eliminate the
potential for double counting. Future research may also enable more complete coverage of changes in
the carbon stock of trees for all settlements land.
As described in the forest land remaining forest land chapter, recent U.S. Forest Service improvements,
specifically the NSVB method and updated carbon coefficients, will need to be reflected in the
estimates for settlement trees. This will likely cause an increase in the volume, biomass, and carbon
stocks associated with settlement trees. Please see Chapter 6.2 and Annex 3.14 for more information.
EPA will provide more details on expected implementation timing in a future Inventory.
In addition to these planned improvements, EPA anticipates updating the photo-interpretation of tree
cover within NLCD developed lands to 2021 and refining carbon sequestration and storage rates with
Urban Forest Inventory and Analysis values as they become available.
N20 Emissions from Settlement Soils (Source Category
4E1)
Of the synthetic N fertilizers applied to soils in the United States, approximately 1 to 2 percent are
currently applied to lawns, golf courses, and other landscaping within settlement areas, and contributes
to soil N20 emissions. The area of settlements is considerably smaller than other land uses that are
managed with fertilizer, particularly cropland soils, and therefore, settlements account for a smaller
proportion of total synthetic fertilizer application in the United States. In addition to synthetic N
fertilizers, a portion of surface applied biosolids (i.e., treated sewage sludge) is used as an organic
fertilizer in settlement areas, and drained organic soils (i.e., soils with high organic matter content,
known as histosols) also contribute to emissions of soil N20.
N additions to soils result in direct and indirect N20 emissions. Direct emissions occur on-site due to
the N additions in the form of synthetic fertilizers and biosolids as well as enhanced mineralization of N
in drained organic soils. Indirect emissions result from fertilizer and biosolids N that is transformed and
transported to another location in a form other than N20 (i.e., volatilization of ammonia [NH3] and
nitrogen oxide [NOx], and leaching/runoff of nitrate [N03~]), and later converted into N20 at the off-site
location. The indirect emissions are assigned to settlements because the management activity leading
to the emissions occurred in settlements.
Land Use, Land-Use Change, and Forestry 6-209
-------
Total N20 emissions from soils in settlements remaining settlements96 are 2.5 MMT C02 Eq. (10 kt of
N20) in 2023. There is an overall increase of 23.1 percent from 1990 to 2023 due to an expanding
settlement area leading to more synthetic N fertilizer applications that peaked in the mid-2000s. Inter-
annual variability in these emissions is directly attributable to variability in total synthetic fertilizer
consumption, area of drained organic soils, and biosolids applications in the United States. Emissions
from this source are summarized in Table 6-128 and Table 6-129.
Table 6-128: N20 Emissions from Soils in Settlements Remaining Settlements (MMT
C02 Eq.)
1990 |
2005
2019
2020
2021
2022
2023
Direct N2O Emissions from Soils
1.7 I
2.6
2.1
2.2
2.2
2.2
2.2
Synthetic Fertilizers
0.8
1.5
0.8
0.8
0.8
0.8
0.8
Biosolids
0.2
0.2
0.2
0.2
0.2
0.2
0.2
Drained Organic Soils
0.8
1.0
1.2
1.2
1.2
1.2
1.2
Indirect N2O Emissions from Soils
0.3 |
0.5 |
0.3
0.3
0.3
0.3
0.3
Total
2.1 |
3.1 |
2.5
2.5
2.5
2.5
2.5
Note: Totals may not sum due to independent rounding.
Table 6-129: N20 Emissions from Soils in Settlements Remaining Settlements (kt N20)
1990 |
2005
2019
2020
2021
2022
2023
Direct N2O Emissions from Soils
7 I
10
8
8
8
8
8
Synthetic Fertilizers
3
5
3
3
3
3
3
Biosolids
1
1
1
1
1
1
1
Drained Organic Soils
3
4
4
4
4
4
4
Indirect N2O Emissions from Soils
1 |
2 I
1
1
1
1
1
Total
8 I
12 9
9
10
10
10
Note: Totals may not sum due to independent rounding.
Methodology and Time-Series Consistency
For settlement soils, the IPCC Tier 1 approach is used to estimate soil N20 emissions from synthetic N
fertilizer, biosolids additions, and drained organic soils. Estimates of direct N20 emissions from soils in
settlements are based on the amount of N in synthetic commercial fertilizers applied to settlement
soils, the amount of N in biosolids applied to non-agricultural land and surface disposal (see Section 7.2
for a detailed discussion of the methodology for estimating biosolids available for non-agricultural land
application), and the area of drained organic soils within settlements.
Nitrogen applications to settlement soils are estimated using data compiled by the USGS (Brakebill and
Gronberg 2017). The USGS estimated on-farm and non-farm fertilizer use is based on sales records at
the county level from 1987 through 2012 (Brakebill and Gronberg 2017). Non-farm N fertilizer is assumed
to be applied to settlements and forest lands; values for 2013 through 2017 are based on 2012 values
adjusted for total annual total N fertilizer sales in the United States (AAPFCO 2016 through 2022)
because there are no activity data on non-farm application after 2012. Settlement application is
96 Estimates of soil N2O for settlements remaining settlements include emissions from land converted to settlements
because it was not possible to separate the activity data.
6-210 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
calculated by subtracting forest application from total non-farm fertilizer use. Since the total N fertilizer
sales is only available through 2017 (AAPFCO 2022), the amount of synthetic fertilization from 2018 to
2023 is determined using a data splicing method (see Box 6-3 in Section 6.4 cropland remaining
cropland). This method is based on a linear regression model with moving-average (ARMA) errors
(Brockwelland Davis 2016) using the 1990 to 2017 fertilization data. This method is type of a linear
extrapolation, which is a standard data splicing method for estimating emissions at the end of a time
series (IPCC 2006). To estimate direct N20 for the time series, the total amount of fertilizer N applied to
settlements is multiplied by the IPCC default emission factor (1 percent) (IPCC 2006) for 1990 to 2023.
Biosolids applications are derived from national data on biosolids generation, disposition, and N
content (see Section 7.2 for further detail). The total amount of N resulting from these sources is
multiplied by the IPCC default emission factor for applied N (one percent) to estimate direct N20
emissions (IPCC 2006) for 1990 to 2023.
The IPCC (2006) Tier 1 method is also used to estimate direct N20 emissions due to drainage of organic
soils in settlements at the national scale. Estimates of the total area of drained organic soils are
obtained from the 2017 National Resources Inventory (NRI) (USDA-NRCS 2020) using soils data from the
Soil Survey Geographic Database (SSURGO) (Soil Survey Staff 2020). The NRI time series has been
extended through 2020 using the National Land Cover Dataset (Yang et al. 2018). The areas have been
modified through a process in which the Forest Inventory and Analysis (FIA) survey data are harmonized
with the NRI data (Nelson et al. 2020). This process ensures that the land use areas are consistent
across all land use categories (see Section 6.1 for more information). All settlements occurring on
organic soil are assumed to be drained for the purposes of approximating greenhouse gas emissions. To
estimate annual emissions from 1990 to 2020, the total area is multiplied by the IPCC default emission
factor for temperate regions (IPCC 2006). The annual emissions for 2021 to 2023 are estimated using a
data splicing method (see Box 6-3 in Section 6.4 cropland remaining cropland). This Inventory does not
include soil N20 emissions from drainage of organic soils in Alaska and federal lands, although this is a
planned improvement for a future Inventory.
For indirect emissions, the total N applied from fertilizer and biosolids is multiplied by the IPCC default
factors of 10 percent for volatilization and 30 percent for leaching/runoff to calculate the amount of N
volatilized and the amount of N leached/runoff. The amount of N volatilized is multiplied by the IPCC
default factor of one percent for the portion of volatilized N that is converted to N20 off-site and the
amount of N leached/runoff is multiplied by the IPCC default factor of 0.075 percent for the portion of
leached/runoff N that is converted to N20 off-site. The resulting estimates are summed to obtain total
indirect emissions from 1990 to 2023 for biosolids and synthetic fertilization.
In order to ensure time-series consistency, the same methods are applied from 1990 to 2023 for
biosolids. For synthetic fertilizer, a data splicing method is used to approximate fertilizer application for
the remainder of the 2018 to 2023 time series and then used to estimate emissions. For drainage of
organic soils, the methods described above are applied for 1990 to 2020, and a data splicing method is
used to approximate emissions for the remainder of the 2021 to 2023 time series (see Box 6-3 in Section
6.4 Cropland Remaining Cropland). This method is based on a linear regression model with moving-
average (ARMA) errors (Brockwell and Davis 2016)) using the 1990 to 2020 emissions data. This method
is type of a linear extrapolation, which is a standard data splicing method for estimating emissions at the
end of a time series (IPCC 2006). The time series will be recalculated in a future Inventory with the
methods described previously for drainage of organic soils.
Land Use, Land-Use Change, and Forestry 6-211
-------
Uncertainty
The amount of N20 emitted from settlement soils depends not only on N inputs and area of drained
organic soils, but also on a large number of variables that can influence rates of nitrification and
denitrification, including organic C availability; rate, application method, and timing of N input; oxygen
gas partial pressure; soil moisture content; pH; temperature; and irrigation/watering practices. The
effect of the combined interaction of these variables on N20 emissions is complex and highly uncertain.
The IPCC default methodology does not explicitly incorporate these variables, except variation in the
total amount of fertilizer N and biosolids application, which leads to uncertainty in the results.
Uncertainties exist in both the fertilizer N and biosolids application rates in addition to the emission
factors. Uncertainty in fertilizer N application is assigned a default level of ±50 percent.97 For emissions
from drained organic soils, the total uncertainty was quantified with two variance components (Ogle et
al. 2010) that are combined using simple error propagation methods provided by the IPCC (2006), i.e., by
takingthe square root of the sum of the squares of the standard deviations of the uncertain quantities.
The first variance component is associated with uncertainty in the emission factor, and the second
variance component is associated with scaling of the data from the NRI survey to the entire area of
drained organic soils in settlements remaining settlements, and is computed using a standard variance
estimator for a two-stage sample design (Sarndal et al. 1992). There is also additional uncertainty
associated with the fit of the linear regression model for the data splicing methods that was
incorporated into the analysis for the latter part of the time series.
Uncertainty is propagated through the calculations of N20 emissions from fertilizer N and drainage of
organic soils based on a Monte Carlo analysis. The results are combined with the uncertainty in N20
emissions from the biosolids application using simple error propagation methods (IPCC 2006). The
results are summarized in Table 6-130. Direct N20 emissions from soils in settlements remaining
settlements in 2023 are estimated to be between 1.2 and 3.4 MMT C02 Eq. at a 95 percent confidence
level. This indicates a range of 47 percent below to 54 percent above the 2023 emission estimate of 2.2
MMT C02 Eq. Indirect N20 emissions in 2023 are between 0.1 and 1.1 MMT C02 Eq., ranging from 76
percent below to 218 percent above the estimate of 0.3 MMT C02 Eq.
Table 6-130: Quantitative Uncertainty Estimates of N20 Emissions from Soils in
Settlements Remaining Settlements (MMT C02 Eq. and Percent)
Uncertainty Range Relative to Emission Estimate"
2023
(MMT CO2 Eq.)
(%)
Emissions
Lower
Upper
Lower
Upper
Source
Gas
(MMT CO2 Eq.)
Bound
Bound
Bound
Bound
Settlements Remaining Settlements
Direct N2O Emissions from Soils
N2O
2.2
1.2
3.4
-47%
+54%
Indirect N2O Emissions from Soils
N2O
0.3
0.1
1.1
-76%
+218%
a Range of emission estimates is a 95 percent confidence interval.
Note: These estimates include direct and indirect N20 emissions from settlements remaining settlements and land converted to
settlements because it was not possible to separate the activity data.
97 No uncertainty is provided with the USGS fertilizer consumption data (Brakebill and Gronberg 2017) so a conservative
±50 percent is used in the analysis. Biosolids data are also assumed to have an uncertainty of ±50 percent.
6-212 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
QA/QC and Verification
The spreadsheet containing fertilizer, drainage of organic soils, and biosolids applied to settlements and
calculations for N20 and uncertainty ranges have been checked consistent with the U.S. Inventory
QA/QC plan, which is in accordance with Volume 1, Chapter 6 of the 2006IPCC Guidelines (see Annex 8
for more details). An error was identified where drained organic soil emissions were double counted.
This error was corrected.
Recalculations Discussion
There are no recalculations associated with N20 emissions from settlements from 1990 to 2022
compared to the previous Inventory.
Planned Improvements
The following are key planned improvements for N20 emissions from soils in settlements remaining
settlements:
• Estimating soil N20 emissions from drainage of organic soils in Alaska and federal lands in order
to provide a complete inventory of emissions for this category.
• Updating data on fertilizer amounts from 2018 to 2023 after data are released for the latter part
of the time series.
These improvements will be incorporated into a future Inventory, pending prioritization of resources.
Changes in Yard Trimmings and Food Scrap Carbon
Stocks in Landfills (Source Category 4E1)
In the United States, yard trimmings (i.e., grass clippings, leaves, and branches) and food scraps
account for a significant portion of the municipal waste stream, and a large fraction of the collected yard
trimmings and food scraps are put in landfills. A portion of the carbon contained in landfilled yard
trimmings and food scraps can be stored for very long periods.
Carbon storage estimates within the Inventory are associated with particular land uses. For example,
harvested wood products are reported under forest land remaining forest land because these wood
products originated from the forest ecosystem. Similarly, carbon stock changes in yard trimmings and
food scraps are reported under settlements remaining settlements because the bulk of the carbon,
which comes from yard trimmings, originates from settlement areas. While the majority of food scraps
originate from cropland and grassland, in this Inventory they are reported with the yard trimmings in the
settlements remaining settlements section. Additionally, landfills are considered part of the managed
land base under settlements (see Section 6.1), and reporting these carbon stock changes that occur
entirely within landfills fits most appropriately within the settlements remaining settlements section.
The CH4 emissions resulting from anaerobic decomposition of yard trimmings and food scraps in
landfills are reported in the Waste chapter, see Section 7.1.
The estimated amount of yard trimmings collected annually has stagnated since 1990 and the fraction
that is landfilled has been declining since 1990. From 1970 to 1990, yard trimmings collected for
disposal increased by about 51 percent. In 1990, over 53 million metric tons (wet weight) of yard
Land Use, Land-Use Change, and Forestry 6-213
-------
trimmings and food scraps are estimated to have been generated (i.e., put at the curb for collection to be
taken to disposal sites or to composting facilities) (EPA 2020). Since then, programs banning or
discouraging yard trimmings disposal in landfills have led to an increase in backyard composting and
the use of mulching mowers, and consequently a slowing of year-over-year increases in the tonnage of
yard trimmings generated. From 1990 to 2023, yard trimmings collected for disposal are estimated to
have increased 1.1. percent. At the same time, an increase in the number of municipal composting
facilities has reduced the proportion of collected yard trimmings that are discarded in landfills per
year—from 72 percent in 1990 to 30 percent in 2023. The net effect of the slight increase in generation
and the increase in composting is a 58 percent decrease in the quantity of yard trimmings disposed of in
landfills since 1990. Composting trends and emissions estimations are presented in the Waste chapter,
Section 7.3 Composting.
Food scrap generation has grown by an estimated 165 percent since 1990. Though the proportion of
total food scraps generated that are eventually discarded in landfills has decreased from an estimated
82 percent in 1990 to 55 percent in 2023, the tonnage disposed of in landfills has increased
considerably (by an estimated 78 percent) due to the increase in food scrap generation. Although the
total tonnage of food scraps disposed of in landfills has increased from 1990 to 2023, the difference in
the amount of food scraps added from one year to the next generally decreased, and consequently the
annual net changes in carbon stock from food scraps have generally decreased as well (as shown in
Table 6-131 and Table 6-132). Landfilled food scraps decompose over time, producing CH4 and C02.
Decomposition happens at a higher rate initially, then decreases. As decomposition decreases, the
carbon stock becomes more stable. Because the cumulative carbon stock left in the landfill from
previous years is (1) not decomposing as much as the carbon introduced from food scraps in a single
more recent year; and (2) is much larger than the carbon introduced from food scraps in a single more
recent year, the total carbon stock in the landfill is primarily driven by the more stable "older" carbon
stock, thus resulting in decreasing annual changes in later years.
Overall, the decrease in the landfill disposal rate of yard trimmings has more than compensated for the
increase in food scrap disposal in landfills, and the net result is a decrease in the annual net change in
landfill carbon storage from 24.5 MMTC02 Eq. (6.7 MMTC) in 1990 to 11.7 MMTC02 Eq. (3.2 MMTC) in
2023 (Table 6-131 and Table 6-132), a decrease of 52.2 percent over the time series.
Table 6-131: Net Changes in Yard Trimmings and Food Scrap Carbon Stocks in
Landfills (MMTC02 Eq.)
Carbon Pool
1990
2005
2019
2020
2021
2022
2023
Yard Trimmings
(20.1)1
(7.5)
(8.2)
(8.2)
(8.2)
(8.2)
(8.2)
Grass
(1-7)
(0.6)
(0.8)
(0.8)
(0.8)
(0.8)
(0.8)
Leaves
(8.7)
(3.4)
(3.8)
(3.8)
(3.8)
(3.8)
(3.7)
Branches
(9.8)
(3.4)
(3.7)
(3.7)
(3.7)
(3.6)
(3.6)
Food Scraps
(4.4)
(3.9)
(4.8)
(4.5)
(4.3)
(4.1)
(3.5)
Total Net Flux
(24.5) |
(11.4)
(13.1)
(12.8)
(12.5)
(12.3)
(11.7)
Notes: Totals may not sum due to independent rounding. Parentheses indicate net sequestration.
Table 6-132: Net Changes in Yard Trimmings and Food Scrap Carbon Stocks in
Landfills (MMTC)
Carbon Pool
1990 |
2005
2019
2020
2021
2022
2023
6-214 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Yard Trimmings
(5.5)
(2.0)
(2.2)
(2.2)
(2.2)
(2.2)
(2.2)
Grass
(0.5)
(0.2)
(0.2)
(0.2)
(0.2)
(0.2)
(0.2)
Leaves
(2.4)
(0.9)
(1.0)
(1.0)
(1.0)
(1.0)
(1.0)
Branches
(2.7)
(0.9)
(1.0)
(1.0)
(1.0)
(1.0)
(1.0)
Food Scraps
(1.2)
(1.1)
(1.3)
(1.2)
(1.2)
(1.1)
(0.9)
Total Net Flux
(6.7)
(3.1)
(3.6)
(3.5)
(3.4)
(3.4)
(3.2)
Notes: Totals may not sum due to independent rounding. Parentheses indicate net sequestration.
Methodology and Time-Series Consistency
When waste of biogenic origin (such as yard trimmings and food scraps) is landfilled and do not
completely decompose, the carbon that remains is effectively removed from the carbon cycle. Empirical
evidence indicates that yard trimmings and food scraps do not completely decompose in landfills
(Barlaz 1998, 2005, 2008; De la Cruz and Barlaz 2010), and thus the stock of carbon in landfills can
increase, with the net effect being removal of carbon from the atmosphere. Estimates of the net carbon
flux resulting from landfilled yard trimmings and food scraps were developed by estimating the change
in landfilled carbon stocks between inventory years and uses a country-specific methodology based on
the methodology for estimating the amount of harvested wood products stored in solid waste disposal
systems that is provided in the Land Use, Land-Use Change, and Forestry sector in IPCC (2003) and the
2006 IPCC Guidelines for National Greenhouse Gas Inventories (IPCC 2006). Carbon stock estimates
were calculated by determining the mass of landfilled carbon resulting from yard trimmings and food
scraps discarded in a given year; adding the accumulated landfilled carbon from previous years; and
subtracting the mass of carbon thatwas landfilled in previousyears and has since decomposed and
been emitted as C02 and CH4.
To determine the total landfilled carbon stocks for a given year, the following data and factors were
assembled:
1. The composition of the yard trimmings (i.e., the proportion of grass, leaves and branches);
2. The mass of yard trimmings and food scraps discarded in landfills;
3. The carbon storage factor of the landfilled yard trimmings and food scraps; and
4. The rate of decomposition of the degradable carbon.
The composition of yard trimmings was assumed to be 30 percent grass clippings, 40 percent leaves,
and 30 percent branches on a wet weight basis (Oshins and Block 2000). The yard trimmings were
subdivided, because each component has its own unique adjusted carbon storage factor (i.e., based on
differences in moisture content and carbon content) and rate of decomposition. The mass of yard
trimmings and food scraps disposed of in landfills was estimated by multiplying the quantity of yard
trimmings and food scraps discarded by the proportion of discards managed in landfills. Data on
discards (i.e., the amount generated minus the amount diverted to centralized composting facilities) for
both yard trimmings and food scraps were taken primarily from Advancing Sustainable Materials
Management: Facts and Figures 2018 (EPA 2020), which provides data for 1960,1970,1980,1990, 2000,
2005, 2010, 2015, 2017 and 2018. To provide data for some of the missing years, detailed backup data
were obtained from the 2012, 2013, and 2014, 2015, and 2017 versions of the Advancing Sustainable
Materials Management: Facts and Figures reports (EPA 2019), as well as historical data tables that EPA
developed for 1960 through 2012 (EPA 2016). Remaining years in the time series for which data were not
Land Use, Land-Use Change, and Forestry 6-215
-------
provided were estimated using linear interpolation. Since the Advancing Sustainable Materials
Management: Facts and Figures reports for 2019 through 2023 were unavailable, landfilled material
generation, recovery, and disposal data for 2019 through 2023 were proxied equal to 2018 values.
The amount of carbon disposed of in landfills each year, starting in 1960, was estimated by converting
the discarded landfilled yard trimmings and food scraps from a wet weight to a dry weight basis, and
then multiplying by the initial (i.e., pre-decomposition) carbon content (as a fraction of dry weight). The
dry weight of landfilled material was calculated using dry weight to wet weight ratios (Tchobanoglous et
al. 1993, cited by Barlaz 1998) and the initial carbon contents and the carbon storage factors were
determined by Barlaz (1998, 2005, 2008).
The amount of carbon remaining in the landfill for each subsequent year was tracked based on a simple
model of carbon fate based on a laboratory experiment simulating decomposition of landfilled biogenic
materials by methanogenic microbes (Barlaz 1998, 2005, 2008). Carbon remaining in landfilled
materials is expressed as a proportion of initial carbon content, shown in the row labeled "C Storage
Factor, Proportion of Initial C Stored (%)" in Table 6-133.
The modeling approach applied to simulate U.S. landfill carbon flows builds on the findings of Barlaz
(1998, 2005, 2008). The proportion of carbon stored is assumed to persist in landfills. The remaining
portion is assumed to degrade over time, resulting in emissions of CH4 and C02.98 The degradable
portion of the carbon is assumed to decay according to first-order kinetics. The decay rates for each of
the materials are shown in Table 6-133.
The first-order decay rates, k, for each waste component are derived from De la Cruz and Barlaz (2010):
• De la Cruz and Barlaz (2010) calculate first-order decay rates using laboratory data published in
Eleazer et al. (1997), and a correction factor, f, is calculated so that the weighted average decay
rate for all components is equal to the EPA AP-42 default decay rate (0.04) for mixed municipal
solid waste (MSW) for regions that receive more than 25 inches of rain annually (EPA 1995).
Because AP-42 values were developed using landfill data from approximately 1990, De la Cruz
and Barlaz used 1990 waste composition for the United States from EPA's Characterization of
Municipal Solid Waste in the United States: 1990 Update (EPA 1991) to calculate f. De la Cruz
and Barlaz multiplied this correction factor by the Eleazer et al. (1997) decay rates of each waste
component to develop field-scale first-order decay rates.
• De la Cruz and Barlaz (2010) also use other assumed initial decay rates for mixed MSW in place
of the AP-42 default value based on different types of environments in which landfills in the
United States are located, including dry conditions (less than 25 inches of rain annually, k= 0.02)
and bioreactor landfill conditions (moisture is controlled for rapid decomposition, Ar=0.12).
Similar to the methodology in the Landfills section of the Inventory (Section 7.1), which estimates CH4
emissions, the overall MSW decay rate is estimated by partitioning the U.S. landfill population into three
categories based on annual precipitation ranges of: (1) Less than 20 inches of rain per year, (2) 20 to 40
inches of rain per year, and (3) greater than 40 inches of rain per year. These correspond to overall MSW
decay rates of 0.020, 0.038, and 0.057 year"1, respectively. De la Cruz and Barlaz (2010) calculate
component-specific decay rates corresponding to the first value (0.020 year-1), but not for the other two
overall MSW decay rates.
98 The CH4 emissions resulting from anaerobic decomposition of yard trimmings and food scraps in landfills are reported in
the Waste chapter, Section 7.1—Landfills.
6-216 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
To maintain consistency between landfill-related methodologies across the Inventory, EPA developed
correction factors (f) for decay rates of 0.038 and 0.057 year"1 through linear interpolation. A weighted
national average component-specific decay rate is calculated by assuming that waste generation is
proportional to population (the same assumption used in the landfill methane emission estimate),
based on population data from the 2000 U.S. Census. The percent of census population is calculated for
each of the three categories of annual precipitation (noted in the previous paragraph); the population
data are used as a surrogate for the number of landfills in each annual precipitation category.
Precipitation range percentages weighted by population are updated over time as new Census data are
available, to remain consistent with percentages used in the Waste chapter, Section 7.1 landfills. The
component-specific decay rates are shown in Table 6-133.
De la Cruz and Barlaz (2010) also use other assumed initial decay rates for mixed MSW in place of the
AP-42 default value based on different types of environments in which landfills in the United States are
located, including dry conditions (less than 25 inches of rain annually, k=0.02) and bioreactor landfill
conditions (moisture is controlled for rapid decomposition, Ar=0.12).
For each of the four materials (grass, leaves, branches, food scraps), the stock of carbon in landfills for
any given year is calculated according to Equation 6-2:
Equation 6-2: Total Carbon Stockfor Yard Trimmings and Food Scraps in Landfills
t
LFCit = x (!- MCi)x ICCi x {[C5i x /CC*]+ [C1- (CSi x ICCd)x e-W"*)]}
n
where,
t = Year for which carbon stocks are being estimated (year),
/' = Waste type for which carbon stocks are being estimated (grass, leaves,
branches, food scraps),
LFCU = Stock of carbon in landfills in year t, for waste /' (metric tons),
Wj,n = Mass of waste /'disposed of in landfills in year n (metric tons, wet weight),
n = Year in which the waste was disposed of (year, where 1960 < n < t),
MC, = Moisture content of waste/'(percent of water),
CS, = Proportion of initial carbon that is stored for waste /' (percent),
ICC, = Initial carbon content of waste/'(percent),
e = Natural logarithm, and
k = First-order decay rate for waste/', (year-1).
For a given year t, the total stock of carbon in landfills (TLFCt) is the sum of stocks across all four
materials (grass, leaves, branches, food scraps). The annual flux of carbon in landfills (Ft) for year t is
calculated in as the change in carbon stock compared to the preceding year according to Equation 6-3:
Equation 6-3: Carbon Stock Annual Flux for Yard Trimmings and Food Scraps in
Landfills
Ft = TLFCt ~ TLFC(t-i)
Thus, as seen in Equation 6-2, the carbon placed in a landfill in yearn is tracked for each year tthrough
the end of the inventory period. For example, disposal of food scraps in 1960 resulted in depositing
about 1,135,000 metric tons of carbon in landfills. Of this amount, 16 percent (179,000 metric tons) is
persistent; the remaining 84 percent (956,000 metric tons) is degradable. By 1965, more than half of the
Land Use, Land-Use Change, and Forestry 6-217
-------
degradable portion (507,000 metric tons) decomposes, leaving a total of 628,000 metric tons (the
persistent portion, plus the remainder of the degradable portion).
Continuing the example, by 2023, the total food scraps carbon originally disposed of in 1960 had
declined to 179,000 metric tons (i.e., virtually all degradable carbon had decomposed). By summing the
carbon remaining from 1960 with the carbon remaining from food scraps disposed of in subsequent
years (1961 through 2023), the total landfill carbon from food scraps in 2023 was 53.1 million metric
tons. This value is then added to the carbon stock from grass, leaves, and branches to calculate the
total landfill carbon stock in 2023, yielding a value of 296.0 million metric tons (as shown in Table 6-134).
In the same way total net flux is calculated for forest carbon and harvested wood products, the total net
flux of landfill carbon for yard trimmings and food scraps for a given year (Table 6-132) is the difference in
the landfill carbon stock for the following year and the stock in the current year. For example, the net
change in 2023 shown in Table 6-132 (3.2 MMT C with rounding) is equal to the stock in 2024 (299.2 MMT
C) minus the stock in 2023 (296.0 MMT C). The carbon stocks calculated through this procedure are
shown in Table 6-134.
To develop the 2024 carbon stock estimate, estimates of yard trimming and food scrap carbon stocks
were forecasted for 2024, based on data from 1990 through 2023. These forecasted values were used to
calculate net changes in carbon stocks for 2023. Excel's FORECAST.ETS function was used to predict a
2024 value using historical data via an algorithm called "ExponentialTriple Smoothing." This method
determined the overall trend and provided appropriate carbon stock estimates for 2024.
Table 6-133: Moisture Contents, Carbon Storage Factors (Proportions of Initial Carbon
Sequestered), Initial C Contents, and Decay Rates for Yard Trimmings and Food Scraps
in Landfills
Yard Trimmings
Variable
Grass
Leaves
Branches
Food Scraps
Moisture Content (% H2O)
70
30
10
70
C Storage Factor, Proportion of Initial C Stored (%)
53
85
77
16
Initial C Content (%)
45
46
49
51
Decay Rate (year-1)
0.313
0.179
0.015
0.151
Note: The decay rates are presented as weighted averages based on annual precipitation categories and population residing in
each precipitation category.
Table 6-134: Carbon Stocks in Yard Trimmings and Food Scraps in Landfills (MMT C)
Carbon Pool
1990
2005
2019
2020
2021
2022
2023 2024
Yard Trimmings
156.0
203.1
233.9
236.1
238.4
240.6
242.9 245.1
Branches
14.6
18.1
20.9
21.1
21.3
21.6
21.8 22.0
Leaves
66.7
87.4
101.5
102.5
103.6
104.6
105.6 106.7
Grass
74.7
97.7
111.5
112.5
113.5
114.5
115.5 116.5
Food Scraps
17.9
33.2
48.3
49.6
50.9
52.0
53.1 54.1
Total Carbon Stocks
173.9
236.3
282.2
285.7
289.2
292.6
296.0 299.2
a 2024 C stock estimate was forecasted using 1990 to 2023 data.
Note: Totals may not sum due to independent rounding.
Methodological recalculations were applied to the entire time series to ensure time-series consistency
from 1990 through 2023. When available, the same data source was used across the entire time series
6-218 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
for the analysis. When data were unavailable, missing values were estimated using linear interpolation
or forecasting, as noted above.
Uncertainty
The uncertainty analysis for landfilled yard trimmings and food scraps includes an evaluation of the
effects of uncertainty for the following data and factors: disposal in landfills per year (tons of carbon),
initial carbon content, moisture content, decay rate, and proportion of carbon stored. The carbon
storage landfill estimates are also a function of the composition of the yard trimmings (i.e., the
proportions of grass, leaves and branches in the yard trimmings mixture). There are respective
uncertainties associated with each of these factors.
A Monte Carlo (Approach 2) uncertainty analysis was applied to estimate the overall uncertainty of the
sequestration estimate for 2023. The results of the Approach 2 quantitative uncertainty analysis are
summarized in Table 6-135. Total yard trimmings and food scraps C02 flux in 2023 was estimated to be
between -18.4 and -5.2 MMT C02 Eq. at a 95 percent confidence level. This indicates a range of 57
percent below to 56 percent above the 2023 flux estimate of -11.7 MMT C02 Eq.
Table 6-135: Approach 2 Quantitative Uncertainty Estimates for C02 Flux from Yard
Trimmings and Food Scraps in Landfills (MMT C02 Eq. and Percent)
2024 Flux
Estimate
Source Gas (MMTC02Eq.)
Uncertainty Range Relative to Flux Estimate"
(MMT CO2 Eq.)
(%)
Lower Upper
Bound Bound
Lower Upper
Bound Bound
Yard Trimmings and Food Scraps CO2 (11-7)
(18.4) (5.2)
-57% +56%
a Range of flux estimates predicted by Monte Carlo stochastic simulation for a 95 percent confidence interval.
Note: Parentheses indicate negative values or net carbon sequestration.
QA/QC and Verification
Tier 1 and Tier 2 QA/QC activities were conducted consistent with the U.S. Inventory QA/QC plan.
Source-specific quality control measures for landfilled yard trimmings and food scraps included
checking that input data were properly transposed within the spreadsheet, checking calculations were
correct, and confirming that all activity data and calculations documentation was complete and
updated to ensure data were properly handled through the inventory process.
Order of magnitude checks and checks of time-series consistency were performed to ensure data were
updated correctly and any changes in emissions estimates were reasonable and reflected changes in
activity data. An annual change trend analysis was also conducted to ensure the validity of the
emissions estimates. Errors that were found during this process were corrected as necessary.
To ensure consistency across the LULUCF and Waste sectors, and the accuracy of emissions, EPA plans
to perform a comparison of the activity data used and carbon inputs between the landfilled yard
trimmings and food scraps, and the Waste chapter, Section 7.1 Landfills.
Recalculations Discussion
No recalculations were performed for the current Inventory.
Land Use, Land-Use Change, and Forestry 6-219
-------
Planned Improvements
EPA notes the following improvements may be implemented or investigated within the next two or three
Inventory cycles pending time and resource constraints:
• MSW data more recent than 2018 have not been released through the Advancing Sustainable
Materials Management reports. EPA will continue to monitor the release schedule for these data
and evaluate data for integration into the Inventory when released. Six new food waste
management pathways were introduced in the 2018 Advancing Sustainable Materials
Management report. Time series data for all of these pathways are not provided prior to 2018 but
EPA plans to investigate potential data sources and/or methods to address time-series
consistency and apply these data to the time series.
• EPA has been made aware of inconsistencies in landfilled food scraps data reported to the EPA
Greenhouse Gas Reporting Program (GHGRP) and will evaluate changes to how landfilled and
energy recovery values for yard trimmings and food scraps are calculated.
EPA notes the following improvements will continue to be investigated as time and resources allow, but
there are no immediate plans to implement these improvements until data are available or identified:
• EPA also plans to continue to investigate updates to the decay rate estimates for food scraps,
leaves, grass, and branches, as well as evaluate using decay rates that vary over time based on
Census population and climate data changes overtime. Currently the inventory calculations
use 2010 U.S. Census data, but 2020 U.S. Census data is available and EPA plans to implement
updates to varying decay rates in future Inventories.
• Other improvements include investigation into yard waste composition to determine if changes
need to be made based on changes in residential practices. A review of available literature was
conducted to determine if there are changes in the allocation of yard trimmings, however new
data was not available for the current Inventory. For example, leaving grass clippings in place is
becoming a more common practice, thus reducing the percentage of grass clippings in yard
trimmings disposed in landfills. In addition, agronomists may be consulted for determining the
mass of grass per acre on residential lawns to provide an estimate of total grass generation for
comparison with Inventory estimates. EPA will continue to monitor for new sources of yard
waste data.
• EPA will continue to evaluate data from recent peer-reviewed literature that may modify the
default carbon storage factors, initial carbon contents, and decay rates for yard trimmings and
food scraps in landfills - particularly updates to population precipitation ranges used to
calculate k values. Based upon this evaluation, changes may be made to the default values. EPA
is currently developing a method to update precipitation ranges based on available U.S.
precipitation data from NOAA.
Finally, EPA will continue to review possible sources for available data to ensure all types of landfilled
yard trimmings and food scraps are being included in the Inventory estimates, such as debris from road
construction and commercial food waste not included in other Inventory estimates.
6-220 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
6.11 Land Converted to Settlements (Source
Category 4E2)
Land converted to settlements includes all settlements in an inventory year that had been in another
land use(s) during the previous 20 years (USDA-NRCS 2015)." For example, cropland, grassland or
forest land converted to settlements during the past 20 years would be reported in this category.
Converted lands are retained in this category for 20 years as recommended by IPCC (2006).
Land use change can lead to large losses of carbon to the atmosphere, particularly conversions from
forest land (Houghton et al. 1983). Moreover, conversion of forest to another land use (i.e.,
deforestation) is one of the largest anthropogenic sources of emissions to the atmosphere globally
(Schimel 1995), although this source may be declining globally (Tubiello et al. 2015). IPCC (2006)
recommends quantifying changes in biomass, dead organic matter, and soil organic carbon stocks due
to land-use change. All soil organic carbon stock changes are estimated and reported for land converted
to settlements, but there is limited reporting of other pools in this Inventory. Loss of aboveground and
belowground biomass, dead wood and litter carbon are reported for forest land converted to
settlements and woodlands associated with grasslands converted to settlements. For croplands
converted to settlements, loss of woody perennial biomass and herbaceous live biomass are included
in a total live biomass loss, and loss of herbaceous live biomass is also estimates for other grasslands
(i.e., non-woodlands) converted settlements. Changes in biomass are not estimated for other lands
converted to settlements and wetlands converted to settlements.
There are discrepancies between the current land representation (see Section 6.1) and the area data
that have been used in the Inventory for land converted to settlements. Specifically, this Inventory
includes all settlements in the conterminous United States and Hawaii, but does not include
settlements in Alaska. Areas of drained organic soils in settlements on federal lands are also not
included in this Inventory These differences lead to discrepancies between the managed area in land
converted to settlements and the settlement area included in the inventory analysis (Table 6-140). There
is a planned improvement to include C02 emissions from drainage of organic soils in settlements of
Alaska and federal lands as part of a future Inventory (see Planned Improvements section).
Forest land converted to settlements is the largest source of emissions from 1990 to 2023, accounting
for approximately 75 percent of the average total loss of carbon among all of the land-use conversions in
Land Converted to Settlements. Total losses of aboveground and belowground biomass, dead wood and
litter carbon losses in 2023 for all conversions are 42.0, 6.8,11.1, and 9.6 MMT C02 Eq., respectively
(11.2,1.8, 3.0, and 2.6 MMT C). Mineral and organic soils also lost 9.2 and 1.3 MMT C02 Eq. in 2023 (2.5
and 0.4 MMT C). The total net flux is 79.8 MMT C02 Eq. in 2023 (21.5 MMT C), which is a 14.9 percent
increase in C02 emissions compared to the emissions in the initial inventory year of 1990 (Table 6-136
and Table 6-137). The main driver of net emissions for this source category is the conversion of forest
land to settlements, with large losses of biomass, deadwood and litter carbon.
99 NRI survey locations are classified according to land use histories starting in 1979, and consequently the classifications
are based on less than 20 years from 1990 to 2001. This may have led to an underestimation of land converted to
settlements in the early part of the time series to the extent that some areas are converted to settlements from 1971 to
1978.
Land Use, Land-Use Change, and Forestry 6-221
-------
Table 6-136: Net C02 Flux from Soil, Dead Organic Matter and Biomass Carbon Stock
Changes for Land Converted to Settlements (MMT C02 Eq.)
1990
2005
2019
2020
2021
2022
2023
Cropland Converted to Settlements
5.3
9.9
4.4
4.0
3.9
3.9
4.0
Total Live Biomass1
2.6
1.7
0.9
1.0
1.0
1.0
1.1
Mineral Soils
2.1
6.9
2.9
2.6
2.5
2.5
2.5
Organic Soils
0.5
1.2
0.5
0.4
0.4
0.4
0.4
Forest Land Converted to Settlements
56.1
61.4
67.2
67.4
67.4
67.4
67.4
Aboveground Live Biomass
32.5
35.2
38.6
38.7
38.7
38.7
38.7
Belowground Live Biomass
5.6
6.1
6.7
6.7
6.7
6.7
6.7
Dead Wood
9.1
9.9
10.9
10.9
10.9
10.9
10.9
Litter
7.8
8.5
9.3
9.4
9.4
9.4
9.4
Mineral Soils
1.0
1.5
1.5
1.5
1.5
1.5
1.5
Organic Soils
0.1
0.2
0.2
0.3
0.3
0.3
0.3
Grassland Converted to Settlements
8.5
18.4
10.5
9.6
9.1
9.2
9.2
Other Grassland Conversion Total
Live Biomass1
2.9
2.7
1.6
1.6
1.7
1.7
1.7
Woodland Conversion
Aboveground Live Biomass
0.4
0.4
0.5
0.5
0.5
0.5
0.5
Woodland Conversion
Belowground Live Biomass
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Dead Wood
0.2
0.2
0.2
0.2
0.2
0.2
0.2
Litter
0.2
0.2
0.2
0.2
0.2
0.2
0.2
Mineral Soils
4.3
13.7
7.3
6.5
6.0
6.0
6.0
Organic Soils
0.5
1.3
0.6
0.5
0.5
0.6
0.6
Other Lands Converted to Settlements
(0.4)
(1.4)
(1.0)
(0.8)
(0.8)
(0.8)
(0.8)
Mineral Soils
(0.4)
(1.5)
(1.0)
(0.9)
(0.8)
(0.8)
(0.8)
Organic Soils
+
0.1
+
+
+
+
+
Wetlands Converted to Settlements
+
0.6
0.3
0.1
+
0.1
0.1
Mineral Soils
+
0.1
+
+
+
+
+
Organic Soils
+
0.6
0.3
+
+
+
+
Total Aboveground Biomass Flux1
38.4
40.0
41.6
41.8
41.9
41.9
42.0
Total Belowground Biomass Flux
5.7
6.2
6.8
6.8
6.8
6.8
6.8
Total Dead Wood Flux
9.2
10.1
11.1
11.1
11.1
11.1
11.1
Total Litter Flux
8.0
8.7
9.5
9.6
9.6
9.6
9.6
Total Mineral Soil Flux
7.0
20.7
10.8
9.8
9.2
9.2
9.2
Total Organic Soil Flux
1.2
3.4
1.6
1.3
1.2
1.3
1.3
Total Net Flux
69.5
89.0
81.4
80.3
79.7
79.8
79.8
+ Absolute value does not exceed 0.05 MMT C02 Eq.
1 Biomass C stock changes for Cropland Converted to Settlements and Other Grasslands Converted to Settlements are not
disaggregated into aboveground and belowground biomass in the Tier 1 method, and are summed in the Total Aboveground
Biomass Flux.
Notes: Totals may not sum due to independent rounding. Parentheses indicate negative values or net sequestration.
6-222 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Table 6-137: Net C02 Flux from Soil, Dead Organic Matter and Biomass Carbon Stock
Changes for Land Converted to Settlements (MMT C)
1990
2005
2019
2020
2021
2022
2023
Cropland Converted to Settlements
1.4
2.7
1.2
1.1
1.1
1.1
1.1
Live Biomass
0.7
0.5
0.3
0.3
0.3
0.3
0.3
Mineral Soils
0.6
1.9
0.8
0.7
0.7
0.7
0.7
Organic Soils
0.1
0.3
0.1
0.1
0.1
0.1
0.1
Forest Land Converted to Settlements
15.3
16.7
18.3
18.4
18.4
18.4
18.4
Aboveground Live Biomass
8.9
9.6
10.5
10.5
10.6
10.6
10.6
Belowground Live Biomass
1.5
1.7
1.8
1.8
1.8
1.8
1.8
Dead Wood
2.5
2.7
3.0
3.0
3.0
3.0
3.0
Litter
2.1
2.3
2.5
2.6
2.6
2.6
2.6
Mineral Soils
0.3
0.4
0.4
0.4
0.4
0.4
0.4
Organic Soils
+
0.1
0.1
0.1
0.1
0.1
0.1
Grassland Converted to Settlements
2.3
5.0
2.9
2.6
2.5
2.5
2.5
Non-woodland Herbaceous Live Biomass
0.8
0.7
0.4
0.4
0.5
0.5
0.5
Woodland Aboveground Live Biomass
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Woodland Belowground Live Biomass
+
+
+
+
+
+
+
Dead Wood
+
+
0.1
0.1
0.1
0.1
0.1
Litter
+
+
0.1
0.1
0.1
0.1
0.1
Mineral Soils
1.2
3.7
2.0
1.8
1.6
1.6
1.6
Organic Soils
0.1
0.3
0.2
0.1
0.1
0.2
0.2
Other Lands Converted to Settlements
(0.1)
(0.4)
(0.3)
(0.2)
(0.2)
(0.2)
(0.2)
Mineral Soils
(0.1)
(0.4)
(0.3)
(0.2)
(0.2)
(0.2)
(0.2)
Organic Soils
+
+
+
+
+
+
+
Wetlands Converted to Settlements
+
0.2
0.1
+
+
+
+
Mineral Soils
+
+
+
+
+
+
+
Organic Soils
+
0.2
0.1
+
+
+
+
Total Aboveground Biomass Flux
9.8
10.4
11.1
11.1
11.1
11.1
11.2
Total Belowground Biomass Flux
1.6
1.7
1.8
1.8
1.8
1.8
1.8
Total Dead Wood Flux
2.5
2.7
3.0
3.0
3.0
3.0
3.0
Total Litter Flux
2.2
2.4
2.6
2.6
2.6
2.6
2.6
Total Mineral Soil Flux
1.9
5.6
2.9
2.7
2.5
2.5
2.5
Total Organic Soil Flux
0.3
0.9
0.4
0.4
0.3
0.3
0.4
Total Net Flux
18.2
23.8
21.9
21.6
21.5
21.5
21.5
+ Absolute value does not exceed 0.05 MMT C.
1 Biomass C stock changes for Cropland Converted to Settlements and Other Grasslands Converted to Settlements are not
disaggregated into aboveground and belowground biomass in the Tier 1 method, and are summed in the Total Aboveground
Biomass Flux.
Notes: Totals may not sum due to independent rounding. Parentheses indicate negative values or net sequestration
Methodology and Time-Series Consistency
The following section includes a description of the methodology used to estimate carbon stock changes
for land converted to settlements, including (1) loss of aboveground and belowground biomass, dead
wood and litter carbon with conversion to settlements from forest lands and woodlands designated in
the grassland, (2) loss of total live biomass with conversion from cropland to settlements and from other
Land Use, Land-Use Change, and Forestry 6-223
-------
grassland (i.e., non-woodlands) to settlements as well as (3) the impact from all land-use conversions to
settlements on soil organic carbon stocks in mineral and organic soils.
Biomass, Dead Wood, and Litter Carbon Stock Changes
The IPCC Tier 1 approach is used to estimate biomass carbon stock changes for croplands and
grasslands converted to settlements, according to land use histories recorded in the 2017 USDA NRI
survey for non-federal lands (USDA-NRCS 2020). Forfederal lands, the land use history is derived from
land cover changes in the National Land Cover Dataset ((Yang et al. 2018; Fry et al. 2011; Homer et al.
2007, 2015). The areas have been modified through a process in which the Forest Inventory and Analysis
(FIA) survey data are harmonized with the NRI data (Nelson et al. 2020). This process ensures that the
land use areas are consistent across all land use categories (see Section 6.1 for more information).
Land use and some management information were originally collected for each NRI survey location on a
5-year cycle beginning in 1982. In 1998, the NRI program began collecting annual data, and the annual
data have been incorporated from the NRI into the inventory analysis through 2017 (USDA-NRCS 2020)
(Table 6-138).
Table 6-138: Thousands of hectares of land for total live biomass associated with
grasslands (non-woodland) and croplands converted to settlements
1990
2005
2015
2016
2017
2018-2023
Croplands Converted to Settlements
145.0
94.5
43.9
47.4
50.7
*
Annual Crops
95.0
57.1
33.3
33.3
35.2
~
Non-Woody Crops
38.5
29.8
9.0
12.1
13.2
~
Perennial Woody Crops
11.5
29.8
1.7
2.0
2.3
~
Grasslands Converted to Settlements
166.5
171.5
70.5
81.4
98.3
*
Total: Land Converted to Settlements
311.49
266.08
114.36
128.85
148.96
*
* NRI data have not been incorporated into the Inventory after 2017, designated with asterisks (*). Data splicing methods are
used for the remainder of the time series.
The difference between the stocks is reported as the stock change under the assumption that the
change occurred in the year of the conversion. Biomass carbon losses from croplands and grasslands
converted to settlements include aboveground and belowground herbaceous biomass carbon and
aboveground woody biomass from perennial croplands. Biomass carbon stock factors are assigned for
each land use subcategory (e.g., annual or perennial cropland), perennial type, and maturity class.
Biomass carbon stocks in settlements were assumed to be zero (IPCC 2006). Conversions from
wetlands to settlements and other lands to settlements are also assumed to be 0 (IPCC 2006) so they
are not reported. The total area of each cropland and grassland subcategory converted to settlements
was multiplied by applicable factors from IPCC (2006 and 2019) (Table 6-134 and Table 6-135). For
conversions from non-woodland grasslands, the biomass values are disaggregated by climate zones
(IPCC 2006). For perennial croplands, factors vary by climate domain, perennial type, and maturity class
as indicated in IPCC (2019).
6-224 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Table 6-135: Carbon stock factors for total live biomass associated with conversions
from grassland (non-woodland) and croplands to settlements.
Land Use
Climate
Subcategory/
Type
Maturity
Biomass C
Stock (t ha-1)*
Source
Settlements
-
-
-
0
IPCC 2006
Perennial Woody Crops
Temperate
Orchards
Immature
0.43* ± 46%
IPCC 2019
Table 5.3)
Vineyards
Immature
0.28* ± 26%
IPCC 2019
Table 5.3)
Orchards
Mature
8.50 ±19%
IPCC 2019
Table 5.3)
Vineyards
Mature
5.50 ± 18%
IPCC 2019
Table 5.3)
Non-Woody Crops
-
Annual Crops
-
4.70 ± 75%
IPCC 2019
Table 8.4)
Cold Temperate - Dry
Hay
-
3.07 ± 75%
IPCC 2006
Table 6.4)
Cold Temperate - Moist
-
6.39 ± 75%
IPCC 2006
Table 6.4)
Warm Temperate - Dry
-
2.87 ± 75%
IPCC 2006
Table 6.4)
Warm Temperate - Moist
-
6.35 ± 75%
IPCC 2006
Table 6.4)
Tropical - Dry
-
4.09 ± 75%
IPCC 2006
Table 6.4)
Tropical - Moist &Wet
-
7.57 ± 75%
IPCC 2006
Table 6.4)
Grasslands
Cold Temperate - Dry
-
-
3.07 ± 75%
IPCC 2006
Table 6.4)
Cold Temperate - Moist
-
-
6.39 ± 75%
IPCC 2006
Table 6.4)
Warm Temperate - Dry
-
-
2.87 ± 75%
IPCC 2006
Table 6.4)
Warm Temperate - Moist
-
-
6.35 ± 75%
IPCC 2006
Table 6.4)
Tropical - Dry
-
-
4.09 ± 75%
IPCC 2006
Table 6.4)
Tropical - Moist &Wet
-
-
7.57 ± 75%
IPCC 2006
Table 6.4)
tBiomass C Stock for Hay and Grasslands obtained by multiplying biomass values by 0.47 carbon fraction (IPCC 2019, Table 8.4).
'Biomass C stock from immature perennial woody crops converted to Settlements was obtained by multiplying annual
aboveground biomass C accumulation rate (IPCC 2019, Table 5.3) by the age of the stand.
A Tier 2 method is applied to estimate biomass, dead wood, and litter carbon stock changes for forest
land converted to settlements. Estimates are calculated in the same way as those in the forest land
remaining forest land category using data from the USDA Forest Service, Forest Inventory and Analysis
(FIA) program (USDA Forest Service 2023) however there is no country-specific data for settlements, so
the biomass, litter, and dead wood carbon stocks were assumed to be zero. The difference between the
stocks is reported as the stock change under the assumption that the change occurred in the year of the
conversion. Details for each of the carbon attributes described below are available in Domke et al.
(2022) and Westfall et al. (2024).
If FIA plots include data on individual trees, aboveground and belowground carbon density estimates are
based on Woodall et al. (2011) and Westfall et al. (2024). Aboveground and belowground biomass
estimates also include live understory which is a minor component of biomass defined as all biomass of
undergrowth plants in a forest, including woody shrubs and trees less than 2.54 cm dbh. For this
Inventory, it was assumed that 10 percent of total understory carbon mass is belowground (Smith et al.
2006). Estimates of carbon density are based on information in Birdsey (1996) and biomass estimates
from Jenkins et al. (2003).
This Inventory also includes estimates of change in dead organic matter for standing dead, deadwood
and litter. If FIA plots include data on standing dead trees, standing dead tree carbon density is
estimated following the basic method applied to live trees (Woodall et al. 2011 and Westfall et al. 2024)
with additional modifications for woodland species to account for decay and structural loss (Domke et
al. 2011; Harmon et al. 2011). If FIA plots include data on downed dead wood, downed dead wood
Land Use, Land-Use Change, and Forestry 6-225
-------
carbon density is estimated based on measurements of a subset of FIA plots for downed dead wood
(Domke et al. 2013; Woodall and Monleon 2008). Downed dead wood is defined as pieces of dead wood
greater than 7.5 cm diameter, at transect intersection, that are not attached to live or standing dead
trees. This includes stumps and roots of harvested trees. To facilitate the downscaling of downed dead
wood carbon estimates from the state-wide population estimates to individual plots, downed dead
wood models specific to regions and forest types within each region are used. See Chapter 6.2 for more
information. Litter carbon is the pool of organic carbon (also known as duff, humus, and fine woody
debris) above the mineral soil and includes woody fragments with diameters of up to 7.5 cm. A subset of
FIA plots is measured for litter carbon. If FIA plots include litter material, a modeling approach using
litter carbon measurements from FIA plots is used to estimate litter carbon density (Domke et al. 2016).
In order to ensure time-series consistency, the same methods are applied from 1990 to 2023 for the
forest lands and woodlands converted to settlements so that changes reflect anthropogenic activity and
not methodological adjustments. See Annex 3.14 for more information about reference carbon density
estimates for forest land and woodlands, and the compilation system used to estimate carbon stock
changes from forest land. For croplands and other grasslands (non-woodland conversion) converted to
settlements, the same methods are applied from 1990 to 2018, and a data splicing method is used to
estimate biomass carbon loss for the remainder of the 2018 to 2023 time series (see Box 6-3 in Section
6.4 cropland remaining cropland). Specifically, a linear regression model with moving-average (ARMA)
errors (Brockwell and Davis 2016) is used to impute the missing C stock changes using trends from 1990
to 2018. This method is type of a linear extrapolation, which is a standard data splicing method for
estimating emissions at the end of a time series (IPCC 2006). The time series will be recalculated in a
future Inventory with the methods described previously for biomass carbon stock changes.
Soil Carbon Stock Changes
Soil organic carbon stock changes are estimated for land converted to settlements according to land
use histories recorded in the 2017 USDA NRI survey for non-federal lands (USDA-NRCS 2020) and
extended through 2020 using the USDA-NASS Crop Data Layer Product (USDA-NASS 2021; Johnson and
Mueller 2010) and National Land Cover Dataset (NLCD) (Yanget al. 2018; Fry et al. 2011; Homer et al.
2007, 2015)). For federal lands, the land use history is derived from land cover changes in the NLCD. The
areas have been modified through a process in which the Forest Inventory and Analysis (FIA) survey data
are harmonized with the NRI data (Nelson et al. 2020). This process ensures that the land use areas are
consistent across all land use categories (see Section 6.1 for more information). Land use and some
management information were originally collected for each NRI survey location on a 5-year cycle
beginning in 1982. In 1998, the NRI program began collecting annual data, and the annual data have
been incorporated from the NRI into the inventory analysis through 2017 (USDA-NRCS 2020).
NRI survey locations are classified as land converted to settlements in a given year between 1990 and
2020 if the land use is settlements but had been classified as another use during the previous 20 years.
NRI survey locations are classified according to land use histories starting in 1979, and consequently
the classifications are based on less than 20 years from 1990 to 1998. This may have led to an
underestimation of land converted to settlements in the early part of the time series to the extent that
some areas are converted to settlement between 1971 and 1978.
6-226 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Soil Carbon Stock Changes for Mineral Soils
An IPCC Tier 2 method (Ogle et al. 2003) is applied to estimate carbon stock changes for mineral soils
on land converted to settlements from 1990 to 2020. Data on climate, soil types, land use, and land
management activity are used to classify land area and apply appropriate stock change factors (Ogle et
al. 2003, 2006). Reference carbon stocks are estimated using the National Soil Survey Characterization
Database (USDA-NRCS 1997) with cultivated cropland as the reference condition, rather than native
vegetation as used in IPCC (2006). Soil measurements under agricultural management are much more
common and easily identified in the National Soil Survey Characterization Database (USDA-NRCS 1997)
than are soils under a native condition, and therefore cultivated cropland provide a more robust sample
for estimating the reference condition. Country-specific carbon stock change factors are derived from
published literature to determine the impact of management practices on soil organic carbon storage
(Ogle et al. 2003; Ogle et al. 2006). However, there are insufficient data to estimate a set of land use,
management, and input factors for settlements. Moreover, the 2017 NRI survey data (USDA-NRCS 2020)
do not provide the information needed to assign different land use subcategories to settlements, such
as turf grass and impervious surfaces, which is needed to apply the Tier 1 factors from the IPCC
guidelines (2006). Therefore, the United States has adopted a land use factor of 0.7 to represent a net
loss of soil organic carbon with conversion to settlements under the assumption that there are
additional soil organic carbon losses with land clearing, excavation and other activities associated with
development. More specific factor values can be derived in future Inventories as data become available.
See Annex 3.13 for additional discussion of the Tier 2 methodology for mineral soils.
In order to ensure time-series consistency, the same methods are applied from 1990 to 2020 so that
changes reflect anthropogenic activity and not methodological adjustments. Soil organic carbon stock
changes from 2021 to 2023 are estimated using a linear extrapolation method described in Box 6-3 of
the Methodology section in Cropland Remaining Cropland. The extrapolation is based on a linear
regression model with moving-average (ARMA) errors using the 1990 to 2020 emissions data. This
method is type of a linear extrapolation, which is a standard data splicing method for estimating
emissions at the end of a time series (IPCC 2006). The Tier 2 method described previously will be
applied to recalculate the 2021 to 2023 emissions in a future Inventory.
Soil Carbon Stock Changes for Organic Soils
Annual carbon emissions from drained organic soils in land converted to settlements are estimated
using the Tier 2 method provided in IPCC (2006). The Tier 2 method assumes that organic soils are losing
carbon at a rate similar to croplands, and therefore uses the country-specific values for cropland (Ogle
et al. 2003). To estimate C02 emissions from 1990 to 2020, the area of organic soils in land converted to
settlements is multiplied by the Tier 2 emission factor, which is 11.2 MT C per ha in cool temperate
regions, 14.0 MT C per ha in warm temperate regions and 14.3 MT C per ha in subtropical regions (see
Annex 3.13 for more information).
In order to ensure time-series consistency, the same methods are applied from 1990 to 2020, and a data
splicing method is used to approximate emissions for the remainder of the 2021 to 2023 time series (see
Box 6-3 of the Methodology section in Cropland Remaining Cropland). Specifically, the data splicing
method is a linear regression model with moving-average (ARMA) errors using trends from 1990 to 2018
in the C stock change data. This method is type of a linear extrapolation, which is a standard data
splicing method for estimating emissions at the end of a time series (IPCC 2006). Estimates will be
recalculated in future Inventories when new activity data are incorporated into the analysis.
Land Use, Land-Use Change, and Forestry 6-227
-------
Uncertainty
The uncertainty analysis for carbon losses with forest land converted to settlements is conducted in the
same way as the uncertainty assessment for forest ecosystem carbon flux in the forest land remaining
forest land category Sample and model-based error are combined using simple error propagation
methods provided by the IPCC (2006), i.e., by taking the square root of the sum of the squares of the
standard deviations of the uncertain quantities. For additional details, see the Uncertainty Analysis in
Annex3.14.
Sources of uncertainty for biomass C stock changes with cropland converted to settlements and other
grasslands (non-woodland) conversions to settlements, mineral soil organic carbon stock changes and
annual carbon emission estimates from drained organic soils include emission factors and variance
associated with the NRI sample. The total uncertainty was quantified with two variance components
(Ogle et a I. 2010) that are combined using simple error propagation methods provided by the IPCC
(2006), i.e., by taking the square root of the sum of the squares of the standard deviations of the
uncertain quantities. For the first variance component, a Monte Carlo analysis was used to propagate
uncertainties in the Tier 1 and Tier 2 methods for the land use area and the country-specific factors or
mineral and organic soils. The second variance component is associated with scaling of the data from
the NRI survey to the entire area of land converted to settlements, and is computed using a standard
variance estimator for a two-stage sample design (Sarndal et a 1.1992).
Uncertainty estimates are presented in Table 6-139 for each sub-source (i.e., biomass carbon, dead
wood, litter, soil organic carbon in mineral soils and organic soils) and the method applied in the
inventory analysis (i.e., Tier 1, Tier 2 and Tier 3). Uncertainty estimates are combined from the forest land
converted to settlements and other land use conversions to settlements using the simple error
propagation methods provided by the IPCC (2006). There are also additional uncertainties propagated
through the analysis associated with the data splicing methods applied to estimate non-woodland
grassland conversions and cropland conversions to settlements, and also soil organic carbon stock
changes for all land use conversions from 2021 to 2023. The combined uncertainty for total carbon
stock changes in land converted to settlements ranges from 36 percent below to 36 percent above the
2023 stock change estimate of 79.8 MMT C02 Eq.
Table 6-139: Approach 2 Quantitative Uncertainty Estimates for Soil, Dead Organic
Matter and Biomass Carbon Stock Changes occurring within Land Converted to
Settlements (MMT C02 Eq. and Percent)
Source
2023 Flux
Estimate
(MMTCO2
Eq.)
Uncertainty Range Relative to Flux Estimate"
(MMTCO2 Eq.)
(%)
Lower
Bound
Upper
Bound
Lower
Bound
Upper
Bound
Cropland Converted to Settlements
4.0
(0.3)
8.3
-109%
+109%
Total Live biomass
1.1
(2.1)
4.2
-293%
+293%
Mineral Soil C Stocks
2.5
(0.4)
5.4
-117%
+ 117%
Organic Soil C Stocks
0.4
(0.2)
1.1
-148%
+ 148%
Forest Land Converted to Settlements
67.4
41.4
93.4
-39%
+39%
Aboveground Biomass C Stocks
38.7
14.6
62.8
-62%
+62%
Belowground Biomass C Stocks
6.7
2.5
10.9
-62%
+62%
6-228 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
2023 Flux
Estimate
Uncertainty Range Relative to Flux Estimate"
(MMTCO2 Eq.)
(%)
Source
(MMTCO2
Eq.)
Lower
Bound
Upper
Bound
Lower
Bound
Upper
Bound
Dead Wood
10.9
4.1
17.7
-62%
+62%
Litter
9.4
3.6
15.2
-62%
+62%
Mineral Soil C Stocks
1.5
1.0
1.9
-28%
+28%
Organic Soil C Stocks
0.3
(+)
0.5
-115%
+ 115%
Grassland Converted to Settlements
9.2
3.1
15.2
-66%
+66%
Other grassland (non-woodland) Total Live Biomass
1.7
-1.8
5.3
-208%
+208%
Woodland Aboveground Biomass C Stocks
0.5
0.2
0.8
-62%
+61%
Woodland Belowground Biomass C Stocks
0.1
+
0.1
-44%
+0%
Dead Wood
0.2
0.1
0.3
-62%
+70%
Litter
0.2
0.1
0.3
-61%
+57%
Mineral Soil C Stocks
6.0
1.2
10.8
-80%
+80%
Organic Soil C Stocks
0.6
-0.3
1.4
-157%
+ 157%
Other Lands Converted to Settlements
-0.8
(1.5)
(+)
-99%
+99%
Mineral Soil C Stocks
-0.8
(1.6)
(0.1)
-93%
+93%
Organic Soil C Stocks
+
(0.1)
0.2
-591%
+0%
Wetlands Converted to Settlements
0.1
(0.5)
0.6
-910%
+126%
Mineral Soil C Stocks
+
(+)
0.1
-126%
+ 126%
Organic Soil C Stocks
+
(0.5)
0.6
-1831%
+ 1831%
Total: Land Converted to Settlements
79.8
51.1
108.6
-36%
+36%
Aboveground Biomass C Stocks
42.0
17.4
66.5
-58%
+58%
Belowground Biomass C Stocks
6.8
2.6
11.0
-62%
+62%
Dead Wood
11.1
4.3
17.9
-61%
+61%
Litter
9.6
3.7
15.4
-61%
+61%
Mineral Soil C Stocks
9.2
3.5
14.8
-62%
+62%
Organic Soil C Stocks
1.3
(8.6)
11.1
-766%
+766%
+ Does not exceed 0.05 MMT C02 Eq.
a Range of C stock change estimates is a 95 percent confidence interval.
Notes: Totals may not sum due to independent rounding. Parentheses indicate negative values or net sequestration.
QA/QC and Verification
Quality control measures included checking input data, model scripts, and results to ensure data are
properly handled throughout the inventory process. Inventory forms and text are reviewed and revised as
needed to correct transcription errors. The land use conversions from cropland to settlements and other
grasslands (non-woodlands) had errors in identifying the C stocks for lands converted between federal
and non-federal ownership. A unit conversion error was identified in the uncertainty analysis for
perennial woody biomass. Additionally, a unit conversion was identified in spreadsheet for the
uncertainty analysis results. All errors were corrected.
Recalculations Discussion
Recalculations are associated with new FIA data from 1990 to 2023 on biomass, dead wood and litter
carbon stocks in forest land converted to settlements and woodland conversion associated with
Land Use, Land-Use Change, and Forestry 6-229
-------
grassland converted to settlements. Additional recalculations are associated with incorporating Tier 1
methodology to estimate biomass carbon stocks for cropland and other grassland (non-woodland)
conversions to settlements. As a result, land converted to settlements has an estimated larger carbon
loss of 12.2 MMT C02 Eq. on average over the time series. This represents a 17.0 percent decrease in
carbon stock changes for land converted to settlements compared to the previous Inventory.
Planned Improvements
Key planned improvements for estimates of land converted to settlements will resolve most of the
differences between the managed land base for land converted to settlements and the amount of area
currently included in the Inventory for land converted to settlements (see Table 6-140). These include:
• Developing an inventory of mineral soil organic carbon stock changes in Alaska and losses of
carbon from drained organic soils in federal lands.
• Incorporating new land representation area data into the next Inventory. The current Inventory
for land converted to grassland is based on the land representation from the previous Inventory.
The following are also planned over the next 2 to 3 years, pending resources:
• Improving classification of trees in settlements and to include transfer of biomass from forest
land to those areas in this category.
• Extending the Inventory to include carbon losses associated with drained organic soils in
settlements occurring on federal lands.
Table 6-140: Area of Managed Land in Land Converted to Settlements that is not
included in the current Inventory (Thousand Hectares)
Area (Thousand Hectares)
Year
LCS Managed Land Area
(Section 6.1)
LCS Area Included in Inventory
LCS Area Not Included in
Inventory
1990
2,858
2,865
-7
1991
3,207
3,213
-5
1992
3,568
3,573
-5
1993
4,137
4,138
0
1994
4,700
4,702
-2
1995
5,253
5,261
-8
1996
5,820
5,832
-12
1997
6,397
6,408
-11
1998
6,919
6,928
-9
1999
7,430
7,446
-15
2000
7,937
7,952
-14
2001
8,359
8,361
-2
2002
8,694
8,695
-1
2003
8,703
8,704
0
2004
8,716
8,708
8
2005
8,732
8,724
8
2006
8,693
8,688
5
6-230 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Area (Thousand Hectares)
Year
LCS Managed Land Area
(Section 6.1)
LCS Area Included in Inventory
LCS Area Not Included in
Inventory
2007
8,671
8,668
3
2008
8,499
8,497
2
2009
8,313
8,305
8
2010
8,124
8,124
0
2011
7,926
7,925
1
2012
7,712
7,711
0
2013
7,315
7,318
-4
2014
6,934
6,935
-1
2015
6,521
6,523
-2
2016
6,107
6,105
3
2017
5,711
5,708
3
2018
5,197
5,194
3
2019
4,681
4,689
-7
2020
4,167
4,168
-1
2021
3,752
~
~
2022
3,417
~
~
2023
3,089
~
~
NRI data have not been incorporated into the Inventory after 2020, designated with asterisks (*).
6.12 Other Land Remaining Other Land
(Source Category 4F1)
Land use is constantly occurring, and areas under a number of differing land-use types remain in their
respective land-use type each year, just as other land can remain as other land. While the magnitude of
other land remaining other land is known (see Table 6-4), research is ongoing to track carbon pools in
this land use. Until such time that reliable and comprehensive estimates of carbon for other land
remaining other land can be produced, it is not possible to estimate C02, CH4 or N20 fluxes on other
land remaining other land at this time.
6.13 Land Converted to Other Land
(Source Category 4F2)
Land-use change is constantly occurring, and areas under a number of differing land-use types are
converted to other land each year, just as other land is converted to other uses. While the magnitude of
these area changes is known (see Table 6-4), research is ongoing to track carbon across other land
remaining other land and land converted to other land. Until such time that reliable and comprehensive
estimates of carbon across these land-use and land-use change categories can be produced, it is not
Land Use, Land-Use Change, and Forestry 6-231
-------
possible to separate C02, CH4 or N20 fluxes on land converted to other land from fluxes on other land
remaining other land at this time.
6-232 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
7
Waste
-------
7 Waste
Waste management and treatment activities are sources of greenhouse gas emissions (see Figure 7-1
and Figure 7-2). Landfills were the third largest source of anthropogenic methane (CH4) in the United
States in 2023, accounting for approximately 17.4 percent of total U.S. CH4 emissions. Additionally,
wastewater treatment and discharge, composting of organic waste, and anaerobic digestion at biogas
facilities accounted for approximately 3.1 percent, 0.4 percent, and less than 0.1 percent of U.S. CH4
emissions, respectively. Nitrous oxide (N20) emissions result from the discharge of wastewater
treatment effluents into aquatic environments were estimated, along with the wastewater treatment
process itself and composting. Together, these waste activities account for 5.9 percent of total U.S. N20
emissions. Nitrogen oxides (NOx), carbon monoxide (CO), and non-CH4 volatile organic compounds
(NMVOCs) are emitted by waste activities and are addressed separately at the end of this chapter. A
summary of greenhouse gas emissions from the Waste sector is presented in Table 7-1 and Table 7-2.
Overall, in 2023, waste activities generated emissions of 165.8 MMT C02 Eq., or 2.7 percent of total U.S.
greenhouse gas emissions.
Emissions from landfills contributed 72.0 percent of Waste sector emissions in 2023 (see Figure 7-1)
and are primarily composed of CH4 emissions from municipal solid waste landfills. Landfill emissions
increased by 0.8 MMT C02 Eq. (0.7 percent) since 2022. Emissions from wastewater treatment were the
second largest source of waste-related emissions in 2023, accounting for 25.3 percent of sector
emissions. Wastewater emissions remained relatively unchanged at just 0.1 percent below 2022 levels.
The remaining two sources of emissions, composting and anaerobic digestion at biogas facilities,
account for 2.7 percent and less than 0.1 percent of Waste sector emissions in 2023, respectively.
Figure 7-1: 2023 Waste Sector Greenhouse Gas Sources
Wastewater T reatment
Anaerobic Digestion at
Biogas Facilities
Composting
Landfills
0 10 20 30 40 50 60 70 80 90 100 110 120 130
MMT CO2 Eq.
7-2 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Figure 7-2: Trends in Waste Sector Greenhouse Gas Sources
250
200
150
O
u
100
£3 fM
^ fo Si 0
cm q
CM O
I Anaerobic Digestion at Biogas Facilities
I Composting
I Wastewater Treatment
I Landfills
o
O LO Tj"
(N CTi
vD
£? £
h r^
co
r-s
n >> m ^
H ^ ^ ^
50
OiHrvjro^-invDr^ooCT*OT-ifMro^-irj^Dr*vcoo^o-^H(Nro"^-LoiDr^oooiOi-Hrvjro
cr>cho^cr>o^cr»cr>CT>cr»cr>ooooooooooi-HiHTHTH-MTHi-iT-iTHT-H(N(N(NfN
cncncncncr.cncncr.cncnoooooooooooooooooooooooo
HHHHrHHrHHHH(NfM(N(N(N(\(NfMfNJ(\fMfMfMfM(\(NfM(\fMfM(NfMfM(N
Table 7-1: Emissions from Waste (MMT C02 Eq.)
Gas/Source
1990
2005
2019
2020
2021
2022
2023
CH4
220.9
172.4
151.9
146.1
143.9
142.2
143.2
Landfills
197.8
147.7
128.2
122.6
120.7
118.7
119.5
Wastewater Treatment
22.7
22.7
21.1
21.0
20.7
20.9
21.1
Composting
0.4
2.1
2.5
2.6
2.6
2.6
2.6
Anaerobic Digestion at Biogas Facilities
+
+
+
+
+
+
+
N2O
15.1
19.5
22.9
23.6
23.1
22.9
22.6
Wastewater Treatment
14.8
18.1
21.1
21.8
21.3
21.1
20.8
Composting
0.3
1.5
1.8
1.8
1.8
1.8
1.8
Total
235.9
192.0
174.8
169.7
167.0
165.1
165.8
+ Does not exceed 0.5 kt.
Note: Totals may not sum due to independent rounding.
Waste 7-3
-------
Table 7-2: Emissions from Waste (kt)
Gas/Source
1990
2005 |
2019
2020
2021
2022
2023
CH4
7,889
6,159
5,424
5,219
5,140
5,078
5,114
Landfills
7,063
5,275
4,578
4,379
4,310
4,238
4,266
Wastewater Treatment
811
809
755
748
738
747
755
Composting
15
75
91
92
92
92
93
Anaerobic Digestion at Biogas Facilities
+
+
1
1
1
1
1
N20
57
74
87
89
87
86
85
Wastewater Treatment
56
68
80
82
80
80
79
Composting
1
6 I
7
7
7
7
7
+ Does not exceed 0.5 kt.
Note: Totals by gas may not sum due to independent rounding.
Carbon dioxide (C02), CH4, and N20 emissions from the incineration of waste are accounted for in the
Energy sector rather than in the Waste sector because almost all incineration of municipal solid waste
(MSW) in the United States occurs at waste-to-energy facilities where useful energy is recovered.
Similarly, the Energy sector also includes an estimate of emissions from burning waste tires and
hazardous industrial waste, because virtually all of the combustion occurs in industrial and utility
boilers that recover energy. The incineration of waste in the United States in 2023 resulted in 12.8 MMT
C02 Eq. emissions, more than half of which is attributable to the combustion of plastics. For more
details on emissions from the incineration of waste, see Section 7.5. Greenhouse gas precursor
emissions from the Waste sector are presented in Section 7.6.
Each year, some emission and sink estimates in the Inventory are recalculated and revised with
improved methods and/or data. In general, recalculations are made to the U.S. greenhouse gas
emission estimates either to incorporate new methodologies or, most commonly, to update recent
historical data. These improvements are implemented consistently across the previous Inventory's time
series (i.e., 1990 to 2022) to ensure that the trend is accurate. For the current Inventory, minor
improvements were implemented beyond routine activity data updates, including changes to MSW and
industrial waste landfill activity data, updates to production activity affecting wastewater influent, and
updates to the amount of waste processed by anaerobic digesters. In total, the methodological and
historic data improvements made to the Waste sector in this Inventory resulted in an average decrease
in greenhouse gas emissions across the time series by 0.3 MMT C02 Eq. (0.19 percent). For more
information on specific methodological updates, please see the Recalculations Discussion section for
each category in this chapter.
The estimates for the waste sector are largely complete. Emissions associated with sludge generated
from the treatment of industrial wastewater are not estimated due to the likely insignificant level of
emissions and the lack of relevant data. Emissions reported in the Waste chapter for landfills,
wastewater treatment, and anaerobic digestion at biogas facilities include those from all 50 states,
including Hawaii and Alaska, the District of Columbia, U.S. Territories, and tribal lands to the extent they
are occurring. Emissions from landfills include modern, managed sites in most U.S. Territories except for
outlying Pacific Islands. Emissions from domestic wastewater treatment include most U.S. Territories
except for outlying Pacific Islands. Those emissions are likely insignificant as those outlying Pacific
Islands (e.g., Baker Island) have no permanent population. No industrial wastewater treatment
emissions are estimated for U.S. Territories, due to lack of data availability. However, industrial
wastewater treatment emissions are not expected for outlying Pacific Islands and assumed to be small
7-4 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
for other U.S. Territories. Emissions for composting include Puerto Rico and all states except Alaska.
Some composting operations in Alaska are known, but these consist of aerated composting facilities.
Composting emissions are not included from the remaining U.S. Territories, and these are assumed to
be small. Similarly, EPA is not aware of any anaerobic digestion at biogas facilities in U.S. Territories but
will review this on an ongoing basis to include these emissions if they are occurring. See Annex 5 for
more information on EPA's assessment of the sources not included in this Inventory.
Box 7-1: Use of Greenhouse Gas Reporting Data in Waste Sector
EPA also collects greenhouse gas emissions data from individual facilities and suppliers of certain fossil
fuels and industrial gases through its Greenhouse Gas Reporting Program (GHGRP). The GHGRP applies
to direct greenhouse gas emitters, fossil fuel suppliers, industrial greenhouse gas suppliers, and
facilities that inject C02 underground for sequestration or other reasons and requires reporting by
sources or suppliers in 46 industrial categories. Annual reporting is at the facility level, except for certain
suppliers of fossil fuels and industrial greenhouse gases. In general, the threshold for reporting is 25,000
metric tons or more of C02 Eq. per year. See Annex 9 "Use of EPA Greenhouse Gas Reporting Program in
Inventory" for more information.
Waste Data from EPA's Greenhouse Gas Reporting Program
EPA uses annual GHGRP facility-level data in the Landfills category to compile the national estimate of
emissions from Municipal Solid Waste (MSW) landfills (see Section 7.1 of this chapter for more
information). EPA uses directly reported GHGRP data for net CH4 emissions from MSW landfills for the
years 2010 to 2023 of the Inventory. MSW landfills subject to the GHGRP began collecting data in 2010.
These data are also used to recalculate emissions from MSW landfills for the years 2005 to 2009 to
ensure time-series consistency.
7.1 Landfills (Source Category 5A1)
In the United States, solid waste is managed by landfilling, recovery through recycling or composting,
and combustion through waste-to-energy facilities. Disposing of solid waste in modern, managed
landfills is the most used waste management technique in the United States. More information on how
solid waste is collected and managed in the United States is provided in Box 7-3. The municipal solid
waste (MSW) and industrial waste landfills referred to in this section are all modern landfills that must
comply with a variety of regulations as discussed in Box 7-2. Disposing of waste in illegal dumping sites
is not considered to have occurred in years later than 1980 and these sites are not considered to
contribute to net emissions in this section for the timeframe of 1990 to the current Inventory year. MSW
landfills, or sanitary landfills, are sites where MSW is managed to prevent or minimize health, safety, and
environmental impacts. Waste is deposited in different cells and covered daily with soil; many have
environmental monitoring systems to track performance, collect leachate, and collect landfill gas.
Industrial waste landfills are constructed in a similar way as MSW landfills, but are used to dispose of
industrial solid waste, such as RCRA Subtitle D wastes (e.g., non-hazardous industrial solid waste
defined in Title 40 of the Code of Federal Regulations [CFR] in section 257.2), commercial solid wastes,
or conditionally exempt small-quantity generator wastes (EPA 2016a).
Waste 7-5
-------
After being placed in a landfill, organic waste (such as paper, food scraps, and yard trimmings) is initially
decomposed by aerobic bacteria. After the oxygen has been depleted, the remaining waste is available
for consumption by anaerobic bacteria, which break down organic matter into substances such as
cellulose, amino acids, and sugars. These substances are further broken down through fermentation
into gases and short-chain organic compounds that form the substrates for the growth of methanogenic
bacteria. These CH4 producing anaerobic bacteria convert the fermentation products into stabilized
organic materials and biogas consisting of approximately 50 percent biogenic carbon dioxide (C02) and
50 percent CH4, by volume. Landfill biogas also contains trace amounts of non-methane organic
compounds (NMOC) and volatile organic compounds (VOC) that either result from decomposition
byproducts or volatilization of biodegradable wastes (EPA 2008).
Box 7-2: Description of a Modern, Managed Landfill in the United States
Modern, managed landfills are well-engineered facilities that are located, designed, operated, and
monitored to ensure compliance with federal, state, and tribal regulations. A modern, managed landfill
is EPA's interpretation of the IPCC's terminology of a managed solid waste disposal site. Municipal solid
waste (MSW) landfills must be designed to protect the environment from contaminants which may be
present in the solid waste stream. Additionally, many new landfills collect and destroy landfill gas
through flares or landfill gas-to-energy projects. Requirements for affected MSW landfills may include:
• Siting requirements to protect sensitive areas (e.g., airports, floodplains, wetlands, fault areas,
seismic impact zones, and unstable areas);
• Design requirements for new landfills to ensure that Maximum Contaminant Levels (MCLs) will not
be exceeded in the uppermost aquifer (e.g., composite liners and leachate collection systems);
• Leachate collection and removal systems;
• Operating practices (e.g., daily and intermediate cover, receipt of regulated hazardous wastes, use
of landfill cover material, access options to prevent illegal dumping, use of a collection system to
prevent stormwater run-on/run-off, record-keeping);
• Air monitoring requirements (explosive gases);
• Groundwater monitoring requirements;
• Closure and post-closure care requirements (e.g., final cover construction); and
• Corrective action provisions.
Specific federal regulations that affected MSW landfills must comply with include the 40 CFR Part 258
(Subtitle D of RCRA), or equivalent state regulations, the NSPS 40 CFR Part 60 Subparts WWW and XXX,
and the EG 40 CFR Part 62 Subpart Cf.1 Additionally, state and tribal requirements may exist.
Methane and C02 are the primary constituents of landfill gas generation and emissions. Net carbon
dioxide flux from carbon stock changes of materials of biogenic origin in landfills are estimated and
reported under the Land Use, Land-Use Change, and Forestry (LULUCF) sector (see Chapter 6 of this
Inventory). Nitrous oxide (N20) emissions from the disposal and application of sewage sludge on
landfills are also not explicitly modeled as part of greenhouse gas emissions from landfills. Nitrous
1 For more information regarding federal MSW landfill regulations, see https://www.epa.gov/landfills/municipal-solid-
waste-landfills#regs and https://www.epa.gov/stationarv-sources-air-pollution/municipal-solid-waste-landfills-national-
emission-standards.
7-6 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
oxide emissions from sewage sludge applied to landfills as a daily cover or for disposal are expected to
be relatively small because the microbial environment in an anaerobic landfill is not very conducive to
the nitrification and denitrification processes that result in N20 emissions. Furthermore, the 2006IPCC
Guidelines did not include a methodology for estimating N20 emissions from solid waste disposal sites
"because they are not significant." Therefore, only CH4 generation and emissions are estimated for
landfills under the Waste sector.
Methane generation and emissions from landfills are a function of several factors, including: (1) the total
amount and composition of waste-in-place, which is the total waste landfilled annually over the
operational lifetime of a landfill; (2) the characteristics of the landfill receiving waste (e.g., size, climate,
cover material); (3) the amount of CH4 that is recovered and either flared or used for energy purposes;
and (4) the amount of CH4 oxidized as the landfill gas-that is not collected by a gas collection system-
passes through the cover material into the atmosphere. Each landfill has unique characteristics, but all
managed landfills employ similar operating practices, including the application of a daily and
intermediate cover material over the waste being disposed of in the landfill to prevent odor and reduce
risks to public health. Based on recent literature, the specific type of cover material used can affect the
rate of oxidation of landfill gas (RTI 2011). The most used cover materials are soil, clay, and sand. Some
states also permit the use of green waste, tarps, waste derived materials, sewage sludge or biosolids,
and contaminated soil as a daily cover. Methane production typically begins within the first year after the
waste is disposed of in a landfill and will continue for 10 to 50 or more years as the degradable waste
decomposes overtime.
In 2023, landfill CH4 emissions were approximately 119.5 MMT C02 Eq. (4,266 kt), representing the third
largest source of CH4 emissions in the United States, behind enteric fermentation and natural gas
systems. Emissions from MSW landfills accounted for approximately 84 percent of total landfill
emissions (100.6 MMT C02 Eq.), while industrial waste landfills accounted for the remainder (18.9 MMT
C02 Eq.). Nationally, there are significantly less industrial waste landfills compared to MSW landfills,
which contributes to the lower national estimate of CH4 emissions for industrial waste landfills.
Additionally, the average organic content of waste streams disposed in industrial waste landfills is lower
than MSW landfills. Present estimates of operational (open and accepting waste) MSW landfills in the
United States have ranged from 1,200 to 1,750 facilities (EPA 2024a; EPA 2024b; EPA 2020b; EPA 2019;
Waste Business Journal [WBJ] 2016; WBJ 2010). The Environment Research & Education Foundation
(EREF) conducted a nationwide analysis of MSW management and counted 1,540 operational MSW
landfills in 2013 (EREF 2016). Conversely, there are approximately 3,200 MSW landfills in the United
States that have been closed since 1980 (for which a closure data is known, (EPA 2024b; WBJ 2010).
While the number of active MSW landfills has decreased significantly over the past 20 years, from
approximately 6,326 in 1990 to as few as 1,540 in 2013, the average landfill size has increased (EPA
2024b; EREF 2016; BioCycle 2010). Larger landfills may have deeper cells where a greater amount of
area will be anaerobic (more CH4 is generated in anaerobic versus aerobic areas) and larger landfills
tend to generate more CH4 compared to a smaller landfill (assuming the same waste composition and
age of waste). Regarding industrial waste landfills, the WBJ database includes approximately 1,100
landfills accepting industrial and/or construction and demolition debris for 2021 (WBJ 2021). Only 169
facilities with industrial waste landfills met the reporting threshold under Subpart TT (Industrial Waste
Landfills) in the first year (2011) of EPA's Greenhouse Gas Reporting Program for this subpart (GHGRP
codified in 40 CFR Part 98), indicating that there may be several hundred industrial waste landfills that
are not required to report under EPA's GHGRP. Less industrial waste landfills meet the GHGRP eligibility
Waste 7-7
-------
threshold because they typically accept waste streams with low to no organic content, which will not
decompose and generate CH4 when disposed.
The annual amount of MSW generated and subsequently disposed of in MSW landfills varies annually
and depends on several factors (e.g., the economy, consumer patterns, recycling and composting
programs, inclusion in a garbage collection service). The estimated annual quantity of waste placed in
MSW landfills increased 10 percent from approximately 205 MMT in 1990 to 226 MMT in 2000, then
decreased by 11 percent to 202 MMT in 2010, and then increased by 8 percent to approximately 218
MMT in 2023 (see Annex 3.15, Table A-213). Emissions decreased between 1990 to 2023 largely because
of increased use of landfill gas collection and control systems, closure of older landfills, better
management practices, and increased diversion of organics through state and local policy and
regulations. The total amount of MSW generated is expected to increase as the U.S. population
continues to grow. The quantities of waste landfilled for 2014 to 2023 (presented in Annex 3.15) are
extrapolated based on population growth and the last comprehensive national assessment of MSW
landfilled from 2013 (EREF 2016). Net CH4 emissions from MSW landfills have decreased since 1990
(see Table 7-3 and Table 7-4).
The estimated quantity of waste placed in industrial waste landfills (from the pulp and paper, and food
processing sectors) has remained relatively steady since 1990, ranging from 9.7 MMT in 1990 to 10.7
MMT in 2023 (see Annex 3.15, Table A-213). CH4 emissions from industrial waste landfills have also
remained at similar levels recently, ranging from 16.1 MMT C02 Eq. in 2005 to 18.9 MMT C02 Eq. in 2023
when accounting for both CH4 generation and oxidation. The EPA has focused the industrial waste
landfills source category on industrial sectors known to generate and dispose of by-products that are
organic and contribute to CH4 generation, which are the pulp and paper and food processing sectors.
Construction and demolition (C&D) landfills, another type of industrial waste landfill, may accept waste
that could degrade (e.g., treated wood), but these waste streams are unlikely to generate significant
amounts of CH4 and are therefore not as relevant to the purpose of national greenhouse gas emissions
estimate. There is also a general lack of data on annual quantities of waste disposed in industrial waste
landfills, and the GHGRP Subpart TT (Industrial Waste Landfills) dataset has confirmed C&D landfills, for
example, are insignificant CH4 generators.
EPA's Landfill Methane Outreach Program (LMOP) collects information on landfill gas energy projects
currently operational or under construction throughout the United States. LMOP's Landfill and Landfill
Gas Energy Database contains information on the gas collection and control systems in place at
landfills provided by organizations that are a part of the program, which can include the amount of
landfill gas collected and flared. In 2023, LMOP identified 22 new landfill gas-to-energy (LFGE) projects
(EPA 2024a) that began operation.
Landfill gas collection and control is not accounted for at industrial waste landfills in this chapter (see
the Methodology discussion for more information).
Table 7-3: CH4 Emissions from Landfills (MMT C02 Eq.)
Activity
1990 | 2005 2019
2020
2021
2022
2023
MSW Cm Generation®
230.0 I
303.7 I
340.3
339.4
334.1
329.3
335.2
Industrial ChU Generation
13.6
17.9
20.9
21.0
21.0
21.0
21.0
MSW CH4 Recovered®
(23.8)
(148.4)
(201.5)
(206.4)
(203.4)
(200.0)
(204.9)
MSW CH4 Oxidized0
(20.6)
(23.6)
(29.4)
(29.3)
(28.9)
(29.5)
(29.8)
7-8 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Industrial ChU Oxidized
(1.4)
(1.8)
(2.1)
(2.1)
(2.1)
(2.1)
(2.1)
MSW net ChU Emissions
185.5
131.6
109.4
103.7
101.8
99.8
100.6
Industrial ChU Emissionsb
12.2
16.1
18.8
18.9
18.9
18.9
18.9
Total
197.8
147.7
128.2
122.6
120.7
118.7
119.5
a For years 1990 to 2004, the Inventory methodology for MSW landfills uses the first order decay methodology. A
methodological change occurs in year 2005. For years 2005 to 2023, directly reported net CH4 emissions from the GHGRP
data plus a scale-up factor are used to account for emissions from landfill facilities that are not subject to the GHGRP. More
details on the scale-up factor and howitwas developed can be found in Annex 3.15. These data incorporate CH4 recovered
and oxidized for MSW landfills. As such, CH4 generation, CH4 oxidation, and CH4 recovery are not calculated separately and
totaled to net CH4 emissions. See the Methodology and Time-Series Consistency section of this chapter for more information.
b Methane recovery is not calculated for industrial landfills because this is not a common practice in the United States. Only 3
of 221 landfills that report to Subpart TT (Industrial Waste Landfills) of the GHGRP had an active gas collection and control
system during the year 2023 (EPA 2024b).
Notes: Parentheses indicate negative values. Totals may not sum due to independent rounding.
Table 7-4: CH4 Emissions from Landfills (kt CH4)
Activity
1990
2005
2019
2020
2021
2022
2023
MSW ChU Generation®
8,214
10,845
12,153
12,120
11,931
11,759
11,973
Industrial ChU Generation
484
638
745
748
750
751
750
MSW CH4 Recovered®
(851)
(5,301)
(7,197)
(7,370)
(7,264)
(7,141)
(7,318)
MSW CH4 Oxidized0
(736)
(843)
(1,048)
(1,045)
(1,032)
(1,055)
(1,063)
Industrial CH4 Oxidized
(48)
(64)
(74)
(75)
(75)
(75)
(75)
MSW net CH4 Emissions
6,627
4,701
3,907
3,705
3,635
3,563
3,592
Industrial net CH4 Emissionsb
436
574
670
673
675
675
675
Total
7,063
5,275
4,578
4,379
4,310
4,238
4,266
a For years 1990 to 2004, the Inventory methodology for MSW landfills uses the first order decay methodology. A
methodological change occurs in year 2005. For years 2005 to 2023, directly reported net CH4 emissions from the GHGRP
data plus a scale-up factor are used to account for emissions from landfill facilities that are not subject to the GHGRP. More
details on the scale-up factor and howitwas developed can be found in Annex 3.15. These data incorporate CH4 recovered
and oxidized for MSW landfills. As such, CH4 generation, CH4 oxidation, and CH4 recovery are not calculated separately and
totaled to net CH4 emissions. See the Methodology and Time-Series Consistency section of this chapter for more information.
b Methane recovery is not calculated for industrial landfills because this is not a common practice in the United States. Only 3
of 221 landfills that report to Subpart TT (Industrial Waste Landfills) of the GHGRP had an active gas collection and control
system during the year 2023 (EPA 2024b).
Notes: Parentheses indicate negative values. Totals may not sum due to independent rounding.
Methodology and Time-Series Consistency
Methodology Applied for MSW Landfills
A combination of IPCC Tier 2 and 3 approaches (IPCC 2006) are used over the reported time series to
calculate emissions from MSW Landfills, usingtwo primary methods in accordance with IPCC
methodological decision trees based on available data. The first method uses the first order decay
(FOD) model as described by the 2006 IPCC Guidelines to estimate CH4 generation. The amount of CH4
recovered and combusted from MSW landfills is subtracted from the CH4 generation and is then
adjusted with an oxidation factor. The oxidation factor represents the amount of CH4 in a landfill that is
oxidized to C02 as it passes through the landfill cover (e.g., soil, clay, geomembrane). This method is
presented below.
Waste 7-9
-------
Equation 7-1: Landfill Methane Emissions
CH4,msw = (Gch4 - UH-M * (1 - ox)
where,
Net CH4 emissions from solid waste
CH4 generation from MSW landfills, using emission factors for DOC, k, MCF, F
from IPCC (2006) and other peer-reviewed sources
CH4 recovered and combusted
CH4 oxidized from MSW landfills before release to the atmosphere, using Ox
values from IPCC (2006) and other peer-reviewed or scientifically validated
literature (40 CFR Part 98)
R
Ox
The second method used to calculate CH4 emissions from landfills, also called the back-calculation
method, is based on directly measured amounts of recovered CH4 from the landfill gas and is expressed
below and by Equation HH-8 in 40 CFR Part 98.343. The two parts of the equation consider the portion of
CH4in the landfill gas that is not collected by the landfill gas collection system, and the portion that is
collected. First, the recovered CH4is adjusted with the collection efficiency of the gas collection and
control system and the fraction of hours the recovery system operated in the calendar year. This quantity
represents the amount of CH4in the landfill gas that is not captured by the collection system; this
amount is then adjusted for oxidation. The second portion of the equation adjusts the portion of CH4in
the collected landfill gas with the efficiency of the destruction device(s), and the fraction of hours the
destruction device(s) operated during the year.
The current Inventory uses both methods to estimate CH4 emissions across the time series within EPA's
Waste Model, as summarized in Figure 7-3 below. This chapter provides a summary of the methods,
activity data, and parameters used. Additional stepwise explanations to generate the net emissions are
provided in Annex 3.15.
Equation 7-2: Net Methane Emissions from MSW Landfills
CH4,Solid Waste = [(gg ~ R) x(l - OX) + R X (l - (DE X fDest))]
,CE X fREC
where,
CH 4,solid waste — Net CH4 emissions from solid waste
R
CE
= Quantity of recovered CH4from Equation HH-4 of EPA's GHGRP
= Collection efficiency estimated at the landfill, considering system coverage,
operation, and cover system materials from Table HH-3 of EPA's GHGRP. If area
by soil cover type information is not available, the default value of 0.75 should be
used (percent)
= fraction of hours the recovery system was operating (percent)
= oxidation factor (percent)
= destruction efficiency (percent)
= fraction of hours the destruction device was operating (fraction)
7-10 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Figure 7-3: Methodologies Used Across the Time Series to Compile the Emission
Estimates for MSW Landfills
1990 - 2004
2005 - 2009
2010 - 2016
2017 - Present
U.S.-specific first-order
decay (FOD) model
Back-casted EPA
GHGRP reported net
methane emissions
EPA GHGRP
reported net
methane emissions
EPA GHGRP
reported net
methane emissions
Annex Steps 1-3
Annex Step 4
Annex Step 5
Annex Step 6
IPCC2006 Emission Factors:
• DOC = 0.20
• MCF = 1
• DOC, = 0.5
• OX = 0.10
• DE = 0.99
Activity Data:
• National waste generation
data multiplied by the
national disposal factor
Back-casted GHGRP
emissions plus a 9%
scale-up factor1'2
• Recovery calculated from
fourCH4 recovery
databases
• Back-calculated CH4
generation3
• Weighted average
oxidation factor based on
GHGRP data3
• Net GHGRP emissions
plus a 9% scale-up
factor2
• GHGRP CH4 recovery plus
a 9% scale-up factor
• Back-calculated CH4
generation3
• Weighted average
oxidation factor based on
GHGRP data 3
• Net GHGRP emissions
plus an 11% scale-up
factor2
• GHGRP CH4 recovery plus
an 11% scale-up factor
• Back-calculated CH4
generation3
• Weighted average
oxidation factor based on
GHGRP data 3
1 The intent of the scale-up factor is to estimate emissions from landfills that do not report to the GHGRP. More details on the
scale-up factor and how it was developed can be found in Annex 3.15. The back-casted emissions are calculated using
directly reported net methane emissions for GHGRP reporting years 2010 to 2016, The back-casted emissions are subject to
change in each Inventory based on new reporting year reports and resubmitted greenhouse gas reports for previous years.
This method is compatible with the 2006IPCC Guidelines because facilities reporting to the GHGRP either use the FOD
method, or directly measured methane recovery data with default emission factors either directly included in the 2006 IPCC
Guidelines or scientifically validated through peer review.
2 Emission factors used by facilities reporting to GHGRP Subpart HH are facility-specific defaults derived from peer-reviewed
literature and the 2006 IPCC Guidelines.
3 Methane generation is back-calculated from the net MSW emissions, estimated methane recovery data, and the weighted
average oxidation factor based on GHGRP Subpart HH reported data of 0.18 between 2010 to 2016, and 0.22 between 2017 to
2021, and 0.23 in 2022 and 2023.
The Waste Model is a spreadsheet developed by the IPCC for purposes of estimating methane
emissions from solid waste disposal sites, adapted to the United States by the inclusion and usage of
U.S.-specific parameters. The Waste Model contains activity and waste generation information from
both the MSW and Industrial landfill sectors and estimates the amount of CH4 emissions from each
sector for each year of the time series, using both methods. Prior to the 1990 through 2015 Inventory,
only the FOD method was used. Methodological changes were made to the 1990 through 2015 Inventory
to incorporate higher tier data (i.e., CH4 emissions as directly reported to EPA's GHGRP), which cannot
be directly applied to earlieryears in the time series without significant bias. The technique used to
merge the directly reported GHGRP data with the previous methodology is described as the overlap
technique in the Time-Series Consistency chapter of the 2006 IPCC Guidelines. Additional details on
the technique used are included in Annex 3.15, and a technical memorandum (RTI 2017).
Supporting information, including details on the techniques used to ensure time-series consistency by
incorporating the directly reported GHGRP emissions is presented in Annex 3.15.
Waste 7-11
-------
Methodology Applied for Industrial Waste Landfills
Emissions from industrial waste landfills are estimated using a Tier 2 approach (IPCC 2006) and a
tailored (country-specific) IPCC waste model in accordance with IPCC methodological decision trees
based on available data. Activity data used are industrial production data (ERG 2023) for two sectors
(pulp and paper manufacturing, and food and beverage manufacturing) to which country-specific
default waste disposal factors are applied (a separate disposal factor for each sector). The disposal
factors, as described below, are based on scientifically reviewed data, and are the same across the
entire time series. The emission factors are based on those recommended by the 2006 IPCC Guidelines
and are the same across the entire time series.
The FOD equation from IPCC (2006) is used via the waste model to estimate methane emissions:
Equation 7-3: Net Methane Emissions from Industrial Waste Landfills
CH4,IND = (gCH4 - ZLM * (l - ox)
where,
CH4,ind = Net CH4 emissions from solid waste
Gch4 = CH4 generation from industrial waste landfills, using production data
multiplied by a disposal factor and emission factors for DOC, k, MCF, F
(IPCC 2006)
R = CH4 recovered and combusted (no recovery is assumed for industrial
waste landfills)
OX = CH4 oxidized from industrial waste landfills before release to the
atmosphere (using the 2006 IPCC Guidelines value for OX of 0.10)
The activity data used in the emission calculations are production data (e.g., the amount of meat,
poultry, vegetables processed; the amount of paper produced) versus disposal data. There are currently
no facility-specific data sources that track and report the amount and type of waste disposed of in the
universe of industrial waste landfills in the United States. Based on this limited information, the
Inventory methodology assumes most of the organic waste placed in industrial waste landfills originates
from the food processing (meat, vegetables, fruits) and pulp and paper sectors, thus estimates of
industrial landfill emissions focused on these two sectors.
Awaste disposal factor is applied to the annual quantities of keyfood products generated. A waste
disposal factor of 4.86 percent is used for 1990 to 2009 and a factor of 6 percent is used for 2010 to the
current year. The 4.86 percent disposal factor is based on available data from a 1993 Report to Congress
(EPA 1993). The 6 percent waste disposal factor is derived from recent surveys of the food and beverage
industry where approximately 94 percent of food waste generated is repurposed (FWRA2016). The
composition of waste disposed of in industrial waste landfills is expected to be more consistent in terms
of composition and quantity than that disposed of in MSW landfills. The amount of waste landfilled is
assumed to be a fraction of production that is held constant over the time series as explained in Annex
3.15.
Landfill CH4 recovery is not accounted for in industrial waste landfills and is believed to be minimal
based on available data collected under EPA's GHGRP for industrial waste landfills (Subpart TT), which
shows that only three of the 221 facilities, or 1 percent of facilities, have active gas collection systems
(EPA 2024b). The amount of CH4 oxidized by the landfill cover at industrial waste landfills is assumed to
7-12 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
be 10 percent of the CH4 generated (IPCC 2006; Mancinelli and McKay 1985; Czepiel et al. 1996) for all
years.
Additionally, the 2019 Refinement to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories
(IPCC 2019) were reviewed to determine if any revisions were required to emission factors,
methodologies, and assumptions underlying emission estimates for all source categories. None of the
2019 refinements are applicable to the country-specific methodology applied for the landfills source
category.
Box 7-3: Nationwide Municipal Solid Waste Data Sources
Municipal solid waste (MSW) generated in the United States can be managed through a variety of
methods. MSW that is not recycled, composted, combusted with energy recovery, or digested is
assumed to be landfilled. In addition to these management pathways, waste or excess food from the
food manufacturing and processing sector may be disposed through the sewerage network, used for
animal feed, land application, donated for human consumption, and rendered or recycled into biofuels
in the case of animal by-products, fats, oils and greases.
There have been three main sources for nationwide solid waste management data in the United States
that the Inventory has used (see Annex 3.15, Box A-3 for comparison of estimates from these data
sources):
• The BioCycle and Earth Engineering Center of Columbia University's SOG in America surveys [no
longer published];
• The EPA's Advancing Sustainable Materials Management: Facts and Figures reports; and
• The EREF's MSW Generation in the United States reports.
The SOG surveys and, most recently EREF, collected state-reported data on the amount of waste
generated and the amount of waste managed via different management options: landfilling, recycling,
composting, and combustion. These data sources used a 'bottom-up' method. The survey asked for
actual tonnages instead of percentages in each waste category (e.g., residential, commercial,
industrial, construction and demolition, organics, tires) for each waste management option. If such a
breakdown was not available, the survey asked for total tons landfilled. The data were adjusted for
imports and exports across state lines so that the principles of mass balance were adhered to for
completeness, whereby the amount of waste managed did not exceed the amount of waste generated.
The SOG and EREF reports present survey data aggregated to the state level.
The EPk Advancing Sustainable Materials Management: Facts and Figures report characterizes national
post-consumer municipal solid waste (MSW) generation and management using a top-down materials
flow (mass balance) methodology. It captures an annual snapshot of MSW generation and management
in the United States for specific products. Data are gathered from U.S. Government (e.g., U.S. Census
Bureau and U.S. Department of Commerce), state environmental agencies, industry and trade groups,
and sampling studies. The materials flow methodology develops MSW waste generation estimates of
quantities of MSW products in the marketplace (using product sales and replacement data) and
assessing waste generation by component material based on product lifespans. The data are used to
estimate tons of materials and products generated, recycled, combusted with energy recovery,
managed via other food waste management pathways, or landfilled nationwide. MSW that is not
Waste 7-13
-------
recycled or composted is assumed to be combusted or landfilied, except for wasted food, which uses a
different methodology and includes nine different management pathways. The 2018 Facts and Figures
Report (EPA 2020) uses a methodology that expanded the number of management pathways to include:
animal feed; bio-based materials and/or biochemical processing (i.e., rendering); co-digestion and/or
anaerobic digestion; composting/aerobic processes; combustion; donation; land application; landfill;
and sewer or wastewater treatment.
In this Inventory, emissions from solid waste management are presented separately by waste disposal
option, except for recycling of waste materials.
• Recycling: Emissions from recycling are attributed to the stationary combustion of fossil fuels that
may be used to power on-site recycling machinery and are presented in the stationary combustion
chapter in the Energy sector. The emissions estimates for recycling are not called out separately.
• Landfill Disposal: Emissions from solid waste disposal in landfills and the composting of solid
waste materials are presented in the Landfills and Composting sections in the Waste sector of this
report.
• Anaerobic Digestion: Emissions from anaerobic digesters are presented in three different sections
depending on the digester category:
Emissions from on-farm digesters are included in the Agriculture sector.
Emissions from digesters at wastewater treatment plants are included in the Waste sector, and
Emissions from stand-alone digesters are also included in the Waste sector.
Waste Incineration: Emissions from waste incineration are accounted for in the Incineration chapter of
the Energy sector of this report because, in the United States, almost all incineration of MSW occurs at
waste-to-energy (WTE) facilities or industrial facilities where useful energy is recovered.
Uncertainty
Several types of uncertainty are associated with the estimates of CH4 emissions from MSW and
industrial waste landfills when the FOD method is applied directly for 1990 to 2004 in the Waste Model
and, to some extent, in the GHGRP methodology. The approach used in the MSW emission estimates
assumes that the CH4 generation potential (L0) and the rate of decay that produces CH4from MSW, as
determined from several studies of CH4 recovery at MSW landfills, are representative of conditions at
U.S. MSW landfills. When this top-down approach is applied at the nationwide level, the uncertainties
are assumed to be less than when applying this approach to individual landfills and then aggregating the
results to the national level. In other words, the FOD method as applied in this Inventory is not facility-
specific modeling and while this approach may over- or underestimate CH4 generation at some landfills
if used at the facility-level, the result is expected to balance out because it is being applied nationwide.
There is a high degree of uncertainty associated with the FOD model, particularly when a homogeneous
waste composition and hypothetical decomposition rates are applied to heterogeneous landfills (IPCC
2006). There is less uncertainty in EPA's GHGRP data because this methodology is facility-specific, uses
directly measured CH4 recovery data (when applicable), and allows for a variety of landfill gas collection
efficiencies, destruction efficiencies, and/or oxidation factors to be used.
7-14 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Uncertainty also exists in the scale-up factors (both 9 percent and 11 percent) applied for years 2005 to
2016 and 2017 to 2023, respectively, and in the back-casted emissions estimates for 2005 to 2009. As
detailed in RTI (2018), limited information is available for landfills that do not report to the GHGRP. RTI
developed an initial list of landfills that do not report to the GHGRP with the intent of quantifying the
total waste-in-place for these landfills that would add up to the scale-up factor. Input was provided by
industry, LMOP, and additional EPA support. However, many gaps existed in the initial development of
this Non-Reporting Landfills Database. Assumptions were made for hundreds of landfills to estimate
their waste-in-place and the subsequent scale-up factors. The waste-in-place estimated for each
landfill is likely not 100 percent accurate and should be considered a reasonable estimate. Additionally,
a simple methodology was used to back-cast emissions for 2005 to 2009 using the GHGRP-reported
emissions from 2010 to 2023. This methodology does not factor in annual landfill to landfill changes in
landfill CH4 generation and recovery. Because of this, an uncertainty factor of 25 percent is applied to
the scale-up factor and years (emission estimates) the scale-up factor is applied to.
Aside from the uncertainty in estimating landfill CH4 generation, uncertainty also exists in the estimates
of the landfill gas oxidized at MSW landfills. Facilities directly reporting to EPA's GHGRP can use
oxidation factors ranging from 0 to 35 percent, depending on their facility-specific CH4 flux. As
recommended by the 2006IPCC Guidelines for managed landfills, a 10 percent default oxidation factor
is applied in the Inventory for both MSW landfills (those not reporting to the GHGRP and for the years
1990 to 2004 when GHGRP data are not available) and industrial waste landfills regardless of climate,
the type of cover material, and/or presence of a gas collection system.
Another significant source of uncertainty lies with the estimates of CH4 recovered by flaring and gas-to-
energy projects at MSW landfills that are sourced from the Inventory's CH4 recovery databases (used for
years 1990 to 2004). Four CH4 recovery databases are used to estimate nationwide CH4 recovery for
MSW landfills for 1990 to 2009. The GHGRP MSW landfills database was added as a fourth recovery
database starting with the 1990 to 2013 Inventory report (two years before the full GHGRP data set
started being used for net CH4 emissions for the Inventory). Relying on multiple databases for a
complete picture introduces uncertainty because the coverage and characteristics of each database
differs, which increases the chance of double counting avoided emissions. The methodology and
assumptions that go into each database differ. For example, the flare database assumes the midpoint of
each flare capacity at the time it is sold and installed at a landfill; the flare may be achieving a higher
capacity, in which case the flare database would underestimate the amount of CH4 recovered.
Additionally, two databases, the EIA database and flare vendor database, could no longer be updated for
the entire time series due to external factors. For example, the EIA database has not been updated since
2006 because the EIA stopped collecting landfill recovery data. The EIA database has, for the most part,
been replaced by the GHGRP MSW landfills database. The flare database was populated annually until
2015, but decreasing, voluntary participation from flare vendors sharing their flare sales data for several
years prior to 2015.
To avoid double counting and to use the most relevant estimate of CH4 recovery for a given landfill, a
hierarchical approach is used among the four databases. GHGRP data and the EIA data are given
precedence because facility data were directly reported; the LFGE data are given second priority
because CH4 recovery is estimated from facility-reported LFGE system characteristics; and the flare
data are given the lowest priority because this database contains minimal information about the flare,
no site-specific operating characteristics, and includes smaller landfills not included in the other three
databases (Bronstein et al. 2012). The coverage provided across the databases most likely represents
Waste 7-15
-------
the complete universe of landfill CH4 gas recovery; however, the number of unique landfills between the
four databases does differ.
The 2006IPCC Guidelines default value of 10 percent for uncertainty in recovery estimates was used for
two of the four recovery databases in the uncertainty analysis where metering of landfill gas was in place
(for about 64 percent of the CH4 estimated to be recovered). This 10 percent uncertainty factor applies
to the LFGE database; 12 percent to the EIA database; and 1 percent for the GHGRP MSW landfills
dataset because of the supporting information provided and rigorous verification process. For flaring
without metered recovery data (the flare database), a much higher uncertainty value of 50 percent is
used. The compounding uncertainties associated with the four databases in addition to the
uncertainties associated with the FOD method and annual waste disposal quantities leads to the large
upper and lower bounds for MSW landfills presented in Table 7-5.
The lack of landfill-specific information regarding the number and type of industrial waste landfills in the
United States is a primary source of uncertainty with respect to the industrial waste generation and
emission estimates. The approach used here assumes that most of the organic waste disposed of in
industrial waste landfills that would result in CH4 emissions consists of waste from the pulp and paper
and food processing sectors. However, because waste generation and disposal data are not available in
an existing data source for all U.S. industrial waste landfills, a straight disposal factor is applied over the
entire time series to the amount produced to determine the amounts disposed. Industrial waste
facilities reporting under EPA's GHGRP do report detailed waste stream information, and these data
have been used to improve, for example, the DOC value used in the Inventory methodology for the pulp
and paper sector. A 10 percent oxidation factor is also applied to CH4 generation estimates for industrial
waste landfills and carries the same amount of uncertainty as with the factor applied to CH4 generation
for MSW landfills. The specified probability density functions (PDFs) are assumed to be normal for most
activity data and emission factors, and due to lack of data, are based on expert judgement (RTI 2004).
The results of the 2006 IPCC Guidelines Approach 2 quantitative uncertainty analysis are summarized in
Table 7-5. There is considerable uncertainty for the MSW landfills estimates due to the many data
sources used, each with its own uncertainty factor.
Table 7-5: Approach 2 Quantitative Uncertainty Estimates for CH4 Emissions from
Landfills (MMT C02 Eq. and Percent)
Source
Gas
2023 Emission
Estimate
(MMTCO2 Eq.)
Uncertainty Range Relative to Emission Estimate"
(MMT CO2 Eq.)
(%)
Lower Bound
Upper Bound
Lower Bounc
Upper Bound
Total Landfills
CH4
119.5
108.9
136.1
-9% +14%
MSW
ch4
100.6
98.7
120.5
-2%
+20%
Industrial
ch4
18.9
16.1
26.2
-15%
+39%
a Range of emission estimates predicted by Monte Carlo stochastic simulation for a 95 percent confidence interval. Individual
uncertainty factors are applied to activity data and emission factors in the Monte Carlo analysis.
QA/QC and Verification
General quality assurance/quality control (QA/QC) procedures were applied consistent with the U.S.
Inventory QA/QC Plan, which is in accordance with Vol. 1, Chapter 6 of the 2006 IPCC Guidelines (see
Annex 8 for more details). QA/QC checks are performed for the transcription of the published data set
7-16 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
(e.g., EPA's GHGRP dataset) used to populate the Inventory data set in terms of completeness and
accuracy against the reference source. Additionally, all datasets used for this category have been
checked to ensure they are of appropriate quality and are representative of U.S. conditions. The primary
calculation spreadsheet is tailored from the 2006IPCC Guidelines waste model and has been verified
previously using the original, peer-reviewed IPCC waste model. All model input values and calculations
were verified by secondary QA/QC review. Stakeholder engagements sessions in 2016 and 2017 were
used to gather input on methodological improvements and facilitate an external expert review on the
methodology, activity data, and emission factors.
Category-specific checks include the following:
Evaluation of the secondary data sources used as inputs to the Inventory dataset to ensure they
are appropriately collected and are reliable;
Cross-checking the data (activity data and emissions estimates) with previous years to ensure
the data are reasonable, and that any significant variation can be explained;
Conducting literature reviews to evaluate the appropriateness of country-specific emission
factors (e.g., DOC values, precipitation zones with respect to the application of the k values)
given findings from recent peer-reviewed studies; and
Reviewing secondary datasets to ensure they are nationally complete and supplementing where
necessary (e.g., using a scale-up factor to account for emissions from landfills that do not report
to EPA's GHGRP).
A primary focus of the QA/QC checks in past Inventories was to ensure that CH4 recovery estimates
were not double-counted and that all LFGE projects and flares were included in the respective project
databases. QA/QC checks performed in the past for the recovery databases were not performed in this
Inventory, because new data were not added to the recovery databases in this Inventoryyear.
For the GHGRP data, EPA verifies annual facility-level reports through a multi-step process (e.g.,
combination of electronic checks and manual reviews by staff) to identify potential errors and ensure
that data submitted to EPA are accurate, complete, and consistent.2 Based on the results of the
verification process, EPA follows up with facilities to resolve mistakes that may have occurred. The post-
submittals checks are consistent with several general and category-specific QC procedures, including
range checks, statistical checks, algorithm checks, and year-to-year checks of reported data and
emissions. For the MSW Landfills sector, under Subpart HH of the GHGRP, MSW Landfills with gas
collection are required to report emissions from their site using both a forward- (using a first order decay
model as a basis) and back-calculating (using parameters specific to the landfill itself, such as
measured recovery and collection efficiency of the landfill gas) methodology. Details on the forward-
and back-calculation approach can be found in Annex 3.15 and 40 CFR Subpart HH of Part 98.
Reporters can choose which of these two methodologies they believe best represents the emissions at
their landfill and are required to submit that value as their total Subpart HH emissions. Facilities are
generally not expected to switch between the two equations each year, as the emissions calculated
using each method can vary greatly and can have a significant effect on emission trends for that landfill,
and potentially the entire MSW Landfill sector under the GHGRP. Key checks are in place to assure that
emissions are trending in a sensible way year overyear for each reporting landfill.
2 See https://www.epa.gov/sites/prodijction/files/9015-07/documents/ghgrp verification factsheet.pdf.
Waste 7-17
-------
Recalculations Discussion
Revisions to the individual facility reports submitted to EPA's GHGRP can be made at any time and a
portion of facilities have revised their reports since 2010 for various reasons, resulting in changes to the
total net CH4 emissions for MSW landfills. Each Inventoryyear, the back-casted emissions for 2005 to
2009 will be recalculated using the most recently verified data from the GHGRP. Changes in these data
result in changes to the back-casted emissions. The impact of the revisions to the GHGRP Subpart HH
annual greenhouse gas reports resubmitted for 2020 to 2022 slightly increased total Subpart HH
reported net emissions and Subpart HH total recovery by an average of 0.9 percent and 0.5 percent
respectively, which decreased net MSW emissions by an average of 1.2 percent. A change in net Subpart
HH reported emissions results in the same percentage change in the Inventory emissions for that year.
Slight revisions were also made to the food and beverage sector production data for nearly every year of
the time series. The production data increased by 1 to 3 MMT peryear between 1990 and 2017 and
decreased by a few MMT peryear between 2018 and 2023. These revisions very slightly increased net
emissions from industrial waste landfills by less than 0.003 percent between 1990 and 2010. Emissions
decreased slightly between 2011 and 2021 (ranging from 0.005 percent in 2011 to a percent in 2021) and
increased by 0.01 percent in 2022.
The combined changes to the MSW and industrial waste landfills activity data resulted in annual
increases ranging from 0.001 percent in 2011 to 0.004 percent to net emissions between 2011 and 2016.
Emissions increased between 0.5 percent in 2017 to 0.9 percent in 2022.
The U.S. Census Bureau data indicates population changes in the United States from 2022 to 2023 for
reporting years 2020 through 2022. In 2020 and 2021, the U.S. population grew slightly by approximately
0.008 percent and 0.047 percent, respectively, but declined by around 0.005 percent in 2022. For Puerto
Rico, population changes also occurred during the same period, with a 4 percent increase in 2020,
followed by decreases of 0.027 percent in 2021 and 0.052 percent in 2022. These shifts impacted the
solid waste generated in U.S. tons and waste territories by similar percentages, likely contributing to the
observed decrease in net emissions from 2020 through 2022 in the previous (1990 through 2022)
Inventory estimates compared to the current (1990 through 2023) Inventory estimates.
Over the time series, the total emissions on average decreased by 0.1 percent (0.2 MMT C02 eq.) from
the previous Inventory.
Planned Improvements
The EPA received recommendations from industry stakeholders to revise the DOC and decay rate (k)
values used in Subpart HH of the GHGRP to better reflect recent trends in waste composition at MSW
landfills. In response, the EPA developed a multivariate analysis using publicly available GHGRP data to
optimize DOC and k values for over 1,100 reporting landfills. The analysis informed a rulemaking
revision to 40 CFR Part 98 that was finalized in April 2024. This includes key revisions to the default DOC
and k values in Subpart HH, which will take effect on January 1, 2025.
The revisions, detailed in the final rule, include updated DOC values and k rates to enhance the
accuracy of methane generation modeling from landfills. DOC values are revised for different waste
characterization options, including adjustments from 0.20 to 0.17 for bulk waste and 0.31 to 0.27 for
bulk MSW without inerts. The k values are also updated based on average optimal values, with new
ranges for uncharacterized MSW. These changes are expected to impact future national and state MSW
7-18 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
inventories starting with the 2025 reporting year, with data published in 2026. The revised factors will be
applied retroactively for disposal years 2010 and later to better account for changes in waste
composition, improving the accuracy of landfill methane emissions estimates in the U.S. national and
state inventories.
In recent years, a few studies have measured emissions at a large number of landfills (200+) from plane-
based and satellite-based sensors. These studies (Cusworth 2024 and Nesser 2024) indicate
underestimating of MSW landfill emissions at the facility level and national level. There are a variety of
potential uses of data from new studies, including replacing a previous estimate or factor, verifying or
QA of an existing estimate or factor, and identifying areas for updates. For the present, this indicates a
need to assess emission factors, equations, and methodologies used to calculate emissions from MSW
landfills, in particular the emission factors and equations used in GHGRP Subpart HH. As discussed
above, revisions were finalized in a 2024 rulemaking to improve the data collected. These changes to
emission factors and equations will be examined in the next year to see what potential impact the
changes will have on national emission estimates produced in the Inventory. Additionally, these
estimates will be compared to the recent methane measurement studies as a QA check for the
emission factors, equations, and methodologies used in the Inventory.
With respect to the scale-up factor, EPA received comments on revisions made to the scale-up for the
1990 to 2020 Inventory from a total waste-in-place approach to a time-based threshold of 50 years.
Commenters noted that this time-based threshold approach does not adjust for the non-linearity of
methane production of landfill gas. In response, EPA will further investigate how best to account for
emissions from MSW landfills that do not report to the GHGRP, including using the FOD model for these
landfills based on estimated annual waste disposed for this subset of landfills between 2005 to 2023,
reverting to the total waste-in-place approach, or modifying the time-based threshold approach.
Relatedly, EPA will periodically assess the impact to the waste-in-place and emissions data from
GHGRP facilities that have resubmitted annual reports during any reporting years, are new reporting
facilities, and from facilities that have stopped reporting to the GHGRP to ensure national estimates are
as complete as possible. Facilities may stop reporting to the GHGRP when they meet the "off-ramp"
provisions (reported less than 15,000 metric tons of C02 equivalent emissions for 3 consecutive years or
less than 25,000 metric tons of C02 equivalent emissions for 5 consecutive years). If warranted, EPA will
revise the scale-up factor to reflect newly acquired information to ensure completeness of the
Inventory. EPA considered public comments received on the 1990 through 2019 Inventory specific to
using a time-based threshold to calculate the scale-up factor instead of a total waste-in-place
approach. The rationale supporting the comments was that older, closed landfills with large quantities
of waste-in-place are driving up the scale-up factor but have little impact on total methane generation.
EPA assessed two time-based scenarios for developing the scale-up factor - one scenario looking at the
past 30 years of waste disposed, and the second looking at the past 50 years of waste disposed. The 50-
year time-based threshold was applied and resulted in the 11 percent scale-up factor used between
2017 and 2023.
EPA is planning to account for unmanaged landfills in Puerto Rico and other U.S. Territories to the landfill
emissions estimates. Data limitations for historical waste received at these sites make this challenging.
Presently, emissions from managed sites in Puerto Rico and Guam are accounted for in 2005 to present
as part of the GHGRP Subpart HH dataset.
Waste 7-19
-------
Box 7-4: Overview of U.S. Solid Waste Management Trends
As shown in Figure 7-4 and Figure 7-5 landfilling of MSW is currently and has been the most common
waste management practice. A large portion of materials in the waste stream are recovered for recycling
and composting, which is becoming an increasingly prevalent trend throughout the country. Materials
that are composted and recycled would have previously been disposed in a landfill.
Table 7-6 presents the national-level material composition of waste disposed across typical MSW
landfills in the United States over time. It is important to note that the actual composition of waste
entering each landfill will vary from that presented in Table 7-6.
Understanding how the waste composition changes over time, specifically for the degradable waste
types (i.e., those types known to generate CH4 as they break down in a modern MSW landfill), is
important for estimating greenhouse gas emissions. Increased diversion of degradable materials so that
they are not disposed of in landfills reduces the CH4 generation potential and CH4 emissions from
landfills. For certain degradable waste types (i.e., paper and paperboard), the amounts discarded have
decreased over time due to an increase in waste diversion through recycling and composting (see Table
7-6 and Figure 7-6). As shown in Figure 7-6, the diversion of food scraps has been consistently low since
1990 because most cities and counties do not practice curbside collection of these materials, although
the quantity has been slowly increasing in recent years. Neither Table 7-6 nor Figure 7-6 reflect the
frequency of backyard composting of yard trimmings and food waste because this information is largely
not collected nationwide and is hard to estimate.
Figure 7-4: Management of Municipal Solid Waste in the United States, 2018
Notes: 2018 is the latest year of available data. Data taken from Table 35 of EPA (2020a). MSW to WTE is combustion with energy
recovery (WTE = waste-to-energy).
Source: EPA (2020b)
7-20 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Figure 7-5: MSW Management Trends from 1990 to 2018
160
140
120
„ ioo
c
|2
J 80
60
40
20
0
1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018
Notes: 2018 is the latest year of available data. Only one year of data (2018) is available for the "Other Food Management"
category.
Source: EPA (2020b). The EPA Advancing Sustainable Materials Management reports only present data for select years, thus
several reports were used in the compilation of this figure. All data were taken from Table 35 in EPA 2020b for 1990, 2000, 2015,
2017 and 2018. Data were taken from Table 35 in EPA (2019) for 2010 and 2016. Data were taken from EPA (2018) for 2014. Data
were taken from Table 35 of EPA (2016b) for 2012 and 2013. Data were taken from Table 30 of EPA (2014) for 2008 and 2011. The
reports with data available for years prior to EPA (2012) can be provided upon request but are no longer on the EPA's Advancing
Sustainable Materials Managementweb site.3
Table 7-6: Materials Discarded in the Municipal Waste Stream by Waste Type from
1990 to 2018 (Percent)
Waste Type
1990
2005
2015
2016
2017
2018
Paper and Paperboard
30.0%
24.7%
13.3%
12.7%
13.1%
11.8%
Glass
6.0%
5.8%
5.0%
4.9%
4.9%
5.2%
Metals
7.2%
7.9%
9.5%
9.8%
9.9%
9.5%
Plastics
9.5%
16.4%
18.9%
18.9%
19.2%
18.5%
Rubber and Leather
3.2%
2.9%
3.3%
3.4%
3.5%
3.4%
Textiles
2.9%
5.3%
7.7%
8.0%
8.0%
7.7%
Wood
6.9%
7.5%
8.0%
8.8%
8.7%
8.3%
Other
1.4%
1.8%
2.2%
2.2%
2.2%
2.0%
Food Scraps
13.6%
18.5%
22.0%
22.1%
22.0%
24.1%
Yard Trimmings
17.6%
7.0%
7.8%
6.9%
6.2%
7.2%
Miscellaneous Inorganic Wastes
1.7%
2.2%
2.3%
2.3%
2.3%
2.3%
Source: EPA (2020b)
Landfilling
Recycling
3 See https://www.epa.gov/facts-and-figures-about-materials-waste-and-recvcling/advancing-sustainable-materials-
management.
Waste 7-21
-------
Figure 7-6: Percent of Degradable Materials Diverted from Landfills from 1990 to 2018
(Percent)
90%
1 80%
c
JS
~ 70%
2
IS 60%
¦5
| 50%
in
—
& 40%
ts
i 30%
0
1 20%
c
O)
cD 10%
CL
0%
Paper arid Paperboard
Yard Trimmings (Composted)
Other Food Management (Non-disposal)
Food Scraps (Composted)
1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018
Note: The data shown in this chart are for recycling of paper and paperboard, composting of food scraps and yard trimmings, and
alternative management pathways for the Other Food Management (non-disposal) category. The Other Food Management (non-
disposal) category is a new addition and only one year of data are available for 2018 (28 percent of the food waste generated was
beneficially reused or managed using a method that was not landfilling, recycling, or composting). The Other Food Management
pathways include animal feed, bio-based materials/biochemical processing, co-digestion/anaerobic digestion, donation, land
application, and sewer/wastewater treatment.
Source: EPA (2020b). The EPA Advancing Sustainable Materials reports only present data for select years, thus several reports
were used in the compilation of this figure. All data were taken from Table 35 in EPA (2020b) for 1990, 2000, 2015, 2017 and 2018.
Data were taken from Table 35 in EPA (2019) for 2010 and 2016. Data were taken from EPA (2018) for 2014. Data were taken from
Table 35 of EPA (2016b) for 2012 and 2013. Data were taken from Table 30 of EPA (2014) for 2008 and 2011. The reports with data
available for years prior to EPA (2012) can be provided upon request, but are no longer on the EPA's Advancing Sustainable
Materials Management website.4
7.2 Wastewater Treatment and Discharge
(Source Category 5D)
Wastewater treatment and discharge processes are sources of anthropogenic methane (CH4) and
nitrous oxide (N20) emissions. Wastewater from domestic and industrial sources is treated to remove
soluble organic matter, suspended solids, nutrients, pathogens, and chemical contaminants.5 In the
United States, approximately 19 percent of domestic wastewater is treated in septic systems or other
on-site systems, while the rest is collected and treated centrally (U.S. Census Bureau 2023a). Treatment
of industrial wastewater may occur at the industrial plant using package or specially designed treatment
plants or be collected and transferred off site for co-treatment with domestic wastewater in centralized
4 See https://www.epa.gov/facts-and-figures-about-materials-waste-and-recvcling/advancing-sustainable-materials-
management.
5 Throughout the Inventory, emissions from domestic wastewater also include any commercial and industrial wastewater
collected and co-treated with domestic wastewater.
7-22 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
treatment systems. Therefore, in the United States, domestic wastewater entering a centralized
wastewater treatment system can consist of municipal, commercial and institutional, as well as a
portion of industrial wastewater.
Centralized Treatment. Centralized wastewater treatment systems use sewer systems to collect and
transport wastewater to the treatment plant. Sewer collection systems provide an environment
conducive to the formation of CH4, which can be substantial depending on the configuration and
operation of the collection system (Guisasola et al. 2008). Recent research has shown that at least a
portion of CH4 formed within the collection system enters the centralized system where it contributes to
CH4 emissions from the treatment system (Foley et al. 2015).
Centralized treatment can include a variety of processes, including physical separation of material that
readily settles out (primary treatment), biological processes to convert and remove contaminants
(secondary treatment), and advanced treatment to remove targeted pollutants, such as nutrients
(tertiary treatment). Not all wastewater treatment plants conduct primary treatment prior to secondary
treatment, and not all plants conduct advanced or tertiary treatment (EPA 2010).
Secondary treatment generally removes soluble organic matter. Microorganisms can biodegrade soluble
organic material in wastewater under aerobic or anaerobic conditions, where the latter condition
produces CH4. The resulting biomass (sludge) is removed from the wastewater (effluent) prior to
discharge to the receiving stream and may be further biodegraded under aerobic or anaerobic
conditions, such as anaerobic sludge digestion. Sludge can be produced from both primary and
secondary treatment operations. In some cases, facilities further process this sludge onsite via
anaerobic sludge digesters. These digesters still emit CH4, due to unintentional leakages, but at a greatly
reduced rate than anaerobic treatment without recovery (either CH4 capture or flaring). The resulting
biogas and digestate may be beneficially reused. Constructed wetlands are coupled anaerobic-aerobic
systems more commonly used as a final treatment step following settling and biological treatment (i.e.,
tertiary treatment after primary and secondary treatment) and in limited cases used as the sole method
of wastewater treatment (<0.1 percent of centrally treated wastewater (ERG 2016). Constructed
wetlands develop natural processes that involve vegetation, soil, and associated microbial assemblages
to treat incoming contaminants (IPCC 2014). Constructed wetlands do not produce secondary sludge
(sewage sludge). Emissions from flooded lands or constructed waterbodies (not used for wastewater
treatment) and lands converted to flooded lands (not used for wastewater treatment) are estimated and
reported in Chapter 6, under Sections 6.8 Wetlands Remaining Wetlands and 6.9 Lands Converted to
Wetlands.
Nitrous oxide is generated as a by-product of nitrification, or as an intermediate product of
denitrification of the nitrogen (N) present, usually in the form of urea, proteins, and ammonia in
wastewater. Ammonia N is converted to nitrate (N03) through the aerobic process of nitrification.
Denitrification occurs under anoxic/anaerobic conditions, whereby anaerobic or facultative organisms
reduce oxidized forms of nitrogen (e.g., nitrite, nitrate) in the absence of free oxygen to produce nitrogen
gas (N2). No matter where N20 is formed it is typically stripped (i.e., transferred from the liquid stream to
the air and ultimately emitted into the atmosphere) in aerated parts of the treatment process. Stripping
also occurs in non-aerated zones at rates lower than in aerated zones. More recent research has
revealed that emissions from nitrification or nitrification-denitrification processes at wastewater
treatment, previouslyjudged to be a minor source, may in fact result in more substantial emissions
(IPCC 2019).
Waste 7-23
-------
On-site Treatment. Most on-site systems in the United States are septic systems composed of a septic
tank, generally buried in the ground, and a soil dispersion system. Solids and dense materials contained
in the incoming wastewater (influent) settle in the septic tank as sludge. Floatable material (scum) is
also retained in the tank. The sludge that settles on the bottom of the tank undergoes anaerobic
digestion. Partially treated water is discharged in the soil dispersal system. The solid fraction
accumulates and remains in the tank for several years, during which time it degrades anaerobically. The
gas produced from anaerobic sludge digestion (mainly CH4 and biogenic C02) rises to the liquid surface
and is typically released through vents. The gas produced in the effluent dispersal system (mainly N20
and biogenic C02) is released through the soil.
Discharge. Dissolved CH4 and N20 present in wastewater discharge to aquatic environments has the
potential to be released into the atmosphere (Short et al. 2014; Short et al. 2017). In addition, the
presence of organic matter or nitrogen in wastewater discharges is generally expected to increase CH4
and N20 emissions from these aquatic environments. Where organic matter is released to slow-moving
aquatic systems, such as lakes, estuaries, and reservoirs, CH4 emissions are expected to be higher.
Similarly, in the case of discharge to nutrient-impacted or hypoxic waters, N20 emissions can be
significantly higher (IPCC 2019).
In summary, the principal factor in determining the CH4 generation potential of wastewater is the
amount of degradable organic material in the wastewater. Common parameters used to measure the
organic component of the wastewater are the biochemical oxygen demand (BOD) and chemical oxygen
demand (COD). Under the same conditions, wastewater with higher COD (or BOD) concentrations will
generally yield more CH4 than wastewater with lower COD (or BOD) concentrations. BOD represents the
amount of oxygen that would be required to completely consume the organic matter contained in the
wastewater through aerobic decomposition processes, while COD measures the total material
available for chemical oxidation (both biodegradable and non-biodegradable). The BOD value is
commonly expressed in milligrams of oxygen consumed per liter of sample during 5 days of incubation
at 20°C, or BOD5. Throughout the rest of this chapter, the term "BOD" refers to BOD5. IPCC (2006)
indicates that because BOD is an aerobic parameter, it may be preferable to use COD to estimate CH4
production which occurs in anaerobic conditions; however, the IPCC authors recognized that in most
countries, operations predominately measure BOD for domestic wastewater and COD for industrial
wastewater, so likelihood of available data were considered for the methodological development.
Where present, biogas recovery and flaring operations reduce the amount of CH4 generated that is
actually emitted. Per IPCC guidelines (IPCC 2019), emissions from anaerobic sludge digestion, including
biogas recovery and flaring operations, where the digester's primary use is for treatment of wastewater
treatment solids, are estimated and reported under wastewater treatment. The principal factor in
determining the N20 generation potential of wastewater is the amount of N in the wastewater. The
variability of N in the influent to the treatment system, as well as the operating conditions of the
treatment system itself, also impact the N20 generation potential. The methods and underlying data
sources to estimate emissions from are described in further detail in the "Methodology and Time-Series
Consistency" section below for treatment of domestic and industrial wastewater.
Total methane (CH4) emissions from wastewater treatment and discharge were 21.1 MMT C02 Eq. (755
kt CH4) in 2023, including 14.0 MMT C02 Eq. (500 kt CH4) from domestic wastewater treatment and
discharge and 7.1 MMT C02 Eq. (255 kt CH4) from industrial wastewater treatment and discharge (Table
7-7, Table 7-8). Methane emissions from domestic wastewater remained fairly steady from 1990 through
2002 but have decreased since that time due to decreasing use of anaerobic systems, including
7-24 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
reduced use of on-site septic systems and central anaerobic treatment systems (EPA 1992, 1996, 2000,
and 2004; U.S. Census Bureau 2023a). Methane emissions from industrial wastewater emissions have
generally increased across the time series through 1999 and then fluctuated up and correspond with
production changes from the pulp and paper manufacturing, meat and poultry processing, fruit and
vegetable processing, starch-based ethanol production, petroleum refining, and brewery industries.
Industrial wastewater emissions have generally seen an uptick since 2016.
Nitrous oxide (N20) emissions from wastewater treatment and discharge in 2023 totaled 20.8 MMT C02
Eq. (79 kt N20), including 20.3 MMT C02 Eq. (77 kt N20) from domestic wastewater and 0.5 MMT C02 Eq.
(1.7 kt N20) from industrial wastewater (Table 7-7, Table 7-8). N20 emissions from domestic wastewater
have gradually increased across the time series because of an increasing U.S. population and protein
consumption. Industrial emission sources have gradually increased across the time series with
production changes associated with the treatment of wastewater, namely from meat and poultry
processing and petroleum refining, but also with contributions from pulp and paper manufacturing and
brewery industries. Overall, treatment of wastewater emitted 41.9 MMT C02 Eq. in 2023.
Table 7-7: CH4 and N20 Emissions from Domestic and Industrial Wastewater
Treatment (MMT C02 Eq.)
Activity
1990 |
2005
2019
2020
2021
2022
2023
CH4
22.7
22.7
21.1
21.0
20.7
20.9
21.1
Domestic Treatment
15.1
14.6
11.9
11.7
11.4
11.7
12.0
Domestic Effluent
1.4
1.4
2.0
2.1
2.1
2.0
2.0
Industrial Treatment®
5.5
6.1
6.6
6.6
6.7
6.7
6.6
Industrial Effluent®
0.7
0.6
0.5
0.5
0.5
0.5
0.5
n2o
14.8
18.1
21.1
21.8
21.3
21.1
20.8
Domestic Treatment
10.5
13.7
16.2
16.8
16.5
16.4
16.3
Domestic Effluent
3.9
3.9
4.4
4.5
4.3
4.2
4.1
Industrial Treatment
0.3
0.4
0.4
0.4
0.4
0.4
0.4
Industrial Effluentb
0.1 |
0.1
0.1
0.1
0.1
0.1
0.1
Total
37.5 |
40.7
42.3
42.7
41.9
42.0
41.9
a Industrial activity for ChU includes the pulp and paper manufacturing, meat and poultry processing, fruit and vegetable
processing, starch-based ethanol production, petroleum refining, and breweries industries.
b Industrial activity for N20 includes the pulp and paper manufacturing, meat and poultry processing, starch-based ethanol
production, and petroleum refining.
Note: Totals by gas may not sum due to independent rounding.
Table 7-8: CH4 and N20 Emissions from Domestic and Industrial Wastewater
Treatment (kt)
Activity
1990 |
2005 I
2019
2020
2021
2022
2023
CH4
811 I
809
755
748
738
747
755
Domestic Treatment
539
521
426
419
407
418
430
Domestic Effluent
49
49
73
74
74
72
70
Industrial Treatment®
196
216
236
236
238
238
236
Industrial Effluent®
27
22
19
19
19
19
19
n2o
56
68
80
82
80
80
79
Domestic Treatment
40 |
52 I
61
63
62
62
61
Waste 7-25
-------
Activity
1990
2005 |
2019
2020
2021
2022
2023
Domestic Effluent
15
15
17
17
16
16
15
Industrial Treatmentb
1
1
2
1
1
1
1
Industrial Effluentb
+ I +l
+
+
+
+
+
+ Does not exceed 0.5 kt.
a Industrial activity for ChU includes the pulp and paper manufacturing, meat and poultry processing, fruit and vegetable
processing, starch-based ethanol production, petroleum refining, and breweries industries.
b Industrial activity for N20 includes the pulp and paper manufacturing, meat and poultry processing, starch-based ethanol
production, and petroleum refining.
Note: Totals by gas may not sum due to independent rounding.
Methodology and Time-Series Consistency
The methodologies presented in IPCC (2019) form the basis of the CH4 and N20 emission estimates for
both domestic and industrial wastewater treatment and discharge.6 Domestic wastewater treatment
follows the IPCC Tier 2 methodology for significant pathways, and IPCC Tier 1 methodologies for some
pathways in accordance with IPCC methodological decision trees based on available data (i.e.,
centralized treatment (CH4), centralized (anerobic) treatment (N20), and septic systems (N20)).
Domestic wastewater discharge follows IPCC Tier 2 discharge methodology and emission factors in
accordance with IPCC methodological decision trees based on available data. Default factors from
IPCC (2019) or IPCC (2006) are used when there are insufficient U.S.-specific data to develop a U.S.-
specific factor, though IPCC default factors are often based in part on data from or representative of U.S.
wastewater treatment systems. Industrial wastewater treatment follows IPCC Tier 1 and industrial
wastewater treatment discharge follows Tier 1 or Tier 2 methodologies, again in accordance with
methodological decision trees and available data. Similar to domestic wastewater, IPCC default factors
are used when there are insufficient U.S.-specific data to develop a U.S.-specific factor. EPA will
continue to investigate opportunities to implement the Tier 2 discharge methodology for more industries
as data and resource constraints allow (see the Planned Improvements section below). This section
presents a summary of the methodologies used to estimate CH4 and N20 emissions from wastewater
treatment and discharge.
Methodological approaches were applied to the entire time series to ensure consistency in emissions
from 1990 through 2023. See Annex 3.16 for more detailed information on the methodologies (including
detailed formulas and emission factors), data used to calculate CH4 and N20 emissions, and emission
results (including input variables) from wastewater treatment and discharge.
Refer to the Recalculations Discussion section below for details on updates implemented to improve
accuracy, consistency and/or completeness of the time series.
Domestic Wastewater CH4 Emissions
Domestic wastewater CH4 emissions originate from both septic systems and centralized treatment
systems. Within centralized systems, CH4 emissions can arise from aerobic systems that liberate
6 IPCC (2019) updates, supplements, and elaborates the 2006 IPCC Guidelines where gaps or out-of-date science have
been identified. EPA used these methodologies to improve completeness and include sources of greenhouse gas
emissions that have not been estimated prior to the 1990 to 2019 Inventory, such as N20 emissions from industrial
wastewater treatment, and to improve emission estimates for other sources, such as emissions from wastewater
discharge and centralized wastewater treatment.
7-26 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
dissolved CH4 that formed within the collection system or that are designed with periods of anaerobic
activity (e.g., constructed wetlands and facultative lagoons), anaerobic systems (anaerobic lagoons and
anaerobic reactors), and from anaerobic sludge digesters when the captured biogas is not completely
combusted. Emissions also result from the discharge of treated effluent from centralized wastewater
plants to waterbodies where carbon accumulates in sediments (typically slow-moving systems, such as
lakes, reservoirs, and estuaries). The systems with emissions estimates include:
• Septic systems (A);
Centralized treatment (B), including aerobic systems (B1, other than constructed wetlands)
constructed wetlands only (B2), and constructed wetlands used as tertiary treatment (B3);
Centralized anaerobic systems (C);
Anaerobic sludge digesters (D); and
Centralized wastewater treatment effluent (E).
Total domestic wastewater CH4 emissions are the sum of A through E. Methodological equations for
each of these systems are presented in the subsequent subsections.
Methodology for Septic Systems (CH4) Emissions (A):
Methane emissions from septic systems were estimated by multiplying the U.S. population by the
percent of wastewater treated in septic systems and a country-specific emission factor and then
converting the result to kt/year. The method was selected in accordance with IPCC methodological
decision trees based on available data for septic systems.
U.S. population data were taken from historic U.S. Census Bureau national population totals data and
include the populations of the United States and Puerto Rico (U.S. Census Bureau 2002; U.S. Census
Bureau 2011; U.S. Census Bureau 2021 and 2023b; Instituto de Estadi'sticas de Puerto Rico 2021).
Population data for American Samoa, Guam, Northern Mariana Islands, and the U.S. Virgin Islands were
taken from the U.S. Census Bureau International Database (U.S. Census Bureau 2024). The fraction of
the U.S. population using septic systems or centralized treatment systems is based on data from the
American HousingSun/eys (U.S. Census Bureau 2023a). See Annex 3.16 for more details on methods
and inputs, including data and emission factor values applied in calculations.
Methodology for Centrally Treated Aerobic and Anaerobic Systems (CH4) Emissions (B, C):
Methane emissions from POTWs depend on the total organics in wastewater (TOW). The TOW was
determined using BOD generation rates per capita weighted average both with and without kitchen
scraps as well as an estimated percent of housing units that utilize kitchen garbage disposals.
Households with garbage disposals (with kitchen scraps or ground up food scraps) typically have
wastewater with higher BOD than households without garbage disposals due to increased organic
matter contributions (ERG 2018a).
Methane emissions from POTWs were estimated by multiplying the total organics in centrally treated
wastewater (total BOD5) produced per capita in the United States by the percent of wastewater treated
centrally, or percent collected, the correction factor for additional industrial BOD discharged to the
sewer system, the relative percentage of wastewater treated by aerobic systems (other than
Waste 7-27
-------
constructed wetlands), constructed wetlands only, and anaerobic systems, and the emission factor7 for
aerobic systems, constructed wetlands only, and anaerobic systems.
In the United States, the removal of sludge8 from wastewater reduces the BOD of the wastewater that
undergoes aerobic treatment. The amount of this reduction (S) is estimated using the default IPCC
(2019) methodology and multiplying the amount of sludge removed from wastewater treatment in the
United States by the default factors in IPCC (2019) to estimate the amount of BOD removed based on
whether the treatment system has primary treatment with no anaerobic sludge digestion (assumed to
be zero by expert judgment), primary treatment with anaerobic sludge digestion, or secondary treatment
without primary treatment. The organic component removed from anaerobic wastewater treatment and
the amount of CH4 recovered or flared from both aerobic and anaerobic wastewater treatment were set
equal to the IPCC default of zero. The IPCC (2019) default emission factors used account for the
dissolved CH4 entering the centralized treatment systems.
Constructed wetlands provide aerobic treatment but also exhibit partially anaerobic conditions;
however, they are referred to in this chapter as aerobic systems. Emissions from all constructed wetland
systems for wastewater treatment were included in the estimates of emissions from centralized
wastewater treatment plant processes and effluent from these plants. Methane emissions equations
from constructed wetlands used as sole treatment were previously described. Methane emissions from
constructed wetlands used as tertiary treatment were estimated by multiplying the flow from treatment
to constructed wetlands, wastewater BOD concentration entering tertiary treatment, constructed
wetlands emission factor, and then converting to kt/year.
For constructed wetlands, an IPCC default emission factor for surface flow wetlands was used. This is
the most conservative factor for constructed wetlands and was recommended by IPCC (2014) when the
type of constructed wetland is not known. A median BOD5 concentration of 9.1 mg/L was used for
wastewater entering constructed wetlands used as tertiary treatment based on U.S. secondary
treatment standards for POTWs. This median value is based on plants generally utilizing simple settling
and biological treatment (EPA 2013). Constructed wetlands do not have secondary sludge removal.
Data sources and methodologies for centrally treated anaerobic systems are similar to those described
for aerobic systems, other than constructed wetlands. See discussion above and see Annex 3.16 for
more details on methods and inputs, including data and emission factor values applied in calculations.
Methodology for Anaerobic Sludge Digester (CH4) Emissions (D):
Total CH4 emissions from anaerobic sludge digesters were estimated by multiplying the wastewater
influent flow to POTWs with anaerobic sludge digesters, the cubic feet of digester gas generated per
person per day divided by the flow to POTWs, the fraction of CH4 in biogas, the density of CH4, one
minus the destruction efficiency from burning the biogas in an energy/thermal device and then
converting the results to kt/year. See Annex 3.16 for more details on methods and inputs, including data
and emission factorvalues applied in calculations.
7 Emission factors are calculated by multiplying the maximum ChU-producing capacity of domestic wastewater (Bo, 0.6 kg
ChU/kg BOD) and the appropriate methane correction factors (MCF) for aerobic (0.03) and anaerobic (0.8) systems (IPCC
2019, Table 6.3) and constructed wetlands (0.4) (IPCC 2014, Table 6.4).
8 Throughout this document, the term "sludge" refers to the solids separated during the treatment of municipal
wastewater. The definition includes domestic septage. "Biosolids" refers to treated sewage sludge that meets the EPA
pollutant and pathogen requirements for land application and surface disposal.
7-28 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Methodology for Discharge of Centralized Treatment Effluent (CH4) Emissions (E):
Methane emissions from the discharge of wastewater treatment effluent were estimated by multiplying
the total BOD of the discharged wastewater effluent by an emission factor associated with the location
of the discharge. The BOD in treated effluent was determined by multiplying the total organics in
centrally treated wastewater by the percent of wastewater treated in primary, secondary, and tertiary
treatment, and the fraction of organics remaining after primary treatment (one minus the fraction of
organics removed from primary treatment, secondary treatment, and tertiary treatment). The percent of
wastewater effluent discharged to slow-moving versus other water bodies (e.g., rivers) was determined
via the methodology presented in ERG (2021a). Most wastewater effluent is discharged to other water
bodies (62 percent in 2023). See Annex 3.16 for more details on methods and inputs, including data and
emission factorvalues applied in calculations.
Industrial Wastewater CH4 Emissions
Industrial wastewater CH4 emissions originate from on-site treatment systems, typically comprised of
biological treatment operations. The collection systems at an industrial plant are not as extensive as
domestic wastewater sewer systems; therefore, it is not expected that dissolved CH4 will form during
collection. However, some treatment systems are designed to have anaerobic activity (e.g., anaerobic
reactors or lagoons), or may periodically have anaerobic conditions form (facultative lagoons or large
stabilization basins). Emissions will also result from discharge of treated effluent to waterbodies where
carbon accumulates in sediments (typically slow-moving systems, such as lakes, reservoirs, and
estuaries).
Industry categories that are likely to produce significant CH4 emissions from wastewater treatment were
identified and included in this Inventory. The main criteria used to identify U.S. industries likely to
generate CH4 from wastewater treatment are whether an industry generates high volumes of
wastewater, whether there is a high organic wastewater load, and whether the wastewater is treated
using methods that result in CH4 emissions. The top six industries that meet these criteria are pulp and
paper manufacturing; meat and poultry processing; vegetables, fruits, and juices processing; starch-
based ethanol production; petroleum refining; and breweries. Further discussion of wastewater
treatment for each industry is included below. Total industrial wastewater CH4 emissions are the sum of
A through L.
Table 7-9: Industrial Wastewater Treatment Systems with (CH4) Emissions
Industry
Treatment
Aerobic
Anaerobic
Other
Effluent
Pulp and paper manufacturing
A1
A2
G
Red meat processing
B1
H
Poultry processing
B2
Vegetables, fruits, and juices processing
C
I
Starch-based Ethanol Production, Dry Milling
D1, D3
D2, D4
J
Starch-based Ethanol Production, Wet Milling
D5
D6
Petroleum Refining
E
K
Non-craft Breweries
F1, F2
L
Craft Breweries F3, F4
Waste 7-29
-------
Methodology for Industrial Wastewater Treatment Systems (CH4) Emissions (A through F):
The general IPCC equation (Equation 6.4, IPCC 2019) is used to estimate CH4 emissions from each type
of treatment system used for each industrial category, the TOWi, minus the organic component removed
from aerobic wastewater treatment (Si), is multiplied by the system-specific emissions factor (kg CH4/kg
COD) all minus any methane recovered for the industrial sector. The TOW for each industrial category,
were estimated by multiplying the total industrial product (i.e., production) by the wastewater outflow by
the amount of COD in a unit of wastewater. For some industries, U.S.-specific data on organics loading
is reported as BOD rather than COD. In those cases, an industry-specific COD:BOD ratio is used to
convert the organics loading to COD.
The amount of organics treated in each type of wastewater treatment system was determined using the
percent of wastewater in the industry that is treated on site and whether the treatment system is
anaerobic, aerobic or partially anaerobic. For CH4 emissions, wastewater treated in anaerobic lagoons
or reactors was categorized as "anaerobic", wastewater treated in aerated stabilization basins or
facultative lagoons were classified as "ASB" (meaning there may be pockets of anaerobic activity), and
wastewater treated in aerobic systems such as activated sludge systems were classified as
"aerobic/other." See Annex 3.16 for details on the industrial wastewater treatment systems in place.
The amount of organic component removed from aerobic wastewater treatment as a result of sludge
removal (Saerobic) was either estimated as an industry-specific percent removal, if available, or as an
estimate of sludge produced by the treatment system and IPCC default factors for the amount of
organic component removed (Krem). See Annex 3.16 for more details on methods and inputs, including
data and emission factor values applied in calculations, for each industry.
Methodology for Discharge of Industrial Wastewater Treatment Effluent (CH4) Emissions (G through
L):
Methane emissions from discharge of industrial wastewater treatment effluent are estimated via a Tier 1
method for all industries except for pulp, paper, and paperboard in accordance with IPCC
methodological decision trees in based on available data for treatment and discharge. Emissions from
discharge of pulp, paper, and paperboard treatment effluent is estimated via a Tier 2 method and is
described in the industry-specific data section. Tier 1 emissions from effluent are estimated by
multiplying the total organic content of the discharged wastewater effluent by an emission factor
associated with the discharge.
The COD or BOD in industrial treated effluent (TOWeffluent.ind) was determined by multiplying the total
organics in the industry's untreated wastewater that is treated on site by an industry-specific percent
removal where available or a more general percent removal based on biological treatment for other
industries. See Annex 3.16 for more details on methods and inputs, including data and emission factor
values applied in calculations, for each industry.
Domestic Wastewater N2O Emissions
Domestic wastewater N20 emissions originate from both septic systems and POTWs. Within these
centralized systems, N20 emissions can result from aerobic systems, including systems like
constructed wetlands. Emissions will also result from discharge of centrally treated wastewater to
waterbodies with nutrient-impacted/eutrophic conditions. The systems with emission estimates are:
• Septic systems (A);
7-30 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Centralized treatment (B), including aerobic systems (other than constructed wetlands) (B1),
constructed wetlands only (B2), and constructed wetlands used as tertiary treatment (B3);
Centralized anaerobic systems (C); and
Centralized wastewater treatment effluent (D).
The total domestic wastewater N20 emissions are the sum of A through D. Methodological equations for
each of these systems are presented in the subsequent subsections while detailed information on
methods and inputs, including data and emission factor values applied in calculations is included in
Annex 3.16 of this report; total domestic N20 emissions are summarized as follows:
Methodology for Septic Systems (N20) Emissions (A):
Nitrous oxide emissions from domestic treatment depend on the nitrogen present, in this case, in the
form of protein. Per capita protein consumption (kg protein/person/year) was determined by multiplying
per capita annual food availability data and its protein content. Those data are then adjusted using a
factor to account for the fraction of protein actually consumed.
Nitrous oxide emissions from septic systems were estimated by multiplying the U.S. population by the
percent of wastewater treated in septic systems (about 19 percent in 2023; U.S. Census Bureau 2023a),
consumed protein per capita (kg protein/person/year), the fraction of N in protein, the correction factor
for additional nitrogen from household products, the factor for industrial and commercial co-discharged
protein into septic systems, the factor for non-consumed protein added to wastewater and an emission
factor and then converting the result to kt/year. The method selected is in accordance with IPCC
methodological decision trees and available data. All factors were obtained from IPCC (2019).
Methodology for Centrally Treated Aerobic and Anaerobic Systems (N20) Emissions (B, C):
Nitrous oxide emissions from POTWs depend on the total nitrogen entering centralized wastewater
treatment. The total nitrogen entering centralized wastewater treatment was estimated by multiplying
the U.S. population by the percent of wastewater collected for centralized treatment (about 81 percent
in 2023), the consumed protein per capita, the fraction of N in protein, the correction factor for
additional N from household products, the factor for industrial and commercial co-discharged protein
into wastewater treatment, and the factor for non-consumed protein added to wastewater.
Nitrous oxide emissions from POTWs were estimated by multiplying the total nitrogen entering
centralized wastewater treatment, the relative percentage of wastewater treated by aerobic systems
(other than constructed wetlands) and anaerobic systems, aerobic systems with constructed wetlands
as the sole treatment, the respective emission factors for aerobic systems and anaerobic systems, and
the conversion from N2 to N20.
Nitrous oxide emissions from constructed wetlands used as sole treatment include similar data and
processes as aerobic systems other than constructed wetlands. See description above. Nitrous oxide
emissions from constructed wetlands used as tertiary treatment were estimated by multiplying the flow
to constructed wetlands used as tertiary treatment, wastewater N concentration entering tertiary
treatment, constructed wetlands emission factor, and converting to kt/year.
Data sources and methodologies for anaerobic systems are similar to those described for aerobic
systems, other than constructed wetlands. See discussion above and detailed activity data within Annex
3.16 of this report.
Waste 7-31
-------
Methodology for Discharge of Centralized Treatment Effluent (N20) Emissions (D):
Nitrous oxide emissions from the discharge of wastewater treatment effluent were estimated by
multiplying the total nitrogen in centrally treated wastewater effluent by the percent of wastewater
treated in primary, secondary, and tertiary treatment and the fraction of nitrogen remaining after
primary, secondary, or tertiary treatment and then multiplying by the percent of wastewater volume
routed to waterbodies with nutrient-impaired/eutrophic conditions and all other waterbodies (ERG
2021 a) and emission factors for discharge to impaired waterbodies and other waterbodies from IPCC
(2019).
Industrial Wastewater N2O Emissions
Nitrous oxide emission estimates from industrial wastewater and discharge are estimated according to
the Tier 1 methodologies described in the 2019 Refinement. U.S. industry categories that are likely to
produce significant N20 emissions from wastewater treatment were identified based on whether they
generate high volumes of wastewater, whether there is a high nitrogen wastewater load, and whether the
wastewater is treated using methods that result in N20 emissions. The top four industries that meet
these criteria and were added to the Inventory are meat and poultry processing; petroleum refining; pulp
and paper manufacturing; and breweries (ERG 2021a). See Annex 3.16 for more details on the
wastewater treatment and discharge emissions and production data for these sectors for 2023. Total
industrial wastewater N20 emissions are the sum of Athrough H.
Table 7-10: Industrial Wastewater Treatment Systems with N20 Emissions
Industry Treatment Effluent
Pulp and paper manufacturing A E
Meat and poultry processing B F
Petroleum Refining C G
Non-craft and craft breweries D H
Methodology for Industrial Wastewater Treatment Systems (N20) Emissions (Athrough D):
To estimate N20 emissions, the total nitrogen entering aerobic wastewater treatment for each industry
must be calculated. Then, the emission factor provided by the 2019 Refinement \s applied to the portion
of wastewater that undergoes aerobic treatment.
The total nitrogen that enters each industry's wastewater treatment system is a product of the total
amount of industrial product produced, the wastewater generated per unit of product, and the nitrogen
expected to be present in each cubic meter of wastewater (IPCC equation 6.13).
For the four industries of interest, the total production and the total volume of wastewater generated has
already been calculated for CH4 emissions. For these N20 emission estimates, the total nitrogen in the
untreated wastewater was determined by multiplying the annual industry production by the average
wastewater outflow and the nitrogen loading in the outflow. See Annex 3.16 for details on activity data.
Nitrous oxide emissions from industry wastewater treatment are calculated by applying an emission
factor to the percent of wastewater (and therefore nitrogen) that undergoes aerobic treatment (IPCC
Equation 6.11). For each industry, the degree of utilization (Ti,j) is described within Annex 3.16.
7-32 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Methodology for Industrial Wastewater Treatment Effluent (N20) Emissions (E through H):
Nitrous oxide emissions from industrial wastewater treatment effluent are estimated by multiplying the
total nitrogen content of the discharged wastewater effluent by an emission factor associated with the
location of the discharge. Where wastewater is discharged to aquatic environments with nutrient-
impacted/eutrophic conditions (i.e., water bodies which are rich in nutrients and very productive in
terms of aquatic animal and plant life), or environments where carbon accumulates in sediments such
as lakes, reservoirs, and estuaries, the additional organic matter in the discharged wastewater is
expected to increase emissions (IPCC 2019).
The total nitrogen in wastewater effluent was determined through use of a nutrient estimation tool
developed by EPA's Office of Water (EPA 2019a). The Nutrient Tool uses known nutrient discharge data
within defined industrial sectors or subsectors, as reported on Discharge Monitoring Reports, to
estimate nutrient discharges for facilities within that sector or subsector that do not have reported
nutrient discharges but are likely to discharge nutrients. The estimation considers, within each sector or
subsector, elements such as the median nutrient concentration and flow, as well as the percent of
facilities within the sector or subsector that have reported discharges. Data from 2018 are available for
the pulp, paper, and paperboard, meat and poultry processing, and petroleum refining industries. To
complete the time series, an industry-specific percent removal of nitrogen was calculated using the
total nitrogen in untreated wastewater. See EPA (2019b) for a detailed discussion on the methodology
and data sources used within EPA's Nutrient Tool.
Because data for breweries was not available, the removal of nitrogen was assumed to be equivalent to
secondary treatment, or 40 percent (IPCC 2019). The Tier 1 emission factor (0.005 kg N20/kg N) from
IPCC (2019) was used.
Uncertainty
The overall uncertainty associated with both the 2023 CH4 and N20 emission estimates from
wastewater treatment and discharge was calculated using the 2006 IPCC Guidelines Approach 2
methodology (IPCC 2006). Uncertainty associated with the parameters used to estimate CH4 emissions
include that of numerous input variables used to model emissions from domestic wastewater and
emissions from wastewater from pulp and paper manufacturing, meat and poultry processing, fruits
and vegetable processing, ethanol production, petroleum refining, and breweries. Similarly, uncertainty
associated with the parameters used to estimate N20 emissions include that of numerous input
variables used to model emissions from domestic wastewater and emissions from wastewater from
pulp and paper manufacturing, meat and poultry processing, petroleum refining, and breweries. For
example, uncertainty associated with centrally treated constructed wetlands parameters including U.S.
population served by constructed wetlands (±5%), and emission (±79) and conversion factors are from
IPCC (2014), whereas uncertainty associated with POTW flow to constructed wetlands and influent BOD
and nitrogen concentrations were based on expert judgment (±25%, ERG 2021 b). The specified
probability density functions (PDFs) are assumed to be normal for most activity data and emission
factors, and due to lack of data, are based on expert judgement (ERG 2021c).
The results of this Approach 2 quantitative uncertainty analysis are summarized in Table 7-11. For 2023,
methane emissions from wastewater treatment were estimated to be between 15.3 and 27.9 MMT C02
Eq. at the 95 percent confidence level (or in 19 out of 20 Monte Carlo stochastic simulations). This
indicates a range of approximately 28 percent below to 32 percent above the 2023 emissions estimate
Waste 7-33
-------
of 21.1 MMT C02 Eq. Nitrous oxide emissions from wastewater treatment were estimated to be between
13.1 and 60.9 MMT C02 Eq., which indicates a range of approximately 37 percent below to 193 percent
above the 2023 emissions estimate of 20.8 MMT C02 Eq.
Table 7-11: Approach 2 Quantitative Uncertainty Estimates for 2023 Emissions from
Wastewater Treatment (MMT C02 Eq. and Percent)
Uncertainty Range Relative to Emission Estimate"
2023 Emission
(MMT CO2
Eq.)
(%)
Estimate
Lower
Upper
Lower
Upper
Source
Gas
(MMT CO2 Eq.)
Bound
Bound
Bound
Bound
Wastewater Treatment
CH4
21.1
15.3
27.9
-28%
+32%
Domestic
cm
14.0
9.4
19.7
-33%
+41 %
Industrial
CH4
7.1
4.2
11.3
-42%
+58%
Wastewater Treatment
n2o
20.8
13.1
60.9
-37%
+193%
Domestic
n2o
20.3
12.3
60.0
-40%
+ 195%
Industrial
n2o
0.5
0.5
1.4
-2.2%
+ 199%
a Range of emission estimates predicted by Monte Carlo stochastic simulation for a 95 percent confidence interval.
QA/QC and Verification
General QA/QC procedures were applied to activity data, documentation, and emission calculations
consistent with the U.S. Inventory QA/QC Plan, which is in accordance with Vol. 1 Chapter 6 of the 2006
IPCC Guidelines (see Annex 8 for more details). In addition to general or Tier 1 analysis,
calculation-related QC (category-specific, Tier 2) was performed for a portion of the domestic
wastewater treatment discharges methodology. The calculation-related QC included continuing to
assess the publicly availability of similar activity data is still based on the best available data. Confirmed
that gap filling techniques are consistent with other sources used within the Inventory.
All transcription errors identified were corrected and documented. The QA/QC analysis did not reveal
any systemic inaccuracies or incorrect input values.
Recalculations Discussion
Several estimates were recalculated and implemented in developing the current Inventory.
Population data were updated using the same and latest data sources as the state-level
emissions inventory to create consistency across Inventory estimates. These changes affected
the years 2020-2022.
Updated percent of wastewater collected affected 2022 (U.S. Census Bureau 2023a).
Protein data were updated to reflect available protein values available for 2010 through 2022
(FAO 2024b).
Pulp, paper, and paperboard production data were updated to reflect revised values for 2021
and 2022 (FAO 2024a).
Updated red meat production values for 2022, were updated based on revised data (USDA
2024a).
7-34 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Fruits and vegetables production values were updated for the time series (ERG 2022; USDA
2024c).
Ethanol production values for 2021 and 2022 were based on revised data (RFA 2024a; RFA
2024b).
Updated values for non-craft brewery wastewater generation were included for the years 2017
and 2022, affecting the values for 2016, 2018, 2019, and 2021 (BIER 2024).
Compared to the previous Inventory the cumulative effect of all these recalculations had a minor impact
on the overall wastewater treatment emission estimates:
Domestic wastewater treatment and discharge CH4 emissions increased on average 0.1 percent
over the timeseries, with 1990 through 2019 not changing and the largest increase of 4.6 percent
(0.6MMTCO2 Eq.) in 2022.
Domestic wastewater treatment and discharge N20 emissions decreased an average 0.8
percent over the timeseries, with 1990 through 2009 not changing and the largest decrease of
5.4 percent (1.2 MMT C02 Eq.) in 2022.
Industrial wastewater treatment and discharge CH4 emissions decreased on average 0.02
percent over the timeseries, with the smallest decrease of 0.001 percent (0.0 MMT C02 Eq.) in
2020 and largest decrease of 0.4 percent (0.03 MMT C02 Eq.) in 2022.
Industrial wastewater treatment and discharge N20 emissions decreased an average 0.01
percent over the timeseries, with the smallest decrease of 0.0 percent (0.0 MMT C02 Eq.) in
1990 to the largest decrease of 0.3 percent (0.002 MMT C02 Eq.) in 2022.
Over the time series, the total emissions on average decreased by 0.3 percent from the previous
Inventory. The changes ranged from the smallest decrease, 0.001 percent (0.0002 MMT C02 Eq.), in
1992, to the largest decrease, 1.8 percent (0.8 MMT C02 Eq.), in 2021.
Planned Improvements
EPA notes the following improvements will continue to be investigated as time and resources allow, but
there are no immediate plans to implement them until data are available or identified:
Domestic-specific improvements:
Continue to investigate anaerobic sludge digester and biogas data compiled by the Water
Environment Federation (WEF) in collaboration with other entities as a potential source of
updated activity data for biogas production. Due to lack of these data, the United States
continues to use another method for estimating biogas produced. This method uses the
standard 100 gallons/capita/day wastewater generation factor for the United States (Ten-State
Standards), which EPA believes is reasonable to estimate national emissions. However, based
on stakeholder input, some regions of the United States use markedly less water due to water
conservation efforts so EPA plans to investigate updated sources for this method as well.
Investigate additional sources for estimating wastewater volume discharged and discharge
location for both domestic sources with the goal being to provide additional data points along
the time series, while the goal for industrial wastewater would be to update the Tier 1 discharge
methodology to a Tier 2 methodology.
Investigate additional sources for domestic wastewater treatment type in place data.
Waste 7-35
-------
Investigate the literature shared in expert review comments providing independent methods and
estimates:
Song et al. (2023) concluded that CH4 emissions for domestic wastewater are twice as large
as those estimated via an IPCC (2019) Tier 2 method. EPA plans to compare the
methodology laid out in the paper against current inventory methodology(ies). As part of
this, EPA will evaluate if an update to the U.S. AD leakage rate is warranted.
Song et al. (2024) concluded that N20 emissions for domestic wastewater are 75 percent of
those estimated via an IPCC (2019) Tier 2 method. EPA plans to compare the methodology
laid out in the paper against current inventory methodology(ies).
Additional papers may provide updated emissions estimates or measurements that EPA
plans to review for comparison purposes: Moore et al. (2023), Chong et al. (2024), Sieranen
et al. (2024), and Yin et al. (2024).
IPCC (2019) updated the methodology for the organic component removed as sludge from
septic systems to account for both the fraction of the population using septic systems that are
operating in compliance of system recommendations, as well as the fraction of organics in
wastewater removed in sludge. While the country-specific emission factor for septic systems
are based on measured data, EPA plans to compare these methods.
Continue to review whether sufficient data exist to develop U.S.-specific CH4 or N20 emission
factors for centralized domestic wastewater treatment systems, including whether emissions
should be differentiated for systems that incorporate biological nutrient removal operations.
Evaluate the use of POTW BOD effluent discharge data from ICIS-NPDES.9 Currently only half of
POTWs report organics as BOD5so EPA would need to determine a hierarchy of parameters to
appropriately sum all loads. Using these data could potentially improve the current methane
emission estimates from domestic discharge, or at least provide a comparison to the current
method for QA/QC.
Evaluate the use of POTW N effluent discharge data from ICIS-NPDES. Currently only about 80
percent of POTWs report a form of N so EPA would need to determine an appropriate method to
scale to the total POTW population. EPA is aware of and will investigate two methods: one is
specific for industrial sources (so EPA would need to determine if this method is appropriate for
domestic sources), the second was recommended by an expert commenter and includes using
a TN:BOD (35:200) ratio from Metcalf & Eddy (2014). Using these data could potentially improve
the current nitrous oxide emissions estimates from domestic discharge, or at least provide a
comparison to the current method for QA/QC.
Investigate additional data sources for improving the uncertainty of the estimate of N entering
municipal treatment systems.
Investigate methodologies for calculating renewable natural gas (RNG) generated from
wastewater treatment digester biogas and evaluate data sources of RNG generation rates and
assess calculations for potential inclusion into the Inventory.
9 ICIS-NPDES refers to EPA's Integrated Compliance Information System - National Pollutant Discharge Elimination
System.
7-36 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Industrial specific-improvements:
Continue to evaluate literature provided by expert review commenters for potential inclusion as
updates:
Wastewater characteristics for pulp and paper manufacturing, including reviewing an
update to the current emission factor used, as well as review Esmaeeli et al. (2023) for
applicable information.
Potential updates to brewery wastewater characteristics from Babanova et al. (2022).
Potential updates to slaughterhouse wastewater characteristics or treatment from Wang et
al. (2021).
Investigate additional sources for estimating wastewater volume discharged and discharge
location for industrial sources.
7.3 Composting (Source Category 5B1)
Composting of organic waste, such as food waste, garden (yard) and park waste, and wastewater
treatment sludge and/or biosolids, is common in the United States. Composting reduces the amount of
methane-generating waste entering landfills, destroys pathogens in the waste, sequesters carbon, and
provides a source of organic matter. Composting can also generate a saleable product and reduce the
need for chemical fertilizers when the end product is used as a fertilizer or soil amendment. This source
category assumes all composting facilities are commercial, large-scale windrow composting facilities
with yard trimmings as the main waste stream composted, which aligns with findings from full-scale
compost infrastructure survey data published by BioCycle (2017, 2023). Of 200 major food waste
composting facilities in the United States, 75 (38 percent) use the windrow method, 45 (23 percent) use
the aerated static pile method, and the remainder use other methods. The BioCycle 2023 survey
received responses from facilities using composting methods (e.g., aerated static piles, in-vessel
composting) that are operational in the United States. However national estimates of the material
processed by these facilities are not readily available; therefore, emissions estimates by composting
method are not included in this source category. Residential backyard and community composting is
also not included in this source category.
Composting naturally converts a large fraction of the degradable organic carbon in the waste material
into carbon dioxide (C02) through aerobic processes without anthropogenic influence. With
anthropogenic influences (e.g., at commercial or large on-site composting operations), anaerobic
conditions can be created in sections of the compost pile when there is excessive moisture or
inadequate aeration (or mixing) of the compost pile, resulting in the formation of methane (CH4).
Methane in aerobic sections of a windrow pile is generally oxidized by microorganisms, which convert
the CH4 to C02 emissions. Even though C02 emissions are generated, they are not included in net
greenhouse gas emissions for composting. Consistent with the 2006IPCC Guidelines, net C02flux from
carbon stock changes in waste material are estimated and reported under the LULUCF sector. The
estimated CH4 released into the atmosphere ranges from less than 1 percent to a few percent of the
initial carbon content in the material (IPCC 2006). Depending on how well the compost pile is managed,
nitrous oxide (N20) emissions can also be produced. The formation of N20 depends on the initial
nitrogen content of the material and is mostly due to nitrogen oxide (NOx) denitrification during the
Waste 7-37
-------
thermophilic and secondary mesophilic stages of composting (Cornell 2007). Emissions vary and range
from less than 0.5 percent to 5 percent of the initial nitrogen content of the material (IPCC 2006). Animal
manures are typically expected to generate more N20 than, for example, yard waste, however data are
limited.
From 1990 to 2023, the amount of waste composted in the United States increased from 3,810 kt to
23,155 kt (see Table 7-14). There was some fluctuation in the amount of waste composted between
2006 to 2009 where a peak of 20,063 kt composted was observed in 2008, which decreased to 18,838 kt
composted the followingyear, presumably driven by the economic crisis of 2009 (data not shown). Since
2009, the amount of waste composted has gradually increased commensurate with population growth.
When comparing 2010 to 2023, a 26 percent increase in waste composted is observed. Emissions of
CH4 and N20 from composting from 2010 to 2023 have increased by the same percentage.
In 2023, CH4 emissions from composting (see Table 7-12 and Table 7-13) were 2.6 MMT C02 Eq. (93 kt),
and N20 emissions from composting were 1.8 MMT C02 Eq. (7 kt), representing consistent emissions
trends over the past several years. Composted material primarily includes yard trimmings (grass, leaves,
and tree and brush trimmings) and food scraps from the residential and commercial sectors (such as
grocery stores; restaurants; and schools and businesses). The composted waste quantities reported
here do not include small-scale backyard composting, community composting, or agricultural
composting mainly due to the lack of consistent and comprehensive national data. Additionally, it is
assumed that backyard composting tends to be a more naturally managed process with less chance of
generating anaerobic conditions and CH4 and N20 emissions. Agricultural composting is accounted for
in Chapter 5, Section 5.4 (Agricultural Soil Management) of this Inventory, as most agricultural
composting operations are assumed to land-apply the resultant compost to soils.
The growth in composting since the 1990s is largely due to growing legislation by state and local
governments discouraging the disposal of yard trimmings and food waste in landfills and increased
collection of yard trimming. Most bans or diversion laws on the disposal of yard trimmings were initiated
in the early 1990s by state or local governments (U.S. Composting Council 2010). California, for
example, enacted a waste diversion law for organics including yard trimmings and food scraps in 1999
(AB939) that required jurisdictions to divert 50 percent of the waste stream by 2000, or be subjected to
fines. Currently, 20 states representing up to 42 percent of the nation's population have enacted
legislation banningyard waste from landfill disposal (U.S. Composting Council 2022). Additional
initiatives at the metro and municipal level also exist across the United States. Roughly 4,713
composting facilities exist in the United States with most (57.2 percent) composting yard trimmings only
(BioCycle 2017).
In the last decade, bans and diversions for food waste have also become more common. As of 2023, ten
states (California, Connecticut, Maryland, Massachusetts, New Hampshire [new as of 2023], New
Jersey, New York, Oregon, Rhode Island, and Vermont) had implemented organic waste bans or
mandatory recycling laws to help reduce organic waste entering landfills, with most having taken effect
after 2013 (U.S. Composting Council 2024). Five local governments (Austin, TX; Boulder, CO; Hennepin
County, MN; New York City, NY; Seattle, WA) have implemented organics bans; and two cities have
implemented food scrap collection requirements (Portland, OR and San Francisco, CA). In most cases,
organic waste reduction in landfills is accomplished by following recycling guidelines, donating excess
food for human consumption, or by sending waste to organics processing facilities (Harvard Law School
and CET 2019). An example of an organic waste ban as implemented by California is the California
Mandatory Recycling Law (AB1826), which requires companies to comply with organic waste recycling
7-38 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
procedures if they produce a certain amount of organic waste and took effect on January 1, 2015
(Harvard Law School and CET 2019). In 2017, BioCycle released a report in which 27 of 43 states that
responded to their organics recycling survey noted that food waste (collected residential, commercial,
institutional, and industrial food waste) was recycled via anaerobic digestion and/or composting. These
27 states reported an estimated total of 1.8 million tons of food waste diverted from landfills in 2016
(BioCycle 2018b). In 2023, BioCycle released a report on the food waste composting infrastructure in
the United States with estimated range of 5.2 to 8.8 million tons of food waste composted (35 percent of
the total estimated waste composted) in 2022 (BioCycle 2023). A growing number of initiatives to
encourage households and businesses to compost or beneficially reuse food waste also exist.
Table 7-12: CH4 and N20 Emissions from Composting (MMT C02 Eq.)
Activity
1990
2005
2019
2020
2021
2022
2023
ch4
0.4
2.1 I
2.5
2.6
2.6
2.6
2.6
n2o
0.3
1.5
1.8
1.8
1.8
1.8
1.8
Total
0.7
3.6
4.3
4.4
4.4
4.4
4.4
Note: Totals may not sum due to independent rounding.
Table 7-13: CH4 and N2Q Emissions from Composting (kt)
Activity
1990
2005
2019
2020
2021
2022
2023
ch4
15
75
91
92
92
92
93
n2o
1
I 6
7
7
7
7
7
Methodology
Methane and N20 emissions from composting depend on factors such as the type of waste composted,
the amount and type of supporting material (such as wood chips and peat) used, temperature, moisture
content (e.g., wet and fluid versus dry and crumbly), and aeration during the composting process.
The emissions shown in Table 7-12 and Table 7-13 were estimated using the IPCC default (Tier 1)
methodology (IPCC 2006) in accordance with IPCC methodological decision trees and available data.
Using this method, emissions are the product of an emission factor and the mass of organic waste
composted (note: no CH4 recovery is expected to occur at composting operations in the emission
estimates presented):
Equation 7-4: Greenhouse Gas Emission Calculation for Composting
Et=M X EFi
where,
E, = CH4 or N20 emissions from composting, kt CH4or N20
M = mass of organic waste composted in kt
EF, = emission factor for composting, 41 CH4/kt of waste treated (wet basis) and
0.3 t N20/kt of waste treated (wet basis) (IPCC 2006)
i = designates either CH4or N20
Per IPCC Tier 1 methodology defaults, the emission factors for CH4 and N20 assume a moisture content
of 60 percent in the wet waste (IPCC 2006). While the moisture content of composting feedstock can
Waste 7-39
-------
vary significantly by type, composting as a process ideally proceeds between 40 to 65 percent moisture
(University of Maine 2016; Cornell 1996).
Estimates of the quantity of waste composted (M, wet weight as generated) are presented in Table 7-14
for select years. Estimates of the quantity of waste composted for 1990 and 2005 were taken from EPA's
Advancing Sustainable Materials Management: Facts and Figures 2015 (EPA 2018); estimates of the
quantities composted for 2017 to 2018 were taken from EPA's Advancing Sustainable Materials
Management: 2018 Tables and Figures (EPA 2020); the estimate of the quantity composted for 2019 to
2023 were extrapolated using the 2018 quantity composted and a ratio of the U.S. population growth for
each year between 2018 and 2023 (U.S. Census Bureau 2021; U.S. Census Bureau 2024). Estimates of
waste composted by commercial facilities in Puerto Rico were provided for select years by EPA Region 2
(Kijanka 2020). This data includes amount of waste composted at three facilities in Puerto Rico for 2017,
2018, and 2019, ranging from approximately 1,200 kt to a high of 15,000 kt. The average waste
composted for these years was used as the annual amount composted for the respective facility for
years the facility was operational. The annual quantity of composted waste in Puerto Rico was
forecasted for 2020 through 2023 using available data from prior years, assumed metro area population
data near where each facility is located, and the Microsoft Excel FORECAST function to obtain annual
composting estimates.
Table 7-14: U.S. Waste Composted (kt)
Activity
1990 1
2005 1
2019
2020
2021
2022
2023
Waste Composted
3,810 |
18,655 |
22,698
22,919
22,956
23,041
23,155
Uncertainty
The major uncertainty drivers are the assumption that all composting emissions come from commercial
windrow facilities and the use of default emission factors (IPCC 2006) which is tied to a homogenous
mixture of waste processed across the country (largely yard trimmings). Data presented by BioCycle
(BioCycle 2017, 2023) confirm most composting operations use the windrow method and yard
trimmings are the largest share of material composted across the country, but there are other
composting methods used and waste characteristics will vary at a facility level. Additionally, there are
composting operations in Puerto Rico and U.S. territories that are not explicitly included in the national
quantity of material composted as reported in the EPA Sustainable Materials Management Reports
because the methodological scope does not include Puerto Rico and U.S. territories. EPA took steps to
include emissions from Puerto Rico and U.S. Territories beginning in the 1990 to 2020 Inventory and will
continue to seek out additional data in future Inventories.
The estimated uncertainty from the 2006 IPCC Guidelines is ±58 percent for the Tier 1 methodology and
considers the individual emission factors applied to the default emission factors and activity data.
Emissions from composting in 2023 were estimated to range between 1.8 and 7.0 MMT C02 Eq., which
indicates a range of 58 percent below to 58 percent above the 2023 emission estimate of each gas (see
Table 7-15).
7-40 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Table 7-15: Tier 1 Quantitative Uncertainty Estimates for Emissions from Composting
(MMT C02 Eq. and Percent)
Uncertainty Range Relative to Emission Estimate
2023 Emission
(MMT CO2
Eq.)
(%)
Estimate
Lower
Upper
Lower
Upper
Source
Gas
(MMT CO2 Eq.)
Bound
Bound
Bound
Bound
Composting
cm
2.6
1.1
4.1
-58%
+58%
Composting
n2o
1.8
0.8
2.9
-58%
+58%
Composting
Total
4.4
1.8
7.0
-58%
+58%
QA/QC and Verification
General QA/QC procedures were applied to data gathering and input, documentation, and calculations
consistent with the U.S. Inventory QA/QC Plan, which is in accordance with Vol. 1 Chapter 6 of the 2006
IPCC Guidelines (see Annex 8 for more details). No errors were found for the current Inventory.
Recalculations Discussion
The population estimates for 2020 through 2022 were updated using more recent data from the U.S.
Census Bureau for the 1990 through 2023 Inventory. The U.S. population increased by 0.008 percent in
2020 and 0.047 percent in 2021, while it decreased by 0.005 percent in 2022. Consequently, net
emissions in the 1990 through 2023 Inventory increased in 2020 and 2021 and decreased in 2022 by the
same percentages compared to the previous (1990 through 2022) Inventory.
Planned Improvements
EPA completed a literature search in 2021 on emission factors and composting systems and
management techniques that were documented in a draft technical memorandum. The purpose of this
literature review was to compile all published emission factors specific to various composting systems
and composted materials in the United States to determine whether the emission factors used in the
current methodology can be revised or expanded to account for geographical differences and/or
differences in composting systems used. For example, outdoor composting processes in arid regions
typically require the addition of moisture compared to similar composting processes in wetter climates.
In general, there is a lack of facility-specific data on the management techniques and sum of material
composted to enable the use of different emission factors. EPA will continue to seek out more detailed
data on composting facilities to enable this improvement in the future.
Related, EPA has received comments during previous Inventory cycles recommending that calculations
for the composting sector be based on waste subcategories (i.e., leaves, grass and garden debris, food
waste) and category-specific moisture contents. At this time, EPA is not aware of any available datasets
which would enable estimations to be performed at this level of granularity. EPA will continue to search
for data which could lead to the development of subcategory-specific composting emission factors to
be used in future Inventory cycles.
EPA will also continue to seek out activity data including processing capacity and years of operation for
commercial composting facilities in Puerto Rico (for additional years), Guam, and other U.S. Territories
for inclusion in a future Inventory.
Waste 7-41
-------
EPA received some comments to update the discussion of the composting sector to be in line with
current terminology, trends, and methods. EPA will work to refresh this discussion with inputs from
composting experts at EPA.
7.4 Anaerobic Digestion at Biogas Facilities
(Source Category 5B2)
Anaerobic digestion is a series of biological processes in the absence of oxygen in which
microorganisms break down organic matter, producing biogas and digestate. The biogas primarily
consists of CH4, biogenic C02, and trace amounts of other gases such as N20 (IPCC 2006) and is often
combusted to produce heat and power, or further processed into renewable natural gas or for use as a
transportation fuel. Digester gas contains approximately 65 percent CH4(a normal range is 55 percent to
65 percent) and approximately 35 percent C02 (WEF 2012; EPA 1993). Methane emissions may result
from a fraction of the biogas that is lost during the process due to leakages and other unexpected events
(0 to 10 percent of the amount of CH4 generated, IPCC 2006), collected biogas that is not completely
combusted, and entrained gas bubbles and residual gas potential in the digestate. Carbon dioxide
emissions are biogenic in origin and should be reported as an informational item in the Energy Sector
(IPCC 2006). Volume 5 Chapter 4 of the 2006 IPCC Guidelines notes that at biogas plants where
unintentional CH4 emissions are flared, CH4 emissions are likely to be close to zero.
Anaerobic digesters differ based on the operating temperature, feedstock type and moisture content,
and mode of operation. The operating temperature dictates the microbial communities that live in the
digester. Mesophilic microbes are present at temperatures ranging from 85 to 100 degrees Fahrenheit
while thermophilic microbes thrive at temperatures ranging from 122 to 140 degrees Fahrenheit (WEF
2012). Digesters may process one or more types of feedstock, including food waste; municipal
wastewater solids; livestock manure; industrial wastewater and residuals; fats, oils, and grease; and
other types of organic waste streams. Co-digestion (multiple feedstocks) is employed to increase
methane production in cases where an organic matter type does not break down easily. In co-digestion,
various organic wastes are decomposed in a singular anaerobic digester by using a combination of
wastewater solids or manure and food waste from restaurants or food processing industry, a
combination of manure and waste from energy crops or crop residues (EPA 2016), or alternative
combinations of feedstock. The moisture content of the feedstock (wet or dry) impacts the amount of
biogas generation. Wet anaerobic digesters process feedstock with a solids content of less than 15
percent while dry anaerobic digesters process feedstock with a solids content greater than 15 percent
(EPA 2020). Digesters may also operate in batch or continuous mode, which affects the feedstock
loading and removal. Batch anaerobic digesters are manually loaded with feedstock all at once and then
manually emptied while continuous anaerobic digesters are continuously loaded and emptied with
feedstock (EPA 2020).
The three main categories of anaerobic digestion facilities included in national greenhouse gas
inventories include the following:
Anaerobic digestion at biogas facilities, or stand-alone digesters, can be industry-dedicated
digesters that process waste from on industry or industrial facility (typically food of beverage
waste from manufacturing), or multi-source digesters that process feedstocks from various
7-42 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
sources (e.g., municipal food scraps, manure, food processing waste). Some stand-alone
digesters also co-digest other organics such as yard waste.
On-farm digesters manage organic matter and reduce odor generated by farm animals or crops.
On-farm digesters are found mainly at dairy, swine, and poultry farms where there is the highest
potential for methane production to energy conversion. On-farm digesters may also accept food
waste as feedstock for co-digestion.
Digesters at water resource recovery facilities (WRRF) produce biogas through the treatment
and reduction of wastewater solids. Some WRRF facilities may also accept and co-digest food
waste.
This section focuses on stand-alone anaerobic digestion at biogas facilities. Emissions from on-farm
digesters are included Chapter 5 (Agriculture) and AD facilities at WRRFs are included in Section 7.2.
From 1990 to 2023, the estimated amount of waste managed by stand-alone digesters in the United
States increased from approximately 988 kt to 15,094kt, an increase of 1,428 percent. As described in
the Uncertainty section, no data sources present the annual amount of waste managed by these
facilities prior to 2015 when the EPA began a comprehensive data collection survey. Thus, the emission
estimates between 1990 and 2014, and for 2022 to 2023 are general estimates extrapolated from data
collected for years 2015 to 2022 via the EPA surveys (EPA 2018, 2019, 2021, 2023, and 2024). The steady
increase in the amount of waste processed over the time series is likely driven by increasing interest in
using biogas produced from waste as a renewable energy source and other organics diversion goals.
In 2023, emissions from stand-alone anaerobic digestion at biogas facilities were approximately 16,906
MT C02 Eq. (0.60 kt) (see Table 7-16 and Table 7-17).
Table 7-16: CH4 Emissions from Anaerobic Digestion at Biogas Facilities (MT C02 Eq.)
Activity
1990 | 2005
2019
2020
2021
2022
2023
CH4 Generation
22,129
66,388
348,699
328,348
347,874
338,111
338,111
CH4 Recovery
(21,023) |
(63,069)
(331,264)
(311,931)
(330,480)
(321,205)
(321,205)
CH4 Emissions
1,106
3,319
17,435
16,417
17,394
16,906
16,906
Notes: Parentheses indicate negative values. Totals may not sum due to independent rounding.
Table 7-17: CH4
Emissions from Anaerobic Digestion at Biogas Facilities (kt CH4)
Activity
1990 |
2005
2019
2020
2021
2022
2023
CH4 Generation
1 I
2
12
12
12
12
12
CH4 Recovery
(1)
(2)
(12)
(11)
(12)
(11)
(11)
CH4 Emissions
+
+
0.62
0.59
0.62
0.60
0.60
+ Does not exceed 0.5 kt ChU.
Notes: Parentheses indicate negative values. Totals may not sum due to independent rounding.
Methodology
Methane emissions from anaerobic digestion depend on factors such as the type of waste managed, the
amount and type of supporting material (such as wood chips and peat) used, temperature, moisture
content (e.g., wet and fluid versus dry and crumbly), aeration during the digestion process, unintentional
leakages, and how the biogas generated is used/combusted (e.g., flared, used on-site, used off-site).
Waste 7-43
-------
The emissions presented in Table 7-16 were estimated using the IPCC default (Tier 1) methodology
(Volume 5, Chapter 4, IPCC 2006) given in Equation 7-5 below, which applies a default leakage factor of
5 percent to the CH4 generated, which is the product of an emission factor and the mass of organic
waste processed (Equation 7-6). Only CH4 emissions are estimated because N20 emissions are
considered negligible (IPCC 2006). Some Tier 2 data are available (annual quantity of waste digested) for
the later portion of the time series (2015 and later). The methods were selected in accordance with IPCC
methodological decision trees and available data on organic waste processed.
Per the 2006 IPCC Guidelines, emissions of CH4 from anaerobic digestion facilities due to unintentional
leakages during process disturbances or other unexpected events are generally between 0 to 10 percent
of the amount of CH4 generated. When facility-specific information or data are unavailable, a 5 percent
leakage factor is recommended (IPCC 2006).
Equation 7-5: Methane Emissions Calculation for Anaerobic Digestion
CH4 Emissions = L x (GCH4)
where,
CH4 Emissions = total CH4 emissions in inventory year, Gg CH4
L = leakage factor, default assumed 5 percent (IPCC 2006)
Gch4 = total CH4 generation in inventory year, Gg CH4
Equation 7-6: Methane Generation Calculation for Anaerobic Digestion
Gch4 = liM x EF0 x 10"3
where,
M, = mass of organic waste treated by biological treatment type /', Gg, see Table 7-18
EF = emission factor for treatment /', g CH4/kg waste treated, 0.8 Mg/Gg CH4
i = anaerobic digestion
Per IPCC Tier 1 methodology defaults, the emission factor for CH4 assumes a moisture content of 60
percent in the wet waste (IPCC 2006). Both liquid and solid wastes are processed by stand-alone
digesters and the moisture content entering a digester may be higher. One emission factor, 0.8 Mg/Gg
CH4is applied for the entire time series (IPCC 2006 Volume 5, Chapter 4, Table 4.1).
The annual quantity of waste digested is sourced from EPA surveys of anaerobic digestion facilities (EPA
2018, 2019, 2021, 2023, and 2024). The EPA was granted the authority to survey anaerobic digestion
facilities that process food waste annually through an Information Collection Request (ICR No.
2533.01). The scope includes stand-alone and co-digestion facilities (on-farm and water resource
recovery facilities [WRRF]). Five reports with survey results have been published to date:
Anaerobic Digestion Facilities Processing Food Waste in the United States (2015): Survey
Results (EPA 2018)
Anaerobic Digestion Facilities Processing Food Waste in the United States (2016): Survey
Results (EPA 2019)
Anaerobic Digestion Facilities Processing Food Waste in the United States (2017 & 2018): Survey
Results (EPA 2021)
7-44 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Anaerobic Digestion Facilities Processing Food Waste in the United States (2019): Survey
Results (EPA 2023).
Anaerobic Digestion Facilities Processing Food Waste in the United States (2020 &2021) (EPA
2024).
These reports present aggregated survey data including the annual quantity of waste processed by
digester type (i.e., stand-alone, on-farm, and WRRF); waste types accepted; biogas generation and end
use; and more. The amount of waste digested as reported in the survey reports were assumed to be in
wet weight; the majority of stand-alone digesters were found to be wet and mesophilic (EPA 2023).
The aggregated data presented in the EPA reports are underestimates of the actual amount of processed
waste and biogas produced because (1) surveys rarely achieve a 100 percent response rate and some
fraction of facilities in each survey year did not respond to the survey; (2) EPA focused the surveys on
facilities that primarily process food waste, although non-food waste quantities processed were also
collected and reported; and (3) while the EPA has done due diligence to identify all stand-alone
digesters that process food waste, EPA may not have identified all facilities across the United States and
its territories.
The annual quantity of waste digested at stand-alone digesters for 1990 to 2014 (only 1990 and 2005 are
shown in Table 7-18) was estimated by multiplying the count of estimated operating facilities (as
presented in Table 7-19) by the weighted average of waste digested in 2015 to 2019 collected through
EPA's survey data (EPA 2018, 2019, 2021, and 2023). Masked survey responses of food and non-food
waste processed were shared with the Inventory team by the EPA team leading the EPA AD Data
Collection Surveys for 2015 to 2019. Significantly less facilities responded to the survey for 2020 and
2021 compared to prior years and it is unclear whether the reported total waste digested is
representative of all operational stand-along digesters and is therefore why the weighted average does
not include data from 2020 and 2021. This assumption may be revisited when the next iteration of EPA's
survey data is published.
The number of facilities that reported annual quantities of waste digested to the EPA survey varies by
year. The masked data provided by the EPA AD survey data collection team include data for 41, 44, 42,
43, and 18 facilities between 2015 to 2019, respectively. These data were used to calculate the weighted
average of waste digested of 272,294 short tons.
The weighted average applied to the current Inventory is calculated as follows for 1990 to 2014:
Equation 7-7: Weighted Average of Waste Processed
Weighted Average Waste Processed = ^ year — year
/—lyear Sum Of All FaC
where,
year = the year of data for the average waste processed and count of facilities in the numerator
W = total average waste processed in the respective survey year, food and non-food waste
(short tons).
Fac = the number of facilities that reported an amount of waste processed in the respective
survey year. Note the number of facilities that provided an annual quantity of waste
processed data was internally shared and differs from the total number of facilities that
responded to the EPA surveys as presented in EPA (2018, 2019, 2021, and 2023).
Waste 7-45
-------
Estimates of the quantity of waste digested for 1990 to 2014 are calculated by multiplying the weighted
average of waste digested from the masked survey data by the count of operating facilities in each year.
This calculation assumes that each facility operates continuously from the first year of operation for the
remainder of the time series. Additional efforts will be made to quantify the number of operating
facilities and estimates of the total waste digested by year for future Inventories as described in the
Planned Improvements section.
Estimates of the quantity digested for 2015 to 2021 were taken from EPA's AD survey data (EPA 2018,
EPA 2019, EPA 2021, EPA 2023, and EPA 2024). In the 1990 to 2023 Inventory, the quantity of liquid, non-
food waste was converted to tons using a general conversion factor of 3.8 lbs/gallon. Liquid waste was
not incorporated into the EPA survey results prior to 2019 and there is an increase in the amount of
waste digested annually prior to 2019 and 2019 and later years.
The EPA (2023) report provides a significant increase in data granularity for stand-alone digesters
compared to earlier reports because food waste processed by the beverage sector is included as tons of
food waste processed as opposed to gallons of food waste processed in prior survey years. Detail on the
sources and types of the liquid food and non-food waste was not available in the 2015 to 2018 data to
reliably convert the data to tons. However, the 2019 data point provides some assurance that using a
general conversion factor to convert liquid waste to tons yields a more comprehensive estimate of total
waste processed at stand-alone AD facilities. EPA published results from the 2020 and 2021 survey on
their website only (EPA 2024); a separate report with details provided in earlier years has not been
published.
The estimate of waste digested for 2022 to 2023 were extrapolated using the average of the waste
digested from the 2020 to 2021 survey data (EPA 2024) as a proxy. The average did not include data from
2015 and 2016 because there is a drop in the amount of waste digested by nearly 1 million tons between
2016 and 2017. The quantities digested between 2015 and 2016 are similar, and quantities digested
between 2017 and 2018 are similar. The quantity digested for 2019 is nearly twice the amount of prior
EPA survey years because food waste from the beverage sector were able to be accurately converted to
tons. Estimates for 2022 to 2023 will be updated as future EPA survey reports are published.
Table 7-18: Estimated U.S. Waste Digested (kt)
Activity
1990 I
2005 I
2019
2020
2021
2022
2023
Waste Digested
988 |
2,964 |
15,567
14,658
15,530
15,094
15,094
The estimated count of operating facilities is calculated by summing the count of digesters that began
operating by year over the time series. The year a digester began operating is sourced from EPA (2021).
This assumes all facilities are in operation from their first year of operation throughout the remainder of
the time series, including facilities prior to 1990. This is likely an overestimate of facilities operating per
year but does not necessarily translate to an overestimate in the amount of waste processed because a
weighted average of waste processed for the surveyed facilities is applied to these years. The number of
facilities in 1990 to 2014 are directly used in calculating the emissions for those years.
Table 7-19: Estimated Number of Stand-Alone AD Facilities Operating
Year
1990 2005 1
2019
2020
2021
2022
2023
Estimated Count of Operational Facilities
4 12 |
54
54
54
54
54
7-46 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Uncertainty
The methodology applied for the 1990 to 2014 emissions estimates should be considered a starting
point to build on in future years if additional historical data become available. Five years of facility-
provided data are available (2015 to 2019) while the rest of the time series is estimated based on an
assumption of facility counts and the 2015 to 2019 weighted average annual waste digested as
calculated from survey data. The major limitations, and uncertainty drivers in the emissions estimates,
are related to the uncertainty in assumptions to ensure completeness across the time series and the
limitations in the EPA AD survey data, as described below:
1. The EPA AD surveys (EPA 2018; EPA 2019; EPA 2021; EPA 2023; EPA 2024) did not receive a 100
percent response rate, meaning that the survey data represent a portion, albeit the majority, of
stand-alone digesters, and annual waste processed. The methodology applied here did not
attempt to estimate waste digested by facilities that did not respond to the survey, which likely
underestimates the quantity of waste digested and CH4emissions.
2. The EPA AD survey data (EPA 2018; EPA 2019) present both food and non-food waste digested.
The non-food waste was reported as liquid (gallons) and solid (tons). The quantity of liquid
waste managed for 2015 and 2016, which is used as a proxy for 1990 to 2014, was converted to
tons using a general conversion factor of 3.8 lbs/gallon. This may slightly over- or underestimate
the quantity of waste digested and CH4emissions between 1990 to 2018. This conversion was
not made by EPA in the survey report (EPA 2018). However, EPA (2021) did convert the liquid
waste managed to tons for 2017 and 2018 using the general conversion factor of 3.8 lbs/gallon.
3. The assumption required to estimate the activity data for 1990 to 2014 may overestimate the
number of facilities in operation because it assumes that each facility operates from its start
year for the entire time series (i.e. facility closures are not accounted for). This introduces a large
amount of uncertainty in the estimates compared to years where there is directly reported
survey data. It is unclear whether this under- or over-estimates the quantity of waste digested
and CH4emissions.
4. The most recent EPAAD survey data (EPA2023; EPA2024) includes waste processed bythe
beverage sector, which was not presented in prior survey years. No attempts were made to
separately estimate and include this waste stream in years prior to 2019 (i.e. the EPA 2023
survey). This means that annual CH4 estimates for 1990 to 2018 may be underestimated.
The estimated uncertainty from the 2006IPCC Guidelines is ±54 percent for the Approach 1
methodology.
Emissions from anaerobic digestion at stand-alone biogas facilities in 2023 were estimated to be
between 7,802 and 26,009 MT C02 Eq., which indicates a range of 54 percent below to 54 percent above
the 2023 emission estimate of CH4 (see Table 7-20). A ±20 percent uncertainty factor is applied to the
annual amount of material digested (i.e., the activity data), which was developed with expert judgment
(Bronstein 2021). A ±50 percent default uncertainty factor is applied to the CH4 emission factor (IPCC
2006). Using the IPCC's error propagation equation (Equation 3.1 in IPCC 2006 Volume 1, Chapter 3), the
combined uncertainty percentage is ±54 percent.
Waste 7-47
-------
Table 7-20: Approach 1 Quantitative Uncertainty Estimates for Emissions from
Anaerobic Digestion (MT C02 Eq. and Percent)
2023
Emission
Estimate
Source Gas (MT CO2 Eq.)
Uncertainty Range Relative to Emission
Estimate
(MTCO2 Eq.)
(%)
Lower Upper
Bound Bound
Lower Upper
Bound Bound
Anaerobic Digestion at Biogas Facilities CH4 16,906
7,802 26,009
-54% +54%
QA/QC and Verification
General QA/QC procedures were applied to data gathering and input, documentation, and calculations
consistent with the U.S. Inventory QA/QC Plan, which is in accordance with Vol. 1, Chapter 6 of the 2006
IPCC Guidelines (see Annex 8 for more details). No errors were found for the current Inventory.
Recalculations Discussion
The weighted average of waste digested was recalculated for the current Inventory to incorporate EPA
AD survey data from 2019. This impacted the emissions estimates between 1990 to 2014.
In the previous Inventory, the amount of waste processed for 2020 through 2022 was extrapolated based
on available survey data. With the publication of survey data for 2019 (EPA 2023) and 2020 and 2021
(EPA 2024), the extrapolated amounts of waste digested were updated. Recalculations for the amount of
waste processed resulted in increases of 22 percent per year between 1990 to 2014, 37 percent in 2020,
30 percent in 2021, and 26 percent in 2022. Consequently, emissions estimates in the 1990 through
2023 Inventory increased by the same percentages these same years compared to the previous (1990
through 2022) Inventory.
Planned Improvements
EPA will continue to incorporate updated survey data from future EPA AD Data Collection Surveys when
the survey data are published. These revisions may change the estimated emissions for 2022 and 2023.
Additionally, quality control checks on the default emission factor used to determine CH4 generation is
in process.
EPA will also reassess how best to estimate annual waste processed using proxy data for years between
the EPA AD Data Collection Survey reports as needed (e.g., for 2022 and 2023). The methodology
described here assumes the same average amount of waste is processed each year for 2022 and 2023.
EPA continues to seek out data sources to confirm the estimated number of operational facilities by year
prior to 2015 and consider how best to estimate the quantity of waste processed per year by these
facilities with the goal of better estimating the annual quantity of waste digested between 1990 to 2014.
Available data will also be compiled where available for facilities that did not directly respond to the EPA
AD Data Collection surveys for completeness.
7-48 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
7.5 Waste Incineration (Source Category
5C1)
As stated earlier in this chapter, carbon dioxide (C02), nitrous oxide (N20), and methane (CH4) emissions
from the combustion of waste are accounted for in the Energy sector rather than in the Waste sector
because almost all combustion of municipal solid waste (MSW) in the United States occurs at waste-to-
energy facilities where useful energy is recovered. Similarly, the Energy sector also includes an estimate
of emissions from burning waste tires and hazardous industrial waste, because virtually all of the
combustion occurs in industrial and utility boilers that recover energy. The combustion of waste in the
United States in 2023 resulted in 12.8 MMT C02 Eq. of emissions. For more details on emissions from
the combustion of waste, see Section 3.3 of the Energy chapter.
Additional sources of emissions from waste combustion include non-hazardous industrial waste
incineration and medical waste incineration. As described in Annex 5 of this report, data are not readily
available for these sources and emission estimates are not provided.
An analysis of the likely level of medical waste incineration emissions was conducted based on a 2009
study of hospital/ medical/ infectious waste incinerator (HMIWI) facilities in the United States (RTI 2009).
Based on that study's information of waste throughput and an analysis of the fossil-based composition
of the waste, it was determined that annual greenhouse gas emissions for medical waste incineration
would be below 500 kt C02 Eq. per year and considered insignificant for the purposes of national
inventory reporting. More information on this analysis is provided in Annex 5.
Furthermore, an analysis was conducted on the likely level of sewage sludge incineration emissions
based on the total amount of sewage sludge generated and assumed percent incineration. Based on the
assumed amount of sludge incinerated and non-C02 factors for solid biomass it was determined that
annual greenhouse gas emissions for sewage sludge incineration would be below 500 kt C02 Eq. per
year and considered insignificant for the purposes of national inventory reporting. More information on
this analysis is provided in Annex 5.
7.6 Waste Sources of Precursor Greenhouse
Gases
In addition to the main greenhouse gases addressed above, waste generating and handling processes
are also sources of precursors to greenhouse gases. This section summarizes information on precursor
emissions, which include carbon monoxide (CO), nitrogen oxides (NOx), non-methane volatile organic
compounds (NMVOCs), ammonia (NH3), and sulfur dioxide (S02). These gases are not direct greenhouse
gases, but can indirectly impact Earth's radiative balance by altering the concentrations of other
greenhouse gases (e.g., tropospheric ozone) and atmosphere aerosol (e.g., particulate sulfate). Total
emissions of NOx, CO, NMVOCs, NH3, and S02 from waste sources for the years 1990 through 2023 are
provided in Table 7-21.
Waste 7-49
-------
Table 7-21: Emissions of NOx, CO, NMVOC, NH3 and S02 from Waste (kt)
Gas/Source
1990
2005
2019
2020
2021
2022
2023
NOx
84
51
73
76
76
75
74
CO
979
1,178
1,181
1,342
1,343
1,340
1,340
NMVOCs
870
152
156
173
172
171
171
NH3
73
18
19
84
84
83
83
SO2
36
20
23
33
32
31
31
Methodology and Time-Series Consistency
Emission estimates for 1990 through 2023 were obtained from data published on the National
Emissions Inventory (NEI) Air Pollutant Emissions Trends Data website (EPA 2024). For Table 7-21, NEI
reported emissions of CO, NOx, S02, NH3, and NMVOCs were recategorized from NEI Emissions
Inventory System (EIS) sectors. The EIS sectors were mapped to categories more closely aligned with
sectors and categories in this report, based on discussions between the EPA Inventory and NEI staff (see
crosswalk documented in Annex 6.3). EIS sectors mapped to the waste sector categories in this report
include: waste disposal and recycling (landfills; publicly owned treatment works; industrial wastewater;
treatment, storage, and disposal facilities; waste incineration; and other).10 As described in the NEI
Technical Support Documentation (TSD) (EPA 2023), emissions are estimated through a combination of
emissions data submitted directly to the EPA by state, local, and tribal air agencies, as well as additional
information added by the Agency from EPA emissions programs, such as the emission trading program,
Toxics Release Inventory (TRI), and data collected during rule development or compliance testing.
Within the NEI, there is only one EIS sector for waste generating and handling processes, so precursor
estimates are aggregated in Table 7-21 for consistency with NEI reporting. Future presentations of this
data may disaggregate emissions so it better maps to reporting categories in this report.
Methodological recalculations were applied to the entire time series to ensure time-series consistency
from 1990 through 2023, which are described in detail in the NEI's TSD and on EPA's Air Pollutant
Emission Trends website (EPA 2023b; EPA 2024). No quantitative estimates of uncertainty were
calculated for this source category.
10 Precursor emissions from waste incineration were reported in the Energy sector in the previous Inventory but are not
disaggregated from the Waste sector in this report.
7-50 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
8 Other
No greenhouse gas emissions are currently reported under the "Other" inventory sector.
Other 8-1
-------
9 Recalculations and Improvements
Each year, some emission and sink estimates in the Inventory of U.S. Greenhouse Gas Emissions and
Sinks are recalculated and revised, as efforts are made to improve the estimates through the use of
better methods and/or data with the goal of improving inventory quality and reducing uncertainties,
including improving the transparency, completeness, consistency, and overall usefulness of the report.
In this effort, the United States follows the 2006IPCC Guidelines (IPCC 2006) and its
refinements/supplements.
When methodological changes have been implemented, the previous Inventory's time series (i.e., 1990
to 2022) is assessed and recalculated as needed to reflect the change, per guidance in IPCC (2006). The
most common reason for recalculating U.S. greenhouse gas emission estimates is to update recent historical
data. Changes in historical data are often the result of changes in statistical data supplied by other
agencies, and these changes do not necessarily impact the entire time series. Improvements underway
or planned for future reports are discussed under planned improvements sections within each category
across the report.
The results of all methodological changes and historical data updates made in the current Inventory in
calculating C02-equivalent U.S. greenhouse gas emissions and sinks across the Energy, Industrial
Processes and Product Use (IPPU), Agriculture, Land Use, Land Use-Change and Forestry (LULUCF), and
Waste sectors are presented in Figure 9-2, while impacts on both total and net emissions by gas are
summarized in Table 9-1 and Table 9-2. Collectively, all methodological changes and historical data
updates made in the current Inventory resulted in lower estimates of annual net emissions by an annual
average decrease of 56.0 MMT C02 Eq. (0.9 percent). The tables below present results relative to the
previously published Inventory (i.e., the 1990 to 2022 report) in units of MMT C02 Eq.
Recalculations and Improvements 9-1
-------
Figure 9-1: Impacts of Recalculations on Net Emissions
Table 9-1: Overall Impact of Recalculations by Gas Compared to 1990-2022 Inventory
(MMT C02 Eq.)
Change in Gas
1990
2005
2019
2020
2021
2022
Average
Annual
Change
o
o
0.1
+
1.4
1.0
2.9
2.4
0.1
CH4a
1.5
1.7
(1.8)
(4.5)
(4.8)
(5.6)
0.3
N2Oa
(0.3)
5.6
{+)
0.3
0.2
(2.2)
1.8
HFCs
+
3.3
7.6
7.5
7.2
6.8
4.1
PFCs
0.2
0.1
+
+
{+)
(0.1)
0.2
SFe
{+)
{+)
(0.1)
(0.4)
(0.5)
(0.3)
{+)
NF3
(0.1)
{+)
+
+
+
{+)
(0.1)
Change in Total Gross
Emissions (Sources)b
1.4
10.7
7.2
3.9
5.0
0.9
6.4
Change in LULUCF Total Net Flux
(62.3)
(64.1)
(61.0)
(61.4)
(60.4)
(52.1)
(64.7)
Change in LULUCF Emissions
1.1
2.9
5.1
14.3
8.1
1.0
2.2
Change in LULUCF Sector Net
Total0
(61.2)
(61.2)
(55.8)
(47.2)
(52.3)
(51.1)
(62.5)
CH4
1.2
2.3
3.6
9.7
5.7
1.2
1.9
n2o
(0.2)
0.6
1.5
4.6
2.4
(0.2)
0.3
Change in Net Emissions
(Sources and Sinks)d
(59.8)
(50.6)
(48.6)
(43.3)
(47.3)
(50.2)
(56.0)
+ Values do not exceed 0.05 MMT C02 Eq.
a Recalculations impacts for ChU and N20 do not include ChU and N20 emissions from LULUCF.
9-2 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
b Total gross emissions are the overallimpact of recalculations sources of emissions.
cThe LULUCF Sector Net Total is the net sum of all LULUCF ChUand N20 emissions to the atmosphere plus LULUCF net carbon
stock changes.
d Net emissions presented in this table are the overall impact of recalculations on all emissions sources and sinks and include
emissions and removals from LULUCF.
Notes: Parentheses indicate negative values. Totals may not sum due to independent rounding.
Table 9-2: Overall Impact of Recalculations by Sector Compared to 1990-2022
Inventory (MMT C02 Eq.)
Change in Inventory Sector
1990
2005
2019
2020
2021
2022
Average
Annual
Change
Energy
4.7
10.3
4.7
3.5
3.5
1.9
6.4
IPPU
0.1
3.3
9.0
7.4
9.3
6.4
4.2
Agriculture
0.3
0.7
0.7
0.8
1.0
(0.1)
0.6
Waste
(0.0)
0.0 |
(0.9)
(1.9)
(2.1)
(1.8)
(0.3)
Change in Total Gross Emissions
(Sources)3
1.4
10.7
7.2
3.9
5.0
0.9
6.4
Change in LULUCF Sector Net Totalb
(61.2)
(61.2)
(55.8)
(47.2)
(52.3)
(51.1)
(62.5)
Change in Net Emissions (Sources and
Sinks)0
(59.8)
(50.6)
(48.6)
(43.3)
(47.3)
(50.2)
(56.0)
+ Values do not exceed 0.05 MMT C02 Eq.
a Total gross emissions are the overallimpact of recalculations sources of emissions.
bThe LULUCF Sector Net Total is the net sum of all LULUCF ChUand N20 emissions to the atmosphere plus LULUCF net carbon
stock changes.
c Net emissions presented in this table are the overall impact of recalculations on all emissions sources and sinks and include
emissions and removals from LULUCF.
Notes: Parentheses indicate negative values. Totals may not sum due to independent rounding.
Key Recalculations and Improvements for
1990-2023 Inventory
The following source and sink categories underwent the most significant methodological and historical
data changes. A brief summary of the recalculations and/or improvements undertaken are provided for
the most significant categories in Table 9-3. In addition, the current Irvertory includes the following new
categories that were not a part of the previous Irvertory that improve the completeness of the national
estimates: C02 transport, injection, and geological storage, perennial woody biomass carbon stock
changes and biomass carbon stock changes from croplands and lands converted to and from croplands
(e.g., croplands converted to grasslands, grasslands converted to croplands). This Irvertory also now
includes additional gases (NF3 and HFCs) within the SF6 and PFC from other product use category. To
understand the details of any specific recalculation or methodological improvement, see the
Recalculations sections within each source/sink categories' section found in Chapters 3 through 7 of
this report.
Recalculations and Improvements 9-3
-------
Table 9-3: Key Recalculations
Reason for Recalculation or
Sector Category Improvement
Impact of
Recalculation
on 2022
Average Impact over
Time Series
Value
(MMT
Percent CO2 Eq.)
LULUCF Forest Land Completeness. Use of updated NSVB
Remaining Forest methods, updated FIA data, and improved
Land (CO2) forest land classification in interior Alaska.
See details in Chapter 6.2.
(55.4)
-8.0% (71.0)
LULUCF Land Converted to Completeness and Accuracy. Use of
Settlements (CO2) updated FIA data for forests converted to
settlements and implementation of Tier 1
methodology to estimate biomass carbon
stock changes for land conversions. See
details in Chapter 6.11.
11.6
+ 17.3% 12.2
IPPU Substitution of Completeness. Use of an updated
Ozone Depleting Vintaging Model, incorporating latest
Substances (HFCs) activity data. See details in Chapter 4.25.
6.8
+3.2% 4.1
LULUCF Land Converted to Completeness and Accuracy. Use of
Grassland (CO2) updated FIA data for forests converted to
grasslands and implementation of Tier 1
methodology to estimate biomass carbon
stock changes for land conversions. See
details in Chapter 6.7.
(4.7)
-16.5% (4.0)
LULUCF Land Converted to Accuracy. Use of updated FIA data and
Forest Land (CO2) adjustments to NSVB methods. See
details in Chapter 6.3.
(3.5)
-3.5% (3.5)
LULUCF Wetlands Consistency. Use of updated National
Remaining Inventory of Dams (NID) data. See details
Wetlands (CH4) in Chapter 6.8.
1.6
+3.7% 1.6
LULUCF Cropland Completeness. Incorporation of perennial
Remaining woody biomass C stock change estimates
Cropland (CO2) and conversions between crop types. See
details in Chapter 6.4.
0.1
-8.0% (1.4)
Energy Mobile Combustion Accuracy. Use of new EV milage estimates,
(N2O) in addition to updated fuel EFs from GREET
and fleet compositions from MOVES5. See
details in Chapter 3.1.
(0.1)
-4.6% 1.4
LULUCF Forest Land Accuracy. Use of updated WFEIS-based
Remaining Forest data to estimate non-C02 emissions from
Land (Non-C02) forest fires. See details in Chapter 6.2.
(0.5)
-2.6% 0.7
Agriculture Agricultural Soil Accuracy. Use of updated time-series
Management (N2O) activity data and updated data splicing
methods. See details in Chapter 5.4.
1.0
+0.2% 0.6
9-4 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Figure 9-2: Impacts from Recalculations to U.S. Greenhouse Gas Emissions and Sinks
by Sector
40 ¦ Waste
¦ Agriculture
30 "Energy
¦ LULUCF Sector Net Total
2q Industrial Processes and Product Use
™ Change in Net Total Emissions
-90
-100
-110
CTtCT»C*CF»O^G,\O^CriCriO^OOOOOOOOOO'—It—I i—I i—I t-H t-H i—\ 1—I H H fS| (N f\l
cr.cr«a^cno^cr.cno^cr.cnooooooooooooooooooooooo
HHHHHHHHHH(NfM(N(NfNfMrMfNlfNirsirsJ(NrslfMfNfNl,NrslfMfN(NfNirNl
Table 9-4 and Table 9-5 include the quantitative effects of methodological changes and historical data
updates made in the current Inventory in calculating C02-equivalent U.S. greenhouse gas emissions and
sinks by gas across all sectors.
Table 9-4: Revisions to U.S. Greenhouse Gas Emissions (MMT C02 Eq.)
Change in Gas/Source
1990
2005
2019
2020
2021
2022
Average
Annual
Change
N
O
o
0.1
+
1.4
1.0
2.9
2.4
0.1
Fossil Fuel Combustion
+
+
+
0.6
0.4
3.4
0.1
Electric Power Sector
NC
NC
NC
NC
NC
M
(+)
Transportation
+
(+)
+
0.1
+
2.3
0.1
Industrial
+
+
+
1.4
+
(1.4)
+
Residential
NC
NC
NC
NC
+
1.1
+
Commercial
(+)
+
(+)
(+)
+
0.4
+
Recalculations and Improvements 9-5
-------
Change in Gas/Source
1990
2005
2019
2020
2021
2022
Average
Annual
Change
U.S. Territories
NC
NC
NC
(1.0)
0.3
0.9
+
Non-Energy Use of Fuels
NC
NC
+
0.1
0.1
(1.1)
(+)
Natural Gas Systems
0.1
+
0.2
0.1
(+)
(0.1)
+
Cement Production
NC
NC
NC
NC
NC
NC
NC
Lime Production
NC
NC
NC
NC
NC
NC
NC
Other Process Uses of Carbonates
NC
NC
NC
NC
NC
(+)
(+)
Glass Production
NC
NC
NC
NC
NC
NC
NC
Soda Ash Production
NC
NC
NC
NC
NC
NC
NC
Non-EOR Carbon Dioxide Utilization
NC
NC
(2.5)
(2.1)
(2.1)
(2.2)
(0.4)
Incineration of Waste
NC
NC
NC
NC
NC
0.1
+
Titanium Dioxide Production
NC
NC
NC
NC
0.1
0.1
+
Aluminum Production
NC
NC
NC
NC
NC
NC
NC
Iron and Steel Production & Metallurgical
Coke Production
(+)
(+)
3.7
3.0
5.3
4.5
0.5
Ferroalloy Production
NC
NC
NC
NC
(0.1)
NC
(+)
Ammonia Production
NC
NC
(+)
(0.7)
(0.7)
(0.7)
(0.1)
Urea Consumption for Non-Agricultural
Purposes
NC
NC
0.1
0.1
0.1
(1.6)
(+)
Phosphoric Acid Production
NC
NC
NC
NC
NC
(+)
(+)
Petrochemical Production
NC
NC
NC
NC
NC
NC
NC
Carbide Production and Consumption
NC
NC
NC
NC
NC
NC
NC
Lead Production
NC
NC
(+)
+
+
+
+
Zinc Production
NC
NC
NC
NC
NC
NC
NC
Petroleum Systems
+
+
(0.1)
(0.1)
(+)
0.1
+
Abandoned Oil and Gas Wells
NC
NC
NC
NC
NC
NC
NC
Magnesium Production and Processing
NC
NC
NC
NC
(+)
+
(+)
Liming
NC
+
NC
NC
+
(0.1)
(+)
Urea Fertilization
NC
NC
(0.1)
(0.1)
(0.1)
(0.1)
(+)
Coal Mining
+
+
NC
NC
NC
(+)
(+)
Substitution of Ozone Depleting
Substances
NC
NC
+
+
+
+
+
CO2Transport, Injection and Geological
Storage*
0.0
0.0
+
+
0.1
0.1
0.1
Biomass and Biodiesel Consumptions
NC
NC
(1.0)
(1.0)
(1.0)
(1.0)
(0.2)
International Bunker Fuelsb
NC
NC
NC
NC
NC
NC
NC
CH4c
1.5
1.7
(1.8)
(4.5)
(4.8)
(5.6)
0.3
Stationary Combustion
+
+
(0.1)
(0.1)
(0.1)
0.1
(+)
Mobile Combustion
+
0.9
(+)
(+)
(+)
(+)
0.4
Coal Mining
NC
NC
NC
NC
NC
NC
+
Abandoned Underground Coal Mines
NC
NC
NC
NC
(0.1)
(0.2)
(+)
Natural Gas Systems
0.8
0.6
0.3
(0.2)
+
(0.3)
0.2
Petroleum Systems
0.6
0.2
(1.5)
(2.7)
(3.5)
(3.4)
(0.1)
Abandoned Oil and Gas Wells
NC
NC
NC
NC
NC
NC
NC
Petrochemical Production
NC
NC
NC
NC
NC
NC
NC
9-6 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Change in Gas/Source
1990
2005
2019
2020
2021
2022
Average
Annual
Change
Carbide Production and Consumption
NC
NC
NC
NC
NC
NC
NC
Iron and Steel Production & Metallurgical
Coke Production
NC
NC
NC
NC
NC
NC
NC
Ferroalloy Production
NC
NC
NC
NC
{+)
NC
(+)
Enteric Fermentation
NC
NC
NC
NC
NC
NC
NC
Manure Management
NC
NC
NC
NC
NC
NC
NC
Rice Cultivation
NC
NC
NC
NC
0.2
(0.9)
(+)
Field Burning of Agricultural Residues
NC
NC
NC
NC
{+)
+
+
Landfills
NC
+
(0.5)
(1.5)
(1.4)
(1.1)
(0.2)
Wastewater Treatment
(+)
(+)
{+)
+
+
0.1
+
Composting
<+)
(~)
<+)
+
+
(+)
+
Anaerobic Digestion at Biogas Facilities
NC
NC
NC
+
+
+
+
Incineration of Waste
NC
NC
NC
NC
NC
+
+
International Bunker Fuelsb
NC
NC
NC
NC
NC
NC
NC
N2Oc
(0.3)
5.6
(+)
0.3
0.2
(2.2)
1.8
Stationary Combustion
+
+
{+)
{+)
(+)
(2.1)
(0.1)
Mobile Combustion
(0.7)
4.9
(0.4)
(0.1)
+
(0.1)
1.4
AdipicAcid Production
NC
NC
NC
NC
NC
NC
NC
Nitric Acid Production
NC
NC
NC
NC
NC
NC
NC
Manure Management
NC
NC
NC
NC
NC
NC
NC
Agricultural Soil Management
0.3
0.7
0.8
0.9
1.0
1.0
0.6
Field Burning of Agricultural Residues
NC
NC
NC
NC
(+)
+
+
Wastewater Treatment
NC
NC
(0.5)
(0.5)
(0.8)
(0.8)
(0.2)
N2O from Product Uses
NC
NC
NC
NC
NC
NC
NC
Caprolactam, Glyoxal, and Glyoxylic Acid
Production
NC
NC
NC
NC
NC
NC
NC
Incineration of Waste
NC
NC
NC
NC
NC
+
+
Composting
{+)
{+)
{+)
+
+
(+)
+
Electronics Industry
NC
NC
NC
NC
+
(+)
(+)
Natural Gas Systems
+
{+)
{+)
+
(+)
(0.1)
(+)
Petroleum Systems
+
+
{+)
+
+
+
+
International Bunker Fuelsb
NC
NC
NC
NC
NC
NC
NC
HFCs, PFCs, SFs and NF3
0.1
3.3
7.6
7.2
6.7
6.3
4.2
HFCs
+
3.3
7.6
7.5
7.2
6.8
4.1
Substitution of Ozone Depleting
Substances
NC
3.2
7.6
7.5
7.2
6.7
4.1
Fluorochemical Production
+
0.1
+
+
+
+
+
Electronics Industry
NC
{+)
+
+
+
+
+
Magnesium Production and Processing
NC
NC
NC
NC
NC
NC
NC
Other Product Manufacture and Use*
0.0
0.0
0.0
+
0.0
0.0
0.0
PFCs
0.2
0.1
+
+
(+)
(0.1)
0.2
Aluminum Production
NC
NC
NC
NC
NC
+
+
Fluorochemical Production
0.2
+
+
+
(0.1)
(0.2)
0.1
Electronics Industry
NC
+
0.1
0.1
0.1
0.1
+
Recalculations and Improvements 9-7
-------
Change in Gas/Source
1990
2005
2019
2020
2021
2022
Average
Annual
Change
Substitution of Ozone Depleting
Substances
NC
NC
NC
NC
NC
NC
NC
SFe and PFCs from Other Product Use
+
+
{+)
{+)
{+)
(0.1)
+
Electrical Equipment
NC
{+)
NC
NC
NC
NC
{+)
SFs
(+)
(+)
(0.1)
(0.4)
(0.5)
(0.3)
(+)
Electrical Equipment
{+)
{+)
(0.1)
(0.4)
(0.5)
(0.2)
(+)
SFe and PFCs from Other Product Use
{+)
{+)
{+)
{+)
{+)
(0.1)
(+)
Fluorochemical Production
NC
NC
+
+
+
+
+
Electronics Industry
NC
{+)
+
+
+
+
+
Magnesium Production and Processing
NC
NC
NC
NC
+
{+)
+
nf3
(0.1)
(+)
+
+
+
(+)
(0.1)
Electronics Industry
NC
NC
+
+
+
{+)
(+)
Fluorochemical Production
(0.2)
{+)
+
+
+
+
(0.1)
Other Product Manufacture and Use*
+
+
+
+
+
+
0.0
Change in Total Gross Emissions
1.4
10.7
7.2
3.9
5.0
0.9
6.4
Percentage Change in Total Emissions
0.0%
0.1%
0.1%
0.1%
+
+
0.1%
Change in LULUCF Sector Net Totald
(61.2)
(61.2)
(55.8)
(47.2)
(52.3)
(51.1)
(62.5)
Change in Net Emissions (Sources and
Sinks)
(59.8)
(50.6)
(48.6)
(43.3)
(47.3)
(50.2)
(56.0)
Percent Change in Net Emissions
-1.1%
-0.8%
-0.8%
-0.8%
-0.9%
-0.9%
-0.9%
NC (No Change)
+ Absolute value does not exceed 0.05 MMT C02 Eq. or 0.05 percent.
* Indicates a new source or new disaggregation for the current Inventory year. Emissions from new sources are captured in net
emissions and percent change totals.
a Emissions from biomass and biofuel consumption are not included specifically in summing Energy sector totals. Net carbon
fluxes from changes in biogenic carbon reservoirs are accounted for in the estimates for Land Use, Land-Use Change, and
Forestry.
b Emissions from international bunker fuels are not included in totals.
c LULUCF emissions of ChUand N20 are reported separately from gross emissions totals in Table 9-5. LULUCF emissions include
the ChU and N20 emissions reported for peatlands remaining peatlands, forest fires, drained organic soils, grassland fires, and
coastal wetlands remaining coastal wetlands; ChU emissions from land converted to coastal wetlands; and N20 emissions
from forest soils and settlement soils.
dThe LULUCF Sector Net Total is the net sum of allChUand N20 emissions to the atmosphere plus net carbon stock changes.
More details on the impact of recalculations on the LULUCF sector can be found in Table 9-5.
Notes: Net change in total emissions presented without LULUCF. Parentheses indicate negative values. Totals may not sum due
to independent rounding.
Table 9-5: Revisions to U.S. Greenhouse Gas Emissions and Removals (Net Flux) from
Land Use, Land-Use Change, and Forestry (MMT C02 Eq.)
Change in Land-Use Category
1990
2005
2019
2020
2021
2022
Average
Annual
Change
Forest Land Remaining Forest Land
(80.5)
(72.7)
(60.4)
(51.7)
(57.2)
(55.9)
(70.3)
Changes in Forest Carbon Stocks®
(80.1)
(74.0)
(64.0)
(64.4)
(63.7)
(55.4)
(71.0)
Non-C02 Emissions from Forest Firesb
(0.4)
1.3
3.6
12.7
6.5
(0.5)
0.7
N2O Emissions from Forest Soils0
NC
NC
NC
NC
NC
NC
NC
Non-C02 Emissions from Drained Organic Soilsd
NC
NC
NC
NC
NC
NC
NC
Land Converted to Forest Land
(3.4)
(3.4)
(3.6)
(3.6)
(3.5)
(3.5)
(3.5)
9-8 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Change in Land-Use Category
1990
2005
2019
2020
2021
2022
Average
Annual
Change
Changes in Forest Carbon Stockse
(3.4)
(3.4)
(3.6)
(3.6)
(3.5)
(3.5)
(3.5)
Cropland Remaining Cropland
6.1
0.6
0.1
0.1
0.1
0.1
1.4
Changes in Mineral and Organic Soil Carbon
Stocks
6.1
0.6
0.1
0.1
0.1
0.1
1.4
Land Converted to Cropland
3.1
1.0
0.1
(0.1)
+
(0.1)
0.5
Changes in all Ecosystem Carbon Stocks'
3.1
1.0
0.1
(0.1)
+
(0.1)
0.5
Grassland Remaining Grassland
(0.4)
(0.4)
(0.3)
(0.3)
(0.3)
(0.3)
(0.4)
Changes in Mineral and Organic Soil Carbon
Stocks
(0.4)
(0.4)
(0.3)
(0.3)
(0.3)
(0.3)
(0.4)
Non-C02 Emissions from Grassland Firess
NC
NC
NC
NC
NC
NC
NC
Land Converted to Grassland
0.3
0.2
(4.5)
(4.6)
(4.7)
(4.7)
(4.0)
Changes in all Ecosystem Carbon Stocks'
0.3
0.2
(4.5)
(4.6)
(4.7)
(4.7)
(4.0)
Wetlands Remaining Wetlands
1.7
1.5
1.6
1.6
1.6
1.6
1.6
Changes in Organic Soil Carbon Stocks in
Peatlands
NC
NC
NC
NC
(+)
+
+
Changes in Biomass, DOM, and Soil Carbon
Stocks in Coastal Wetlands
NC
NC
NC
NC
NC
NC
NC
CH4 Emissions from Coastal Wetlands Remaining
Coastal Wetlands
NC
NC
NC
NC
NC
NC
NC
N2O Emissions from Coastal Wetlands
Remaining Coastal Wetlands
NC
NC
NC
NC
NC
NC
NC
Non-C02 Emissions from Peatlands Remaining
Peatlands
NC
NC
NC
NC
(+)
+
+
CH4 Emissions from Flooded Land Remaining
Flooded Land
1.7
1.5
1.6
1.6
1.6
1.6
1.6
Land Converted to Wetlands
(0.4)
0.1
(+)
(+)
(+)
(+)
(0.1)
Changes in Biomass, DOM, and Soil Carbon
Stocks in Land Converted to Coastal Wetlands
NC
NC
NC
NC
NC
NC
NC
CH4 Emissions from Land Converted to Coastal
Wetlands
NC
NC
NC
NC
NC
NC
NC
Changes in Land Converted to Flooded Land
(0.2)
0.1
(+)
(+)
(+)
(+)
(0.1)
CH4 Emissions from Land Converted to Flooded
Land
(0.2)
+
(+)
(+)
(+)
(+)
(+)
Settlements Remaining Settlements
0.1
+
0.1
0.1
0.3
0.1
0.1
Changes in Organic Soil Carbon Stocks
NC
NC
NC
NC
0.1
0.6
+
Changes in Settlement Tree Carbon Stocks
0.1
+
0.1
0.1
0.1
0.1
0.1
Changes in Yard Trimming and Food Scrap
Carbon Stocks in Landfills
NC
NC
NC
NC
NC
(0.5)
(+)
N2O Emissions from Settlement Soilsh
NC
NC
NC
(+)
(+)
(+)
(+)
Land Converted to Settlements
12.3
11.9
11.2
11.4
11.5
11.6
12.2
Changes in all Ecosystem Carbon Stocks'
12.3
11.9
11.2
11.4
11.5
11.6
12.2
Change in LULUCF Total Net Flux1
(62.3)
(64.1)
(61.0)
(61.4)
(60.4)
(52.1)
(64.7)
Change in LULUCF Emissions1
1.1
2.9
5.1
14.3
8.1
1.0
2.2
Change in LULUCF Sector Net Totalk
(61.2)
(61.2)
(55.8)
(47.2)
(52.3)
(51.1)
(62.5)
Percent Change in LULUCF Sector Net Total
-6.3%
-6.7%
-6.5%
-5.2%
-5.7%
-6.7%
-6.7%
NC (No Change)
Recalculations and Improvements 9-9
-------
+ Absolute value does not exceed 0.05 MMT C02 Eq. or 0.05 percent.
a Includes the net changes to carbon stocks stored in all forest ecosystem pools and harvested wood products.
b Estimates include ChUand N20 emissions from fires on both forest land remaining forest land and land converted to forest
land.
c Estimates include N20 emissions from N fertilizer additions on both forest land remaining forest land and land
converted to forest land.
d Estimates include ChUand N20 emissions from drained organic soils on both forest land remaining forest land and land
converted to forest land.
e Includes the net changes to carbon stocks stored in all forest ecosystem pools.
f Includes changes in mineral and organic soil carbon stocks for all land use conversions to cropland, grassland, and
settlements, respectively. Also includes aboveground/belowground biomass, dead wood, and litter carbon stock changes for
conversion of forest land to cropland, grassland, and settlements, respectively.
g Estimates include ChUand N20 emissions from fires on both grassland remaining grassland and land converted to grassland.
h Estimates include N20 emissions from N fertilizer additions on both settlements remaining settlements and land
converted to settlements because it is not possible to separate the activity data at this time.
LULUCF Carbon Stock Change includes any C stock gains and losses from all land use and land use conversion categories.
' LULUCF Emissions include the ChUand N20 emissions reported for peatlands remaining peatlands, forest fires, drained organic
soils, grassland fires, and coastal wetlands remaining coastal wetlands; ChU emissions from land converted to coastal
wetlands; and N20 emissions from forest soils and settlement soils.
kThe LULUCF Sector Net Total is the net sum of all LULUCF ChUand N20 emissions to the atmosphere plus LULUCF net carbon
stock changes.
Notes: Parentheses indicate negative values. Totals may not sum due to independent rounding.
9-10 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
10 References and Abbreviations
Executive Summary
BEA (2024) 2023 Comprehensive Revision of the National Income and Product Accounts: Current-
dollar and "real" GDP, 1929-2023. Bureau of Economic Analysis (BEA), U.S. Department of
Commerce, Washington, D.C. Available online at: http://www.bea.goV/national/index.htm#gdp.
EIA (2025) Monthly Energy Review, January 2025. Energy Information Administration, U.S.
Department of Energy, Washington, D.C. DOE/EIA-0035(2025/01).
US EPA (2019). Greenhouse Gas Reporting Program (GHGRP), Electronic Greenhouse Gas
Reporting Tool Frequently Asked Questions Q409. U.S. Environmental Protection Agency,
Washington, D.C. Available online at:
https://ccdsupport.com/confluence/pages/viewpage.actionPpagelch322699300.
IPCC (2021) Climate Change 2021: The Physical Science Basis. Contribution of Working Group I to
the Sixth Assessment Report of the Intergovernmental Panel on Climate Change. [Masson-
Delmotte, V., P. Zhai, A. Pirani, S.L. Connors, C. Pean, S. Berger, N. Caud, Y. Chen, L. Goldfarb,
M.I. Gomis, M. Huang, K. Leitzell, E. Lonnoy, J.B.R. Matthews, T.K. Maycock, T. Waterfield, O.
Yelekgi, R. Yu, and B. Zhou (eds.)]. Cambridge University Press, Cambridge, United Kingdom and
NewYork, NY, USA, 2391 pp. doi:10.1017/9781009157896.
IPCC (2019) 2079 Refinement to the 2006 IPCC Guidelines for National Greenhouse Gas
Inventories. Calvo Buendia, E., Tanabe, K., Kranjc, A., Baasansuren, J., Fukuda, M., Ngarize, S.,
Osako, A., Pyrozhenko, Y., Shermanau, P. and Federici, S. (eds). Published: IPCC, Switzerland.
IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to
the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. [Stocker, T.F.,
D. Qin, G.-K., Plattner, M. Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M.
Midgley (eds.)]. Cambridge University Press, Cambridge, United Kingdom and NewYork, NY,
USA, 1535 pp.
IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to
the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. [S. Solomon,
D. Qin, M. Manning, Z. Chen, M. Marquis, K.B. Averyt, M. Tignorand H.L. Miller (eds.)].
Cambridge University Press. Cambridge, United Kingdom 996 pp.
References and Abbreviations 10-1
-------
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change.
[H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa,
Japan.
National Academies of Sciences, Engineering, and Medicine (2018) Improving characterization of
anthropogenic methane emissions in the United States. Washington, DC: The National
Academies Press. Available online at: https://doi.org/10.17226/24987.
National Research Council (2010) Verifying greenhouse gas emissions: methods to support
international climate agreements. Washington, DC: The National Academies Press. Available
online at: https://doi.org/10.17226/12883.
NOAA/ESRL (2025a) Trends in Atmospheric Carbon Dioxide. Available online at:
https://gml.noaa.gov/ccgg/trends/. January 13, 2025.
NOAA/ESRL (2025b) Trends in Atmospheric Methane. Available online at:
https://gml.noaa.gov/ccgg/trends ch4/. January 13, 2025.
NOAA/ESRL (2025c) Trends in Atmospheric Nitrous Oxide. Available online at:
https://gml.noaa.gov/ccgg/trends n2o/. January 13, 2025.
UNFCCC (2014) Report of the Conference of the Parties on its Nineteenth Session, Held in Warsaw
from 11 to 23 November 2013. (FCCC/CP/2013/10/Add.3). January 31, 2014. Available online at:
http://unfccc.int/resource/docs/2013/cop19/eng/10a03.pdf.
U.S. Census Bureau (2025) U.S. Census Bureau International Database (IDB). Available online at:
https://www.census.gov/programs-surveys/international-programs.html.
Introduction
IPCC (2021) Climate Change 2021: The Physical Science Basis. Contribution of Working Group I to
the Sixth Assessment Report of the Intergovernmental Panel on Climate Change. [Masson-
Delmotte, V., P. Zhai, A. Pirani, S.L. Connors, C. Pean, S. Berger, N. Caud, Y. Chen, L. Goldfarb,
M.I. Gomis, M. Huang, K. Leitzell, E. Lonnoy, J.B.R. Matthews, T.K. Maycock, T. Waterfield, O.
Yelekgi, R. Yu, and B. Zhou (eds.)]. Cambridge University Press, Cambridge, United Kingdom and
NewYork, NY, USA, 2391 pp. doi:10.1017/9781009157896.
IPCC (2014) Climate Change 2014: Mitigation of Climate Change. Contribution of Working Group III
to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change [Edenhofer,
O., R. Pichs-Madruga, Y. Sokona, J. Minx, E. Farahani, S. Kadner, K. Seyboth, A. Adler, I. Baum, S.
Brunner, P. Eickemeier, B. Kriemann, J. Savolainen, S. Schlomer, C. von Stechow, and T. Zwickel
(eds.)]. Cambridge University Press, Cambridge, United Kingdom and NewYork, NY, USA, 1435
pp.
IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to
the Fifth Assessment Report of the Intergovernmental Panel on Climate Change [Stocker, T.F.,
D. Qin, G.-K. Plattner, M. Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M.
10-2 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Midgley (eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY,
USA, 1535 pp.
IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to
the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. [S. Solomon,
D. Qin, M. Manning, Z. Chen, M. Marquis, K.B. Averyt, M. Tignorand H.L. Miller (eds.)].
Cambridge University Press. Cambridge, United Kingdom 996 pp.
IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change.
[H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa,
Japan.
IPCC (2001) Climate Change 2001: The Scientific Basis. Intergovernmental Panel on Climate
Change. [J.T. Houghton, Y. Ding, D.J. Griggs, M. Noguer, P.J. van der Linden, X. Dai, C.A. Johnson,
and K. Maskell (eds.)]. Cambridge University Press. Cambridge, United Kingdom.
IPCC/TEAP (2005) Special Report: Safeguarding the Ozone Layer and the Global Climate System,
Chapter 4: Refrigeration. 2005. Available online at:
https://www.ipcc.ch/site/assets/uploads/2018/03/sroc04-1 .pdf.
Meinhausen, M., et al. (2017) Historical Greenhouse Gas Concentrations for Climate Modeling
(CMIP6). Available online at: https://gmd.copernicus.Org/articles/10/2057/2017/.
NOAA (2017) Vital Signs of the Planet. Available online at: http://climate.nasa.gov/causes/.
Accessed on 9 January 2017.
NOAA/ESRL (2025a) Trends in Atmospheric Carbon Dioxide. Available online at:
https://gml.noaa.gov/ccgg/trends/gr.html. March 4, 2025.
NOAA/ESRL (2025b) Trends in Atmospheric Methane. Available online at:
https://gml.noaa.gov/ccgg/trends ch4/. March 4, 2025.
NOAA/ESRL (2025c) Trends in Atmospheric Nitrous Oxide. Available online at:
https://gml.noaa.gov/ccgg/trends n2o/. March 4, 2025.
NOAA/ESRL (2025d) Trends in Atmospheric Sulfur Hexafluoride. Available online at:
https://gml.noaa.gov/ccgg/trends sf6/. March 4, 2025.
Rigby, M., et al. (2010) History of Atmospheric SF6 from 1973 to 2008. Available online at:
www.atmos-chem-phvs.net/10/10305/2010. January 13, 2025.
UNEP/WMO (1999) Information Unit on Climate Change. Framework Convention on Climate
Change. Available online at: http://unfccc.int.
UNFCCC (2014) Report of the Conference of the Parties on its nineteenth session, held in Warsaw
from 11 to 23 November 2013. (FCCC/CP/2013/10/Add.3). January 31, 2014. Available online
at: http://unfccc.int/resource/docs/2013/cop19/eng/10a03.pdf.
UNFCCC (2023) Report of the Conference of the Parties on its twenty-seventh session, held in
Sharm el-Sheikh from 6 to 20 November 2022. (FCCC/CP/2022/10/Add.1). March 17, 2023.
Available online at: https://unfccc.int/sites/default/files/resource/cp2022 10a01 E.pdf
References and Abbreviations 10-3
-------
USGCRP (2017) Climate Science Special Report: Fourth National Climate Assessment, Volume I.
[Wuebbles, D.J., D.W. Fahey, K.A. Hibbard, D.J. Dokken, B.C. Stewart, andT.K. Maycock (eds.)].
U.S. Global Change Research Program, Washington, DC, USA, 470 pp, doi: 10.7930/J0J964J6.
Available online at: https://science2017.globalchange.gov/.
WMO/UNEP (2018) Scientific Assessment of Ozone Depletion: 2018. Available online at:
https://csl.noaa.gov/assessments/ozone/2018.
Trends
BEA (2024) 2023 Comprehensive Revision of the National Income and Product Accounts: Current-
dollar and "real" GDP, 1929-2023. Bureau of Economic Analysis (BEA), U.S. Department of
Commerce, Washington, D.C. Available online at: http://www.bea.goV/national/index.htm#gdp.
EIA (2025) Monthly Energy Review, January 2025. Energy Information Administration, U.S.
Department of Energy, Washington, D.C. DOE/EIA-0035(2025/01).
EIA (2020) Fuel Oil and Kerosene Sales. Energy Information Administration, U.S. Department of
Energy. Washington, D.C. Available online at: http://www.eia.gov/petroleum/fueloilkerosene.
EPA (2024) "Criteria pollutants National Tier 1 for 1970 - 2023." National Emissions Inventory (NEI)
Air Pollutant Emissions Trends Data. Office of Air Quality Planning and Standards, February
2024. Available online at: https://www.epa.gov/air-emissions-inventories/air-pollutant-
emissions-trends-data.
EPA (2023) 2023 EPA Automotive Trends Report. Office of Transportation and Air Quality, U.S.
Environmental Protection Agency. Available online at: https://www.epa.gov/automotive-
trends/download-automotive-trends-report.
IPCC (2021) Climate Change 2021: The Physical Science Basis. Contribution of Working Group I to
the Sixth Assessment Report of the Intergovernmental Panel on Climate Change. [Masson-
Delmotte, V., P. Zhai, A. Pirani, S.L. Connors, C. Pean, S. Berger, N. Caud, Y. Chen, L. Goldfarb,
M.I. Gomis, M. Huang, K. Leitzell, E. Lonnoy, J.B.R. Matthews, T.K. Maycock, T. Waterfield, O.
Yelekgi, R. Yu, and B. Zhou (eds.)]. Cambridge University Press, Cambridge, United Kingdom and
NewYork, NY, USA, 2391 pp. doi:10.1017/9781009157896.
IPCC (2019) 2079 Refinement to the 2006 IPCC Guidelines for National Greenhouse Gas
Inventories. Calvo Buendia, E., Tanabe, K., Kranjc, A., Baasansuren, J., Fukuda, M., Ngarize, S.,
Osako, A., Pyrozhenko, Y., Shermanau, P. and Federici, S. (eds). Published: IPCC, Switzerland.
IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to
the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. [Stocker, T.F.,
D. Qin, G.-K., Plattner, M. Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M.
Midgley (eds.)]. Cambridge University Press, Cambridge, United Kingdom and NewYork, NY,
USA, 1535 pp.
IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to
the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. [S. Solomon,
10-4 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
D. Qin, M. Manning, Z. Chen, M. Marquis, K.B. Averyt, M. Tignorand H.L. Miller (eds.)].
Cambridge University Press. Cambridge, United Kingdom 996 pp.
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change.
[H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa,
Japan.
U.S. Census Bureau (2025) U.S. Census Bureau International Database (IDB). Available online at:
https://www.census.gov/programs-surveys/international-programs.html.
U.S. Department of Agriculture, National Agricultural Statistics Service (USDA/NASS) (2024) Farm
Production Expenditures Annual Summary. National Agricultural Statistics Service, U.S.
Department of Agriculture, Washington DC. Available online at:
https://usda .libra ry.cornell.edu/concern/publications/qz20ss48r?locale=en.
Energy
Fossil Fuel Combustion
AAR (2008 through 2023) Railroad Facts. Policy and Economics Department, Association of
American Railroads, Washington, D.C. Private communication with Dan Keen.
AISI (2004 through 2021) Annual Statistical Report, American Iron and Steel Institute, Washington,
D.C.
APTA (2007 through 2023) Public Transportation Fact Book. American Public Transportation
Association, Washington, D.C. Available online at:
http://www.apta.com/resources/statistics/Pages/transitstats.aspx.
APTA (2006) Commuter Rail National Totals. American Public Transportation Association,
Washington, D.C.
BEA (2024) 2022 Comprehensive Revision of the National Income and Product Accounts: Current-
dollar and "real" GDP, 1929-2023. Bureau of Economic Analysis (BEA), U.S. Department of
Commerce, Washington, D.C. Available online at: http://www.bea.goV/national/index.htm#gdp.
BEA (1991 through 2015) Unpublished BE-36 survey data. Bureau of Economic Analysis, U.S.
Department of Commerce. Washington, D.C.
Benson, D. (2002 through 2004) Unpublished data. Upper Great Plains Transportation Institute,
North Dakota State University and American Short Line & Regional Railroad Association.
Browning (2022a) Addressing the Time Series Inconsistency in FHWA Data. Memorandum from ICF
to Sarah Roberts, Office of Transportation and Air Quality, U.S. Environmental Protection
Agency. September 2022.
References and Abbreviations 10-5
-------
Browning (2022b) Updated Methodology for Estimating CH4 and N20 Emissions from Highway
Vehicle Alternative Fuel Vehicles. Memorandum from ICF to Sarah Roberts, Office of
Transportation and Air Quality, U.S. Environmental Protection Agency. November 2022.
Browning, L. (2020) GHG Inventory EF Development Using Certification Data. Technical Memo,
September 2020.
Browning, L. (2019) Updated On-highway CH4and N20 Emission Factors for GHG Inventory.
Memorandum from ICF to Sarah Roberts, Office of Transportation and Air Quality, U.S.
Environmental Protection Agency. September 2019.
Browning, L. (2018a) Updated Methodology for Estimating Electricity Use from Highway Plug-In
Electric Vehicles. Technical Memo, October 2018.
Browning, L. (2018b) Updated Non-Highway CH4 and N20 Emission Factors for U.S. GHG Inventory.
Technical Memo, November 2018.
Browning, L. (2017) Updated Methodology for Estimating CH4 and N20 Emissions from Highway
Vehicle Alternative Fuel Vehicles. Technical Memo, October 2017.
Coffeyville Resources Nitrogen Fertilizers (2012) Nitrogen Fertilizer Operations. Available online at:
http://coffeyvillegroup.com/NitrogenFertilizerOperations/index.html.
Dakota Gasification Company (2006) C02 Pipeline Route and Designation Information. Bismarck,
ND.
DHS (2008) Email Communication. Elissa Kay, Department of Homeland Security and Joe Aamidor,
ICF International. January 11, 2008.
DLA Energy (2025) Unpublished data from the Fuels Automated System (FAS). Defense Logistics
Agency Energy, U.S. Department of Defense. Washington, D.C.
DOC (1991 through 2024) Unpublished Report of Bunker Fuel Oil Laden on Vessels Cleared for
Foreign Countries. Form-563. Foreign Trade Division, Bureau of the Census, U.S. Department of
Commerce. Washington, D.C.
DOE (1993 through 2022) Transportation Energy Data Book. Edition 40. Office of Transportation
Technologies, CenterforTransportation Analysis, Energy Division, Oak Ridge National
Laboratory. ORNL-6978. Personal Communication between Stacy Davis (DOE) and Deep Shah
(ICF) for sharing selected tables from the pre-release version.
DOE (2012)2070 Worldwide Gasification Database. National Energy Technology Laboratory and
Gasification Technologies Council. Available online at:
http://www.netl.doe.gov/technologies/coalpower/gasification/worlddatabase/index.html.
Accessed on 15 March 2012.
DOT (1991 through 2025) Airline Fuel Cost and Consumption. U.S. Department of Transportation,
Bureau of Transportation Statistics, Washington, D.C. DAI-10. Available online at:
http://www.transtats.bts.gov/fuel.asp.
Eastman Gasification Services Company (2011) Project Data on Eastman Chemical Company's
Chemicals-from-Coal Complex in Kingsport, TN.
10-6 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
EIA (2025a) Monthly Energy Review, January 2025, Energy Information Administration, U.S.
Department of Energy, Washington, DC. DOE/EIA-0035 (2025/01).
EIA (2025b) Preliminary Monthly Electric Generator Inventory, Energy Information Administration,
U.S. Department of Energy, Washington, DC. based on Form EIA-860M as a supplement to Form
EIA-860 (2025/01).
EIA (2024a) International Energy Statistics 1980-2023. Energy Information Administration, U.S.
Department of Energy. Washington, D.C. Available online at:
https://www.eia.gov/beta/international/.
EIA (2024b) Quarterly Coal Report: January - March 2024 Energy Information Administration, U.S.
Department of Energy. Washington, D.C. DOE/EIA-0121.
EIA (2024c) Natural Gas Annual 2023. Energy Information Administration, U.S. Department of
Energy. Washington, D.C. DOE/EIA-0131(20).
EIA(2024d). Petroleum Supply Annual 2023. Energy Information Administration, U.S. Department of
Energy. Washington, D.C. Available online at: https://www.eia.gov/petroleum/.
EIA (2024e) Form EIA-923 detailed data with previous form data (EIA-906/920), Energy Information
Administration, U.S. Department of Energy. Washington, DC. DOE/EIA. September 2024.
EIA (2024f) Glossary. Energy Information Administration, U.S. Department of Energy, Washington,
D.C. Available online at: https://www.eia.gov/tools/glossary/?id=electricity.
EIA (2024g) "U.S. Energy-Related Carbon Dioxide Emissions, 2023." Energy Information
Administration, U.S. Department of Energy. Washington, D.C. Available online at:
https://www.eia.gov/environment/emissions/carbon/.
EIA (2023a) Annual Coal Report2022. Energy Information Administration, U.S. Department of
Energy. Washington, D.C. DOE/EIA-0584.
EIA (2023b) "Energy use in homes." Use of energy explained. Available online at:
https://www.eia.gov/energyexplained/use-of-energy/homes.php.
EIA (2022a) "Residential and commercial natural gas prices reach multiyear highs in 2022." Energy
Information Administration, U.S. Department of Energy. Washington, D.C. Available online at:
https://www.eia.gov/todayinenergy/detail.php?id=54119.
EIA (2022b) "U.S. natural gas prices spiked in February 2021, then generally increased through
October." Energy Information Administration, U.S. Department of Energy. Washington, D.C.
Available online at: https://www.eia.gov/todayinenergy/detail.php?id=50778.
EIA (2020b) "Natural gas prices, production, consumption, and exports increased in 2019." Today in
Energy. Available online at: https://www.eia.gov/todayinenergy/detail.php?id=37892.
EIA (2018) "Both natural gas supply and demand have increased fromyear-ago levels." Today in
Energy. Available online at: https://www.eia.gov/todayinenergy/detail.php?id=37193.
References and Abbreviations 10-7
-------
EIA (2009a) Emissions of Greenhouse Gases in the United States 2008, Draft Report. Office of
Integrated Analysis and Forecasting, Energy Information Administration, U.S. Department of
Energy. Washington, D.C. DOE-EIA-0573 (2009).
EIA (2009b) Manufacturing Consumption of Energy (MECS) 2006. U.S. Department of Energy, Energy
Information Administration, Washington, D.C. Released July 2009.
EIA (2008) Historical Natural Gas Annual, 1930-2008. Energy Information Administration, U.S.
Department of Energy. Washington, D.C.
EIA (2007) Personal Communication. Joel Lou, Energy Information Administration and Aaron
Beaudette, ICF International. Residual and Distillate Fuel Oil Consumption for Vessel Bunkering
(Both International and Domestic) for American Samoa, U.S. Pacific Islands, and Wake Island.
October 24, 2007.
EIA (2002) Alternative Fuels Data Tables. Energy Information Administration, U.S. Department of
Energy. Washington, D.C. Available online at: https://www.eia.gov/renewable/.
EIA (2001) U.S. Coal, Domestic and International Issues. Energy Information Administration, U.S.
Department of Energy. Washington, D.C. March 2001.
EIA (1990-2001) State Energy Data System. Energy Information Administration, U.S. Department of
Energy. Washington, D.C.
Environment and Climate Change Canada (2025) Personal Communication between Environment
and Climate Change Canada and Vincent Camobreco for imported C02. February 2025.
EPA (2024a) Acid Rain Program Dataset 1996-2022. Office of Air and Radiation, Office of
Atmospheric Programs, U.S. Environmental Protection Agency, Washington, D.C.
EPA (2024b) Motor Vehicle Emissions Simulator (MOVES5). Office of Transportation and Air Quality,
U.S. Environmental Protection Agency, Washington, D.C. Available online at:
https://www.epa.gov/moves.
EPA (2023a) The 2023 EPA Automotive Trends Report: Greenhouse Gas Emissions, Fuel Economy,
and Technology since 1975. Office of Transportation and Air Quality, U.S. Environmental
Protection Agency. Available online at: https://www.epa.gov/automotive-trends.
EPA (2021) The Emissions & Generation Resource Integrated Database (eGRID) 2019 Technical
Support Document. Clean Air Markets Division, Office of Atmospheric Programs, U.S.
Environmental Protection Agency, Washington, D.C. Available Online at:
https://www.epa.gov/sites/default/files/2021-02/documents/egrid2019 technical guide.pdf
EPA (2020) EPA Inventory of U.S. Greenhouse Gas Emissions and Sinks 1990-2019: Updated
Gasoline and Diesel Fuel C02 Emission Factors - Memo.
EPA (2010) Carbon Content Coefficients Developed for EPA's Mandatory Reporting Rule. Office of
Air and Radiation, Office of Atmospheric Programs, U.S. Environmental Protection Agency,
Washington, D.C.
10-8 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Erickson, T. (2003) Plains C02 Reduction (PCOR) Partnership. Presented at the Regional Carbon
Sequestration Partnership Meeting Pittsburgh, Pennsylvania, Energy and Environmental
Research Center, University of North Dakota. November 3, 2003.
FAA (2024) Personal Communication between FAA and John Steller, Mausami Desai, and Vincent
Camobreco for aviation emissions estimates from the Aviation Environmental Design Tool
(AEDT). March 2024.
FHWA (1996 through 2024) Highway Statistics. Federal Highway Administration, U.S. Department of
Transportation, Washington, D.C. Report FHWA-PL-96-023-annual. Available online at:
http://www.fhwa.dot.gov/policy/ohpi/hss/hsspubs.htm.
FHWA (2015) Off-Highway and Public-Use Gasoline Consumption Estimation Models Used in the
Federal Highway Administration, Publication Number FHWA-PL-17-012. Available online at:
https://www.fhwa.dot.gov/policyinformation/pubs/pl17012.pdf.
Fitzpatrick, E. (2002) The Weyburn Project: A Model for International Collaboration.
FRB (2023) Industrial Production and Capacity Utilization. Federal Reserve Statistical Release,
G.17, Federal Reserve Board. Available online at:
http://www.federalreserve.gov/releases/G17/table1 2.htm.
Gaffney, J. (2007) Email Communication. John Gaffney, American Public Transportation Association
and Joe Aamidor, ICF International. December 17, 2007.
IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to
the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. [Stocker, T.F.,
D. Qin, G.K. Plattner, M. Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley
(eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, 1535
pp.
IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to
the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. [S. Solomon,
D. Qin, M. Manning, Z. Chen, M. Marquis, K.B. Averyt, M. Tignorand H.L. Miller (eds.)].
Cambridge University Press. Cambridge, United Kingdom, 996 pp.
IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change.
[H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa,
Japan.Marland, G. and A. Pippin (1990) "United States Emissions of Carbon Dioxide to the
Earth's Atmosphere by Economic Activity." Energy Systems and Policy, 14(4):323.
Nadel, S., Elliot, N., Langer, T. (2015) Energy Efficiency in the United States: 35 Years and Counting.
American Council for Energy-Efficient Economy (ACEEE), pg. iv-vi, 1-5. Available online at:
https://www.aceee.org/sites/default/files/publications/researchreports/e1502.pdf
NREL (2023) "NREL Researchers Reveal How Buildings Across United States Do—and Could—Use
Energy." Available online at: https://www.nrel.gov/news/features/2Q23/nrel-researchers-reveal-
how-buildings-across-the-united-states-do-and-could-use-
References and Abbreviations 10-9
-------
energy. html#:~:text=Buildings%20are%20responsible%20for%2040. building%20stock%20is%
20also%20essential.
SAIC/EIA (2001) Monte Carlo Simulations of Uncertainty in U.S. Greenhouse Gas Emission
Estimates. Final Report. Prepared by Science Applications International Corporation (SAIC)for
Office of Integrated Analysis and Forecasting, Energy Information Administration, U.S.
Department of Energy. Washington, D.C. June 22, 2001.
U.S. Aluminum Association (USAA) (2008 through 2021) U.S. Primary Aluminum Production. U.S.
Aluminum Association, Washington, D.C.
USAF (1998) Fuel Logistics Planning. U.S. Air Force: AFPAM23-221. May 1, 1998.
U.S. Census Bureau (2001 through 2011) Current Industrial Reports Fertilizer Materials and Related
Products: Annual Summary. Available online at: https://www.census.gov/data/tables/time-
series/econ/cir/mq325b.html.
United States Geological Survey (USGS) (2020a) 2020 Mineral Commodity Summaries: Aluminum.
U.S. Geological Survey, Reston, VA.
USGS (2021 b) 2021 Mineral Commodity Summary: Titanium and Titanium Dioxide. U.S. Geological
Survey, Reston, VA.
USGS (2019) 2017Mineral Yearbook: Aluminum. U.S. Geological Survey, Reston, VA
USGS (2014 through 2021a) Mineral Industry Surveys: Silicon. U.S. Geological Survey, Reston, VA.
USGS (2014 through 2021 b) Mineral Commodity Summary, Lead. U.S. Geological Survey, Reston,
VA.
USGS (2014 through 2019) Minerals Yearbook: Nitrogen [Advance Release], Available online at:
http://minerals.usgs.gov/minerals/pubs/commodity/nitrogen/.
USGS (1991 through 2020) Minerals Yearbook - Iron and Steel Scrap. U.S. Geological Survey,
Reston, VA.
USGS (1991 through 2015a) Minerals Yearbook: Manufactured Abrasives Annual Report. U.S.
Geological Survey, Reston, VA. Available online at:
http://minerals.usgs.gov/minerals/pubs/commodity/abrasives/.
USGS (1991 through 2015b) Minerals Yearbook: Titanium. U.S. Geological Survey, Reston, VA.
USGS (1991 through 2015c) Minerals Yearbook: Silicon Annual Report. U.S. Geological Survey,
Reston, VA. Available online at: http://minerals.usgs.gov/minerals/pubs/commodity/silicon/.
USGS (1996 through 2013) Minerals Yearbook: Silicon. U.S. Geological Survey, Reston, VA.
USGS (1995 through 2013) Minerals Yearbook: Lead Annual Report. U.S. Geological Survey, Reston,
VA.
USGS (1995,1998, 2000, 2001, 2002, 2007) Minerals Yearbook: Aluminum Annual Report. U.S.
Geological Survey, Reston, VA.
10-10 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Stationary Combustion (excluding C02)
EIA (2025) Monthly Energy Review, January 2025. Energy Information Administration, U.S.
Department of Energy. Washington, D.C. DOE/EIA-0035(2025/01).
EIA (2024) International Energy Statistics 1980-2023. Energy Information Administration, U.S.
Department of Energy. Washington, D.C. Available online at:
https://www.eia.gov/international/data/world.
EPA (2024a) Acid Rain Program Dataset 1996-2024. Office of Air and Radiation, Office of
Atmospheric Programs, U.S. Environmental Protection Agency, Washington, D.C.
EPA (2024b) Motor Vehicle Emissions Simulator (MOVES5). Office of Transportation and Air Quality,
U.S. Environmental Protection Agency, Washington, D.C. Available online at:
https://www.epa.gov/moves.
EPA (1997) Compilation of Air Pollutant Emission Factors, AP-42. Office of Air Quality Planning and
Standards, U.S. Environmental Protection Agency. Research Triangle Park, NC. October 1997.
FHWA (1996 through 2024) Highway Statistics. Federal Highway Administration, U.S. Department of
Transportation, Washington, D.C. Report FHWA-PL-96-023-annual. Available online at:
http://www.fhwa.dot.gov/policy/ohpi/hss/hsspubs.htm.
IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to
the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. [Stocker, T.F.,
D. Qin, G.K. Plattner, M. Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley
(eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, 1535
pp.
IPCC (2007). Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to
the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. [S. Solomon,
D. Qin, M. Manning, Z. Chen, M. Marquis, K.B. Averyt, M. Tignorand H.L. Miller (eds.)].
Cambridge University Press. Cambridge, United Kingdom, 996 pp.
IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change.
[H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa,
Japan.SAIC/EIA (2001) Monte Carlo Simulations of Uncertainty in U.S. Greenhouse Gas
Emission Estimates. Final Report. Prepared by Science Applications International Corporation
(SAIC) for Office of Integrated Analysis and Forecasting, Energy Information Administration, U.S.
Department of Energy. Washington, D.C. June 22, 2001.
NESCAUM (2024). Criteria, Greenhouse Gas, and Hazardous Air Pollutant Emissions Factors from
Residential Cordwood and Pellet Stoves Using an Integrated Duty Cycle Test Protocol. Nora
Traviss, George Allen, and Mahdi Ahmadi. ACS ES&T Air 2024 1 (9), 1190-1202. DOI:
10.1021/acsestair.4c00135
References and Abbreviations 10-11
-------
Mobile Combustion (excluding C02)
AAR (2008 through 2024) Railroad Facts. Policy and Economics Department, Association of
American Railroads, Washington, D.C. Private communication with Dan Keen.
ANL (2023) The Greenhouse Gases, Regulated Emissions, and Energy Use in Transportation Model
(GREET2023). Argonne National Laboratory. October 2023. Available online at:
https://greet.es.anl.gov.
APTA (2007 through 2023) Public Transportation Fact Book. American Public Transportation
Association, Washington, D.C. Available online at:
http://www.apta.com/resources/statistics/Pages/transitstats.aspx.
APTA (2006) Commuter Rail National Totals. American Public Transportation Association,
Washington, D.C. Available online at: http://www.apta.com/research/stats/rail/crsum.cfm.
BEA (1991 through 2015) Unpublished BE-36 survey data. Bureau of Economic Analysis, U.S.
Department of Commerce. Washington, D.C.
Benson, D. (2002 through 2004) Personal communication. Unpublished data developed by the
Upper Great Plains Transportation Institute, North Dakota State University and American Short
Line & Regional Railroad Association.
Browning, L. (2024) Updated Methodology for Estimating VMT from Highway Electric Vehicles,
August 24, 2024
Browning (2022a) Addressing the Time Series Inconsistency in FHWA Data. Memorandum from ICF
to Sarah Roberts, Office of Transportation and Air Quality, U.S. Environmental Protection
Agency. September 2022.
Browning (2022b) Updated Methodology for Estimating CH4 and N20 Emissions from Highway
Vehicle Alternative Fuel Vehicles. Memorandum from ICF to Sarah Roberts, Office of
Transportation and Air Quality, U.S. Environmental Protection Agency. November 2022.
Browning (2020) GHG Inventory EF Development Using Certification Data. Memorandum from ICF
to Sarah Roberts, Office of Transportation and Air Quality, U.S. Environmental Protection
Agency. September 2020.
Browning, L. (2019) Updated On-highway CH4and N20 Emission Factors for GHG Inventory.
Memorandum from ICF to Sarah Roberts and Justine Geidosch, Office of Transportation and Air
Quality, U.S. Environmental Protection Agency. September 2019.
Browning, L. (2018) Updated Non-Highway CH4 and N20 Emission Factors for U.S. GHG Inventory.
Technical Memorandum from ICF International to Sarah Roberts and Justine Geidosch, Office
of Transportation and Air Quality, U.S. Environmental Protection Agency. November 2018.
Browning, L. (2005) Personal communication with Lou Browning, "Emission control technologies
for diesel highway vehicles specialist," ICF International.
BTS (2024) Amtrak Fuel Consumption and Travel. Bureau of Transportation Statistics, Washington,
DC. Available online at: https://www.bts.gov/content/amtrak-fuel-consumption-and-travel-1.
10-12 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
DLA Energy (2025) Unpublished data from the Defense Fuels Automated Management System
(DFAMS). Defense Energy Support Center, Defense Logistics Agency, U.S. Department of
Defense. Washington, D.C.
DOC (1991 through 2024) Unpublished Report of Bunker Fuel Oil Laden on Vessels Cleared for
Foreign Countries. Form-563. Foreign Trade Division, Bureau of the Census, U.S. Department of
Commerce. Washington, D.C.
DOE (1993 through 2023) Transportation Energy Data Book Edition 40. Office of Transportation
Technologies, CenterforTransportation Analysis, Energy Division, Oak Ridge National
Laboratory. Personal Communication between Stacy Davis (DOE) and Deep Shah (ICF) for
sharing selected tables from the pre-release version.
DOT (1991 through 2023) Airline Fuel Cost and Consumption. U.S. Department of Transportation,
Bureau of Transportation Statistics, Washington, D.C. DAI-10. Available online at:
http://www.transtats.bts.gov/fuel.asp.
EIA (2025) Monthly Energy Review, January 2025, Energy Information Administration, U.S.
Department of Energy, Washington, D.C. DOE/EIA-0035(2025/01).
EIA (2024) Natural Gas Monthly September 2024. Energy Information Administration, U.S.
Department of Energy, Washington, Available online at:
https://www.eia.gov/naturalgas/monthly/pdf/ngm all.pdf.
EIA (1991 through 2022) Fuel Oil and Kerosene Sales. Energy Information Administration, U.S.
Department of Energy. Washington, D.C. Available online at:
http://www.eia.gov/petroleum/fueloilkerosene.
EIA (2007) Personal Communication. Joel Lou, Energy Information Administration and Aaron
Beaudette, ICF International. Residual and Distillate Fuel Oil Consumption for Vessel Bunkering
(Both International and Domestic) for American Samoa, U.S. Pacific Islands, and Wake Island.
October 24, 2007.
EIA (2002) Alternative Fuels Data Tables. Energy Information Administration, U.S. Department of
Energy, Washington, D.C. Available online at: http://www.eia.doe.gov/fuelrenewable.html.
EPA (2024) Annual Certification Test Results Report. Office of Transportation and Air Quality, U.S.
Environmental Protection Agency. Available online at: https://www.epa.gov/compliance-and-
fuel-economy-data/annual-certification-test-data-vehicles-and-engines.
EPA (2024a) Motor Vehicle Emissions Simulator (MOVES5). Office of Transportation and Air Quality,
U.S. Environmental Protection Agency. Available online at: https://www.epa.gov/moves.
EPA (2024b) Confidential Engine Family Sales Data Submitted to EPA by Manufacturers. Office of
Transportation and Air Quality, U.S. Environmental Protection Agency.
EPA (2004) Mobile6.2 Vehicle Emission Modeling Software. Office of Mobile Sources, U.S.
Environmental Protection Agency, Ann Arbor, Michigan.
EPA (1999) Emission Facts: The History of Reducing Tailpipe Emissions. Office of Mobile Sources.
May 1999. EPA420-F-99-017. Available online at: https://www.epa.gov/nscep.
References and Abbreviations 10-13
-------
EPA (1998) Emissions of Nitrous Oxide from Highway Mobile Sources: Comments on the Draft
Inventory of U.S. Greenhouse Gas Emissions and Sinks, 1990-1996. Office of Mobile Sources,
Assessment and Modeling Division, U.S. Environmental Protection Agency. August 1998.
EPA420-R-98-009.
EPA (1994a) Automobile Emissions: An Overview. Office of Mobile Sources. August 1994. EPA 400-
F-92-007. Available online at: https://www.epa.gov/nscep.
EPA (1994b) Milestones in Auto Emissions Control. Office of Mobile Sources. August 1994. EPA400-
F-92-014. Available online at: https://www.epa.gov/nscep.
Esser, C. (2003 through 2004) Personal Communication with Charles Esser, Residual and Distillate
Fuel Oil Consumption for Vessel Bunkering (Both International and Domestic) for American
Samoa, U.S. Pacific Islands, and Wake Island.
FAA (2022) Personal Communication between FAA and John Steller, Mausami Desai and Vincent
Camobreco for aviation emission estimates from the Aviation Environmental Design Tool
(AEDT). March 2022.
FHWA (1996 through 2024) Highway Statistics. Federal Highway Administration, U.S. Department of
Transportation, Washington, D.C. Report FHWA-PL-96-023-annual.
FTA (2024) National Transit Database "Fuel and Energy by Mode and TOS" table. Available online at:
https://data.transportation.gov/.
Gaffney, J. (2007) Email Communication. John Gaffney, American Public Transportation Association
and Joe Aamidor, ICF International. December 17, 2007.
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change.
[H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa,
Japan.
ICF (2006) Revised Gasoline Vehicle EFs for LEV and Tier 2 Emission Levels. Memorandum from ICF
International to John Davies, Office of Transportation and Air Quality, U.S. Environmental
Protection Agency. November 2006.
ICF (2004) Update of Methane and Nitrous Oxide Emission Factors for On-Highway Vehicles. Final
Report to U.S. Environmental Protection Agency. February 2004.
Raillnc (2014 through 2024) Raillnc Short line and Regional Traffic Index. Carloads Originated Year-
to-Date. November 2023. Available online at: https://public.railinc.com/.
Whorton, D. (2006 through 2014) Personal communication, Class II and III Rail energy
consumption, American Short Line and Regional Railroad Association.
Carbon Emitted from Non-Energy Uses of Fossil Fuels
ACC (2024a) "U.S. Resin Production & Sales 2023 vs. 2022." Available online at:
https://www.americanchemistry.com/chemistry-in-america/data-industry-statistics/statistics-
10-14 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
on-the-plastic-resins-industry/resources/pips-resin-sales-and-production-cy-figures-2023-vs-
?0??-
ACC (2024b) Guide to the Business of Chemistry, 2024, American Chemistry Council.
ACC (2023) "U.S. Resin Production & Sales 2022 vs. 2021." Available online at:
https://www.americanchemistry.com/chemistrv-in-america/data-industry-statistics/statistics-
on-the-plastic-resins-industry/resources/pips-resin-sales-and-production-cy-figures-2022-vs-
?0?1-
ACC (2022) "U.S. Resin Production & Sales 2021 vs. 2020." Available online at:
https://www.americanchemistry.com/chemistry-in-america/data-industry-statistics/statistics-
on-the-plastic-resins-industrv.
ACC (2021) "U.S. Resin Production & Sales 2020 vs. 2019." Available online at:
https://www.americanchemistry.com/chemistry-in-america/chemistry-in-everydav-
products/plastics.
ACC (2020) "U.S. Resin Production & Sales 2019 vs. 2018." Available online at:
https://www.americanchemistry.com/chemistry-in-america/chemistry-in-everyday-
products/plastics.
ACC (2019) "U.S. Resin Production & Sales 2018 vs. 2017." Available online at:
https://www.americanchemistry.com/chemistry-in-america/chemistry-in-everydav-
products/plastics.
ACC (2018) "U.S. Resin Production & Sales 2017 vs. 2016." Available online at:
https://www.americanchemistry.com/chemistry-in-america/chemistry-in-everydav-
products/plastics.
ACC (2017) "U.S. Resin Production & Sales 2016 vs. 2015."
ACC (2016) "U.S. Resin Production & Sales 2015 vs. 2014."
ACC (2015) "PIPS Year-End Resin Statistics for 2014 vs. 2013: Production, Sales and Captive Use."
Available online at: https://www.americanchemistry.com/chemistry-in-america/data-industry-
statistics/statistics-on-the-plastic-resins-industrv/resin-report-subscriptions.
ACC (2014) "U.S. Resin Production &Sales: 2013 vs. 2012," American Chemistry Council. Available
online at: http://www.americanchemistry.com/Jobs/EconomicStatistics/Plastics-
Statistics/Production-and-Sales-Data-bv-Resin.pdf.
ACC (2013) "U.S. Resin Production & Sales: 2012 vs. 2011," American Chemistry Council. Available
online at: http://www.americanchemistry.com/Jobs/EconomicStatistics/Plastics-
Statistics/Production-and-Sales-Data-by-Resin.pdf.
ACC (2003-2011) "PIPS Year-End Resin Statistics for 2010: Production, Sales and Captive Use."
Available online at: http://www.americanchemistry.com/Jobs/EconomicStatistics/Plastics-
Statistics/Production-and-Sales-Data-bv-Resin.pdf.
References and Abbreviations 10-15
-------
Bankof Canada (2024) Financial Markets Department Year Average of Exchange Rates. Available
online at: https://www.bankofcanada.ca/rates/exchange/annual-average-exchange-
rates/#download.
Bankof Canada (2023) Financial Markets Department Year Average of Exchange Rates. Available
online at: https://www.bankofcanada.ca/rates/exchange/annual-average-exchange-
rates/#download.
Bankof Canada (2022) Financial Markets Department Year Average of Exchange Rates. Available
online at: https://www.bankofcanada.ca/rates/exchange/annual-average-exchange-
rates/#download.
Bankof Canada (2021) Financial Markets Department Year Average of Exchange Rates. Available
online at: https://www.bankofcanada.ca/rates/exchange/annual-average-exchange-
rates/#download.
Bankof Canada (2020) Financial Markets Department Year Average of Exchange Rates. Available
online at: https://www.bankofcanada.ca/rates/exchange/annual-average-exchange-
rates/#download.
Bankof Canada (2019) Financial Markets Department Year Average of Exchange Rates. Available
online at: https://www.bankofcanada.ca/rates/exchange/annual-average-exchange-
rates/#download.
Bankof Canada (2018) Financial Markets Department Year Average of Exchange Rates. Available
online at: https://www.bankofcanada.ca/rates/exchange/annual-average-exchange-rates/.
Bankof Canada (2017) Financial Markets Department Year Average of Exchange Rates. Available
online at: https://www.bankofcanada.ca/rates/exchange/legacy-noon-and-closing-rates/.
Bankof Canada (2016) Financial Markets Department Year Average of Exchange Rates. Available
online at: https://www.bankofcanada.ca/rates/exchange/legacy-noon-and-closing-rates/.
Bankof Canada (2014) Financial Markets Department Year Average of Exchange Rates. Available
online at: https://www.bankofcanada.ca/rates/exchange/legacy-noon-and-closing-rates/.
Bankof Canada (2013) Financial Markets Department Year Average of Exchange Rates. Available
online at: https://www.bankofcanada.ca/rates/exchange/legacy-noon-and-closing-rates/.
Bankof Canada (2012) Financial Markets Department Year Average of Exchange Rates. Available
online at: https://www.bankofcanada.ca/rates/exchange/legacy-noon-and-closing-rates/.
CIAC (2024) 2024 Economic Review of Chemistry. Available online at:
https://canadianchemistry.ca/wp-content/uploads/2024/10/Economic-Review-of-
Chem sept2024 ENpdf.pdf.
EIA (2025) Monthly Energy Review, January 2025. Energy Information Administration, U.S.
Department of Energy, Washington, D.C. DOE/EIA-0035 (2025/01). Available online at:
https://www.eia.gov/totalenergy/data/monthly/pdf/mer.pdf.
10-16 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
EIA(2024) International Energy Statistics 1980-2023. Energy Information Administration, U.S.
Department of Energy. Washington, D.C. Available online at:
https://www.eia.gov/beta/international/.
EIA (2021) EIA Manufacturing Consumption of Energy (MECS) 2018. U.S. Department of Energy,
Energy Information Administration, Washington, D.C.
EIA (2020) Glossary. Energy Information Administration, U.S. Department of Energy, Washington,
D.C. Available online at: https://www.eia.gov/tools/glossary/index.php?id=N#nat Gas Liquids.
EIA (2019) Personal communication between EIA and ICF on November 11, 2019.
EIA (2017) EIA Manufacturing Consumption of Energy (MECS) 2014. U.S. Department of Energy,
Energy Information Administration, Washington, D.C.
EIA (2013) EIA Manufacturing Consumption of Energy (MECS) 2010. U.S. Department of Energy,
Energy Information Administration, Washington, D.C.
EIA (2010) EIA Manufacturing Consumption of Energy (MECS) 2006. U.S. Department of Energy,
Energy Information Administration, Washington, D.C.
EIA (2005) EIA Manufacturing Consumption of Energy (MECS) 2002. U.S. Department of Energy,
Energy Information Administration, Washington, D.C.
EIA (2001) EIA Manufacturing Consumption of Energy (MECS) 1998. U.S. Department of Energy,
Energy Information Administration, Washington, D.C.
EIA (1997) EIA Manufacturing Consumption of Energy (MECS) 1994. U.S. Department of Energy,
Energy Information Administration, Washington, D.C.
EIA (1994) EIA Manufacturing Consumption of Energy (MECS) 1991. U.S. Department of Energy,
Energy Information Administration, Washington, D.C.
EPA (2025) EPA's Emissions Inventory System (EIS) to National Inventory Report (NIR) Mapping file
EIS_NIR_mapping_2.26.25.xlsx. U.S. Environmental Protection Agency. Washington, D.C.
EPA (2024) Resource Conservation and Recovery Act (RCRA) Info, Biennial Report, GM Form
(Section 2- Onsite Management) and WR Form.
EPA (2021) Resource Conservation and Recovery Act (RCRA) Info, Biennial Report, GM Form
(Section 2- Onsite Management) and WR Form.
EPA (2019) Advancing Sustainable Materials Management: 2016 and 2017 Data Tables. Office of
Land and Emergency Management, U.S. Environmental Protection Agency. Washington, D.C.
Available online at: https://www.epa.gov/sites/production/files/2019-
11/documents/2016 and 2017 facts and figures data tables O.pdf.
EPA (2018a) Advancing Sustainable Materials Management: Facts and Figures 2015, Assessing
Trends in Material Generation, Recycling and Disposal in the United States. Washington, D.C.
EPA (2018b) Resource Conservation and Recovery Act (RCRA) Info, Biennial Report, GM Form
(Section 2- Onsite Management) and WR Form.
References and Abbreviations 10-17
-------
EPA (2017) EPA's Pesticides Industry Sales and Usage, 2008 - 2012 Market Estimates. Available
online at: https://www.epa.gov/sites/production/files/2017-01/documents/pesticides-industry-
sales-usage-2016 O.pdf. Accessed September 2017.
EPA (2016a) Advancing Sustainable Materials Management: 2014 Facts and Figures Fact Sheet.
Office of Solid Waste and Emergency Response, U.S. Environmental Protection Agency,
Washington, D.C. Available online at: https://www.epa.gov/sites/production/files/2016-
11/documents/2014 smmfactsheet 508.pdf.
EPA (2016b) Resource Conservation and Recovery Act (RCRA) Info, Biennial Report, GM Form
(Section 2- Onsite Management) and WR Form.
EPA (2015) Resource Conservation and Recovery Act (RCRA) Info, Biennial Report, GM Form
(Section 2- Onsite Management) and WR Form.
EPA (2014a) Municipal Solid Waste in the United States: 2012 Facts and Figures. Office of Solid
Waste and Emergency Response, U.S. Environmental Protection Agency, Washington, D.C.
Available online at: https://www.epa.gov/sites/default/files/2015-
09/documents/2012 msw dat tbls.pdf.
EPA (2014b) Chemical Data Access Tool (CDAT). U.S. Environmental Protection Agency, June 2014.
Available online at:
https://edg.epa. gov/metadata/catalog/search/resource/details.page?uuid=%7B2D73C764-
6919-404D-8C9B-61869B3330D6%7D. Accessed January 2015.
EPA (2013a) Municipal Solid Waste in the United States: 2011 Facts and Figures. Office of Solid
Waste and Emergency Response, U.S. Environmental Protection Agency, Washington, D.C.
Available online at: http://www.epa.gov/epaoswer/non-hw/muncpl/msw99.htm.
EPA (2013b) Resource Conservation and Recovery Act (RCRA) Info, Biennial Report, GM Form
(Section 2- Onsite Management) and WR Form.
EPA (2011) EPA's Pesticides Industry Sales and Usage, 2006 and 2007 Market Estimates. Available
online at: https://www.epa.gov/pesticides/pesticides-industry-sales-and-usage-2006-and-
2007-market-estimates. Accessed January 2012.
EPA (2009) Biennial Reporting System (BRS) Database. U.S. Environmental Protection Agency,
Envirofacts Warehouse. Washington, D.C. Available online at: https://www.epa.gov/enviro/br-
search. Data for 2001-2007 are current as of Sept. 9, 2009.
EPA (2004) EPA's Pesticides Industry Sales and Usage, 2000 and 2001 Market Estimates. Available
online at: https://nepis.epa.gov/Exe/ZyPURLcgi?Dockey=3000659P.TXT. Accessed September
2006.
EPA (2002) EPA's Pesticides Industry Sales and Usage, 1998 and 1999 Market Estimates, Table 3.6.
Available online at https://nepis.epa.gov/Exe/ZyPURL.cgi?Dockey=200001G5TXT Accessed
July 2003.
EPA (2001) AP 42, Volume I, Fifth Edition. Chapter 11: Mineral Products Industry. Available online at:
http://www.epa.gov/ttn/chief/ap42/ch11/index.html.
10-18 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
EPA (2000a) Biennial Reporting System (BRS). U.S. Environmental Protection Agency, Envirofacts
Warehouse. Washington, D.C. Available online at: https://www.epa.gov/enviro/br-search.
EPA (2000b) Toxics Release Inventory, 1998. U.S. Environmental Protection Agency, Office of
Environmental Information, Office of Information Analysis and Access, Washington, D.C.
Available online at: https://enviro.epa.gov/triexplorer/tri release.chemical.
EPA (1999) EPA's Pesticides Industry Sales and Usage, 1996-1997 Market Estimates. Available
online at: https://nepis.epa.gov/Exe/ZyPURL.cgi?Dockey=200001 IL.TXT.
EPA (1998) EPA's Pesticides Industry Sales and Usage, 1994-1995 Market Estimates. Available
online at: http://www.epa.gov/oppbead1/pestsales/95pestsales/market estimates1995.pdf.
FEB (2013) Fiber Economics Bureau, as cited in C&EN (2013) Lackluster Year for Chemical Output:
Production stayed flat or dipped in most world regions in 2012. Chemical &Engineering News,
American Chemical Society, 1 July. Available online at: http://www.cen-online.org.
FEB (2012) Fiber Economics Bureau, as cited in C&EN (2012) Too Quiet After the Storm: After a
rebound in 2010, chemical production hardly grew in 2011. Chemical & Engineering News,
American Chemical Society, 2 July. Available online at: http://www.cen-online.org.
FEB (2011) Fiber Economics Bureau, as cited in C&EN (2011) Output Ramps up in all Regions.
Chemical Engineering News, American Chemical Society, 4 July. Available online at:
http://www.cen-online.org.
FEB (2010) Fiber Economics Bureau, as cited in C&EN (2010) Output Declines in U.S., Europe.
Chemical & Engineering News, American Chemical Society, 6 July. Available online at:
http://www.cen-online.org.
FEB (2009) Fiber Economics Bureau, as cited in C&EN (2009) Chemical Output Slipped In Most
Regions Chemical & Engineering News, American Chemical Society, 6 July. Available online at:
http://www.cen-online.org.
FEB (2007) Fiber Economics Bureau, as cited in C&EN (2007) Gains in Chemical Output Continue.
Chemical & Engineering News, American Chemical Society. July 2, 2007. Available online at:
http://www.cen-online.org.
FEB (2005) Fiber Economics Bureau, as cited in C&EN (2005) Production: Growth in Most Regions
Chemical & Engineering News, American Chemical Society, 11 July. Available online at:
http://www.cen-online.org.
FEB (2003) Fiber Economics Bureau, as cited in C&EN (2003) Production Inches Up in Most
Countries, Chemical & Engineering News, American Chemical Society, 7 July. Available online
at: http://www.cen-online.org.
FEB (2001) Fiber Economics Bureau, as cited in ACS (2001) Production: slow gains in output of
chemicals and products lagged behind U.S. economy as a whole Chemical & Engineering
News, American Chemical Society, 25 June. Available online at: http://pubs.acs.org/cen.
References and Abbreviations 10-19
-------
Financial Planning Association (2006) Canada/US Cross-Border Tools: US/Canada Exchange Rates.
Available online at: http://www.fpanet.org/global/planners/US Canada ex rates.cfm.
Accessed on August 16, 2006.
Gosselin, Smith, and Hodge (1984) "Clinical Toxicology of Commercial Products." Fifth Edition,
Williams & Wilkins, Baltimore.
ICIS (2016) "Production issues force US melamine plant down" Available online at:
https://www.icis.com/resources/news/2016/05/Q3/9994556/production-issues-force-us-
melamine-plant-down/.
ICIS (2008) "Chemical profile: Melamine" Available online at:
https://www.icis.com/resources/news/2008/12/01/9174886/chemical-profile-melamine/.
Accessed November 2017.
IISRP (2003) "IISRP Forecasts Moderate Growth in North America to 2007" International Institute of
Synthetic Rubber Producers, Inc. New Release. Available online at:
http://www.iisrp.com/press-releases/2003-Press-Releases/IISRP-NA-Forecast-03-07.html.
IISRP (2000) "Synthetic Rubber Use Growth to Continue Through 2004, Says IISRP and RMA"
International Institute of Synthetic Rubber Producers press release.
INEGI (2006) Produccion bruta total de las unidades economicas manufactureras por Subsector,
Rama, Subrama y Clase de actividad. Available online at:
http://www.inegi.gob.mx/est/contenidos/espanol/proyectos/censos/ce2004/tb manufactures,
asp. Accessed on August 15, 2006.
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change.
[H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa,
Japan.
Marland, G., and R.M. Rotty (1984) "Carbon dioxide emissions from fossil fuels: A procedure for
estimation and results for 1950-1982," Tellus 36b:232-261.
NPRA (2002) North American Wax - A Report Card. Available online at:
http://www.npra.org/members/publications/papers/lubes/LW-02-126.pdf.
U.S. Census Bureau (2021) 2017 Economic Census. Available online at:
https://www.census.gov/data/tables/2017/econ/economic-census/naics-sector-31-33.html.
Accessed October 2021.
U.S. Census Bureau (2014) 2012 Economic Census. Available online at:
http://www.census.gov/econ/census/schedule/whats been released.html. Accessed
November 2014.
U.S. Census Bureau (2009) Soap and Other Detergent Manufacturing: 2007.
U.S. Census Bureau (2004) Soap and Other Detergent Manufacturing: 2002. Issued December
2004. EC02-311-325611 (RV). Available online at:
http://www.census.gov/prod/ec02/ec0231 i325611 .pdf.
10-20 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
U.S. Census Bureau (1999) Soap and Other Detergent Manufacturing: 1997. Available online at:
http://www.census.gov/epcd/www/ec97stat.htnn.
U.S. International Trade Commission (2024) "Interactive Tariff and Trade DataWeb: Quick Query."
Available online at: http://dataweb.usitc.gov/. Accessed September 2024.
USTMA (2024) "2023 End-of-Life Tire Management Report." U.S. Tire Manufacturers Association,
Washington, DC. October 2024. Available online at: https://www.ustires.org/system/files/2024-
10/USTMA%27s%202023%20End-of-Life%20Tire%20Management%20Report.pdf.
USTMA (2022) "2021 U.S. Scrap Tire Management Summary." U.S. Tire Manufacturers Association,
Washington, DC. October 2022. Available online at:
https://www.ustires.org/svstem/files/files/2024-
05/21 %20US%20Scrap%20Tire%20Management%20Report%20101722.pdf.
USTMA (2020) "2019 U.S. Scrap Tire Management Summary." U.S. Tire Manufacturers Association,
Washington, DC. October 2020. Available online at:
https://www.ustires.org/sites/default/files/2019%20USTMA%20Scrap%20Tire%20Management
%20Summary%2QReport.pdf.
USTMA (2018) "2017 U.S. Scrap Tire Management Summary." U.S. Tire Manufacturers Association,
Washington, DC. July 2018. Available online at: https://www.tyrepress.com/wp-
content/uploads/2018/07/USTMA scraptire summ 2017 07 11 2018.pdf.
USTMA (2016) "2015 U.S. Scrap Tire Management Summary." U.S. Tire Manufacturers Association.
August 2016. Available online at:
https://www.ustires.org/sites/default/files/MAR 028 USTMA.pdf.
USTMA (2014) "2013 U.S. Scrap Tire Management Summary." U.S. Tire Manufacturers Association.
November 2014. Available online at:
https://www.ustires.org/sites/default/files/MAR 027 USTMA.pdf.
USTMA (2013) "U.S. Scrap Tire Management Summary 2005-2009." U.S. Tire Manufacturers
Association. October 2011; Updated September 2013. Available online at:
https://www.ustires.org/sites/default/files/MAR 025 USTMA.pdf.
USTMA (2012) "Scrap Tire Markets: Facts and Figures - Scrap Tire Characteristics." U.S. Tire
Manufacturers Association. Accessed 18 on January 2012.
Incineration of Waste
ArSova, Ljupka, Rob van Haaren, Nora Goldstein, Scott M. Kaufman, and Nickolas J. Themelis
(2008) "16th Annual BioCycle Nationwide Survey: The State of Garbage in America" BioCycle,
JG Press, Emmaus, PA. December.
Bahor, B (2009) Covanta Energy's public review comments re: Draft Inventory of U.S. Greenhouse
Gas Emissions and Sinks: 1990-2007. Submitted via email on April 9, 2009 to Leif Hockstad,
U.S. EPA.
References and Abbreviations 10-21
-------
De Soete, G.G. (1993) "Nitrous Oxide from Combustion and Industry: Chemistry, Emissions and
Control." In A. R. Van Amstel, (ed.) Proc. of the International Workshop Methane and Nitrous
Oxide: Methods in National Emission Inventories and Options for Control, Amersfoort, NL.
February 3-5,1993.
Energy Recovery Council (2018) Energy Recovery Council. 2018 Directory of Waste to Energy
Facilities. Jed Michaels and Karunya Krishnan. October 2018. Available online at:
http://energyrecoverycouncil.org/wp-content/uploads/2019/10/ERC-2018-directory.pdf.
Energy Recovery Council (2009) "2007 Directory of Waste-to-Energy Plants in the United States."
Accessed on September 29, 2009.
EIA (2019) ElASt. Louis Federal Reserve's Economic Data (FRED) Consumer Price Index for All
Urban Consumers: Education and Communication (CPIEDUSL). Available online at:
https://www.eia.gov/opendata/excel/.
EIA (2017) MSWIncineration for Heating or Electrical Generation, December 2017, Energy
Information Administration, U.S. Department of Energy, Washington, DC. DOE/EIA-0035.
Available online at: https://www.eia.gov/opendata/?src=-f3.
EPA (2024) Greenhouse Gas Reporting Program (GHGRP). 2024 Envirofacts. Available online at:
https://ghgdata.epa.gov/ghgp/main.do.
EPA (2020a) Advancing Sustainable Materials Management: 2018 Data Tables. Office of Land and
Emergency Management, U.S. Environmental Protection Agency. Washington, D.C. Available
online at: https://www.epa.gov/sites/production/files/2020-
11/documents/2018 ff fact sheet.pdf.
EPA (2020b) Greenhouse Gas Reporting Program (GHGRP). 2020 Envirofacts. Available online at:
https://ghgdata.epa.gov/ghgp/main.do.
EPA (2019) Advancing Sustainable Materials Management: 2016 and 2017 Data Tables. Office of
Land and Emergency Management, U.S. Environmental Protection Agency. Washington, D.C.
Available online at: https://www.epa.gov/sites/production/files/2019-
11/documents/2016 and 2017 facts and figures data tables O.pdf.
EPA (2018a) Advancing Sustainable Materials Management: 2015 Data Tables. Office of Land and
Emergency Management, U.S. Environmental Protection Agency. Washington, D.C. Available
online at: https://www.epa.gov/sites/production/files/2018-
07/documents/smm 2015 tables and figures 07252018 fnl 508 O.pdf.
EPA (2018b) Greenhouse Gas Reporting Program Data. Washington, DC: U.S. Environmental
Protection Agency. Available online at: https://www.epa.gov/ghgreporting/ghg-reporting-
program-data-sets.
EPA (2016) Advancing Sustainable Materials Management: 2014 Fact Sheet. Office of Land and
Emergency Management, U.S. Environmental Protection Agency. Washington, D.C. Available
online at: https://www.epa.gov/sites/production/files/2016-
11/documents/2014 smmfactsheet 508.pdf.
10-22 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
EPA (2015) Advancing Sustainable Materials Management: Facts and Figures 2013 -Assessing
Trends in Material Generation, Recycling and Disposal in the United States. Office of Solid
Waste and Emergency Response, U.S. Environmental Protection Agency. Washington, D.C.
Available online at:
http://www3.epa.gov/epawaste/nonhaz/municipal/pubs/2013 advncng smm rpt.pdf.
EPA (2007, 2008, 2011, 2013, 2014) Municipal Solid Waste in the United States: Facts and Figures.
Office of Solid Waste and Emergency Response, U.S. Environmental Protection Agency.
Washington, D.C. Available online at: http://www.epa.gov/osw/nonhaz/municipal/msw99.html.
EPA (2006) Solid Waste Management and Greenhouse Gases: A Life-Cycle Assessment of
Emissions and Sinks. Office of Solid Waste and Emergency Response, U.S. Environmental
Protection Agency. Washington, D.C.
EPA (2000) Characterization of Municipal Solid Waste in the United States: Source Data on the 1999
Update. Office of Solid Waste, U.S. Environmental Protection Agency. Washington, D.C.
EPA530-F-00-024.
Goldstein, N. and C. Madtes (2001) "13th Annual BioCycle Nationwide Survey: The State of Garbage
in America." BioCycle, JG Press, Emmaus, PA. December 2001.
IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to
the Fifth Assessment Report of the Intergovernmental Panel on Climate Change [Stocker, T.F., D.
Qin, G.K. Plattner, M. Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley
(eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, 1535
pp.
IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to
the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. [S. Solomon,
D. Qin, M. Manning, Z. Chen, M. Marquis, K.B. Averyt, M. Tignorand H.L. Miller (eds.)].
Cambridge University Press. Cambridge, United Kingdom, 996 pp.
IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change.
[H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa,
Japan.
Kaufman, et al. (2004) "14th Annual BioCycle Nationwide Survey: The State of Garbage in America
2004" Biocycle, JG Press, Emmaus, PA. January 2004.
Schneider, S. (2007) E-mail between Shelly Schneider of Franklin Associates (a division of ERG) and
Sarah Shapiro of ICF International, January 10, 2007.
Shin, D. (2014) Generation and Disposition of Municipal Solid Waste (MSW) in the United States-A
National Survey. Thesis. Columbia University, Department of Earth and Environmental
Engineering, January 3, 2014.
Simmons, et al. (2006) "15th Nationwide Survey of Municipal Solid Waste Management in the
United States: The State of Garbage in America." BioCycle, JG Press, Emmaus, PA. April 2006.
References and Abbreviations 10-23
-------
USTMA (2024) "2023 End-of-Life Tire Management Report." U.S. Tire Manufacturers Association,
Washington, DC. October 2024. Available online at: https://www.ustires.org/system/files/2024-
10/USTMA%27s%202023%20End-of-Life%20Tire%20Management%20Report.pdf.
USTMA (2022) "2021 U.S. Scrap Tire Management Summary." U.S. Tire Manufacturers Association,
Washington, DC. October 2022. Available online at:
https://www.ustires.org/sites/default/files/21%20US%20Scrap%20Tire%20Management%20Re
port%70101777.pdf.
USTMA (2020) "2019 U.S. Scrap Tire Management Summary." U.S. Tire Manufacturers Association,
Washington, DC. October 2020. Available online at:
https://www.ustires.org/sites/default/files/2019%20USTMA%20Scrap%20Tire%20Management
%20Summary%2QReport.pdf.
USTMA (2018) "2017 U.S. Scrap Tire Management Summary." U.S. Tire Manufacturers Association,
Washington, DC. July 2018. Available online at: https://www.tyrepress.com/wp-
content/uploads/2018/07/USTMA scraptire summ 2017 07 11 2018.pdf.
USTMA (2016) "2015 U.S. Scrap Tire Management Summary." U.S. Tire Manufacturers Association.
August 2016. Available online at:
https://www.ustires.org/sites/default/files/MAR 028 USTMA.pdf.
USTMA (2014) "2013 U.S. Scrap Tire Management Summary." U.S. Tire Manufacturers Association.
November 2014. Available online at:
https://www.ustires.org/sites/default/files/MAR 027 USTMA.pdf.
USTMA (2013) "U.S. Scrap Tire Management Summary 2005-2009." U.S. Tire Manufacturers
Association. October 2011; Updated September 2013. Available online at:
https://www.ustires.org/sites/default/files/MAR 025 USTMA.pdf.
USTMA (2012a) "Rubber FAQs." U.S. Tire Manufacturers Association. Accessed on 19 November
2014.
USTMA (2012b) "Scrap Tire Markets: Facts and Figures - Scrap Tire Characteristics." U.S. Tire
Manufacturers Association. Accessed 18 on January 2012.
van Haaren, Rob, Themelis, N., and Goldstein, N. (2010) "The State of Garbage in America."
BioCycle, October 2010. Volume 51, Number 10, pg. 16-23.
Coal Mining
AAPG (1984) Coalbed Methane Resources of the United States. AAPG Studies in Geology Series
#17.
Creedy, D.R (1993) Methane Emissions from Coal Related Sources in Britain: Development of a
Methodology. Chemosphere, 26: 419-439.
DM ME (2024) DGO Data Information System. Department of Mines, Minerals and Energy of Virginia.
Available online at https://www.dmme.virginia.gov/dgoinquiry/frmmain.aspx.
10-24 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
EIA (2024) Annual Coal Report 2023. Table 1. Energy Information Administration, U.S. Department
of Energy. Washington, D.C. DOE/EIA-0584.
El Paso (2009) Shoal Creek Mine Plan, El Paso Exploration & Production.
EPA (2024) Greenhouse Gas Reporting Program (GHGRP) 2023 Subpart FF: Underground Coal
Mines.
EPA (2005) Surface Mines Emissions Assessment. Draft. U.S. Environmental Protection Agency.
EPA (1996) Evaluation and Analysis of Gas Content and Coal Properties of Major Coal Bearing
Regions of the United States. EPA/600/R-96-065. U.S. Environmental Protection Agency.
ERG (2024). Correspondence between ERG and Buchanan Mine.
Geological Survey of Alabama State Oil and Gas Board (GSA) (2023) Well Records Database.
Available online at http://www.gsa.state.al.us/ogb/database.aspx.
IEA (2023) Coal 2023, International Energy Agency, Paris, License: CC BY 4.0. Available online at:
https://www.iea.org/reports/coal-2023.
IPCC (2019) 2019 Refinement to the 2006 IPCC Guidelines for National Greenhouse Gas
Inventories. Calvo Buendia, E., Tanabe, K., Kranjc, A., Baasansuren, J., Fukuda, M., Ngarize S.,
Osako, A., Pyrozhenko, Y., Shermanau, P. and Federici, S. (eds). Published: IPCC, Switzerland.
IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to
the Fifth Assessment Report of the Intergovernmental Panel on Climate Change [Stocker, T.F.,
D. Qin, G.-K. Plattner, M. Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M.
Midgley (eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY,
USA, 1535 pp.
IPCC (2011) Use of Models and Facility-Level Data in Greenhouse Gas Inventories. Report of IPCC
Expert Meeting on Use of Models and Measurements in Greenhouse Gas Inventories 9-11
August 2010, Sydney, Australia. Eds: Eggleston H.S., Srivastava N., Tanabe K., Baasansuren J.,
Fukuda M. IGES.
IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to
the Fourth Assessment Report of the Intergovernmental Panel on Climate Change [Solomon, S.,
D. Qin, M. Manning, Z. Chen, M. Marquis, K.B. Averyt, M. Tignorand H.L. Miller (eds.)].
Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, 996 pp.
JWR (2010) No. 4&7 Mines General Area Maps. Walter Energy: Jim Walter Resources.
King, Brian (1994) Management of Methane Emissions from Coal Mines: Environmental,
Engineering, Economic and Institutional Implication of Options. Neil and Gunter Ltd.
MSHA (2024) Data Transparency at MSHA. Mine Safety and Health Administration. Available online
at http://www.msha.gov/.
Mutmansky, Jan M. and Yanbei Wang (2000) Analysis of Potential Errors in Determination of Coal
Mine Annual Methane Emissions. Mineral Resources Engineering, 9(4).
References and Abbreviations 10-25
-------
Saghafi, Abouna (2013) Estimation of Fugitive Emissions from Open Cut Coal Mining and
Measurable Gas Content. 13th Coal Operators' Conference, University of Wollongong, The
Australian Institute of Mining and Metallurgy & Mine Managers Association of Australia. 306-
313.
USBM (1986) Results of the Direct Method Determination of the Gas Contents of U.S. Coal Basins.
Circular 9067. U.S. Bureau of Mines.
West Virginia Geological & Economic Survey (WVGES) (2024) Oil & Gas Production Data. Available
online at http://www.wvgs.wvnet.edu/www/datastat/datastat.htm.
Abandoned Underground Coal Mines
ACR (2024) American Carbon Registry at Winrock International. Registry. Available online at:
https://acrcarbon.org/acr-registry.
CMOP (2024) EPA's Coalbed Methane Outreach Program, Map of US Coal Mine Methane Current
Projects and Potential Opportunities. Available online at: https://www.epa.gov/cmop/map-us-
coal-mine-methane-current-projects-and-potential-opportunities.
COGIS (2018) Colorado Oil and Gas Information System. Colorado Oil and Gas Commission,
Department of Natural Resources. Available online at https://cogcc.state.co.us/data.html.
EPA (2004) Methane Emissions Estimates & Methodology for Abandoned Coal Mines in the U.S.
Draft Final Report. Washington, D.C. April 2004.
GMI (2024) Global Methane Initiative, International Coal Mine Methane Project List Database.
Available online at: https://www.globalmethane.org/resources/details.aspx?resourceid=1981.
IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to
the Fifth Assessment Report of the Intergovernmental Panel on Climate Change [Stocker, T.F.,
D. Qin, G.-K. Plattner, M. Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M.
Midgley (eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY,
USA, 1535 pp.
IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to
the Fourth Assessment Report of the Intergovernmental Panel on Climate Change [Solomon, S.,
D. Qin, M. Manning, Z. Chen, M. Marquis, K.B. Averyt, M. Tignorand H.L. Miller (eds.)].
Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, 996 pp.
MSHA (2024) U.S. Department of Labor, Mine Health & Safety Administration, Mine Data Retrieval
System. Available online at: https://www.msha.gov/mine-data-retrieval-system.
Petroleum Systems
API (1992) Global Emissions of Methane from Petroleum Sources. American Petroleum Institute,
Health and Environmental Affairs Department, Report No. DR140, February 1992.
BOEM (2024a) BOEM Platform Structures Online Query. Available online at:
https://www.data.boem.gov/Platform/PlatformStructures/Default.aspx.
10-26 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
BOEM (2024b) BOEM Oil and Gas Operations Reports - Part A (OGOR-A). Production Data for 1947
to 2023. Download "Production Data" online at:
https://www.data.boenn.gov/Main/RawData.aspx.
BOEM (2024c) BOEM Oil and Gas Operations Reports - Part A (OGOR-A). Production Data for 1996
to 2023. Available online at: https://www.data.boenn.gov/Main/OGOR-A.aspx.
BOEM (2024d) BOEM Oil and Gas Operations Reports - Part B (OGOR-B). Flaring volumes for 1996
to 2023. Available online at: https://www.data.boem.gov/Main/OGOR-B.aspx.
EIA (2024) Crude Oil Production. Energy Information Administration.
Enverus (2023) September 2023 Download. Enverus, Inc.
EPA (2024) Greenhouse Gas Reporting Program. U.S. Environmental Protection Agency. Data
reported as of August 16, 2024.
EPA (2017) 2017Nonpoint Oil and Gas Emission Estimation Tool, Version 1.2. Prepared for U.S.
Environmental Protection Agency by Eastern Research Group, Inc. (ERG). October 2019.
EPA (1999) Estimates of Methane Emissions from the U.S. OH Industry (Draft Report). Prepared by
ICF International. Office of Air and Radiation, U.S. Environmental Protection Agency. October
1999.
EPA (1997) Compilation of Air Pollutant Emission Factors, AP-42. Office of Air Quality Planning and
Standards, U.S. Environmental Protection Agency. Research Triangle Park, NO. October 1997.
EPA/GRI (1996) Methane Emissions from the Natural Gas Industry. Prepared by Radian. U.S.
Environmental Protection Agency. April 1996.
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change.
[H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa,
Japan.
Natural Gas Systems
BOEM (2023). Outer Continental Shelf Air Quality System (OCS AQS): Year 2021 Emissions
Inventory Quality Assurance/Quality Control (QA/QC) Study. March 2023. BOEM 2023-023.
https://www.boem.gov/environment/environmental-studies/2021-ocs-emissions-inventory
Enverus (2023) September 2023 Download. Enverus, Inc.
EPA (2024) Greenhouse Gas Reporting Program- Subpart W- Petroleum and Natural Gas Systems.
Environmental Protection Agency. Data reported as of August 16, 2024.
FERC (2024) Form No. 2, Major Natural Gas Pipeline Annual Report. Federal Energy Regulatory
Commission, https://ferc.gov/industries-data/natural-gas/industry-forms.
Fischer et al. (2018) "An Estimate of Natural Gas Methane Emissions from California Homes."
Environmental Science & Technology 2018, 52 (17), 10205-10213.
https://pubs.acs.Org/doi/10.1021/acs.est.8b03217.
References and Abbreviations 10-27
-------
GRI/EPA (1996) Methane Emissions from the Natural Gas Industry. Prepared by Harrison, M., T.
Shires, J. Wessels, and R. Cowgill, eds., Radian International LLC for National Risk Management
Research Laboratory, Air Pollution Prevention and Control Division, Research Triangle Park, NC.
EPA-600/R-96-080a.
GTI (2001) Gas Resource Database: Unconventional Natural Gas and Gas Composition Databases.
Second Edition. GRI-01/0136.
GTI (2009) Field Measurement Program to Improve Uncertainties for Key Greenhouse Gas Emission
Factors for Distribution Sources. November 2009. Gas Technology Institute and Innovative
Environmental Solutions. GTI Project Number 20497. OTD Project Number 7.7.b.
GTI (2019) Classification of Methane Emissions from Industrial Meters, Vintage vs Modern Plastic
Pipe, and Plastic-lined Steel and Cast-iron Pipe. June 2019. Gas Technology Institute and U.S.
Department of Energy GTI Project Number 22070. DOE project Number ED-FE0029061.
IPCC (2019) 2019 Refinement to the 2006 IPCC Guidelines for National Greenhouse Gas
Inventories. Calvo Buendia, E., Tanabe, K., Kranjc, A., Baasansuren, J., Fukuda, M., Ngarize, S.,
Osako, A., Pyrozhenko, Y., Shermanau, P. and Federici, S. (eds). Published: IPCC, Switzerland.
Lamb, et al. (2015) "Direct Measurements Show Decreasing Methane Emissions from Natural Gas
Local Distribution Systems in the United States." Environmental Science & Technology, Vol. 49
5161-5169.
PHMSA (2024) Gas Distribution Annual Data. Pipeline and Hazardous Materials Safety
Administration, U.S. Department of Transportation, Washington, DC. Available online at:
https://www.phmsa.dot.gov/data-and-statistics/pipeline/annual-report-mileage-gas-
distribution-systems.
Zimmerle et al. (2015) "Methane Emissions from the Natural Gas Transmission and Storage System
in the United States." Environmental Science and Technology, Vol. 49 9374-9383.
Zimmerle et al. (2019) "Characterization of Methane Emissions from Gathering Compressor
Stations." October 2019. Available at https://mountainscholar.org/handle/10217/195489.
Abandoned Oil and Gas Wells
Alaska Oil and Gas Conservation Commission, Available online at:
https://www.commerce.alaska.gov/web/aogcc/Data.aspx.
Arkansas Geological & Conservation Commission, "List of Oil & Gas Wells - Data From November
1,1936 to January 1,1955."
The Derrick's Handbook of Petroleum: A Complete Chronological and Statistical Review of
Petroleum Developments From 1859 to 1898 (V.1), (1898-1899) (V.2).
Enverus (2023) October 2023 Download. Enverus, Inc.
Florida Department of Environmental Protection - Oil and Gas Program, Available online at:
https://floridadep.gov/water/oil-gas.
10-28 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Geological Survey of Alabama, Oil & Gas Board, Available online at:
https://www.gsa.state.al.us/ogb/.
GRI/EPA (1996) Methane Emissions from the Natural Gas Industry. Prepared by Harrison, M., T.
Shires, J. Wessels, and R. Cowgill, eds., Radian International LLC for National Risk Management
Research Laboratory, Air Pollution Prevention and Control Division, Research Triangle Park, NC.
EPA-600/R-96-080a.
GTI (2001) Gas Resource Database: Unconventional Natural Gas and Gas Composition Databases.
Second Edition. GRI-01/0136.
Interstate Oil and Gas Compact Commission (2021). IDLE AND ORPHAN OIL AND GAS WELLS:
STATE AND PROVINCIAL REGULATORY STRATEGIES 2021. Available online at:
https://iogcc.ok.goV/sites/g/files/gmc836/f/iogcc idle and orphan wells 2021 final web.pdf.
Kang, etal. (2016) "Identification and characterization of high methane-emitting abandoned oiland
gas wells." PNAS, vol. 113 no. 48,13636-13641, doi: 10.1073/pnas.1605913113.
Oklahoma Geological Survey. "Oklahoma Oil: Past, Present, and Future." Oklahoma Geology Notes,
v. 62 no. 3, 2002 pp. 97-106.
Pennsylvania Department of Environmental Protection, Oil and Gas Reports - Oil and Gas Operator
Well Inventory. Available online at:
http://www.depreportingservices.state.pa.us/ReportServer/Pages/ReportViewer.aspx7/Oil Gas
/OG Well Inventory.
C02 Transport, Injection, and Geologic Storage
ICF (2025a) Expert judgment on uncertainty around C02 Transport, Injection, and Geological
Storage (TIGS), ICF. February 2025
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change.
[H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa,
Japan.SAIC/EIA (2001) Monte Carlo Simulations of Uncertainty in U.S. Greenhouse Gas
Emission Estimates. Final Report. Prepared by Science Applications International Corporation
(SAIC) for Office of Integrated Analysis and Forecasting, Energy Information Administration, U.S.
Department of Energy. Washington, D.C. June 22, 2001.
U.S. Environmental Protection Agency (EPA) (2024a) Greenhouse Gas Reporting Program (GHGRP).
Aggregation of Reported Facility Level Data under Subpart PP for Calendar Years 2010 through
2023. Office of Air and Radiation, Office of Atmospheric Programs, U.S. Environmental
Protection Agency, Washington, D.C.
U.S. Environmental Protection Agency (EPA) (2024b) Greenhouse Gas Reporting Program (GHGRP),
Facility Level Information on Greenhouse gases Tool (FLIGHT), Subpart RR: Geological
Sequestration of C02. Available online at: http://ghgdata.epa.gov/ghgp.
PHMSA (2024) 2010+ Pipeline Miles and Facilities data, Hazardous Liquid Pipeline Miles and Tanks
by Commodity. Pipeline and Hazardous Materials Safety Administration, U.S. Department of
References and Abbreviations 10-29
-------
Transportation, Washington, DC. Available online at: https://www.phmsa.dot.gov/data-and-
statistics/pipeline/data-and-statistics-overview.
International Bunker Fuels
Anderson, B.E.,etal. (2011) Alternative Aviation Fuel Experiment (AAFEX), NASA Technical
Memorandum, in press.
ASTM (1989) Military Specification for Turbine Fuels, Aviation, Kerosene Types, NATO F-34 (JP-8) and
NATO F-35. February 10,1989.
DHS (2008) Personal Communication with Elissa Kay, Residual and Distillate Fuel Oil Consumption
(International Bunker Fuels). Department of Homeland Security, Bunker Report. January 11,
2008.
DLA Energy (2025) Unpublished data from the Defense Fuels Automated Management System
(DFAMS). Defense Energy Support Center, Defense Logistics Agency, U.S. Department of
Defense. Washington, D.C.
DOC (1991 through 2024) Unpublished Report of Bunker Fuel Oil Laden on Vessels Cleared for
Foreign Countries. Form-563. Foreign Trade Division, Bureau of the Census, U.S. Department of
Commerce. Washington, D.C.
DOT (1991 through 2013) Fuel Cost and Consumption. Federal Aviation Administration, Bureau of
Transportation.
Statistics, U.S. Department of Transportation. Washington, D.C. DAI-10.
EIA (2025) Monthly Energy Review, January 2025, Energy Information Administration, U.S.
Department of Energy, Washington, D.C. DOE/EIA-0035(2025/01).
EPA (2020) EPA Inventory of U.S. Greenhouse Gas Emissions and Sinks 1990-2019: Updated
Gasoline and Diesel Fuel C02 Emission Factors - Memo.
FAA (2024) Personal Communication between FAA and John Steller, Mausami Desai, and Vincent
Camobreco for aviation emissions estimates from the Aviation Environmental Design Tool
(AEDT). March 2024.
IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to
the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. [Stocker, T.F.,
D. Qin, G.K. Plattner, M. Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley
(eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, 1535
pp.
IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to
the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. [S. Solomon,
D. Qin, M. Manning, Z. Chen, M. Marquis, K.B. Averyt, M. Tignorand H.L. Miller (eds.)].
Cambridge University Press. Cambridge, United Kingdom, 996 pp.
IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change.
10-30 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
[H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa,
Japan.
USAF (1998) Fuel Logistics Planning. U.S. Air Force pamphlet AFPAM23-221, May 1, 1998.
IPCC/UNEP/OECD/IEA (1997) Revised 1996IPCC Guidelines for National Greenhouse Gas
Inventories. 31 Intergovernmental Panel on Climate Change, United Nations Environment
Programme, Organization for Economic 32 Co-Operation and Development, International
Energy Agency, Paris, France.
Biomass and Biofuels Consumption
EIA (2025) Monthly Energy Review, January 2025. Energy Information Administration, U.S.
Department of Energy. Washington, D.C. DOE/EIA-0035(2025/01).
EIA (2024) Biofuels explained: Use of biomass-based diesel fuel. Energy Information
Administration, U.S. Department of Energy. Washington, D.C. Available online at:
https://www.eia.gov/energyexplained/biofuels/use-of-biodiesel.php.
EPA (2023). Greenhouse Gas Reporting Program (GHGRP). 2022 Envirofacts. Available online at:
https://ghgdata.epa.gov/ghgp/main.do.
EPA (2010) Carbon Content Coefficients Developed for EPA's Mandatory Reporting Rule. Office of
Air and Radiation, Office of Atmospheric Programs, U.S. Environmental Protection Agency,
Washington, D.C.
ICF (2025) Expert judgment on uncertainty around Biomass and Biofuels Consumption, ICF.
February 2025
Lindstrom, P. (2006) Personal Communication. Perry Lindstrom, Energy Information Administration
and Jean Kim, ICF International.
NESCAUM (2024). Criteria, Greenhouse Gas, and Hazardous Air Pollutant Emissions Factors from
Residential Cordwood and Pellet Stoves Using an Integrated Duty Cycle Test Protocol. Nora
Traviss, George Allen, and Mahdi Ahmadi. ACS ES&T Air 2024 1 (9), 1190-1202. DOI:
10.1021/acsestair.4c00135
Energy Sources of Precursor Greenhouse GasesEPA (2024) "Criteria pollutants National Tier 1 for
1970 - 2023." National Emissions Inventory (NEI) Air Pollutant Emissions Trends Data. Office of
Air Quality Planning and Standards, February 2024. Available online at:
https://www.epa.gov/air-emissions-inventories/air-pollutant-emissions-trends-data
EPA (2023a) EPA's Emissions Inventory System (EIS) to National Inventory Report (NIR) Mapping file
EIS_NIR_mapping.xlsx. U.S. Environmental Protection Agency. Washington, D.C.
EPA (2023b) "2020 National Emissions Inventory Technical Support Document: Introduction."
Office of Air Quality Planning and Standards, March 2023. Available online at:
https://www.epa.gov/system/files/documents/2023-
01/NEI2020 TSD Sectionl lntroduction.pdf.
References and Abbreviations 10-31
-------
Industrial Processes and Product Use
Cement Production
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change.
[H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa,
Japan.
U.S. Bureau of Mines (1990 through 1993) Minerals Yearbook: Cement Annual Report. U.S.
Department of the Interior, Washington, D.C.
U.S. Environmental Protection Agency (EPA) (2015) Greenhouse Gas Reporting Program Report
Verification. Available online at https://www.epa.gov/sites/production/files/2015-
07/documents/ghgrp verification factsheet.pdf.
U.S. EPA (2024) Greenhouse Gas Reporting Program (GHGRP). Aggregation of Reported Facility
Level Data under Subpart H -National Level Clinker Production from Cement Production for
Calendar Years 2014 through 2023. Office of Air and Radiation, Office of Atmospheric
Programs, U.S. Environmental Protection Agency, Washington, D.C.
United States Geological Survey (USGS) (2024a) 2022 Minerals Yearbook - Cement (Advance
Release Tables). U.S. Geological Survey, Reston, VA. June 2024.
USGS (2024b) Mineral Commodity Summaries: Cement. U.S. Geological Survey, Reston, VA.
January 2024.
USGS (2024c) Mineral Industry Surveys, Cement in December 2022. U.S. Geological Survey,
Reston, VA. (February 2024).
USGS (1995 through 2014) Minerals Yearbook - Cement. U.S. Geological Survey, Reston, VA.
Van Oss (2013a) 1990 through 2012 Clinker Production Data Provided by Hendrikvan Oss (USGS)
via email on November8, 2013.
Van Oss (2013b) Personal communication. Hendrikvan Oss, Commodity Specialist of the U.S.
Geological Survey and Gopi Manne, Eastern Research Group, Inc. October 28, 2013.
Lime Production
IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change.
[H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa,
Japan.
Males, E. (2003) Memorandum from Eric Males, National Lime Association to William N. Irving &
Leif Hockstad, Environmental Protection Agency. March 6, 2003.
Miner, R. and B. Upton (2002) Methods for estimating greenhouse gas emissions from lime kilns at
kraft pulp mills. Energy. Vol. 27 (2002), p. 729-738.
10-32 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
RTI (2023) Expert judgment, RTI International. March 30, 2023.
Seeger (2013) Memorandum from Arline M. Seeger, National Lime Association to Leif Hockstad,
Environmental Protection Agency. March 15,2013.
U.S. Environmental Protection Agency (EPA) (2024) Greenhouse Gas Reporting Program (GHGRP).
Aggregation of Reported Facility Level Data under Subpart S-National Lime Production for
Calendar Years 2010 through 2023. Office of Air and Radiation, Office of Atmospheric
Programs, U.S. Environmental Protection Agency, Washington, D.C.
United States Geological Survey (USGS) (2024a) 2019 Minerals Yearbook: Lime. U.S. Geological
Survey, Reston, VA (March 2024). Latest edition was updated in 2024 for 2019. Applicable
editions are available at: https://www.usgs.gov/centers/national-minerals-information-
center/lime-statistics-and-information. See "Archive" for editions prior to 1993.
USGS (2024b) 2024 Mineral Commodities Summary: Lime. U.S. Geological Survey, Reston, VA
(January 2024). Latest edition was updated in 2023 for 2022. Applicable editions are available
at: https://www.usgs.gov/centers/national-minerals-information-center/lime-statistics-and-
information.
USGS (2024c) 2022 Minerals Yearbook Annual Tables: Lime. U.S. Geological Survey, Reston, VA
(January 2024). Latest edition was updated in 2024 for 2022 tables. Applicable editions are
available at: https://www.usgs.gov/centers/national-minerals-information-center/lime-
statistics-and-information.
USGS (2021) Mineral Commodities Summary: Lime. U.S. Geological Survey, Reston, VA (January
2021).
USGS (2020a) Personal communication. Lori E. Apodaca, U.S. Geological Survey and Amanda
Chiu, U.S. Environmental Protection Agency. December 17, 2020.
USGS (2020b) 2017 Minerals Yearbook: Lime. U.S. Geological Survey, Reston, VA (June 2020).
USGS (2020c) 2018 Minerals Yearbook Annual Tables: Lime. U.S. Geological Survey, Reston, VA
(November 2020).
USGS (2019) 2016 Minerals Yearbook: Lime. U.S. Geological Survey, Reston, VA (August 2019).
USGS (2018) 2015 Minerals Yearbook: Lime. U.S. Geological Survey, Reston, VA (March 2018).
USGS (2012a) 2012 Mineral Commodities Summary: Lime. U.S. Geological Survey, Reston, VA
(January 2012).
USGS (2012b) 2012 Expert Elicitation. Michael Miller, U.S. Geological Survey (2012).
USGS (2007-2011) Mineral Commodities Summary: Lime. U.S. Geological Survey, Reston, VA.
USGS (2002) 2002 Mineral Commodities Summary: Lime. U.S. Geological Survey, Reston, VA
(January 2002).
USGS (1996) 1996 Mineral Commodities Summary: Lime. U.S. Geological Survey, Reston, VA
(January 1996).
USGS (1991) 1991 Minerals Yearbook: Lime. U.S. Geological Survey, Reston, VA (1991).
References and Abbreviations 10-33
-------
Glass Production
Federal Reserve (2024) Board of Governors of the Federal Reserve System (US), Industrial
Production: Manufacturing: Durable Goods: Glass and Glass Product (NAICS = 3272)
[IPG3272N], retrieved from FRED, Federal Reserve Bank of St. Louis. Available at:
https://fred.stlouisfed.org/series/IPG3272N. Accessed on September 10. 2024.
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change.
[H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa,
Japan.
RTI (2025) Expert judgment, RTI International. February 21, 2025.
RTI (2023) Expert judgment, RTI International. March 30, 2023.
RTI (2022) Expert judgment. Melissa Icenhour, RTI International. November 16, 2022.
U.S. Bureau of Mines (1991 and 1993a) Minerals Yearbook: Crushed Stone Annual Report. U.S.
Department of the Interior. Washington, D.C.
U.S. Department of Energy (DOE) (2002) Glass Industry of the Future: Energy and Environmental
Profile of the U.S. Glass Industry. Office of Industrial Technologies, U.S. Department of Energy.
Washington, D.C.
U.S. Environmental Protection Agency (EPA) (2015) Greenhouse Gas Reporting Program Report
Verification. Available online at https://www.epa.gov/sites/production/files/2015-
07/documents/ghgrp verification factsheet.pdf.
U.S. EPA (2024) Greenhouse Gas Reporting Program (GHGRP). Aggregation of Reported Facility
Level Data under Subpart N -National Glass Production for Calendar Years 2010 through 2023.
Office of Air and Radiation, Office of Atmospheric Programs, U.S. Environmental Protection
Agency, Washington, D.C.
U.S. EPA (2009) Technical Support Document for the Glass Manufacturing Sector: Proposed Rule
for Mandatory Reporting of Greenhouse Gases. U.S. Environmental Protection Agency,
Washington, D.C.
United States Geological Survey (USGS) (1995 through 2015b) Minerals Yearbook: Soda Ash Annual
Report. U.S. Geological Survey, Reston, VA.
USGS (2024) Mineral Industry Surveys: Soda Ash in December 2023. U.S. Geological Survey,
Reston, VA. March 2024. Available online at: https://www.usgs.gov/centers/national-minerals-
information-center/soda-ash-statistics-and-information.
USGS (2017 through 2024) Mineral Industry Surveys: Soda Ash. U.S. Geological Survey, Reston, VA..
Available online at: https://www.usgs.gov/centers/national-minerals-information-center/soda-
ash-statistics-and-information.
10-34 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Other Process Uses of Carbonates
AISI (2018 through 2021) Annual Statistical Report. American Iron and Steel Institute.
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change.
[H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa,
Japan.
Kostick, D. S. (2012) Personal communication. Dennis S. Kostick, U.S. Geological Survey, Soda Ash
Commodity Specialist and Gopi Manne and Bryan Lange of Eastern Research Group, Inc.
October 2012.
McNeece (2024) Personal communication, Steve McNeece, Nevada Department of Environmental
Quality and Amanda Chiu, U.S. Environmental Protection Agency. September 11, 2024.
McNeece (2023) Personal communication, Steve McNeece, Nevada Department of Environmental
Quality and Amanda Chiu, U.S. Environmental Protection Agency. November 28, 2023.
RTI (2023) Expert judgment, RTI International. March 30, 2023.
Simmons (2025) Personal communication, Kristi Simmons, U.S. Geological Survey and Ga-Young
Park, U.S. Environmental Protection Agency. February 12, 2025.
U.S. Bureau of Mines (1991 and 1993a) Minerals Yearbook: Crushed Stone Annual Report. U.S.
Department of the Interior. Washington, D.C.
U.S. Environmental Protection Agency (EPA) (2024). Greenhouse Gas Reporting Program (GHGRP).
Dataset as of August 16, 2024. Available online at: https://ghgdata.epa.gov/ghgp/
United States Geological Survey (USGS) (2024) Mineral Industry Surveys: Soda Ash in September
2024. U.S. Geological Survey, Reston, VA. November 2024.
USGS (2023a) Mineral Industry Surveys: Soda Ash in September 2023. U.S. Geological Survey,
Reston, VA. November 2023.
USGS (2023b) 2027 Minerals Yearbook: Stone, Crushed [Advanced Release]. U.S. Geological
Survey, Reston, VA. June 2023.
USGS (2022a) 2018 Minerals Yearbook: Stone, Crushed [Advanced Release]. U.S. Geological
Survey, Reston, VA. August 2022.
USGS (2022b) Mineral Industry Surveys: Soda Ash in August 2022. U.S. Geological Survey, Reston,
VA. November 2022.
USGS (2022c) 2020 Minerals Yearbook: Stone, Crushed [Advanced Release]. U.S. Geological
Survey, Reston, VA. August 2022.
USGS (2022<3) 2019 Minerals Yearbook: Stone, Crushed [Advanced Release]. U.S. Geological
Survey, Reston, VA. June 2022.
USGS (2022e) 2018 Minerals Yearbook: Magnesium Compounds [Advanced Release]. U.S.
Geological Survey, Reston, VA. May 2022.
References and Abbreviations 10-35
-------
USGS (2022f) 2018 Minerals Yearbook: Clay and Shale [Advanced Release]. U.S. Geological Survey,
Reston, VA. March 2022.
USGS (2022g) 2018 Minerals Yearbook: Soda Ash [Advanced Release]. U.S. Geological Survey,
Reston, VA. January 2022.
USGS (2021a) Minerals Yearbook2019: Soda Ash [Advanced Data Release of the 2019 Annual
Tables], U.S. Geological Survey, Reston, VA. August 2021.
USGS (2021 b) Mineral Industry Surveys: Soda Ash in April 2021. U.S. Geological Survey, Reston, VA.
July 2021.
USGS (2021c) 2017 Minerals Yearbook: Stone, Crushed [Advanced Release]. U.S. Geological
Survey, Reston, VA. June 2021.
USGS (2021 d) 2020 Mineral Commodity Summaries: Stone (Crushed). U.S. Geological Survey,
Reston, VA. January 2021.
USGS (2020a) Minerals Yearbook 2017: Stone, Crushed [Advanced Data Release of the 2017Annual
Tables]. U.S. Geological Survey, Reston, VA. August 2020.
USGS (2020b) Mineral Industry Surveys: Soda Ash in April 2020. U.S. Geological Survey, Reston, VA.
July 2020.
USGS (2020c) 2016 Minerals Yearbook: Stone, Crushed [Advanced Release]. U.S. Geological
Survey, Reston, VA. January 2020.
USGS (2019) Mineral Industry Surveys: Soda Ash in April 2019. U.S. Geological Survey, Reston, VA.
July 2019.
USGS (2018) Mineral Industry Surveys: Soda Ash in February 2018. U.S. Geological Survey, Reston,
VA. 2018.
USGS (2017) Mineral Industry Surveys: Soda Ash in January 2017. U.S. Geological Survey, Reston,
VA. March 2017.
USGS (1995a through 2017) Minerals Yearbook: Crushed Stone Annual Report. U.S. Geological
Survey, Reston, VA.
USGS (1994 through 2015b) Minerals Yearbook: Soda Ash Annual Report. U.S. Geological Survey,
Reston, VA.
USGS (1990 through 2002) Minerals Yearbook: Magnesium Compounds Annual Report. U.S.
Geological Survey, Reston, VA.
USGS (1948) Reports: Magnesite andbrucite deposits atGabbs, Nye County, Nevada. U.S.
Geological Survey, Reston, VA
Willett (2024) Personal communication, Jason Christopher Willett, U.S. Geological Survey and Ga-
Young, Park, U.S. Environmental Protection Agency. November 20, 2024.
Willett (2023) Personal communication, Jason Christopher Willett, U.S. Geological Survey and
Amanda Chiu, U.S. Environmental Protection Agency. November 21, 2023.
10-36 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Willett (2017) Personal communication, Jason Christopher Willett, U.S. Geological Survey and
Mausami Desai and John Steller, U.S. Environmental Protection Agency. March 9, 2017.
Ammonia Production
ACC (2024) Business of Chemistry (Annual Data). American Chemistry Council, Arlington, VA.
Coffeyville Resources Energy, Inc. (CVR) (2008) CVR Energy, Inc. 2008 Annual Report. Available
online at: https://cvrenergy.gcs-web.com/annual-report-and-proxy-archive.
CVR (2023) CVR Energy, Inc. 2023 Annual Report on Form 10-K. Available online at:
https://cvrenergy.gcs-web.com/annual-report-proxy-information.
CVR (2022) CVR Energy, Inc. 2022 Annual Report on Form 10-K. Available online at:
https://cvrenergy.gcs-web.com/annual-report-proxy-information.
CVR (2021) CVR Energy, Inc. 2021 CVI Annual Report on Form 10-K. Available online at:
https://cvrenergy.gcs-web.com/annual-report-and-proxy-archive.
CVR (2020) CVR Energy, Inc. 2020 CV! Annual Report on Form 10-K -Final. Available online at:
https://cvrenergy.gcs-web.com/annual-report-and-proxy-archive.
CVR (2019) CVR Energy, Inc. 2019 CVI Form 10-K-Final. Available online at https://cvrenergy.gcs-
web.com/annual-report-and-proxy-archive.
CVR (2018) CVR Energy, Inc. 2018 CVI Annual Report on Form 10-K -Final. Available online at:
https://cvrenergy.gcs-web.com/annual-report-and-proxy-archive.
CVR (2017) CVR Energy, Inc. 2017 CVI Annual Report on Form 10-K (Web). Available online at:
https://cvrenergy.gcs-web.com/annual-report-and-proxy-archive.
CVR (2016) CVR Energy, Inc. 2016 CVI Annual Report on Form 10-K (Web). Available online at:
https://cvrenergy.gcs-web.com/annual-report-and-proxy-archive.
CVR (2014) CVR Energy, Inc. 2014 Annual Report. Available online at: https://cvrenergv.gcs-
web.com/annual-report-and-proxy-archive.
CVR (2015) CVR Energy, Inc. 2015 Annual Report. Available online at: https://cvrenergv.gcs-
web.com/annual-report-and-proxy-archive.
CVR (2013) CVR Energy, Inc. 2013 Annual Report. Available online at: https://cvrenergv.gcs-
web.com/annual-report-and-proxy-archive.
CVR (2012) CVR Energy, Inc. 2012 Annual Report. Available online at: https://cvrenergv.gcs-
web.com/annual-report-and-proxy-archive.
CVR (2011) CVR Energy, Inc. 2011 Annual Report. Available online at: https://cvrenergy.gcs-
web.com/annual-report-and-proxy-archive.
CVR (2010) CVR Energy, Inc. 2010 Annual Report. Available online at: https://cvrenergy.gcs-
web.com/annual-report-and-proxy-archive.
References and Abbreviations 10-37
-------
CVR (2009) CVR Energy, Inc. 2009 Annual Report Available online at: https://cvrenergv.gcs-
web.com/annual-report-and-proxy-archive.
EFMA (2000) Best Available Techniques for Pollution Prevention and Control in the European
Fertilizer Industry. Booklet No. 5 of 8: Production of Urea and Urea Ammonium Nitrate. Available
online at: http://fertilizerseurope.com/site/index.php?id=390.
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change.
[H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa,
Japan.
RTI (2023) Expert judgment, RTI International. March 30, 2023.
RTI (2025) Expert judgment, RTI International. February 21, 2025.
United States Census Bureau (2011) Current Industrial Reports Fertilizer Materials and Related
Products: 2010 Summary. Available online at:
http://www.census.gov/manufacturing/cir/historical data/mq325b/index.html.
U.S. Census Bureau (2010) Current Industrial Reports Fertilizer Materials and Related Products:
2009 Summary. Available online at:
http://www.census.gov/manufacturing/cir/historical data/mq325b/index.html.
U.S. Census Bureau (2009) Current Industrial Reports Fertilizer Materials and Related Products:
2008 Summary. Available online at:
http://www.census.gov/manufacturing/cir/historical data/mq325b/index.html.
U.S. Census Bureau (2008) Current Industrial Reports Fertilizer Materials and Related Products:
2007Summary. Available online at:
http://www.census.gov/cir/www/325/mq325b/mq325b075.xls.
U.S. Census Bureau (2007) Current Industrial Reports Fertilizer Materials and Related Products:
2006 Summary. Available online at: http://www.census.gOv/industrv/1/mq325b065.pdf.
U.S. Census Bureau (2006) Current Industrial Reports Fertilizer Materials and Related Products:
2005 Summary. Available online at: http://www.census.gov/cir/www/325/mq325b.html.
U.S. Census Bureau (2004, 2005) Current Industrial Reports Fertilizer Materials and Related
Products: Fourth Quarter Report Summary. Available online at:
http://www.census.gov/cir/www/325/mq325b.html.
U.S. Census Bureau (1998 through 2003) Current Industrial Reports Fertilizer Materials and Related
Products: Annual Reports Summary. Available online at:
http://www.census.gov/cir/www/325/mq325b.html.
U.S. Census Bureau (1991 through 1994) Current Industrial Reports Fertilizer Materials Annual
Report. Report No. MQ28B. U.S. Census Bureau, Washington, D.C.
United States Energy Information Administration (EIA) (2023) Monthly Energy Review, February
2023, Energy Information Administration, U.S. Department of Energy, Washington, DC.
DOE/EIA-0035(2023/2).
10-38 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
United States Environmental Protection Agency (EPA) (2024a). Greenhouse Gas Reporting Program.
Dataset as of August 16, 2024. Available online at: https://ghgdata.epa.gov/ghgp/.
U.S. EPA (2024b) Greenhouse Gas Reporting Program. Aggregation of Reported Facility Level Data
under Subpart G -Annual Urea Production from Ammonia Manufacturing for Calendar Years
2017-2023. Office of Air and Radiation, Office of Atmospheric Programs, U.S. Environmental
Protection Agency, Washington, D.C.
U.S. EPA (2024c) Greenhouse Gas Reporting Program. Aggregation of Reported Facility Level Data
under Subpart RR for Calendar Years 2020-2023. Office of Air and Radiation, Office of
Atmospheric Programs, U.S. Environmental Protection Agency, Washington, D.C.
U.S. EPA (2018) Greenhouse Gas Reporting Program. Aggregation of Reported Facility Level Data
under Subpart G -Annual Urea Production from Ammonia Manufacturing for Calendar Years
2011-2016. Office of Air and Radiation, Office of Atmospheric Programs, U.S. Environmental
Protection Agency, Washington, D.C
United States Geological Survey (USGS) (2024) 2024 Mineral Commodity Summaries: Nitrogen
(Fixed) - Ammonia. January 2024. Available online at:
https://pubs.usgs.gov/periodicals/mcs2024/mcs2024-nitrogen.pdf.
USGS (1994-2009) Minerals Yearbook: Nitrogen. Available online at:
http://minerals.usgs.gov/minerals/pubs/commodity/nitrogen/
Urea Consumption for Non-Agricultural Purposes
European Fertilizer Manufacturers'Association (EFMA) (2000) Best Available Techniques for
Pollution Prevention and Control in the European Fertilizer Industry. Booklet No. 5 of 8:
Production of Urea and Urea Ammonium Nitrate.
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change.
[H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa,
Japan.
RTI (2023) Expert judgment, RTI International. March 30, 2023.
The Fertilizer Institute (TFI) (2002) U.S. Nitrogen Imports/Exports Table. The Fertilizer Institute.
Available online at: http://www.tfi.org/statistics/usnexim.asp. August 2002.
United States Census Bureau (2001 through 2011) Current Industrial Reports Fertilizer Materials
and Related Products: Annual Summary. Available online at:
http://www.census.gov/manufacturing/cir/historical data/index.html.
United States Department of Agriculture (2012) Economic Research Service Data Sets, Data Sets,
U.S. Fertilizer Imports/Exports: Standard Tables. Available online at:
http://www.ers.usda.gov/data-products/fertilizer-importsexports/standard-tables.aspx.
United States Environmental Protection Agency (EPA) (2024a) Greenhouse Gas Reporting Program.
Aggregation of Reported Facility Level Data under Subpart G -Annual Urea Production from
References and Abbreviations 10-39
-------
Ammonia Manufacturing for Calendar Years 2017-2023. Office of Air and Radiation, Office of
Atmospheric Programs, U.S. Environmental Protection Agency, Washington, D.C.
United States EPA (2024b). Greenhouse Gas Reporting Program. Dataset as of August 16, 2024.
Available online at: https://ghgdata.epa.gov/ghgp/.
United States EPA (2018) Greenhouse Gas Reporting Program. Aggregation of Reported Facility
Level Data under Subpart G -Annual Urea Production from Ammonia Manufacturing for
Calendar Years 2011 -2016. Office of Air and Radiation, Office of Atmospheric Programs, U.S.
Environmental Protection Agency, Washington, D.C.
United States Geological Survey (USGS) (1994 through 2019) Minerals Yearbook: Nitrogen.
Available online at: http://minerals.usgs.gov/minerals/pubs/commodity/nitrogen/.
USGS (2024) Minerals Commodity Summaries: Nitrogen (Fixed) - Ammonia. January 2024.
Available online at: https://pubs.usgs.gov/periodicals/mcs2024/mcs2024-nitrogen.pdf.
USGS (2024a) Minerals Yearbook: Nitrogen (2022 Tables Only Release). Available online at:
https://d9-wret.s3.us-west-2.amazonaws.com/assets/palladium/production/s3fs-
public/media/files/myb1-2022-nitro-ert.xlsx. March 29, 2024.
USGS (2023) Minerals Yearbook: Nitrogen (2021 Tables Only Release). Available online at:
https://d9-wret.s3.us-west-2.amazonaws.com/assets/palladium/production/s3fs-
public/media/files/mvb1-2021-nitro-advrel.xlsx. September 25, 2024.
USGS (2022) Minerals Yearbook: Nitrogen (2020 Tables Only Release). Available online at:
https://d9-wret.s3.us-west-2.amazonaws.com/assets/palladium/production/s3fs-
public/media/files/myb1-2020-nitro-adv2.xlsx. June 6, 2024.
United States International Trade Commission (ITC) (2002) United States International Trade
Commission Interactive Tariff and Trade DataWeb, Version 2.5.0. Available online at:
http://dataweb.usitc.gov/. August 2002.
Nitric Acid Production
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change.
[H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa,
Japan.
Icenhour (2020) Personal communication, Melissa Icenhour, RTI International and Amanda Chiu,
U.S. Environmental Protection Agency. December 3, 2020.
RTI (2025) Expert judgment, RTI International. February 21, 2025.
United States Census Bureau (2010a) Current Industrial Reports. Fertilizers and Related
Chemicals: 2009. "Table 1: Summary of Production of Principle Fertilizers and Related
Chemicals: 2009 and 2008." June, 2010. MQ325B(08)-5. Available online at:
http://www.census.gov/manufacturing/cir/historical data/mq325b/index.html.
10-40 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
U.S. Census Bureau (2010b) Personal communication between Hilda Ward (of U.S. Census Bureau)
and Caroline Cochran (of ICF International). October 26, 2010 and November 5, 2010.
U.S. Census Bureau (2009) Current Industrial Reports. Fertilizers and Related Chemicals: 2008.
"Table 1: Shipments and Production of Principal Fertilizers and Related Chemicals: 2004 to
2008." June, 2009. MQ325B(08)-5. Available online at:
http://www.census.gov/manufacturing/cir/historical data/mq325b/index.html.
U.S. Census Bureau (2008) Current Industrial Reports. Fertilizers and Related Chemicals: 2007.
"Table 1: Shipments and Production of Principal Fertilizers and Related Chemicals: 2003 to
2007." June, 2008. MQ325B(07)-5. Available online at:
http://www.census.gov/manufacturing/cir/historical data/mq325b/index.html.
U.S. EPA (2024) Greenhouse Gas Reporting Program. Aggregation of Reported Facility Level Data
under Subpart V -National Nitric Acid Production for Calendar Years 2017 through 2023. Office
of Air and Radiation, Office of Atmospheric Programs, U.S. Environmental Protection Agency,
Washington, D.C.
U.S. EPA (2018) Greenhouse Gas Reporting Program. Aggregation of Reported Facility Level Data
under Subpart V - National Nitric Acid Production for Calendar Years 2010 through 2016. Office
of Air and Radiation, Office of Atmospheric Programs, U.S. Environmental Protection Agency,
Washington, D.C.
U.S. EPA (2015) Greenhouse Gas Reporting Program Report Verification. Available online at
https://www.epa.gov/sites/production/files/2015-
07/documents/ghgrp verification factsheet.pdf.
U.S. EPA (2010) Available and Emerging Technologies for Reducing Greenhouse Gas Emissions from
the Nitric Acid Production Industry. Office of Air Quality Planning and Standards, U.S.
Environmental Protection Agency. Research Triangle Park, NC. December 2010. Available
online at: http://www.epa.gov/nsr/ghgdocs/nitricacid.pdf.
U.S. EPA (1998) Compilation of Air Pollutant Emission Factors, AP-42. Office of Air Quality Planning
and Standards, U.S. Environmental Protection Agency. Research Triangle Park, NC. February
1998.
Adipic Acid Production
ACC (2024) Business of Chemistry (Annual Data). American Chemistry Council, Arlington, VA.
Ard (2024) Personal communication, Howard Ard, Florida Department of Environmental Protection
and Amanda Chiu, U.S. Environmental Protection Agency. February 6, 2024.
Ascend (2023) "Ascend reaches GHG abatement milestone." Available online at
https://www.ascendmaterials.com/news/ascend-reaches-ghg-abatement-milestone.
December 1, 2023.
C&EN (1995) "Production of Top 50 Chemicals Increased Substantially in 1994." Chemical &
Engineering News, 73(15):17. April 10,1995.
References and Abbreviations 10-41
-------
C&EN (1994) "Top 50 Chemicals Production Rose Modestly Last Year." Chemical & Engineering
News, 72(15):13. April 11,1994.
C&EN (1993) "Top 50 Chemicals Production Recovered Last Year." Chemical & Engineering News,
71(15):11. April 12,1993.
C&EN (1992) "Production of Top 50 Chemicals Stagnates in 1991." Chemical & Engineering News,
70(15): 17. April 13,1992.
CMR (2001) "Chemical Profile: Adipic Acid." Chemical Market Reporter. July 16, 2001.
CMR (1998) "Chemical Profile: Adipic Acid." Chemical Market Reporter. June 15,1998.
CW (2005) "Product Focus: Adipic Acid." Chemical Week. May 4, 2005.
CW (1999) "Product Focus: Adipic Acid/Adiponitrile." Chemical Week, p. 31. March 10,1999.
Desai (2010, 2011) Personal communication. Mausami Desai, U.S. Environmental Protection
Agency and Adipic Acid Plant Engineers. 2010 and 2011.
ICIS (2007) "Adipic Acid." ICIS Chemical Business Americas. July 9, 2007.
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change.
[H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa,
Japan.
Reimer, R.A., Slaten, C.S., Seapan, M., Koch, T.A. and Triner, V.G. (1999) "Implementation of
Technologies for Abatement of N20 Emissions Associated with Adipic Acid Manufacture."
Proceedings of the 2nd Symposium on Non-C02 Greenhouse Gases (NCGG-2),
Noordwijkerhout, The Netherlands, 8-10 Sept. 1999, Ed. J. van Ham et al., Kluwer Academic
Publishers, Dordrecht, pp. 347-358.
RTI (2023) Expert judgment, RTI International. March 30, 2023.
Thiemens, M.H., and W.C. Trogler (1991) "Nylon production; an unknown source of atmospheric
nitrous oxide." Science 251:932-934.
United States Environmental Protection Agency (EPA) (2021 through 2024) Greenhouse Gas
Reporting Program. Subpart E Data. Office of Air and Radiation, Office of Atmospheric
Programs, U.S. Environmental Protection Agency, Washington, D.C. Available online at:
https://www.epa.gov/ghgreporting/ghg-reporting-program-data-sets.
U.S. EPA (2019, 2020) Greenhouse Gas Reporting Program. Subpart E, S-CEMS, BB, CC, LL Data Set
(XLSX) (Adipic Acid Tab). Office of Air and Radiation, Office of Atmospheric Programs, U.S.
Environmental Protection Agency, Washington, D.C. Available online at:
https://www.epa.gov/ghgreporting/ghg-reporting-program-data-sets.
U.S. EPA (2015) Greenhouse Gas Reporting Program Report Verification. Available online at
https://www.epa.gov/sites/production/files/2015-
07/documents/ghgrp verification factsheet.pdf.
10-42 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
U.S. EPA (2014 through 2018) Greenhouse Gas Reporting Program. Subpart E, S-CEMS, BB, CC, LL
Data Set (XLSX) (Adipic Acid Tab). Office of Air and Radiation, Office of Atmospheric Programs,
U.S. Environmental Protection Agency, Washington, D.C. Available online at:
http://www2.epa.gov/ghgreporting/ghg-reporting-program-data-sets.
U.S. EPA (2010 through 2013) Analysis of Greenhouse Gas Reporting Program data - Subpart E
(Adipic Acid), Office of Air and Radiation, Office of Atmospheric Programs, U.S. Environmental
Protection Agency, Washington, D.C.
Caprolactam, Glyoxaland Glyoxylie Acid Production
American Chemistry Council (ACC) (2024) Business of Chemistry (Annual Data). American
Chemistry Council, Arlington, VA.
AdvanSix (2024) AdvanSix's Hopewell Facility Fact Sheet. Retrieved from:
https://www.advansix.com/hopewell/about-us/on September 23, 2024.
BASF (2024) Welcome to BASF in Freeport Texas. Retrieved from https://www.basf.com/us/en/who-
we-are/organization/locations/featured-sites/Freeport.html on September 23, 2024.
ChemView (2021). Compilation of data submitted under TSCA in 2012 and 2016. Accessed April
2021. Available at https://chemview.epa.gov/chemview.
Cline, D. (2019) Firm to Clean Up and Market Former Fibrant Site. The Augusta Chronicle.
September 9, 2019. Retrieved from https://www.augustachronicle.com.
Ecofys, et al. (2009). Methodology for the free allocation of emission allowances in the EU ETSpost
2012: Sector Report for the Chemical Industry. Prepared by Ecofys, Fraunhofer Institute for
Systems and Innovation Research, and Oko-lnstitutforthe European Commission. November
2009. Available at https://ec.europa.eu/clima/system/files/2016-11/bm study-
chemicals en.pdf.
ICIS (2006) Chemical Profile - Caprolactam. October 15, 2006.
ICIS (2004) Chemical Profile - Caprolactam. January 5, 2004.
IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to
the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. [Stocker, T.F.,
D. Qin, G.-K. Plattner, M. Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M.
Midgley (eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY,
USA, 1535 pp.
IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to
the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. [S. Solomon,
D. Qin, M. Manning, Z. Chen, M. Marquis, K.B. Averyt, M. Tignorand H.L. Miller (eds.)].
Cambridge University Press. Cambridge, United Kingdom 996 pp.
IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change.
References and Abbreviations 10-43
-------
[H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa,
Japan.
RTI (2023) Expert judgment, RTI International. March 30, 2023.
Shaw Industries Group, Inc. (Shaw) (2015) "Shaw Carpet Recycling Facility Successfully Processes
Nylon and Polyester". July 13, 2015. Available online at: https://shawinc.com/Newsroom/Press-
Releases/Shaw-Carpet-Recycling-Facility-Successfully-Proces/.
Textile World (2000) "Evergreen Makes Nylon Live Forever". Textile World. October 1, 2000. Available
online at: https://www.textileworld.com/textile-world/textile-news/2000/1Q/evergreen-makes-
nylon-live-forever/.
U.S. EPA (2023) Revised Technical Support Document For Caprolactam, Glyoxal, And Glyoxylic Acid
Production: Supplemental Proposed Rule For The Greenhouse Gas Reporting Program. U.S.
Environmental Protection Agency. April 1, 2023.
Carbide Production and Consumption
IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to
the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. [Stocker, T.F.,
D. Qin, G.-K. Plattner, M. Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M.
Midgley (eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY,
USA, 1535 pp.
IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to
the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. [S. Solomon,
D. Qin, M. Manning, Z. Chen, M. Marquis, K.B. Averyt, M. Tignorand H.L. Miller (eds.)].
Cambridge University Press. Cambridge, United Kingdom 996 pp.
IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change.
[H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa,
Japan.
Biscay, Nicolas & Henry, Lucile & Adschiri, Tadafumi & Yoshimura, Masahiro & Aymonier, Cyril.
(2021). Behavior of Silicon Carbide Materials under Dry to Hydrothermal Conditions.
Nanomaterials. 11. 1351. doi: 10.3390/nano11051351.
Environment and Climate Change Canada (ECCC) (2022), Personal Communication between
Genevieve Leblanc-Power, Environment and Climate Change Canada and Mausami Desai and
Amanda Chiu, U.S. Environmental Protection Agency. April 12, 2022.
RTI (2023) Expert judgment, RTI International. March 30, 2023.
United States Census Bureau (1990 through 2023) USITC Trade DataWeb. Available online at:
http://dataweb.usitc.gov/.
10-44 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
USGS (1991a through 2021) Minerals Yearbook: Manufactured Abrasives Annual Report. U.S.
Geological Survey, Reston, VA. Available online at: https://www.usgs.gov/centers/national-
minerals-information-center/manufactured-abrasives-statistics-and-information.
USGS (1991b through 2021) Minerals Yearbook: Silicon Annual Report. U.S. Geological Survey,
Reston, VA. Available online at: http://minerals.usgs.gov/nninerals/pubs/connnnodity/silicon/.
USGS (2023a) 2022 Minerals Yearbook: Manufactured Abrasives (2022 advanced-release tables),
September 27, 2023. U.S. Geological Survey, Reston, VA. Available online at:
https://www.usgs.gov/centers/national-minerals-information-center/manufactured-abrasives-
statistics-and-information
USGS (2023b) 2022 Minerals Yearbook: Silicon (2022 advanced-release tables), November 27,
2023. U.S. Geological Survey, Reston, VA. Available online at:
https://www.usgs.gov/centers/national-minerals-information-center/silicon-statistics-and-
information
USGS (2024) Mineral Commodity Summaries: Abrasives (Manufactured). U.S. Geological Survey,
Reston, Va. January 2024.
USGS (2021 a) Mineral Commodity Summaries: Abrasives (Manufactured). U.S. Geological Survey,
Reston, Va. January 2021. Available online at:
https://pubs.usgs.gov/periodicals/mcs2021/mcs2021-abrasives.pdf.
Washington Mills (2023), North Grafton, MA. Available online at:
https://www.washingtonmills.com/silicon-carbide/sic-industries. Accessed on April 4, 2023.
Titanium Dioxide Production
Gambogi, J. (2002) Telephone communication. Joseph Gambogi, Commodity Specialist, U.S.
Geological Survey and Philip Groth, ICF International. November 2002.
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, Intergovernmental Panel on Climate Change. [H.S.
Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
RTI (2023) Expert judgment, RTI International. March 30, 2023.
United States Geological Survey (USGS) (2023) 2020 Minerals Yearbook: Titanium, 2020 tables-only
release, Table 1. U.S. Geological Survey, Reston, Va. March 2023.
USGS (2024) Mineral Commodity Summaries: Titanium and Titanium Dioxide. U.S. Geological
Survey, Reston, Va. January 2024.
USGS (1994 through 2022) Minerals Yearbook: Titanium. U.S. Geological Survey, Reston, VA.
Soda Ash Production
IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change.
References and Abbreviations 10-45
-------
[H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa,
Japan.
RTI (2023) Expert judgment, RTI International. March 30, 2023.
United States Geological Survey (USGS) (2024a) Mineral Commodity Summary: Soda Ash. U.S.
Geological Survey, Reston, VA. January 2024.
USGS (2024b) Mineral Industry Surveys: Soda Ash in June 2023. U.S. Geological Survey, Reston, VA.
August 2024.
USGS (2023a) Mineral Commodity Summary: Soda Ash. U.S. Geological Survey, Reston, VA.
January 2023.
USGS (2023b) Mineral Industry Surveys: Soda Ash in June 2023. U.S. Geological Survey, Reston, VA.
August 2023.
United States Geological Survey (USGS) (2022a) Mineral Commodity Summary: Soda Ash. U.S.
Geological Survey, Reston, VA. January 2022.
USGS (2022b) Mineral Industry Surveys: Soda Ash in June 2022. U.S. Geological Survey, Reston, VA.
August 2022.
USGS (2021) Mineral Industry Surveys: Soda Ash in April 2021. U.S. Geological Survey, Reston, VA.
July 2021.
USGS (2020) Mineral Industry Surveys: Soda Ash in April 2020. U.S. Geological Survey, Reston, VA.
July 2020.
USGS (2019) Mineral Industry Surveys: Soda Ash in April 2019. U.S. Geological Survey, Reston, VA.
July 2019.
USGS (2018a) Mineral Industry Surveys: Soda Ash in February 2018. U.S. Geological Survey, Reston,
VA. Accessed September 2018.
USGS (2017) Mineral Industry Surveys: Soda Ash in January 2017. U.S. Geological Survey, Reston,
VA. March 2017.
USGS (2016) Mineral Industry Surveys: Soda Ash in November 2016. U.S. Geological Survey,
Reston, VA. January 2017.
USGS (2015a) Mineral Industry Surveys: Soda Ash in July 2015. U.S. Geological Survey, Reston, VA.
September 2015.
USGS (1994 through 2015b, 2018b) Minerals Yearbook: Soda Ash Annual Report. U.S. Geological
Survey, Reston, VA.
USGS (1995c) Trona Resources in the Green River Basin, Southwest Wyoming. U.S. Department of
the Interior, U.S. Geological Survey. Open-File Report 95-476. Wiig, Stephen, Grundy, W.D.,
Dyni, John R.
10-46 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Petrochemical Production
ACC (2024) Business of Chemistry (Annual Data). American Chemistry Council, Arlington, VA.
AN (2014) About Acrylonitrile: Production. AN Group, Washington, D.C.
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change.
[H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa,
Japan.
Johnson, G. L. (2005 through 2010) Personal communication. Greg Johnson of Liskow& Lewis, on
behalf of the International Carbon Black Association (ICBA) and Caroline Cochran, ICF
International. September 2010.
Johnson, G. L. (2003) Personal communication. Greg Johnson of Liskow & Lewis, on behalf of the
International Carbon Black Association (ICBA) and Caren Mintz, ICF International. November
2003.
United States Environmental Protection Agency (EPA) (2024) Greenhouse Gas Reporting Program.
Aggregation of Reported Facility Level Data under Subpart X -National Petrochemical
Production for Calendar Years 2010 through 2023. Office of Air and Radiation, Office of
Atmospheric Programs, U.S. Environmental Protection Agency, Washington, D.C.
U.S. EPA (2015) Greenhouse Gas Reporting Program Report Verification. Available online at
https://www.epa.gov/sites/production/files/2015-
07/documents/ghgrp verification factsheet.pdf.
U.S. EPA (2008) Technical Support Document for the Petrochemical Production Sector: Proposed
Rule for Mandatory Reporting of Greenhouse Gases. U.S. Environmental Protection Agency.
September 2008.
U.S. EPA (2000) Economic Impact Analysis for the Proposed Carbon Black Manufacturing NESHAP,
U.S. Environmental Protection Agency. Research Triangle Park, NC. EPA-452/D-00-003. May
2000.
HCFC-22 Production
ARAP (2010) Electronic mail communication from Dave Stirpe, Executive Director, Alliance for
Responsible Atmospheric Policy to Deborah Ottingerof the U.S. Environmental Protection
Agency. September 10, 2010.
ARAP (2009) Electronic mail communication from Dave Stirpe, Executive Director, Alliance for
Responsible Atmospheric Policy to Deborah Ottingerof the U.S. Environmental Protection
Agency. September 21, 2009.
ARAP (2008) Electronic mail communication from Dave Stirpe, Executive Director, Alliance for
Responsible Atmospheric Policy to Deborah Ottingerof the U.S. Environmental Protection
Agency. October 17, 2008.
References and Abbreviations 10-47
-------
ARAP (2007) Electronic mail communication from Dave Stirpe, Executive Director, Alliance for
Responsible Atmospheric Policy to Deborah Ottingerof the U.S. Environmental Protection
Agency. October 2, 2007.
ARAP (2006) Electronic mail communication from Dave Stirpe, Executive Director, Alliance for
Responsible Atmospheric Policy to Sally Rand of the U.S. Environmental Protection Agency. July
11,2006.
ARAP (2005) Electronic mail communication from Dave Stirpe, Executive Director, Alliance for
Responsible Atmospheric Policy to Deborah Ottingerof the U.S. Environmental Protection
Agency. August 9, 2005.
ARAP (2004) Electronic mail communication from Dave Stirpe, Executive Director, Alliance for
Responsible Atmospheric Policy to Deborah Ottingerof the U.S. Environmental Protection
Agency. June 3, 2004.
ARAP (2003) Electronic mail communication from Dave Stirpe, Executive Director, Alliance for
Responsible Atmospheric Policy to Sally Rand of the U.S. Environmental Protection Agency.
August 18, 2003.
ARAP (2002) Electronic mail communication from Dave Stirpe, Executive Director, Alliance for
Responsible Atmospheric Policy to Deborah Ottingerof the U.S. Environmental Protection
Agency. August 7, 2002.
ARAP (2001) Electronic mail communication from Dave Stirpe, Executive Director, Alliance for
Responsible Atmospheric Policy to Deborah Ottingerof the U.S. Environmental Protection
Agency. August 6, 2001.
ARAP (2000) Electronic mail communication from Dave Stirpe, Executive Director, Alliance for
Responsible Atmospheric Policy to Sally Rand of the U.S. Environmental Protection Agency.
August 13, 2000.
ARAP (1999) Facsimile from Dave Stirpe, Executive Director, Alliance for Responsible Atmospheric
Policy to Deborah Ottinger Schaefer of the U.S. Environmental Protection Agency. September
23, 1999.
ARAP (1997) Letterfrom Dave Stirpe, Director, Alliance for Responsible Atmospheric Policyto
Elizabeth Dutrow of the U.S. Environmental Protection Agency. December 23,1997.
EPA (2015) Greenhouse Gas Reporting Program Report Verification. Available online at
https://www.epa.gov/sites/production/files/2015-
07/documents/ghgrp verification factsheet.pdf.
IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to
the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. [S. Solomon,
D. Qin, M. Manning, Z. Chen, M. Marquis, K.B. Averyt, M. Tignorand H.L. Miller (eds.)].
Cambridge University Press. Cambridge, United Kingdom 996 pp.
IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change.
10-48 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
[H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa,
Japan.
RTI (2008) "Verification of Emission Estimates of HFC-23 from the Production of HCFC-
22:Emissions from 1990 through 2006." Report prepared by RTI International for the Climate
Change Division. March 2008.
RTI (1997) "Verification of Emission Estimates of HFC-23 from the Production of HCFC-22:
Emissions from 1990 through 1996." Report prepared by Research Triangle Institute for the
Cadmus Group. November 25,1997; revised February 16,1998.
UNFCCC (2014) Report of the Conference of the Parties on its nineteenth session, held in Warsaw
from 11 to 23 November 2013. United Nations Framework Convention on Climate Change,
Warsaw. (FCCC/CP/2013/10/Add.3). January 31, 2014. Available online at:
http://unfccc.int/resource/docs/2013/cop19/eng/10a03.pdf.
Production of Fluorochemicals Other Than HCFC-22
3M (2024) 3M Global EHS Laboratory Response to EPA Data Request on Fluorochemical Emissions.
February 2024.
Daikin (2013) Major Source Operating Permit, Daikin America, Alabama Department of
Environmental Management, August 1, 2013.
http://lf.adem.alabama.gov/WebLink/DocView.aspx?id=29951882&dbid=0. (p. 11-1).
Honeywell (2012) Part 70 Operating Permit, Geismar Plant, Honeywell International Inc., Louisiana,
Louisiana Department of Environmental Quality, Page 13, January 28, 2011.
https://edms.deq.louisiana.gov/app/doc/view?doc=7812895.
Honeywell (2011) Part 70 Operating Permit, Baton Rouge Plant Honeywell International Inc.,
Louisiana Department of Environmental Quality, Page 25, October 16, 2012.
https://edms.deq.louisiana.gov/app/doc/view?doc=8579001.
ICI Americas (1993) New Permit, KLEA- 134a Plant, ICI Americas, St. Gabriel, Louisiana, Louisiana
Department of Environmental Quality, Page 44, May 28,1993.
https://edms.deq.louisiana.gov/app/doc/view?doc=1309650.
IPCC (2021) Climate Change 2021: The Physical Science Basis. Contribution of Working Group I to
the Sixth Assessment Report of the Intergovernmental Panel on Climate Change. [Masson-
Delmotte, V., P. Zhai, A. Pirani, S.L. Connors, C. Pean, S. Berger, N. Caud, Y. Chen, L. Goldfarb,
M.I. Gomis, M. Huang, K. Leitzell, E. Lonnoy, J.B.R. Matthews, T.K. Maycock, T. Waterfield, O.
Yelekgi, R. Yu, and B. Zhou (eds.)]. Cambridge University Press, Cambridge, United Kingdom and
New York, NY, USA, 2391 pp. doi:10.1017/9781009157896. Available from www.ipcc.ch. The
AR6 GWPs are listed in Table 7.SM.7, which appears on page 16 of the Supplementary Material.
IPCC (2019) 2079 Refinement to the 2006 IPCC Guidelines for National Greenhouse Gas
Inventories, Calvo Buendia, E., Tanabe, K., Kranjc, A., Baasansuren, J., Fukuda, M., Ngarize S.,
Osako, A., Pyrozhenko, Y., Shermanau, P. and Federici, S. (eds). Published: IPCC, Switzerland.
References and Abbreviations 10-49
-------
IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to
the Fifth Assessment Report of the Intergovernmental Panel on Climate Change [Stocker, T.F.,
D. Qin, G.-K. Plattner, M. Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M.
Midgley (eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY,
USA, 1535 pp. The GWPs are listed in Table 8.A.1 of Appendix 8.A: Lifetimes, Radiative
Efficiencies and Metric Values, which appears on pp. 731-737 of Chapter 8, "Anthropogenic and
Natural Radiative Forcing."
IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to
the Fourth Assessment Report of the Intergovernmental Panel on Climate Change [Core Writing
Team, Pachauri, R.K and Reisinger, A. (eds.)]. IPCC, Geneva, Switzerland, 104 pp.
McKenna (2022) A 3M Plant in Illinois Was the Country's Worst Emitter of a Climate-Killing
'Immortal' Chemical in 2021, Phil McKenna, Inside Climate News, December 29, 2022. [3M
Cordova ILfacility.] https://insideclimatenews.org/news/29122022/3m-cordova-illinois-pfas-
cf4-pollution/.
Perkins (1982) Perkins, B. L., Evaluation of Environmental Control Technologies for Commercial
Nuclear Fuel Conversion (UF6) Facilities, LA-9397-MS, October 1982 [030000442.pdf].
Rand (2007) 2004-2006 SF6 Data Summary, Project Memorandum Prepared by D. Knopman and K.
Smythe, RAND Corporation, for the National Electrical Manufacturers Association, June 2007.
SRI Consulting (2004) Chemical Economics Handbook (CEH) Market Research Report:
Fluorocarbons, R. Will, A. Kishi, S. Schlag. SRI Consulting, 2004.
U.S. EPA (2008) Survey of Producers of HFCs, PFCs, SF6 and NF3, 2008. Office of Air and Radiation,
Office of Atmospheric Programs, U.S. Environmental Protection Agency.
U.S. EPA (2023a) GHGRP Data Relevant to the AIM Act, Greenhouse Gas Reporting Program.
https://www.epa.gov/ghgreporting/ghgrp-data-relevant-aim-act. Last accessed 11/16/2023.
U.S. EPA (2023b) Vintaging Model for HFCs. 2023. Office of Atmospheric Programs, U.S.
Environmental Protection Agency.
U.S. EPA (2023c) Estimated layer-weighted substrate production by the semiconductor industry.
Office of Atmospheric Programs, Office of Atmospheric Programs, U.S. Environmental
Protection Agency.
U.S. EPA (2023d) SF6 Consumption by Users. 2023. SF6 consumption for 3 industries, Electric
Transmission and Distribution, Semiconductor Manufacturing, and Magnesium Production.
Office of Atmospheric Protection, U.S. Environmental Protection Agency.
U.S. EPA (2015) Greenhouse Gas Reporting Program Report Verification. Available online at
https://www.epa.gov/sites/production/files/2015-
07/documents/ghgrp verification factsheet.pdf.
U.S. EPA (1995) Protocol for Equipment Leak Emission Estimates. Office of Air and Radiation, Office
of Air Quality Planning and Standards, U.S. Environmental Protection Agency. Publication No.
EPA-453/R-95-017. November 1995.
10-50 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Vollmer et al. (2019) Vollmer, M. K., Bernard, F., Mitrevski, B., Steele, L. P., Trudinger, C. M.,
Reimann, S., Langenfelds, R. L., Krummel, P. B., Fraser, P. J., Etheridge, D. M., Curran, M. A. J.,
and Burkholder, J. B.: Abundances, emissions, and loss processes of the long-lived and potent
greenhouse gas octafluorooxolane (octafluorotetrahydrofuran, C-C4F80) in the atmosphere,
Atmos. Chem. Phys., 19, 3481-3492, https://doi.org/10.5194/acp-19-3481 -2019 , 2019.
Non-EOR Carbon Dioxide Utilization
ARI (1990 through 2010) C02 Use in Enhanced Oil Recovery. Deliverable to ICF International under
Task Order 102, July 15, 2011.
ARI (2007) C02-EOR: An Enabling Bridge forthe Oil Transition. Presented at"Modelingthe Oil
Transition—a DOE/EPA Workshop on the Economic and Environmental Implications of Global
Energy Transitions." Washington, D.C. April 20-21, 2007.
ARI (2006) CO2-EOR: An Enabling Bridge forthe Oil Transition. Presented at"Modelingthe Oil
Transition—a DOE/EPA Workshop on the Economic and Environmental Implications of Global
Energy Transitions." Washington, D.C. April 20-21, 2006.
Broadhead (2003) Personal communication. Ron Broadhead, Principal Senior Petroleum Geologist
and Adjunct faculty, Earth and Environmental Sciences Department, New Mexico Bureau of
Geology and Mineral Resources, and Robin Petrusak, ICF International. September 5, 2003.
COGCC (2014) Monthly C02 Produced by County (1999-2009). Available online at:
http://cogcc.state.co.us/COGCCReports/production.aspx?id=MonthlyCQ2ProdByCounty.
Accessed October 2014.
Denbury Resources Inc. (2002 through 2010) Annual Report: 2001 through 2009, Form 10-K.
Available online at: http://www.denbury.com/investor-relations/SEC-Filings/SEC-Filings-
Details/default.aspx?Filingld=9823015. Accessed September 2014.
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change.
[H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa,
Japan.
New Mexico Bureau of Geology and Mineral Resources (2006) Natural Accumulations of Carbon
Dioxide in New Mexico and Adjacent Parts of Colorado and Arizona: Commercial Accumulation
of C02. Available online at: http://geoinfo.nmt.edU/staff/broadhead/CQ2.html#commercial.
RTI (2023) Expert judgment, RTI International. March 30, 2023.
U.S. Environmental Protection Agency (EPA) (2024) Greenhouse Gas Reporting Program (GHGRP).
Aggregation of Reported Facility Level Data under Subpart PP -National Level C02 Transferred
for Food & Beverage Applications for Calendar Years 2010 through 2023. Office of Air and
Radiation, Office of Atmospheric Programs, U.S. Environmental Protection Agency, Washington,
D.C.
References and Abbreviations 10-51
-------
U.S. EPA (2015) Greenhouse Gas Reporting Program Report Verification. Available online at
https://www.epa.gov/sites/production/files/2Q15-
07/documents/ghgrp verification factsheet.pdf.
Phosphoric Acid Production
EFMA (2000) "Production of Phosphoric Acid." Best Available Techniques for Pollution Prevention
and Control in the European Fertilizer Industry. Booklet 4 of 8. European Fertilizer
Manufacturers Association. Available online at:
http://www.efnna.org/Publications/BAT%202000/Bat04/section04.asp.
Florida Institute of Phosphate Research (FIPR) (2003a) "Analyses of Some Phosphate Rocks."
Facsimile Gary Albarelli, Florida Institute of Phosphate Research, Bartow, Florida, to Robert
Lanza, ICF International. July 29, 2003.
FIPR (2003b) Florida Institute of Phosphate Research. Personal communication between Michael
Lloyd (Laboratory Manager, FIPR, Bartow, Florida) and Robert Lanza (ICF International) on
August 2003.
Golder Associates and M3 Engineering, Bayovar 12 Phosphate Project: N143-101 Updated Pre-
Feasibility Study, Issued June 28, 2016. Available at:
https://www.sec.gOv/Archives/edgar/data/1471603/000121716016000634/focusjune2016bayo
var techrep.htm. Accessed on October 7, 2020.
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change.
[H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa,
Japan.
NCDENR (2013) North Carolina Department of Environment and Natural Resources, Title V Air
Permit Review for PCS Phosphate Company, Inc. - Aurora. Available online at:
http://www.ncair.org/permits/permit reviews/PCS rev 08282012.pdf. Accessed on January 25,
2013.
Research Triangle Institute (RTI) (2024). Expert engineering determination.
United States Environmental Protection Agency (EPA) (2024) Greenhouse Gas Reporting Program.
Review of Reported Facility Level Data under Subpart Z -Annual Phosphoric Acid Production
from Phosphate Rock for Calendar Year 2023. Office of Air and Radiation, Office of Atmospheric
Programs, U.S. Environmental Protection Agency, Washington, D.C.
United States Geological Survey (USGS) (2024) Mineral Commodity Summaries: Phosphate Rock
2024. January 2024. U.S. Geological Survey, Reston, VA. Available online at:
https://www.usgs.gov/centers/nmic/phosphate-rock-statistics-and-information
USGS (2024b) Personal communication between Stephen Jasinski (USGS) and Mausami Desai
(EPA) on December 2, 2024.
10-52 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
USGS (2023) Mineral Commodity Summaries: Phosphate Rock 2023. January 2023. U.S. Geological
Survey, Reston, VA. Available online at: https://www.usgs.gov/centers/nmic/phosphate-rock-
statistics-and-information
USGS (2022) Mineral Commodity Summaries: Phosphate Rock 2022. January 2022. U.S. Geological
Survey, Reston, VA. Available online at: https://www.usgs.gov/centers/nmic/phosphate-rock-
statistics-and-information
USGS (2021a) Mineral Commodity Summaries: Phosphate Rock 2021. January 2021. U.S.
Geological Survey, Reston, VA. Available online at:
https://www.usgs.gov/centers/nmic/phosphate-rock-statistics-and-information.
USGS (2021 b) Personal communication between Stephen Jasinski (USGS) and Amanda Chiu (EPA)
on August 25,2021.
USGS (2020) Mineral Commodity Summaries: Phosphate Rock 2020. January 2020. U.S. Geological
Survey, Reston, VA. Available online at: https://www.usgs.gov/centers/nmic/phosphate-rock-
statistics-and-information.
USGS (2019a) Mineral Commodity Summaries: Phosphate Rock 2019. February 2019. U.S.
Geological Survey, Reston, VA. Available online at:
https://www.usgs.gov/centers/nmic/phosphate-rock-statistics-and-information.
USGS (2019b) Communication between Stephen Jasinski (USGS) and John Steller (EPA) on
November 15, 2019.
USGS (2018) Mineral Commodity Summaries: Phosphate Rock2018. January 2018. U.S. Geological
Survey, Reston, VA. Available online at: https://www.usgs.gov/centers/nmic/phosphate-rock-
statistics-and-information.
USGS (2017) Mineral Commodity Summaries: Phosphate Rock 2017. January 2017. U.S. Geological
Survey, Reston, VA. Available online at: https://www.usgs.gov/centers/nmic/phosphate-rock-
statistics-and-information.
USGS (2016) Mineral Commodity Summaries: Phosphate Rock 2016. January 2016. U.S. Geological
Survey, Reston, VA. Available online at: https://www.usgs.gov/centers/nmic/phosphate-rock-
statistics-and-information.
USGS (1994 through 2015b) Minerals Yearbook. Phosphate Rock Annual Report. U.S. Geological
Survey, Reston, VA.
USGS (2012) Personal communication between Stephen Jasinski (USGS) and Mausami Desai (EPA)
on October 12, 2012.
Iron and Steel Production and Metallurgical Coke
Production
American Coke and Coal Chemicals Institute (ACCCI) (2021) U.S. Coke Plants as of November
2021, ACCCI, Washington, D.C. November 2021.
References and Abbreviations 10-53
-------
American Iron and Steel Institute (AISI) (2004 through 2023) Annual Statistical Report, American
Iron and Steel Institute, Washington, D.C.
Carroll (2017) Personal communication, Colin P. Carroll, Director of Environment, Health and
Safety, American Iron and Steel Institute and John Steller, U.S. Environmental Protection
Agency, November 2017.
Carroll (2016) Personal communication, Colin P. Carroll, Director of Environment, Health and
Safety, American Iron and Steel Institute and Mausami Desai, U.S. Environmental Protection
Agency, December 2016.
IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to
the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. [Stocker, T.F.,
D. Qin, G.-K. Plattner, M. Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M.
Midgley (eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY,
USA, 1535 pp.
IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to
the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. [S. Solomon,
D. Qin, M. Manning, Z. Chen, M. Marquis, K.B. Averyt, M. Tignorand H.L. Miller (eds.)].
Cambridge University Press. Cambridge, United Kingdom, 996 pp.
IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change.
[H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa,
Japan.
IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change.
[H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa,
Japan.
IPCC/UNEP/OECD/IEA (1995) "Volume 3: Greenhouse Gas Inventory Reference Manual. Table 2-2."
IPCC Guidelines for National Greenhouse Gas Inventories. Intergovernmental Panel on Climate
Change, United Nations Environment Programme, Organization for Economic Co-Operation
and Development, International Energy Agency. IPCCWG1 Technical Support Unit, United
Kingdom.
RTI (2024) Expert judgment, RTI International. April 2024.
RTI (2023) Expert judgment, RTI International. March 30, 2023.
Steiner (2008) Personal communication, Bruce Steiner, Technical Consultant with the American
Iron and Steel Institute and Mausami Desai, U.S. Environmental Protection Agency, November
2008.
Tuck (2024) Personal communication, Candice Tuck, Commodity Specialist, U.S. Geological Survey
and Amanda Chiu, U.S. Environmental Protection Agency, November 29, 2024.
Tuck (2023a) Personal communication, Candice Tuck, Commodity Specialist, U.S. Geological
Survey and Amanda Chiu, U.S. Environmental Protection Agency, January 24, 2023.
10-54 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Tuck (2023b) Personal communication, Candice Tuck, Commodity Specialist, U.S. Geological
Survey and Amanda Chiu, U.S. Environmental Protection Agency, November 30, 2023.
United States Department of Energy (DOE) (2000) Energy and Environmental Profile of the U.S. Iron
and Steel Industry. Office of Industrial Technologies, U.S. Department of Energy. August 2000.
DOE/EE-0229.EIA.
United States Energy Information Administration (EIA) (1998 through 2019) Quarterly Coal Report:
October-December, Energy Information Administration, U.S. Department of Energy,
Washington, D.C.
U.S. EIA (2021 through 2024) Quarterly Coal Report: January - March, Energy Information
Administration, U.S. Department of Energy. Washington, D.C.
U.S. EIA (2020) Natural Gas Annual 2019. Energy Information Administration, U.S. Department of
Energy. Washington, D.C. September 2020.
U.S. EIA (2017b) Monthly Energy Review, December 2017, Energy Information Administration, U.S.
Department of Energy, Washington, D.C. DOE/EIA-0035(2015/12).
U.S. EIA (1992) Coal and lignite production. EIA State Energy Data Report 1992, Energy Information
Administration, U.S. Department of Energy, Washington, D.C.
United States Environmental Protection Agency (EPA) (2024). Greenhouse Gas Reporting Program.
Dataset as of August 16, 2024. Available online at: https://ghgdata.epa.gov/ghgp/.
EPA (2010) Carbon Content Coefficients Developed for EPA's Mandatory Reporting Rule. Office of
Air and Radiation, Office of Atmospheric Programs, U.S. Environmental Protection Agency,
Washington, D.C.
United States Geological Survey (USGS) (2024a) 2022 Mineral Commodities Summaries: Iron and
Steel. U.S. Geological Survey, Reston, VA. January 2023.
United States Geological Survey (USGS) (2024b) 2022 USGS Minerals Yearbook-Iron and Steel
Scrap (tables-only release). U.S. Geological Survey, Reston, VA.
United States Geological Survey (USGS) (2023a) 2022 Mineral Commodities Summaries: Iron and
Steel. U.S. Geological Survey, Reston, VA. January 2023.
USGS (2023b) 2027 USGS Minerals Yearbook - Iron and Steel Scrap (tables-only release). U.S.
Geological Survey, Reston, VA.
USGS (2022) 2020 USGS Minerals Yearbook - Iron and Steel Scrap (tables-only release). U.S.
Geological Survey, Reston, VA.
USGS (2021 a) 2021 Mineral Commodities Summaries: Iron and Steel. U.S. Geological Survey,
Reston, VA. January 2021.
USGS (2021 b) 2019 USGS Minerals Yearbook- Iron and Steel Scrap (tables-only release). U.S.
Geological Survey, Reston, VA.
USGS (2020a) 2018 USGS Minerals Yearbook-Iron and Steel Scrap (tables-only release). U.S.
Geological Survey, Reston, VA.
References and Abbreviations 10-55
-------
USGS (2020b) 207 7 USGS Minerals Yearbook - Iron and Steel. U.S. Geological Survey, Reston, VA.
USGS (1991 through 2020) USGS Minerals Yearbook- Iron and Steel Scrap. U.S. Geological Survey,
Reston, VA.
Ferroalloy Production
Environmental Protection Agency (EPA) (2024). Facility Level Information on Greenhouse gases Tool
(FLIGHT). Dataset as of August 16, 2024. Available online at: https://ghgdata.epa.gov/ghgp/.
IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to
the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. [Stocker, T.F.,
D. Qin, G.-K. Plattner, M. Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M.
Midgley (eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY,
USA, 1535 pp.
IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to
the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. [S. Solomon,
D. Qin, M. Manning, Z. Chen, M. Marquis, K.B. Averyt, M. Tignorand H.L. Miller (eds.)].
Cambridge University Press. Cambridge, United Kingdom 996 pp.
IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change.
[H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa,
Japan.
Onder, H., and E.A. Bagdoyan (1993) Everything You've Always Wanted to Know about Petroleum
Coke. Allis Mineral Systems.
RTI (2023) Expert judgment, RTI International. March 30, 2023.
United States Geological Survey (USGS) (2024a) Mineral Industry Survey: Silicon in December 2023.
U.S. Geological Survey, Reston, VA. March 2024.
USGS (2024b) 2021 Minerals Yearbook: Ferroalloys (tables-only release). U.S. Geological Survey,
Reston, VA. September 2024.
USGS (2023) 2022 Minerals Yearbook: Silicon (tables-only release). U.S. Geological Survey, Reston,
VA. November 2023.
USGS (2022b) 2020 Minerals Yearbook: Ferroalloys (tables-only release). U.S. Geological Survey,
Reston, VA. May 2023.
USGS (2022) 2021 Mineral Commodity Summaries. Silicon. U.S. Geological Survey, Reston, VA.
January 2022.
USGS (2021) 2020 Mineral Commodity Summaries: Silicon. U.S. Geological Survey, Reston, VA.
January 2021.
USGS (2013) 2013 Minerals Yearbook: Chromium. U.S. Geological Survey, Reston, VA. March 2016.
USGS (1996 through 2023) Minerals Yearbook: Silicon. U.S. Geological Survey, Reston, VA.
10-56 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Aluminum Production
EPA (2024) Greenhouse Gas Reporting Program (GHGRP). Envirofacts, Subpart: F Aluminum
Production. Available online at: https://www.epa.gov/enviro/greenhouse-gas-subpart-f
EPA (2015) Greenhouse Gas Reporting Program Report Verification. Available online at
https://www.epa.gov/sites/production/files/2015-
07/documents/ghgrp verification factsheet.pdf.
IPCC (2019) 207 9 Refinement to the 2006IPCC Guidelines for National Greenhouse Gas
Inventories. The National Greenhouse Gas Inventories Programme, The Intergovernmental
Panel on Climate Change. [Calvo Buendia, E., Tanabe, K., Kranjc, A., Baasansuren, J., Fukuda,
M., Ngarize, S., Osako, A., Pyrozhenko, Y., Shermanau, P. and Federici, S. (eds.)]. Switzerland.
IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change.
[H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa,
Japan.
USGS (2024) 2023 Minerals Commodity Summaries: Aluminum. U.S. Geological Survey, Reston,
VA.
USGS (2023) Minerals Yearbook: Aluminum Annual Report. U.S. Geological Survey, Reston, VA.
USGS (2022) Mineral Commodity Summaries 2022. U.S. Geological Survey, Reston VA.
USGS (2021) 2020 Mineral Commodity Summaries: Aluminum. U.S. Geological Survey, Reston, VA.
USGS (2021) 2019 Mineral Commodity Summaries: Aluminum. U.S. Geological Survey, Reston, VA.
USGS (2020) Minerals Yearbook: Aluminum Annual Report. U.S. Geological Survey, Reston, VA.
USGS (2020) Mineral Industry Surveys: Aluminum in December 2020. U.S. Geological Survey,
Reston VA. December 2020
USGS (2020) 2019 Mineral Commodity Summaries: Aluminum. U.S. Geological Survey, Reston, VA.
USGS (2019) Minerals Yearbook: Aluminum Annual Report. U.S. Geological Survey, Reston, VA.
USGS (2018) Minerals Yearbook: Aluminum Annual Report. U.S. Geological Survey, Reston, VA.
USGS (2017) Minerals Yearbook: Aluminum Annual Report. U.S. Geological Survey, Reston, VA.
USGS (2016) Minerals Yearbook: Aluminum Annual Report. U.S. Geological Survey, Reston, VA.
USGS (2015) Minerals Yearbook: Aluminum Annual Report. U.S. Geological Survey, Reston, VA.
USGS (2014) Minerals Yearbook: Aluminum Annual Report. U.S. Geological Survey, Reston, VA.
USGS (2013) Minerals Yearbook: Aluminum Annual Report. U.S. Geological Survey, Reston, VA.
USGS (2012) Minerals Yearbook: Aluminum Annual Report. U.S. Geological Survey, Reston, VA.
USGS (2011) Minerals Yearbook: Aluminum Annual Report. U.S. Geological Survey, Reston, VA.
USGS (2010) Minerals Yearbook: Aluminum Annual Report. U.S. Geological Survey, Reston, VA.
References and Abbreviations 10-57
-------
USGS
2009
Minerals
Yearbook: Aluminum Annual Report.
U.S. Geological Survey, Reston, VA.
USGS
2008
Minerals
Yearbook: Aluminum Annual Report.
U.S. Geological Survey, Reston, VA.
USGS
2007
Minerals
Yearbook: Aluminum Annual Report.
U.S. Geological Survey, Reston, VA.
USGS
2006
Minerals
Yearbook: Aluminum Annual Report.
U.S. Geological Survey, Reston, VA.
USGS
2005
Minerals
Yearbook: Aluminum Annual Report.
U.S. Geological Survey, Reston, VA.
USGS
2004
Minerals
Yearbook: Aluminum Annual Report.
U.S. Geological Survey, Reston, VA.
USGS
2003
Minerals
Yearbook: Aluminum Annual Report.
U.S. Geological Survey, Reston, VA.
USGS
2002
Minerals
Yearbook: Aluminum Annual Report.
U.S. Geological Survey, Reston, VA.
USGS
2001
Minerals
Yearbook: Aluminum Annual Report.
U.S. Geological Survey, Reston, VA.
USGS
2000
Minerals
Yearbook: Aluminum Annual Report.
U.S. Geological Survey, Reston, VA.
USGS
1999
Minerals
Yearbook: Aluminum Annual Report.
U.S. Geological Survey, Reston, VA.
USGS
1998
Minerals
Yearbook: Aluminum Annual Report.
U.S. Geological Survey, Reston, VA.
USGS
1997
Minerals
Yearbook: Aluminum Annual Report.
U.S. Geological Survey, Reston, VA.
USGS
1996
Minerals
Yearbook: Aluminum Annual Report.
U.S. Geological Survey, Reston, VA.
USGS
1995
Minerals
Yearbook: Aluminum Annual Report.
U.S. Geological Survey, Reston, VA.
USGS
1994
Minerals
Yearbook: Aluminum Annual Report.
U.S. Geological Survey, Reston, VA.
USGS
1993
Minerals
Yearbook: Aluminum Annual Report.
U.S. Geological Survey, Reston, VA.
USGS
1992
Minerals
Yearbook: Aluminum Annual Report.
U.S. Geological Survey, Reston, VA.
USGS
1991
Minerals
Yearbook: Aluminum Annual Report.
U.S. Geological Survey, Reston, VA.
USGS
1990
Minerals
Yearbook: Aluminum Annual Report.
U.S. Geological Survey, Reston, VA.
Magnesium Production
ARB (2015) "Magnesium casters successfully retool for a cleaner future." California Air Resources
Board News Release. Release # 15-07. February 5, 2015. Accessed October 2017. Available
online at: https://www.arb.ca.gov/newsrel/newsrelease.php?id=704.
Bartos S., C. Laush, J. Scharfenberg, and R. Kantamaneni (2007) "Reducing greenhouse gas
emissions from magnesium die casting." Journal of Cleaner Production, 15: 979-987, March.
EPA (2020) Envirofacts. Greenhouse Gas Reporting Program (GHGRP), Subpart T: Magnesium
Production and Processing. Available online at: https://www.epa.gov/enviro/greenhouse-gas-
customized-search. Accessed on October 2020.
EPA (2015) Greenhouse Gas Reporting Program Report Verification. Available online at
https://www.epa.gov/sites/production/files/2015-
07/documents/ghgrp verification factsheet.pdf.
10-58 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Gjestland, H. and D. Magers (1996) "Practical Usage of Sulphur [Sulfur] Hexafluoride for Melt
Protection in the Magnesium Die Casting Industry." #13, 1996 Annual Conference Proceedings,
International Magnesium Association. Ube City, Japan.
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change.
[H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa,
Japan.
RAND (2002) RAND Environmental Science and Policy Center, "Production and Distribution of SF6
by End-Use Applications" Katie D. Smythe. International Conference on SF6 and the
Environment: Emission Reduction Strategies. San Diego, CA. November 21-22, 2002.
USGS (1995 through 2024) Minerals Yearbook: Magnesium Annual Report. U.S. Geological Survey,
Reston, VA. Available online at:
http://minerals.usgs.goV/minerals/pubs/commodity/magnesium/index.html#mis.
USGS (2010b) Mineral Commodity Summaries: Magnesium Metal. U.S. Geological Survey, Reston,
VA. Available online at: http://minerals.usgs.gov/minerals/pubs/commodity/magnesium/mcs-
2010-mgmet.pdf.
Lead Production
Dutrizac, J.E., V. Ramachandran, and J.A. Gonzalez (2000) Lead-Zinc 2000. The Minerals, Metals,
and Materials Society.
IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change.
[H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa,
Japan.
Morris, D., F.R. Steward, and P. Evans (1983) Energy Efficiency of a Lead Smelter. Energy 8(5):337-
349.
RTI (2023) Expert judgment, RTI International. March 30, 2023.
Sjardin, M. (2003) C02 Emission Factors for Non-Energy Use in the Non-Ferrous Metal, Ferroalloys
and Inorganics Industry. Copernicus Institute. Utrecht, the Netherlands.
Ullman (1997) Ullman's Encyclopedia of Industrial Chemistry: Fifth Edition. Volume A5. John Wiley
and Sons.
United States Geological Survey (USGS) (2024a) 2023 Mineral Commodity Summary, Lead. U.S.
Geological Survey, Reston, VA. January 2024.
USGS (2024b) 2022 Minerals Yearbook, Lead -Advance Data Release. U.S. Geological Survey,
Reston, VA. August 2024.
USGS (2015) 2014 Mineral Commodity Summary, Lead. U.S. Geological Survey, Reston, VA. January
2015.
References and Abbreviations 10-59
-------
USGS (1994 - 2023) Mineral Commodity Summary, Lead. U.S. Geological Survey, Reston, VA.
USGS (1994-2023) Minerals Yearbook: Lead Annual Report. U.S. Geological Survey, Reston, VA.
Zinc Production
American Zinc Recycling (AZR) (2021) Summary of Company History. Available online at
https://azr.com/our-historv/. Accessed on March 16, 2021.
AZR (2020) Personal communication. Erica Livingston, American Zinc Recycling and Amanda Chiu,
U.S. Environmental Protection Agency. October 29, 2020.
American Zinc Products (AZP) (2021) American Zinc Products Marks First Anniversary of Zinc
Production. Available online at https://americanzincproducts.com/american-zinc-products-
marks-first-anniversarv-of-zinc-production/. Accessed on March 1, 2022.
Befesa (2024) Personal communication. Eric Hunsberger, Befesa Zinc US Inc. and Amanda Chiu,
U.S. Environmental Protection Agency. September 13 and 25, 2024.
Befesa (2023) Personal communication. Eric Hunsberger, Befesa Zinc US Inc. and Amanda Chiu,
U.S. Environmental Protection Agency. September 19, 2023.
Befesa (2022) Personal communication. Eric Hunsberger, Befesa Zinc US Inc. and Amanda Chiu,
U.S. Environmental Protection Agency. November 8, 2022.
Horsehead Corp. (2016) Form 10-K, Annual Report for the Fiscal Year Ended December 31, 2015.
Available online at:
https://www.sec.gov/Archives/edgar/data/1385544/000119312516725704/d236839d10k.htm.
Submitted on January 25, 2017.
Horsehead Corp. (2015) Form 10-K, Annual Report forthe FiscalYear Ended December31, 2014.
Available online at:
http://www.sec.gov/Archives/edgar/data/1385544/0001385544150000Q5/zinc-
2014123110k.htm. Submitted on March 2, 2015.
Horsehead Corp. (2014) Form 10-K, Annual Report forthe FiscalYear Ended December31, 2013.
Available online at:
http://www.sec.gOV/Archives/edgar/data/1385544/000138554414000003/zinc-
2013123110k.htm. Submitted on March 13, 2014.
Horsehead Corp. (2013) Form 10-K, Annual Report for the FiscalYear Ended December 31, 2012.
Available online at:
http://www.sec.gOv/Archives/edgar/data/1385544/000119317513110431/0001193175-13-
110431-index.htm. Submitted March 18, 2013.
Horsehead Corp. (2012a) Form 10-K, Annual Report forthe FiscalYear Ended December 31, 2011.
Available online at:
http://www.sec.gOv/Archives/edgar/data/1385544/000119312512107345/d293011 d10k. htm.
Submitted on March 9, 2012.
10-60 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Horsehead Corp. (2012b) Horsehead's NewZinc Plant and its Impact on the Zinc Oxide Business.
February 22, 2012. Available online at:
http://www.horsehead.net/downloadAttachmentNDO.php?ID=118. Accessed on September
10, 2015.
Horsehead Corp. (2011) 10-KAnnual Report forthe Fiscal Year Ended December31, 2010.
Available online at: http://google.brand.edgar-online.com/default.aspx?synn=zinc. Submitted
on March 16, 2011.
Horsehead Corp. (2010a) 10-KAnnual Reportforthe FiscalYear Ended December31, 2009.
Available online at: http://google.brand.edgar-online.com/default.aspx?sym=zinc. Submitted
on March 16, 2010.
Horsehead Corp. (2010b) Horsehead Holding Corp. Provides Update on Operations at its Monaca,
PA Plant. July 28, 2010. Available online at:
http://www. horsehead .net/press releases. php?showall=no&news=&l D=65.
Horsehead Corp (2009) 10-KAnnual Reportforthe FiscalYear Ended December31, 2008. Available
online at:
https://www.sec.gov/Archives/edgar/data/1385544/000095015209002674/l35087ae10vk.htm.
Submitted on March 16, 2009.
Horsehead Corp (2008) 10-KAnnual Reportforthe FiscalYear Ended December31, 2007. Available
online at: http://google.brand.edgar-online.com/default.aspx?sym=zinc. Submitted on March
31,2008.
Horsehead Corp (2007) Registration Statement (General Form) S-1. Available online at
http://google.brand.edgar-online.com/default.aspx?sym=zinc. Submitted on April 13, 2007.
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change.
[H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa,
Japan.
PIZO (2021) Personal communication. Thomas Rheaume, Arkansas Department of Energy and
Environment and Amanda Chiu, U.S. Environmental Protection Agency. February 16, 2021.
PIZO (2012) Available online at http://pizotech.com/index.html. Accessed on October 10, 2012.
Recycling Today (2020) "AZR to restart for zinc recycling plant in North Carolina." March 6, 2020.
https://www.recyclingtoday.com/article/american-zinc-recycling-restarting-north-carolina-
plant-2020/. Accessed October 10, 2020.
Recycling Today (2017) "Horsehead announces corporate name change to American Zinc
Recycling." May 3, 2017. https://www.recyclingtoday.com/article/horsehead-changes-name-
american-zinc-recycling/. Accessed September 19, 2022.
Steel Dust Recycling (SDR) (2024) Personal communication. Jeremy Whitten, Steel Dust Recycling
LLC and Amanda Chiu, U.S. Environmental Protection Agency. September 13 and 19, 2023.
References and Abbreviations 10-61
-------
Steel Dust Recycling (SDR) (2023) Personal communication. Jeremy Whitten, Steel Dust Recycling
LLC and Amanda Chiu, U.S. Environmental Protection Agency. September 20 and 25, 2023.
SDR (2022) Personal communication. Jeremy Whitten, Steel Dust Recycling LLC and Amanda Chiu,
U.S. Environmental Protection Agency. October 10, 2022.
SDR (2021) Personal communication. Jeremy Whitten, Steel Dust Recycling LLC and Amanda Chiu,
U.S. Environmental Protection Agency. January 8, 2021.
SDR (2018) Personal communication. Jeremy Whitten, Steel Dust Recycling LLC and John Steller,
U.S. Environmental Protection Agency. October 25, 2018.
SDR (2017) Personal communication. Jeremy Whitten, Steel Dust Recycling LLC and John Steller,
U.S. Environmental Protection Agency. January 26, 2017.
Electronics Industry
Burton, C.S., and R. Beizaie (2001)"EPA's PFC Emissions Model (PEVM)v. 2.14: Description and
Documentation" prepared for Office of Global Programs, U. S. Environmental Protection
Agency, Washington, DC. November 2001.
Citigroup Smith Barney (2005) Global Supply/Demand Model for Semiconductors. March 2005.
DisplaySearch (2010) DisplaySearch Q4'09 Quarterly FPD Supply/Demand and Capital Spending
Report. DisplaySearch, LLC.
Doering, R. and Nishi, Y (2000) "Handbook of Semiconductor ManufacturingTechnology", Marcel
Dekker, New York, USA, 2000.
EPA (2024) Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2022. U.S. Environmental
Protection Agency, Washington, DC. EPA-430-R-23-002.
EPA (2010) Technical Support Document for Process Emissions from Electronics Manufacture (e.g.,
Micro-Electro-Mechanical Systems, Liquid Crystal Displays, Photovoltaics, and
Semiconductors). U.S. Environmental Protection Agency, Washington, DC.
EPA (2006) Uses and Emissions of Liquid PFC Heat Transfer Fluids from the Electronics Sector. U.S.
Environmental Protection Agency, Washington, DC. EPA-430-R-06-901.
EPA Greenhouse Gas Reporting Program (GHGRP) Envirofacts. Subpart I: Electronics Manufacture.
Available online at: https://enviro.epa.gov/facts/ghg/search.html.
IPCC (2019) 207 9 Refinement to the 2006IPCC Guidelines for National Greenhouse Gas
Inventories. The National Greenhouse Gas Inventories Programme, The Intergovernmental
Panel on Climate Change. Calvo Buendia, E., Tanabe, K., Kranjc, A., Baasansuren, J., Fukuda,
M., Ngarize, S., Osako, A., Pyrozhenko, Y., Shermanau, P. and Federici, S. (eds). Published:
IPCC, Switzerland.
IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to
the Fifth Assessment Report of the Intergovernmental Panel on Climate Change [Stocker, T.F., D.
Qin, G.-K. Plattner, M. Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley
10-62 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
(eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, 1535
pp.
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change.
[H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa,
Japan.
ITRS (2007, 2008, 2011, 2013) International Technology Roadmap for Semiconductors: 2006
Update, January 2007; International Technology Roadmap for Semiconductors: 2007 Edition,
January 2008; International Technology Roadmap for Semiconductors: 2011, January 2012;
Update, International Technology Roadmap for Semiconductors: 2013 Edition, Available online
at: https://www.semiconductors.org/resources/2007-international-technology-roadmap-for-
semiconductors-itrs/. These and earlier editions and updates are available online at:
https://www.semiconductors.org/resources/7fwp resource types=utilization-
reports&fwp paged=2. Information about the number of interconnect layers foryears 1990-
2010 is contained in Burton and Beizaie, 2001. PEVM is updated using new editions and
updates of the ITRS, which are published annually.
Platzer, Michaela D. (2015) U.S. Solar Photovoltaic Manufacturing: Industry Trends, Global
Competition, Federal Support. Congressional Research Service. January 27, 2015.
https://fas.org/sgp/crs/misc/R42509.pdf.
SEMI-Semiconductor Equipment and Materials Industry (2023) World Fab Forecast, September
2023 Edition
SEMI - Semiconductor Equipment and Materials Industry (2021) World Fab Forecast, June 2021
Edition.
SEMI - Semiconductor Equipment and Materials Industry (2018) World Fab Forecast, June 2018
Edition.
SEMI - Semiconductor Equipment and Materials Industry (2017) World Fab Forecast, August2018
Edition.
SEMI - Semiconductor Equipment and Materials Industry (2016) World Fab Forecast, May 2017
Edition.
SEMI - Semiconductor Equipment and Materials Industry (2013) World Fab Forecast, May 2013
Edition.
SEMI - Semiconductor Equipment and Materials Industry (2012) World Fab Forecast, August2012
Edition.
Semiconductor Industry Association (SIA) (2009-2011) STATS: SICAS Capacity and Utilization Rates
Q1-Q4 2008, Q1-Q4 2009, Q1-Q4 2010. Available online at:
http://www.semiconductors.org/industry statistics/semiconductor capacity utilization sicas
reports/.
References and Abbreviations 10-63
-------
United States Census Bureau (USCB) (2011, 2012, 2014, 2015, 2016, 2017, 2018, 2019, 2020, 2021,
2022, 2023) Historical Data: Quarterly Survey of Plant Capacity Utilization. Available online at:
https://www.census.gov/progranns-surveys/qpc.htnnl.
VLSI Research, Inc. (2012) Worldwide Silicon Demand. August 2012
Substitution of Ozone Depleting Substances
Carrier (2023) New Carrier AquaSnap 30RC Air Cooled Chiller Helps Maximize Building Space while
Delivering Efficiency and Sustainability, January 10, 2023.
EPA (2025) Assessment of Foams Post-Life Emissions Charge Size Calculation Correction Within
the Vintaging Model. Prepared for U.S. EPA's Stratospheric Protection Division by ICF under EPA
Contract Number 68HERH19D0029.
EPA (2024a) Summary of Updates to the Window Units End-use in the Vintaging Model. Prepared for
U.S. EPA's Stratospheric Protection Division by ICF under EPA Contract Number
68HERH19D0029.
EPA (2024b) Proposed Updates to the Vintaging Model Cold Storage End-Use in the Vintaging
Model. Prepared for U.S. EPA's Stratospheric Protection Division by ICF under EPA Contract
Number 68HERH19D0029.
EPA (2018) EPA's Vintaging Model of ODS Substitutes: A Summary of the 2017 Peer Review. Office of
Air and Radiation. Document Number EPA-400-F-18-001. Available online at:
https://www.epa.gov/sites/production/files/2018-09/documents/epas-vintaging-model-of-ods-
substitutes-peer-review-factsheet.pdf.
IPCC (2019) 207 9 Refinement to the 2006IPCC Guidelines for National Greenhouse Gas
Inventories. The National Greenhouse Gas Inventories Programme, The Intergovernmental
Panel on Climate Change. Calvo Buendia, E., Tanabe, K., Kranjc, A., Baasansuren, J., Fukuda,
M., Ngarize, S., Osako, A., Pyrozhenko, Y., Shermanau, P. and Federici, S. (eds). Published:
IPCC, Switzerland.
IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to
the Fifth Assessment Report of the Intergovernmental Panel on Climate Change [Stocker, T.F., D.
Qin, G.-K. Plattner, M. Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley
(eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, 1535
pp.
IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to
the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. [S. Solomon,
D. Qin, M. Manning, Z. Chen, M. Marquis, K.B. Averyt, M. Tignorand H.L. Miller (eds.)].
Cambridge University Press. Cambridge, United Kingdom 996 pp.
IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change.
[H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa,
Japan.
10-64 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Electrical Equipment
CARB (2023). California Greenhouse Gas Emission Inventory-2023 Edition. Accessed March 2024.
Available online at: https://ww2.arb.ca.gov/ghg-inventory-data.
EPA (2022) Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020. EPA430-R-22-003.
Available online at: https://www.epa.gov/ghgemissions/inventory-us-greenhouse-gas-
emissions-and-sinks-1990-2020.
Harnisch and Eisenhauer, "Natural CF4 and SF6 on Earth," GEOPHYSICAL RESEARCH LETTERS,
VOL. 25, NO.13, PAGES 2401-2404, JULY 1, 1998.
https://agupubs.onlinelibrary.wiley.com/doi/pdf/10.1029/98GL01779.
U.S. Department of Homeland Security (2019) Homeland Infrastructure Foundation-Level Data
(HIFLD). Accessed March 2021. Available online at: https://hifld-
geoplatform.hub.arcgis.com/maps/bd24d1 a282c54428b024988d32578e59
U.S. Department of Homeland Security (2020) Homeland Infrastructure Foundation-Level Data
(HIFLD). Accessed October 2021. Available online at: https://hifld-
geoplatform.hub.arcgis.com/maps/bd24d1 a282c54428b024988d32578e59
U.S. Department of Homeland Security (2021) Homeland Infrastructure Foundation-Level Data
(HIFLD). Accessed September 2022. Available online at: https://hifld-
geoplatform.hub.arcgis.com/maps/bd24d1 a282c54428b024988d32578e59
U.S. Department of Homeland Security (2022) Homeland Infrastructure Foundation-Level Data
(HIFLD). Accessed September 2023. Available online at: https://hifld-
geoplatform.hub.arcgis.com/maps/bd24d1 a282c54428b024988d32578e59
U.S. Department of Homeland Security (2023) Homeland Infrastructure Foundation-Level Data
(HIFLD). Accessed August 2024. Available online at: https://hifld-
geoplatform.hub.arcgis.com/maps/bd24d1 a282c54428b024988d32578e59
Hu, L., Ottinger, D., Bogle, S., Montzka, S., DeCola, P., Dlugokencky, E., Andrews, A., Thoning, K.,
Sweeney, C., Dutton, G., Aepli, L., and Crotwell, A. (2022) "Declining, seasonal-varying
emissions of sulfur hexafluoride from the United States point to a new mitigation opportunity."
EGUsphere [preprint]. Available online at: https://doi.org/10.5194/egusphere-2022-862.
IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to
the Fifth Assessment Report of the Intergovernmental Panel on Climate Change [Stocker, T.F., D.
Qin, G.-K. Plattner, M. Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley
(eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, 1535
pp.
IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to
the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. S. Solomon,
D. Qin, M. Manning, Z. Chen, M. Marquis, K.B. Averyt, M. Tignorand H.L. Miller (eds.).
Cambridge University Press. Cambridge, United Kingdom 996 pp.
References and Abbreviations 10-65
-------
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change. H.S.
Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Hayama, Kanagawa, Japan.
IPCC (1996) Climate Change 1995: The Science of Climate Change. Intergovernmental Panel on
Climate Change, J.T. Houghton, L.G. Meira Filho, B.A. Callander, N. Harris, A. Kattenberg, and K.
Maskell (eds.). Cambridge University Press. Cambridge, United Kingdom.
Levin et al. (2010) "The Global SF6Source Inferred from Long-term High Precision Atmospheric
Measurements and its Comparison with Emission Inventories." Atmospheric Chemistry and
Physics, 10: 2655-2662.
Middleton, B. (2000) Cold Weather Applications of Gas Mixture (SFs/Ak, SFs/CF4) Circuit Breakers: A
User Utility's Perspective [Conference Presentation]. The U.S. Environmental Protection
Agency's Conference on SF6 and the Environment: Emission Reduction Strategies, San Diego,
CA, United States. Available online at: https://www.epa.gov/sites/default/files/2016-
02/documents/conf00 middleton.pdf
O'Connell, P., F. Heil, J. Henriot, G. Mauthe, H. Morrison, L. Neimeyer, M. Pittroff, R. Probst, J.P.
Tailebois (2002) SFs in the Electric Industry, Status 2000, CIGRE. February 2002.
Ottinger D, Averyt, M. & Harris, D. (2014). Trends in emissions offluorinated GHGs reported under
the Greenhouse Gas Reporting Program: Patterns and potential causes. Submitted to the
Seventh International Symposium on Non-C02 Greenhouse Gases (NCGG-7), Amsterdam,
Netherlands.
RAND (2004) "Trends in SF6 Sales and End-Use Applications: 1961-2003," Katie D. Smythe.
International Conference on SF6 and the Environment: Emission Reduction Strategies. RAND
Environmental Science and Policy Center, Scottsdale, AZ. December 1 -3, 2004.
UDI (2017) 2017 UDI Directory of Electric Power Producers and Distributors, 125th Edition, Platts.
UDI (2013)2073 UDI Directory of Electric Power Producers and Distributors^ 21st Edition, Platts.
UDI (2010) 2070 UDI Directory of Electric Power Producers and Distributors, 118th Edition, Platts.
UDI (2007) 2007 UDI Directory of Electric Power Producers and Distributors, 115th Edition, Platts.
UDI (2004) 2004 UDI Directory of Electric Power Producers and Distributors, 112th Edition, Platts.
UDI (2001)2007 UDI Directory of Electric Power Producers and Distributors, 109th Edition, Platts.
SF6 and PFCs from Other Product Use
E-3 Sentry (AWACS). (2015, September). U.S. Air Force. https://www.af.mil/About-Us/Fact-
Sheets/Display/Article/104504/e-3-sentrv-awacs/.
IPCC (2019) 2079 Refinement to the 2006 IPCC Guidelines for National Greenhouse Gas
Inventories. The National Greenhouse Gas Inventories Programme, The Intergovernmental
Panel on Climate Change. Calvo Buendia, E., Tanabe, K., Kranjc, A., Baasansuren, J., Fukuda,
10-66 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
M., Ngarize, S., Osako, A., Pyrozhenko, Y., Shermanau, P. and Federici, S. (eds). Published:
IPCC, Switzerland.
IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change.
[H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa,
Japan.
Workman, R.L. et al. Particle Data Group). Prog. Theor. Exp. Phys. 2022, 083C01. Available online at:
https://pdg.lbl.gov/2022/reviews/rpp2022-rev-particle-detectors-accel.pdf
U.S. Department of Energy (DOE). 2022. Federal Energy Management Program (FEMP). Available
online at: https://www.energy.gov/femp/federal-comprehensive-annual-energy-reporting-
requirements.
Nitrous Oxide from Product Uses
CGA (2003) "CGA Nitrous Oxide Abuse Hotline: CGA/NWSA Nitrous Oxide Fact Sheet."
Compressed Gas Association. November 3, 2003.
CGA (2002) "CGA/NWSA Nitrous Oxide Fact Sheet." Compressed Gas Association. March 25, 2002.
Heydorn, B. (1997) "Nitrous Oxide—North America." Chemical Economics Handbook, SRI
Consulting. May 1997.
IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to
the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. [Stocker, T.F.,
D. Qin, G.K. Plattner, M. Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley
(eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, 1535
pp.
IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to
the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. [S. Solomon,
D. Qin, M. Manning, Z. Chen, M. Marquis, K.B. Averyt, M. Tignorand H.L. Miller (eds.)].
Cambridge University Press. Cambridge, United Kingdom 996 pp.
IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change.
[H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa,
Japan.
Ottinger (2024) Personal communication. Deborah Ottinger, U.S. Environmental Protection Agency
and Karen Schaffner, RTI. December 3, 2024.
RTI (2023) Expert judgment, RTI International. March 30, 2023.
Tupman, M. (2002) Personal communication. Martin Tupman, Airgas Nitrous Oxide and Laxmi
Palreddy, ICF International. July 3, 2002.
References and Abbreviations 10-67
-------
Industrial Processes and Product Use Sources of Precursor
Greenhouse Gases
EPA (2024) "Criteria pollutants National Tier 1 for 1970 - 2024." National Emissions Inventory (NEI)
Air Pollutant Emissions Trends Data. Office of Air Quality Planning and Standards, February
2024. Available online at: https://www.epa.gov/air-emissions-inventories/air-pollutant-
emissions-trends-data.
EPA (2023a) EPA's Emissions Inventory System (EIS) to National Inventory Report (NIR) Mapping file
EIS_NIR_mapping.xlsx. U.S. Environmental Protection Agency. Washington, D.C."
EPA (2023b) "2020 National Emissions Inventory Technical Support Document: Introduction."
Office of Air Quality Planning and Standards, March 2023. Available online at:
https://www.epa.gov/system/files/documents/2023-
01/NEI2020 TSD Sectionl lntroduction.pdf.
Agriculture
Enteric Fermentation
Archibeque, S. (2011) Personal Communication. Shawn Archibeque, Colorado State University, Fort
Collins, Colorado and staff at ICF International.
Crutzen, P.J., I. Aselmann, and W. Seiler (1986) Methane Production by Domestic Animals, Wild
Ruminants, Other Herbivores, Fauna, and Humans. Tellus, 38B:271-284.
Donovan, K. (1999) Personal Communication. Kacey Donovan, University of California at Davis and
staff at ICF International.
Donovan, K. and L. Baldwin. (1999) "Results of the AAMOLLY Model Runs for the Enteric
Fermentation Model"; University of California, Davis, 1999.
Doren, P.E., J. F. Baker, C. R. Long and T. C. Cartwright (1989) Estimating Parameters of Growth
Curves of Bulls, J Animal Science 67:1432-1445.
Enns, M. (2008) Personal Communication. Dr. Mark Enns, Colorado State University and staff at ICF
International.
EPA (2002) Quality Assurance/Quality Control and Uncertainty Management Plan for the U.S.
Greenhouse Gas Inventory: Procedures Manual for Quality Assurance/Quality Control and
Uncertainty Analysis, U.S. Greenhouse Gas Inventory Program, U.S. Environmental Protection
Agency, Office of Atmospheric Programs, EPA430-R-02-007B, June 2002.
ERG (2021) Updated Other Animal Population Distribution Methodology. ERG, Lexington, MA.
ERG (2016) Development of Methane Conversion Rate Scaling Factor and Diet-Related Inputs to
the Cattle Enteric Fermentation Model for Dairy Cows, Dairy Heifers, and Feedlot Animals. ERG,
Lexington, MA. December 2016.
10-68 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Galyean and Gleghorn (2001) Summary of the 2000 Texas Tech University Consulting Nutritionist
Survey. Texas Tech University. Available online at
http://www.depts.ttu.edu/afs/burnett center/progress reports/bc12.pdf. June 2009.
Holstein Association (2010) History of the Holstein Breed (website). Available online at:
http://www.holsteinusa.com/holstein breed/breedhistorv.html. Accessed September 2010.
ICF (2006) Cattle Enteric Fermentation Model: Model Documentation. Prepared by ICF
International for the Environmental Protection Agency. June 2006.
ICF (2003) Uncertainty Analysis of 2001 Inventory Estimates of Methane Emissions from Livestock
Enteric Fermentation in the U.S. Memorandum from ICF International to the Environmental
Protection Agency. May 2003.
IFEEDER and Decision Innovation Solutions (2021) Animal Feed/Food Consumption and COVID-19
Impact Analysis. Available online at:
https://www.ifeeder.org/ifeeder/assets/file/public/research/210301-final-report-ifeeder-
animal-feed-food-consumption-covid-19.pdf.
IPCC (2019) 207 9 Refinement to the 2006IPCC Guidelines for National Greenhouse Gas
Inventories. The National Greenhouse Gas Inventories Programme, The Intergovernmental
Panel on Climate Change. [CalvoBuendia, E., Tanabe, K., Kranjc, A., Baasansuren, J., Fukuda,
M., Ngarize S., Osako, A., Pyrozhenko, Y., Shermanau, P. and Federici, S. (eds)]. Switzerland.
IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to
the Fifth Assessment Report of the Intergovernmental Panel on Climate Change [Stocker, T.F., D.
Qin, G.-K. Plattner, M. Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley
(eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, 1535
pp.
IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to
the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. S. Solomon,
D. Qin, M. Manning, Z. Chen, M. Marquis, K.B. Averyt, M. Tignorand H.L. Miller (eds.).
Cambridge University Press. Cambridge, United Kingdom 996 pp.
IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change. H.S.
Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Hayama, Kanagawa, Japan.
Johnson, D. (2002) Personal Communication. Don Johnson, Colorado State University, Fort Collins,
and ICF International.
Johnson, D. (1999) Personal Communication. Don Johnson, Colorado State University, Fort Collins,
and David Conneely, ICF International.
Kebreab E., K. A. Johnson, S. L. Archibeque, D. Pape, and T. Wirth (2008) Model for estimating
enteric methane emissions from United States dairy and feedlot cattle. J. Anim. Sci. 86: 2738-
2748.
Lippke, H., T. D. Forbes, and W. C. Ellis. (2000) Effect of supplements on growth and forage intake by
stocker steers grazing wheat pasture. J. Anim. Sci. 78:1625-1635.
References and Abbreviations 10-69
-------
NASEM (2021) National Academies of Sciences, Engineering, and Medicine. 2021. Nutrient
Requirements of Dairy Cattle: Eighth Revised Edition. Washington, DC: The National Academies
Press. Available online at: https://doi.org/10.17226/25806.
National Bison Association (1999) Total Bison Population—1999. Report provided during personal
email communication with Dave Carter, Executive Director, National Bison Association, July 19,
2011.
Pinchak, W.E., D. R. Tolleson, M. McCloy, L. J. Hunt, R. J. Gill, R. J. Ansley, and S. J. Bevers (2004)
Morbidity effects on productivity and profitability of stocker cattle grazing in the southern
plains. J. Anim. Sci. 82:2773-2779.
Platter, W. J., J. D. Tatum, K. E. Belk, J. A. Scanga, and G. C. Smith (2003) Effects of repetitive use of
hormonal implants on beef carcass quality, tenderness, and consumer ratings of beef
palatability. J. Anim. Sci. 81:984-996.
Preston, R.L. (2010) What's The Feed Composition Value of That Cattle Feed? Beef Magazine, March
1, 2010. Available at: http://beefmagazine.com/nutrition/feed-composition-tables/feed-
composition-value-cattle-0301.
Skogerboe, T. L., L. Thompson, J. M. Cunningham, A. C. Brake, V. K. Karle (2000) The effectiveness of
a single dose of doramectin pour-on in the control of gastrointestinal nematodes in yearling
stocker cattle. Vet. Parasitology 87:173-181.
Soliva, C.R. (2006) Report to the attention of IPCC about the data set and calculation method used
to estimate methane formation from enteric fermentation of agricultural livestock population
and manure management in Swiss agriculture. On behalf of the Federal Office for the
Environment (FOEN), Berne, Switzerland.
Tedeschi, L.O. and D.G. Fox (2020) The Ruminant Nutrition System, Volume 1: An Applied Model for
Predicting Nutrient Requirements and Feed Utilization in Ruminants: Third Edition. Ann Arbor,
Ml: XanEdu. Available online at: https://nutritionmodels.com/rns.html.
U.S. Department of Agriculture (USDA) (2024) Quick Stats: Agricultural Statistics Database.
National Agriculture Statistics Service, U.S. Department of Agriculture. Washington, D.C.
Available online at http://quickstats.nass.usda.gov/. Accessed June 2024.
U.S. Department of Agriculture (USDA) (2023) Quick Stats: Agricultural Statistics Database.
National Agriculture Statistics Service, U.S. Department of Agriculture. Washington, D.C.
Available online at http://quickstats.nass.usda.gov/. Accessed May-June 2023.
USDA (2021) Economic Research Service Dairy Data. Available online at:
https://www.ers.usda.gov/data-products/dairy-data/. Accessed May 2021.
USDA (2019) 1987, 1992, 1997, 2002, 2007,2012, and 2017 Census of Agriculture. National
Agriculture Statistics Service, U.S. Department of Agriculture. Washington, D.C. Available
online at: https://www.nass.usda.gov/AgCensus/index.php. May 2019.
USDA (1996) Beef Cow/Calf Health and Productivity Audit (CHAPA): Forage Analyses from Cow/Calf
Herds in 18 States. National Agriculture Statistics Service, U.S. Department of Agriculture.
Washington, D.C. Available online at http://www.aphis.usda.gov/vs/ceah/cahm. March 1996.
10-70 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
USDA:APHIS:VS (2010) Beef 2007-08, Part V: Reference of Beef Cow-calf Management Practices in
the United States, 2007-08. USDA-APHIS-VS, CEAH. Fort Collins, CO.
USDA:APHIS:VS (2002) Reference of 2002 Dairy Management Practices. USDA-APHIS-VS, CEAH.
Fort Collins, CO. Available online at http://www.aphis.usda.gov/vs/ceah/cahm.
USDA:APHIS:VS (1998) Beef '97, Parts I-IV. USDA-APHIS-VS, CEAH. Fort Collins, CO. Available
online at http://www.aphis.usda.gov/animal health/nahms/beefcowcalf/index.shtml#beef97.
USDA:APHIS:VS (1996) Reference of 1996 Dairy Management Practices. USDA-APHIS-VS, CEAH.
Fort Collins, CO. Available online at http://www.aphis.usda.gov/vs/ceah/cahm.
USDA:APHIS:VS (1994) Beef Cow/Calf Health and Productivity Audit. USDA-APHIS-VS, CEAH. Fort
Collins, CO. Available online at http://www.aphis.usda.gov/vs/ceah/cahm.
USDA:APHIS:VS (1993) Beef Cow/Calf Health and Productivity Audit. USDA-APHIS-VS, CEAH. Fort
Collins, CO. August 1993. Available online at http://www.aphis.usda.gov/vs/ceah/cahm.
Vasconcelos and Galyean (2007) Nutritional recommendations of feedlot consulting nutritionists:
The 2007 Texas Tech University Study. J. Anim. Sci. 85:2772-2781.
Manure Management
ASAE (1998) ASAE Standards 1998, 45th Edition. American Society of Agricultural Engineers. St.
Joseph, Ml.
Bryant, M.P., V.H. Varel, R.A. Frobish, and H.R. Isaacson (1976) In H.G. Schlegel (ed.)]; Seminar on
Microbial Energy Conversion. E. Goltz KG. Gottingen, Germany.
Bush, E. (1998) Personal communication with Eric Bush, Centers for Epidemiology and Animal
Health, U.S. Department of Agriculture regarding National Animal Health Monitoring System's
(NAHMS) Swine '95 Study.
EPA (2023) AgSTAR Anaerobic Digester Database. Available online at:
https://www.epa.gov/agstar/livestock-anaerobic-digester-database. Accessed August 2023.
EPA (2008) Climate Leaders Greenhouse Gas Inventory Protocol Offset Project Methodology for
Project Type Managing Manure with Biogas Recovery Systems.
EPA (2005) National Emission Inventory—Ammonia Emissions from Animal Agricultural Operations,
Revised Draft Report. U.S. Environmental Protection Agency. Washington, D.C. April 22, 2005.
EPA (2002a) Development Document for the Final Revisions to the National Pollutant Discharge
Elimination System (NPDES) Regulation and the Effluent Guidelines for Concentrated Animal
Feeding Operations (CAFOS). U.S. Environmental Protection Agency. EPA-821-R-03-001.
December 2002.
EPA (2002b) Cost Methodology for the Final Revisions to the National Pollutant Discharge
Elimination System Regulation and the Effluent Guidelines for Concentrated Animal Feeding
Operations. U.S. Environmental Protection Agency. EPA-821-R-03-004. December 2002.
References and Abbreviations 10-71
-------
EPA (1992) Global Methane Emissions from Livestock and Poultry Manure, Office of Air and
Radiation, U.S. Environmental Protection Agency. February 1992.
ERG (2023) Summary of Data Processing and Proposed Integration of 2018 Beef Feedlot and Poultry
Waste Management System Data into the Manure Management Greenhouse Gas Inventory.
Memorandum to EPA from ERG, December 2023.
ERG (2021) Updated Other Animal Population Distribution Methodology. Memorandum to EPA from
ERG.
ERG (2019) "Incorporation of USDA2016 ARMS Dairy Data into the Manure Management
Greenhouse Gas Inventory." Memorandum to USDAOCE and EPA from ERG, December 2019.
ERG (2018) "Incorporation of USDA 2009 ARMS Swine Data into the Manure Management
Greenhouse Gas Inventory." Memorandum to USDAOCE and EPA from ERG, November 2018.
ERG (2010a) "Typical Animal Mass Values for Inventory Swine Categories." Memorandum to EPA
from ERG. July 19, 2010.
ERG (2010b) Telecon with William Boyd of USDA NRCS and Cortney Itle of ERG Concerning
Updated VS and Nex Rates. August 8, 2010.
ERG (2010c) "Updating Current Inventory Manure Characteristics new USDA Agricultural Waste
Management Field Handbook Values." Memorandum to EPA from ERG. August 13, 2010.
ERG (2008) "Methodology for Improving Methane Emissions Estimates and Emission Reductions
from Anaerobic Digestion System for the 1990-2007 Greenhouse Gas Inventory for Manure
Management." Memorandum to EPA from ERG. August 18, 2008.
ERG (2003a) "Methodology for Estimating Uncertainty for Manure Management Greenhouse Gas
Inventory." Contract No. GS-10F-0036, Task Order 005. Memorandum to EPA from ERG,
Lexington, MA. September 26, 2003.
ERG (2003b) "Changes to Beef Calves and Beef Cows Typical Animal Mass in the Manure
Management Greenhouse Gas Inventory." Memorandum to EPA from ERG, October 7, 2003.
ERG (2001) Summary of development of MDP Factor for methane conversion factor calculations.
ERG, Lexington, MA. September 2001.
ERG (2000a) Calculations: Percent Distribution of Manure for Waste Management Systems. ERG,
Lexington, MA. August 2000.
ERG (2000b) Discussion of Methodology for Estimating Animal Waste Characteristics (Summary of
Bo Literature Review). ERG, Lexington, MA. June 2000.
Groffman, P.M., R. Brumme, K. Butterbach-Bahl, K.E. Dobbie, A.R. Mosier, D. Ojima, H. Papen, W.J.
Parton, K.A. Smith, and C. Wagner-Riddle (2000) "Evaluating annual nitrous oxide fluxes at the
ecosystem scale." Global Biogeochemical Cycles, 14(4):1061-1070.
Hashimoto, A.G. (1984) "Methane from Swine Manure: Effect of Temperature and Influent Substrate
Composition on Kinetic Parameter (k)." Agricultural Wastes, 9:299-308.
10-72 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Hashimoto, A.G., V.H. Varel, and Y.R. Chen (1981) "Ultimate Methane Yield from Beef Cattle
Manure; Effect of Temperature, Ration Constituents, Antibiotics and Manure Age." Agricultural
Wastes, 3:241-256.
Hill, D.T. (1984) "Methane Productivity of the Major Animal Types." Transactions of the ASAE,
27(2):530-540.
Hill, D.T. (1982) "Design of Digestion Systems for Maximum Methane Production." Transactions of
the ASAE, 25(1 ):226-230.
IPCC (2019) 207 9 Refinement to the 2006IPCC Guidelines for National Greenhouse Gas
Inventories. The National Greenhouse Gas Inventories Programme, The Intergovernmental
Panel on Climate Change. [CalvoBuendia, E., Tanabe, K., Kranjc, A., Baasansuren, J., Fukuda,
M., Ngarize S., Osako, A., Pyrozhenko, Y., Shermanau, P. and Federici, S. (eds)]. Switzerland.
IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to
the Fifth Assessment Report of the Intergovernmental Panel on Climate Change [Stocker, T.F., D.
Qin, G.-K. Plattner, M. Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley
(eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, 1535
pp.IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change.
[H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa,
Japan.
IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change. H.S.
Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Hayama, Kanagawa, Japan.
Morris, G.R. (1976) Anaerobic Fermentation of Animal Wastes: A Kinetic and Empirical Design
Fermentation. M.S. Thesis. Cornell University.
National Bison Association (1999) Total Bison Population—1999. Report provided during personal
email communication with Dave Carter, Executive Director, National Bison Association July 19,
2011.
Ott, S.L. (2000) Dairy '96 Study. Stephen L. Ott, Animal and Plant Health Inspection Service, U.S.
Department of Agriculture. June 19, 2000.
Robertson, G. P. and P. M. Groffman (2015) Nitrogen transformations. Soil Microbiology, Ecology,
and Biochemistry, pages 421-446. Academic Press, Burlington, Massachusetts, USA.
Safley, L.M., Jr. (2000) Personal Communication. Deb Bartram, ERG and L.M. Safley, President, Agri-
Waste Technology. June and October 2000.
Sweeten, J. (2000) Personal Communication. John Sweeten, Texas A&M University and Indra Mitra,
ERG. June 2000.
UEP (1999) Voluntary Survey Results—Estimated Percentage Participation/Activity. Caged Layer
Environmental Management Practices, Industry data submissions for EPA profile development,
United Egg Producers and National Chicken Council. Received from John Thorne, Capitolink.
June 2000.
References and Abbreviations 10-73
-------
USDA (2023a) Quick Stats: Agricultural Statistics Database. National Agriculture Statistics Service,
U.S. Department of Agriculture. Washington, D.C. Available online at:
http://quickstats.nass.usda.gov/.
USDA (2023b) Chicken and Eggs 2022 Summary. National Agriculture Statistics Service, U.S.
Department of Agriculture. Washington, D.C. February 2023. Available online at:
https://www.nass.usda.gov/Publications/.
USDA (2023c) Poultry - Production and Value 2022 Summary. National Agriculture Statistics
Service, U.S. Department of Agriculture. Washington, D.C. April 2023. Available online at:
https://www.nass.usda.gov/Publications/.
USDA (2021 a) Chicken and Eggs 2020 Summary. National Agriculture Statistics Service, U.S.
Department of Agriculture. Washington, D.C. February 2021. Available online at:
http://www.nass.usda.gov/Publications/index.asp.
USDA (2021 b) Poultry - Production and Value 2020 Summary. National Agriculture Statistics
Service, U.S. Department of Agriculture. Washington, D.C. April 2021. Available online at:
http://www.nass.usda.gov/Publications/index.asp.
USDA (2019a) Chicken and Eggs 2018 Summary. National Agriculture Statistics Service, U.S.
Department of Agriculture. Washington, D.C. February 2019. Available online at:
http://www.nass.usda.gov/Publications/index.php.
USDA (2019b) Poultry - Production and Value 2018 Summary. National Agriculture Statistics
Service, U.S. Department of Agriculture. Washington, D.C. April 2019. Available online at:
http://www.nass.usda.gov/Publications/index.php.
USDA (2019c) Chicken and Eggs 2013-2017 Summary. National Agriculture Statistics Service, U.S.
Department of Agriculture. Washington, D.C. June 2019. Available online at:
http://www.nass.usda.gov/Publications/index.php.
USDA (2019d) 1987,1992, 1997, 2002, 2007, 2012, and 2017 Census of Agriculture. National
Agriculture Statistics Service, U.S. Department of Agriculture. Washington, D.C. Available
online at: https://www.nass.usda.gov/AgCensus/index.php. May 2019.
USDA (2018) Poultry - Production and Value 2017 Summary. National Agriculture Statistics Service,
U.S. Department of Agriculture. Washington, D.C. April 2018. Available online at:
http://www.nass.usda.gov/Publications/index.php.
USDA (2017) Poultry - Production and Value 2016 Summary. National Agriculture Statistics Service,
U.S. Department of Agriculture. Washington, D.C. April 2017. Available online at:
http://www.nass.usda.gov/Publications/index.php.
USDA (2016) Poultry - Production and Value 2015 Summary. National Agriculture Statistics Service,
U.S. Department of Agriculture. Washington, D.C. April 2016. Available online at:
http://www.nass.usda.gov/Publications/index.php.
USDA (2015) Poultry - Production and Value 2014 Summary. National Agriculture Statistics Service,
U.S. Department of Agriculture. Washington, D.C. April 2015. Available online at:
http://www.nass.usda.gov/Publications/index.php.
10-74 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
USDA (2014) Poultry - Production and Value 2013 Summary. National Agriculture Statistics Service,
U.S. Department of Agriculture. Washington, D.C. April 2014. Available online at:
http://www.nass.usda.gov/Publications/index.php.
USDA (2013a) Chicken and Eggs 2012 Summary. National Agriculture Statistics Service, U.S.
Department of Agriculture. Washington, D.C. February 2013. Available online at:
http://www.nass.usda.gov/Publications/index.php.
USDA (2013b) Poultry - Production and Value 2012 Summary. National Agriculture Statistics
Service, U.S. Department of Agriculture. Washington, D.C. April 2013. Available online at:
http://www.nass.usda.gov/Publications/index.php.
USDA (2012a) Chicken and Eggs 2011 Summary. National Agriculture Statistics Service, U.S.
Department of Agriculture. Washington, D.C. February 2012. Available online at:
http://www.nass.usda.gov/Publications/index.php.
USDA (2012b) Poultry - Production and Value 2011 Summary. National Agriculture Statistics
Service, U.S. Department of Agriculture. Washington, D.C. April 2012. Available online at:
http://www.nass.usda.gov/Publications/index.php.
USDA (2011 a) Chicken and Eggs 2010 Summary. National Agriculture Statistics Service, U.S.
Department of Agriculture. Washington, D.C. February 2011. Available online at:
http://www.nass.usda.gov/Publications/index.php.
USDA (2011 b) Poultry - Production and Value 2010 Summary. National Agriculture Statistics
Service, U.S. Department of Agriculture. Washington, D.C. April 2011. Available online at:
http://www.nass.usda.gov/Publications/index.php.
USDA (2010a) Chicken and Eggs 2009 Summary. National Agriculture Statistics Service, U.S.
Department of Agriculture. Washington, D.C. February 2010. Available online at:
http://www.nass.usda.gov/Publications/index.php.
USDA (2010b) Poultry - Production and Value 2009 Summary. National Agriculture Statistics
Service, U.S. Department of Agriculture. Washington, D.C. April 2010. Available online at:
http://www.nass.usda.gov/Publications/index.php.
USDA (2009a) Chicken and Eggs 2008 Summary. National Agriculture Statistics Service, U.S.
Department of Agriculture. Washington, D.C. February 2009. Available online at:
http://www.nass.usda.gov/Publications/index.php.
USDA (2009b) Poultry - Production and Value 2008 Summary. National Agriculture Statistics
Service, U.S. Department of Agriculture. Washington, D.C. April 2009. Available online at:
http://www.nass.usda.gov/Publications/index.php.
USDA (2009c) Chicken and Eggs - Final Estimates 2003-2007. National Agriculture Statistics
Service, U.S. Department of Agriculture. Washington, D.C. March 2009. Available online at:
https://www.nass.usda.gov/Publications/index.php.
USDA (2009d) Poultry Production and Value—Final Estimates 2003-2007. National Agriculture
Statistics Service, U.S. Department of Agriculture. Washington, D.C. May 2009. Available online
at: https://www.nass.usda.gov/Publications/index.php.
References and Abbreviations 10-75
-------
USDA (2008) Agricultural Waste Management Field Handbook, National Engineering Handbook
(NEH), Part 651. Natural Resources Conservation Service, U.S. Department of Agriculture.
USDA (2004a) Chicken and Eggs—Final Estimates 1998-2003. National Agriculture Statistics
Service, U.S. Department of Agriculture. Washington, D.C. April 2004. Available online at:
https://www.nass.usda.gov/Publications/index.php.
USDA (2004b) Poultry Production and Value—Final Estimates 1998-2002. National Agriculture
Statistics Service, U.S. Department of Agriculture. Washington, D.C. April 2004. Available online
at: https://www.nass.usda.gov/Publications/index.php.
USDA (1999) Poultry Production and Value—Final Estimates 1994-97. National Agriculture
Statistics Service, U.S. Department of Agriculture. Washington, D.C. March 1999. Available
online at: https://www.nass.usda.gov/Publications/index.php.
USDA (1998) Chicken and Eggs—Final Estimates 1994-97. National Agriculture Statistics Service,
U.S. Department of Agriculture. Washington, D.C. December 1998. Available online at:
https://www.nass.usda.gov/Publications/index.php.
USDA (1996) Agricultural Waste Management Field Handbook, National Engineering Handbook
(NEH), Part 651. Natural Resources Conservation Service, U.S. Department of Agriculture. July
1996.
USDA (1995a) Poultry Production and Value—Final Estimates 1988-1993. National Agriculture
Statistics Service, U.S. Department of Agriculture. Washington, D.C. March 1995. Available
online at: https://www.nass.usda.gov/Publications/index.php.
USDA (1995b) Chicken and Eggs—Final Estimates 1988-1993. National Agriculture Statistics
Service, U.S. Department of Agriculture. Washington, D.C. December 1995. Available online at:
https://www.nass.usda.gov/Publications/index.php.
USDA (1994) Sheep and Goats—Final Estimates 1989-1993. National Agriculture Statistics Service,
U.S. Department of Agriculture. Washington, D.C. January 31,1994. Available online at:
https://www.nass.usda.gov/Publications/index.php.
USDA APHIS (2003) Sheep 2001, Part I: Reference of Sheep Management in the United States, 2001
and Part IV: Baseline Reference of 2001 Sheep Feedlot Health and Management. USDA-APHIS-
VS. Fort Collins, CO. #N356.0702. Available online at
http://www.aphis.usda.gov/animal health/nahms/sheep/index.shtml#sheep2001.
USDAAPHIS (2000) Layers '99—Part II: References of 1999Table Egg Layer Management in the U.S.
USDA-APHIS-VS. Fort Collins, CO. Available online at
http://www.aphis.usda.gov/animal health/nahms/poultry/downloads/layers99/Layers99 dr Pa
rtll.pdf.
USDAAPHIS (1996) Swine '95: Grower/Finisher Part II: Reference of 1995 U.S. Grower/Finisher
Health & Management Practices. USDA-APHIS-VS. Fort Collins, CO. Available online at:
http://www.aphis.usda.gov/animal health/nahms/swine/downloads/swine95/Swine95 dr Part
ll.pdf.
10-76 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Rice Cultivation
Baicich, P. (2013) The Birds and Rice Connection. Bird Watcher's Digest. Available online at:
http://www.usarice.com/doclib/194/6867.pdf.
Brockwell, P.J., and R.A. Davis (2016) Introduction to time series and forecasting. Springer.
Cantens, G. (2004 through 2005) Personal Communication. Janet Lewis, Assistant to Gaston
Cantens, Vice President of Corporate Relations, Florida Crystals Company and ICF
International.
Cheng, K., S.M. Ogle, W.J. Parton, G. Pan. (2014) "Simulating greenhouse gas mitigation potentials
for Chinese croplands usingthe DAYCENT ecosystem model." Global Change Biology 20:948-
962.
Cheng, K., S.M. Ogle, W.J. Parton and G. Pan. (2013) "Predicting methanogenesis from rice paddies
usingthe DAYCENT ecosystem model." Ecological Modelling 261 -262:19-31.
Del Grosso, S.J., S.M. Ogle, W.J. Parton, and F.J. Breidt (2010) "Estimating Uncertainty in N20
Emissions from U.S. Cropland Soils." Global Biogeochemical Cycles, 24, GB1009,
doi: 10.1029/2009GB003544.
Deren, C. (2002) Personal Communication and Dr. Chris Deren, Everglades Research and
Education Centre at the University of Florida and Caren Mintz, ICF International. August 15,
2002.
Fitzgerald, G.J., K. M. Scow & J. E. Hill (2000) "Fallow Season Straw and Rice Management Effects on
Methane Emissions in California Rice." Global biogeochemical cycles, 14 (3), 767-776.
Fleskes, J.P., Perry, W.M., Petrik, K.L., Spell, R., and Reid, F. (2005) Change in area of winter-flood
and dry rice in the northern Central Valley of California determined by satellite imagery.
California Fish and Game, 91: 207-215.
Gonzalez, R. (2007 through 2014) Email correspondence. Rene Gonzalez, Plant Manager, Sem-Chi
Rice Company and ICF International.
Hardke, J.T. (2015) Trends in Arkansas rice production, 2014. B.R. Wells Arkansas Rice Research
Studies 2014. Norman, R.J. and Moldenhauer, K.A.K. (Eds.). Research Series 626, Arkansas
Agricultural Experiment Station, University of Arkansas.
Hardke, J. (2014) Personal Communication. Dr. Jarrod Hardke, Rice Extension Agronomist at the
University of Arkansas Rice Research and Extension Center and Kirsten Jaglo, ICF International.
September 11, 2014.
Hardke, J. (2013) Email correspondence. Dr. Jarrod Hardke, Rice Extension Agronomist at the
University of Arkansas Rice Research and Extension Center and Cassandra Snow, ICF
International. July 15, 2013.
Hardke, J.T., and Wilson, C.E. Jr., (2014) Trends in Arkansas rice production, 2013. B.R. Wells
Arkansas Rice Research Studies 2013. Norman, R.J., and Moldenhauer, K.A.K., (Eds.). Research
Series 617, Arkansas Agricultural Experiment Station, University of Arkansas.
References and Abbreviations 10-77
-------
Hardke, J.T., and Wilson, C.E. Jr., (2013) Trends in Arkansas rice production. B.R. Wells Arkansas
Rice Research Studies 2012. Norman, R.J., and Moldenhauer, K.A.K., (Eds.). Research Series
609, Arkansas Agricultural Experiment Station, University of Arkansas.
Hollier, C. A. (ed), (1999) Louisiana rice production handbook. Louisiana State University
Agricultural Center. LCES Publication Number 2321.116pp.
Holzapfel-Pschorn, A., R. Conrad, and W. Seiler (1985) "Production, Oxidation, and Emissions of
Methane in Rice Paddies." FEMS Microbiology Ecology, 31:343-351.
IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to
the Fifth Assessment Report of the Intergovernmental Panel on Climate Change [Stocker, T.F., D.
Qin, G.-K. Plattner, M. Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley
(eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, 1535
pp.
IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change.
[H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa,
Japan.
Johnson, D.M., and R. Mueller (2010) The 2009 Cropland Data Layer. Photogrammetric engineering
and remote sensing 76:1201-1205.
Kirstein, A. (2003 through 2004, 2006) Personal Communication. Arthur Kirstein, Coordinator,
Agricultural Economic Development Program, Palm Beach County Cooperative Extension
Service, FL and ICF International.
Klosterboer, A. (1997,1999 through 2003) Personal Communication. Arlen Klosterboer, retired
Extension Agronomist, Texas A&M University and ICF International. July 7, 2003.
Lindau, C.W. and P.K. Bollich (1993) "Methane Emissions from Louisiana First and Ratoon Crop
Rice." So/7 Science, 156:42-48.
Linquist, B.A., M.A. Adviento-Borbe, C.M. Pittelkow, C.v. Kessel, etal. (2012) Fertilizer management
practices and greenhouse gas emissions from rice systems: A quantitative review and analysis.
Field Crops Research, 135:10-21.
Linscombe, S. (1999, 2001 through 2014) Email correspondence. Steve Linscombe, Professor with
the Rice Research Station at Louisiana State University Agriculture Center and ICF
International.
LSU, (2015) Louisiana ratoon crop and conservation: Ratoon & Conservation Tillage Estimates.
Louisiana State University, College of Agriculture AgCenter. Online at: www.lsuagcenter.com.
Miller, M.R., Garr, J.D., and Coates, P.S., (2010) Changes in the status of harvested rice fields in the
Sacramento Valley, California: Implications for wintering waterfowl. Wetlands, 30: 939-947.
Nelson, Mark D.; Riitters, Kurt H.; Coulston, John W.; Domke, Grant M.; Greenfield, Eric J.; Langner,
Linda L.; Nowak, David J.; O'Dea, Claire B.; Oswalt, Sonja N.; Reeves, Matthew C.; Wear, David
N. 2020. Defining the United States land base: a technical document supporting the USDA
10-78 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Forest Service 2020 RPA assessment. Gen. Tech. Rep. NRS-191. Madison, Wl: U.S. Department
of Agriculture, Forest Service, Northern Research Station. 70 p. https://doi.org/10.2737/NRS-
GTR-191.
Neue, H.U., R. Wassmann, H.K. Kludze, W. Bujun, and R.S. Lantin (1997) "Factors and processes
controlling methane emissions from rice fields." Nutrient Cycling in Agroecosystems 49:111-
117.
Ogle, S.M., F.J. Breidt, M. Easter, S. Williams and K. Paustian. (2007) "An empirically based
approach for estimating uncertainty associated with modeling carbon sequestration in soils."
Ecological Modelling 205:453-463.
Ogle, S.M., F.J. Breidt, M. Easter, S. Williams, K. Killian, and K. Paustian (2010) "Scale and
uncertainty in modeled soil organic carbon stock changes for U.S. croplands using a process-
based model." Global Change Biology 16:810-822.
Ogle, S.M., S. Spencer, M. Hartman, L. Buendia, L. Stevens, D. du Toit, J. Witi (2016) "Developing
national baseline GHG emissions and analyzing mitigation potentials for agriculture and
forestry using an advanced national GHG inventory software system." In Advances in
Agricultural Systems Modeling 6, Synthesis and Modeling of Greenhouse Gas Emissions and
Carbon Storage in Agricultural and Forestry Systems to Guide Mitigation and Adaptation, S. Del
Grosso, L.R. Ahuja and W.J. Parton (eds.), American Society of Agriculture, Crop Society of
America and Soil Science Society of America, pp. 129-148.
Parton, W.J., M.D. Hartman, D.S. Ojima, and D.S. Schimel (1998) "DAYCENT: Its Land Surface
Submodel: Description and Testing". Glob. Planet. Chang. 19: 35-48.
Parton, W.J., D.S. Schimel, C.V. Cole, D.S. Ojima (1987) "Analysis of factors controlling soil organic
matter levels in Great Plains grasslands." Soil Science Society of America Journal 51:1173-1179.
Sarndal C-E, Swensson B, Wretman, J (1992) Model Assisted Survey Sampling. Springer, New York.
Sass, R. L. (2001) CH4 Emissions from Rice Agriculture. Good Practice Guidance and Uncertainty
Management in National Greenhouse Gas Inventories. 399-417. Available online at:
http://www.ipcc-nggip.iges.or.jp/public/gp/bgp/4 7 CH4 Rice Agriculture.pdf.
Sass, R.L., F.M. Fisher, P.A. Harcombe, and F.T. Turner (1990) "Methane Production and Emissions in
a Texas Rice Field." Global Biogeochemical Cycles, 4:47-68.
Sass, R.L., F.M. Fisher, S.T. Lewis, M.F. Jund, and F.T. Turner. (1994) "Methane emissions from rice
fields: effect of soil texture." Global Biogeochemical Cycles 8:135-140.
Schueneman, T. (1997,1999 through 2001) Personal Communication. Tom Schueneman,
Agricultural Extension Agent, Palm Beach County, FL and ICF International.
Slaton, N. (1999 through 2001) Personal Communication. Nathan Slaton, Extension Agronomist—
Rice, University of Arkansas Division of Agriculture Cooperative Extension Service and ICF
International.
References and Abbreviations 10-79
-------
Stansel, J. (2004 through 2005) Email correspondence. Dr. Jim Stansel, Resident Director and
Professor Emeritus, Texas A&M University Agricultural Research and Extension Center and ICF
International.
TAMU (2015) Texas Rice Crop Survey. Texas A&M AgriLIFE Research Center at Beaumont. Online at:
https://beaumont.tamu.edu/.
Texas Agricultural Experiment Station (2007 through 2014) Texas Rice Acreage by Variety.
Agricultural Research and Extension Center, Texas Agricultural Experiment Station, Texas A&M
University System. Available online at:
http://beaumont.tamu.edu/CropSurvey/CropSurveyReport.aspx.
Texas Agricultural Experiment Station (2006) 2005 - Texas Rice Crop Statistics Report. Agricultural
Research and Extension Center, Texas Agricultural Experiment Station, Texas A&M University
System, p. 8. Available online at: http://beaumont.tamu.edu/eLibrary/TRRFReport default.htm.
University of California Cooperative Extension (UCCE) (2015) Rice Production Manual. Revised
(2015) UCCE, Davis, in collaboration with the California Rice Research Board.
USDA (2005 through 2015) Crop Production Summary. National Agricultural Statistics Service,
Agricultural Statistics Board, U.S. Department of Agriculture, Washington, D.C. Available online
at: http://usda.mannlib.cornell.edu.
USDA (2012) Summary of USDA-ARS Research on the Interrelationship of Genetic and Cultural
Management Factors That Impact Grain Arsenic Accumulation in Rice. News and Events.
Agricultural Research Service, U.S. Department of Agriculture, Washington, D.C. Available
online at: http://www.ars.usda.gov/is/pr/2012/120919.htm. September 2013.
USDA (2003) Field Crops, Final Estimates 1997-2002. Statistical Bulletin No. 982. National
Agricultural Statistics Service, Agricultural Statistics Board, U.S. Department of Agriculture,
Washington, D.C. Available online at:
http://usda.mannlib.cornell.edu/usda/reports/general/sb/. September 2005.
USDA (1998) Field Crops Final Estimates 1992-1997. Statistical Bulletin Number 947 a. National
Agricultural Statistics Service, Agricultural Statistics Board, U.S. Department of Agriculture,
Washington, D.C. Available online at: http://usda.mannlib.cornell.edu/. July 2001.
USDA (1994) Field Crops Final Estimates 1987-1992. Statistical Bulletin Number 896. National
Agricultural Statistics Service, Agricultural Statistics Board, U.S. Department of Agriculture,
Washington, D.C. Available online at: http://usda.mannlib.cornell.edu/. July 2001.
USDA-NASS (2021) Published crop data layer. Available at
https://nassgeodata.gmu.edu/CropScape/. Accessed July 2021, USDA-NASS, Washington, DC.
USDA-NRCS (2020) Summary Report: 2017 National Resources Inventory. Natural Resources
Conservation Service, Washington, DC, and Center for Survey Statistics and Methodology, Iowa
State University, Ames, Iowa.
https://www.nrcs.usda.gov/wps/portal/nrcs/main/national/technical/nra/nri/results/.
van Bodegom, P.M., R. Wassmann, T.M. Metra-Corton (2001) "A process based model for methane
emission predictions from flooded rice paddies." Global Biogeochemical Cycles 15: 247-263.
10-80 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Wang, J.J., S.K. Dodla, S. Viator, M. Kongchum, S. Harrison, S. D. Mudi, S. Liu, Z. Tian (2013)
Agriculture Field Management Practices and Greenhouse Gas Emissions from Louisiana Soils.
Louisiana Agriculture, Spring 2013: 8-9. Available online at:
http://www.lsuagcenter.com/NR/rdonlyres/78D8B61A-96A8-49E1-B2EF-
BA1D4CE4E698/93016/v56no2Spring2013.pdf.
Wassmann, R. H.U. Neue, R.S. Lantin, K. Makarim, N. Chareonsil5, L.V. Buendia, and H.
Rennenberg (2000a) Characterization of methane emissions from rice fields in Asia II.
Differences among irrigated, rainfed, and deepwater rice." Nutrient Cycling in Agroecosystems,
58(1):13-22.
Wassmann, R., R.S. Lantin, H.U. Neue, L.V. Buendia, et al. (2000b) "Characterization of Methane
Emissions from Rice Fields in Asia. III. Mitigation Options and Future Research Needs." Nutrient
Cycling in Agroecosystems, 58(1 ):23-36.
Way, M.O., McCauley, G.M., Zhou, X.G., Wilson, L.T., and Morace, B. (Eds.), (2014) 2014Texas Rice
Production Guidelines. Texas A&M AgriLIFE Research Center at Beaumont.
Wilson, C. (2002 through 2007, 2009 through 2012) Personal Communication. Dr. Chuck Wilson,
Rice Specialist at the University of Arkansas Cooperative Extension Service and ICF
International.
Wilson, C.E. Jr., and Branson, J.W., (2006) Trends in Arkansas rice production. B.R. Wells Arkansas
Rice Research Studies 2005. Norman, R.J., Meullenet, J.-F., and Moldenhauer, K.A.K., (Eds.).
Research Series 540, Arkansas Agricultural Experiment Station, University of Arkansas.
Wilson, C.E. Jr., and Branson, J.W., (2005) Trends in Arkansas rice production. B.R. Wells Arkansas
Rice Research Studies 2004. Norman, R.J., Meullenet, J.-F., and Moldenhauer, K.A.K., (Eds.).
Research Series 529, Arkansas Agricultural Experiment Station, University of Arkansas.
Wilson, C.E. Jr., Runsick, S.K., and Mazzanti, R., (2010) Trends in Arkansas rice production. B.R.
Wells Arkansas Rice Research Studies 2009. Norman, R.J., and Moldenhauer, K.A.K., (Eds.).
Research Series 581, Arkansas Agricultural Experiment Station, University of Arkansas.
Wilson, C.E. Jr., Runsick, S.K., Mazzanti, R., (2009) Trends in Arkansas rice production. B.R. Wells
Arkansas Rice Research Studies (2008) Norman, R.J., Meullenet, J.-F., and Moldenhauer, K.A.K.,
(Eds.). Research Series 571, Arkansas Agricultural Experiment Station, University of Arkansas.
Wilson, C.E. Jr., and Runsick, S.K., (2008) Trends in Arkansas rice production. B.R. Wells Arkansas
Rice Research Studies 2007. Norman, R.J., Meullenet, J.-F., and Moldenhauer, K.A.K., (Eds.).
Research Series 560, Arkansas Agricultural Experiment Station, University of Arkansas.
Wilson, C.E. Jr., and Runsick, S.K., (2007) Trends in Arkansas rice production. B.R. Wells Arkansas
Rice Research Studies 2006. Norman, R.J., Meullenet, J.-F., and Moldenhauer, K.A.K., (Eds.).
Research Series 550, Arkansas Agricultural Experiment Station, University of Arkansas.
Yan, X., H. Akiyana, K. Yagi, and H. Akimoto (2009) "Global estimations of the inventory and
mitigation potential of methane emissions from rice cultivation conducted using the 2006
Intergovernmental Panel on Climate Change Guidelines." Global Biogeochemical Cycles, 23,
DOI: 0.1029/2008GB003299.
References and Abbreviations 10-81
-------
Yang, L., et al. (2018). "A new generation of the United States National Land Cover Database:
Requirements, research priorities, design, and implementation strategies." ISPRS Journal of
Photogrammetry and Remote Sensing 146:108-123.
Young, M. (2013) Rice and Ducks. Ducks Unlimited, Memphis, TN. Available online at:
http://www.ducks.org/conservation/farm-bill/rice-and-ducks—by-matt-young.
Agricultural Soil Management
AAPFCO (2008 through 2022) Commercial Fertilizers: 2008-2017. Association of American Plant
Food Control Officials. University of Missouri. Columbia, MO.
AAPFCO (1995 through 2000a, 2002 through 2007) Commercial Fertilizers: 1995-2007. Association
of American Plant Food Control Officials. University of Kentucky. Lexington, KY.
Brockwell, Peter J., and Richard A. Davis (2016) Introduction to time series and forecasting.
Springer.
Cibrowski, P. (1996) Personal Communication. Peter Cibrowski, Minnesota Pollution Control
Agency and Heike Mainhardt, ICF Incorporated. July 29,1996.
Cheng, B., and D.M. Titterington (1994) "Neural networks: A review from a statistical perspective."
Statistical science 9: 2-30.
Claassen, R., M. Bowman, J. McFadden, D. Smith, and S. Wallander (2018) Tillage intensity and
conservation cropping in the United States, EIB 197. United States Department of Agriculture,
Economic Research Service, Washington, D.C.
CTIC (2004) 2004 Crop Residue Management Survey. Conservation Technology Information Center.
Available at http://www.ctic.purdue.edu/CRM/.
Del Grosso, S.J., T. Wirth, S.M. Ogle, W.J. Parton (2008) Estimating agricultural nitrous oxide
emissions. EOS 89, 529-530.
Del Grosso, S.J., A.R. Mosier, W.J. Parton, and D.S. Ojima (2005) "DAYCENT Model Analysis of Past
and Contemporary Soil N20 and Net Greenhouse Gas Flux for Major Crops in the USA." So/7
Tillage and Research, 83: 9-24. doi: 10.1016/j.still.2005.02.007.
Del Grosso, S.J., W.J. Parton, A.R. Mosier, M.D. Hartman, J. Brenner, D.S. Ojima, and D.S. Schimel
(2001) "Simulated Interaction of Carbon Dynamics and Nitrogen Trace Gas Fluxes Using the
DAYCENT Model." In Schaffer, M., L. Ma, S. Hansen, (eds.). Modeling Carbon and Nitrogen
Dynamics for Soil Management. CRC Press. Boca Raton, Florida. 303-332.
Del Grosso, S.J., S.M. Ogle, W.J. Parton, and F.J. Breidt (2010) "Estimating Uncertainty in N20
Emissions from U.S. Cropland Soils." Global Biogeochemical Cycles, 24, GB1009,
doi: 10.1029/2009GB003544.
Del Grosso, S.J., W.J. Parton, C.A. Keough, and M. Reyes-Fox. (2011) Special features of the
DAYCENT modeling package and additional procedures for parameterization, calibration,
validation, and applications, in Methods of Introducing System Models into Agricultural
10-82 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Research, L.R. Ahuja and Liwang Ma, editors, p. 155-176, American Society of Agronomy, Crop
Science Society of America, Soil Science Society of America, Madison, Wl. USA.
Del Grosso, S. J., S. M. Ogle, C. Nevison, R. Gurung, W. J. Parton, C. Wagner-Riddle, W. Smith, W.
Winiwarter, B. Grant, M. Tenuta, E. Marx, S. Spencer, and S. Williams. 2022. A gap in nitrous
oxide emission reporting complicates long-term climate mitigation. Proceedings of the National
Academy of Sciences 119:e2200354119.
Delgado, J.A., S.J. Del Grosso, and S.M. Ogle (2009) "15N isotopic crop residue cycling studies and
modeling suggest that IPCC methodologies to assess residue contributions to N20-N emissions
should be reevaluated." Nutrient Cycling in Agroecosystems, DO110.1007/s10705-009-9300-9.
Edmonds, L., N. Gollehon, R.L. Kellogg, B. Kintzer, L. Knight, C. Lander, J. Lemunyon, D. Meyer, D.C.
Moffitt, and J. Schaeffer (2003) "Costs Associated with Development and Implementation of
Comprehensive Nutrient Management Plans." Part 1. Nutrient Management, Land Treatment,
Manure and Wastewater Handling and Storage, and Recordkeeping. Natural Resource
Conservation Service, U.S. Department of Agriculture.
EPA (2003) Clean Watersheds Needs Survey 2000—Report to Congress, U.S. Environmental
Protection Agency. Washington, D.C. Available online at:
http://www.epa.gov/owm/mtb/cwns/2000rtc/toc.htm.
EPA (1999) Biosolids Generation, Use and Disposal in the United States. Office of Solid Waste, U.S.
Environmental Protection Agency. Available online at:
http://biosolids.policy.net/relatives/18941 .PDF.
EPA (1993) Federal Register. Part II. Standards for the Use and Disposal of Sewage Sludge; Final
Rules. U.S. Environmental Protection Agency, 40 CFR Parts 257, 403, and 503.
Firestone, M. K., and E.A. Davidson, Ed. (1989) Microbiological basis of NO and N20 production and
consumption in soil. Exchange of trace gases between terrestrial ecosystems and the
atmosphere. New York, John Wiley & Sons.
Friedman, J.H. (2001) "Greedy function approximation: A gradient boosting machine." Ann. Statist.
29(5) 1189-1232.
Hagen, S. C., G. Delgado, P. Ingraham, I. Cooke, R. Emery, J. P. Fisk, L. Melendy, T. Olson, S. Patti, N.
Rubin, B. Ziniti, H. Chen, W. Salas, P. Elias, and D. Gustafson. 2020. Mapping Conservation
Management Practices and Outcomes in the Corn Belt Using the Operational Tillage
Information System (OpTIS) and the Denitrification-Decomposition (DNDC) Model. Land 9:408.
ILENR (1993) Illinois Inventory of Greenhouse Gas Emissions and Sinks: 1990. Office of Research
and Planning, Illinois Department of Energy and Natural Resources. Springfield, IL.
IPCC (2014) 2073 Supplement to the 2006 IPCC Guidelines for National Greenhouse Gas
Inventories: Wetlands. The Intergovernmental Panel on Climate Change. [T, Hiraishi, T. Krug, K.
Tanabe, N. Srivastava, B. Jamsranjav, M. Fukuda and T. Troxler (eds.)]. Hayama, Kanagawa,
Japan.
IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to
the Fifth Assessment Report of the Intergovernmental Panel on Climate Change [Stocker, T.F., D.
References and Abbreviations 10-83
-------
Qin, G.-K. Plattner, M. Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley
(eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, 1535
pp.
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change.
[H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa,
Japan.
Johnson, D.M., and R. Mueller (2010) The 2009 Cropland Data Layer. Photogrammetric engineering
and remote sensing 76:1201-1205.
Little, R. (1988) "Missing-data adjustments in large surveysJournal of Business and Economic
Statistics 6: 287-296.
McFarland, M.J. (2001) Biosolids Engineering, New York: McGraw-Hill, p. 2.12.
McGill, W.B., and C.V. Cole (1981) Comparative aspects of cycling of organic C, N, S and P through
soil organic matter. Geoderma 26:267-286.
Metherell, A.K., L.A. Harding, C.V. Cole, and W.J. Parton (1993) "CENTURY Soil Organic Matter Model
Environment." Agroecosystem version 4.0. Technical documentation, GPSR Tech. Report No. 4,
USDA/ARS, Ft. Collins, CO.
Mosier A., C. Kroeze, C. Nevison, O. Oenema, S. Seitzinger, and O. van Cleemput, Closing the global
atmospheric N20 budget: Nitrous oxide emissions through the agricultural nitrogen cycle,
Nutrient Cycling in Agroecosystems, 52, 225-248,1998.
NEBRA (2007) A National Biosolids Regulation, Quality, End Use & Disposal Survey. North East
Biosolids and Residuals Association, July 21, 2007.
Nelson, Mark D.; Riitters, Kurt H.; Coulston, John W.; Domke, Grant M.; Greenfield, Eric J.; Langner,
Linda L.; Nowak, David J.; O'Dea, Claire B.; Oswalt, Sonja N.; Reeves, Matthew C.; Wear, David
N. 2020. Defining the United States land base: a technical document supporting the USDA
Forest Service 2020 RPA assessment. Gen. Tech. Rep. NRS-191. Madison, Wl: U.S. Department
of Agriculture, Forest Service, Northern Research Station. 70 p. https://doi.org/10.2737/NRS-
GTR-191.
Noller, J. (1996) Personal Communication. John Noller, Missouri Department of Natural Resources
and Heike Mainhardt, ICF Incorporated. July 30,1996.
Ogle, S.M., F.J. Breidt, M. Easter, S. Williams and K. Paustian (2007) "Empirically-Based Uncertainty
Associated with Modeling Carbon Sequestration Rates in Soils." Ecological Modeling 205:453-
463.
Ogle, S.M., F.J. Breidt, M. Easter, S. Williams, K. Killian, and K. Paustian (2010) "Scale and
uncertainty in modeled soil organic carbon stock changes for U.S. croplands using a process-
based model." Global Change Biology 16:810-822.
Oregon Department of Energy (1995) Report on Reducing Oregon's Greenhouse Gas Emissions:
Appendix D Inventory and Technical Discussion. Oregon Department of Energy. Salem, OR.
10-84 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Parton, W.J., M.D. Hartman, D.S. Ojima, and D.S. Schimel (1998) "DAYCENT: Its Land Surface
Submodel: Description and Testing." Glob. Planet. Chang. 19: 35-48.
Potter, C., S. Klooster, A. Huete, and V. Genovese (2007) Terrestrial carbon sinks for the United
States predicted from MODIS satellite data and ecosystem modeling. Earth Interactions 11,
Article No. 13, DO110.1175/EI228.1.
Potter, C. S., J.T. Randerson, C.B. Fields, P.A. Matson, P.M. Vitousek, H.A. Mooney, and S.A. Klooster
(1993) "Terrestrial ecosystem production: a process model based on global satellite and
surface data." Global Biogeochemical Cycles 7:811-841.
PRISM Climate Group (2022) PRISM Climate Data, Oregon State University,
http://prism.oregonstate.edu. downloaded January 2022.
Pukelsheim, F. (1994) "The 3-Sigma-Rule." American Statistician 48:88-91.
Ruddy B.C., D.L. Lorenz, and D.K. Mueller (2006) County-level estimates of nutrient inputs to the
land surface of the conterminous United States, 1982-2001. Scientific Investigations Report
2006-5012. U.S Department of the Interior.
Sarndal C-E, Swensson B, Wretman, J (1992) Model Assisted Survey Sampling. Springer, New York.
Scheer, C., S.J. Del Grosso, W.J. Parton, D.W. Rowlings, P.R. Grace (2013) Modeling Nitrous Oxide
Emissions from Irrigated Agriculture: Testing DAYCENT with High Frequency Measurements,
Ecological Applications, in press. Available online at: http://dx.doi.Org/10.1890/13-0570.1.
Soil Survey Staff (2020) Gridded Soil Survey Geographic (gSSURGO) Database for the
Conterminous United States. United States Department of Agriculture, Natural Resources
Conservation Service, Accessed February 2020 (FY2020 official release), Available online at
https://gdg.sc.egov.usda.gov/.
Towery, D. (2001) Personal Communication. Dan Towery regarding adjustments to the CTIC (1998)
tillage data to reflect long-term trends, Conservation Technology Information Center, West
Lafayette, IN, and Marlen Eve, National Resource Ecology Laboratory, Fort Collins, CO. February
2001.
TVA (1991 through 1992a, 1993 through 1994) Commercial Fertilizers. Tennessee Valley Authority,
Muscle Shoals, AL.
USDA-ERS (2020) Agricultural Resource Management Survey (ARMS) Farm Financial and Crop
Production Practices: Tailored Reports. Available online at: https://www.ers.usda.gov/data-
products/arms-farm-financial-and-crop-production-practices/.
USDA-ERS (1997) Cropping Practices Survey Data—1995. Economic Research Service, United
States Department of Agriculture. Available online at:
http://www.ers.usda.gov/data/archive/93018/.
USDA-NASS (2023) Quick Stats. National Agricultural Statistics Service, United States Department
of Agriculture, Washington, D.C., Accessed August 2023, http://quickstats.nass.usda.gov/.
USDA-NASS (2021) Published crop data layer. Available at
https://nassgeodata.gmu.edu/CropScape/. Accessed July 2021, USDA-NASS, Washington, DC.
References and Abbreviations 10-85
-------
USDA-NASS (2017) 2017 Census of Agriculture. USDA National Agricultural Statistics Service,
Complete data available at http://www.nass.usda.gov/AgCensus.
USDA-NASS (2012) 2012 Census of Agriculture. USDA National Agricultural Statistics Service,
Complete data available at http://www.nass.usda.gov/AgCensus.
USDA-NASS (2004) Agricultural Chemical Usage: 2003 Field Crops Summary. Report AgChl (04)a.
National Agricultural Statistics Service, U.S. Department of Agriculture, Washington, D.C.
Available online at: http://usda.mannlib.cornell.edu/reports/nassr/other/pcu-
bb/agcs0504.pdfH.
USDA-NASS (1999) Agricultural Chemical Usage: 1998 Field Crops Summary. Report AgCH1 (99).
National Agricultural Statistics Service, U.S. Department of Agriculture, Washington, DC.
Available online at: http://usda.mannlib.cornell.edu/reports/nassr/other/pcu-bb/agch0599.pdf.
USDA-NASS (1992) Agricultural Chemical Usage: 1991 Field Crops Summary. Report AgChl (92).
National Agricultural Statistics Service, U.S. Department of Agriculture, Washington, D.C.
Available online at: http://usda.mannlib.cornell.edu/reports/nassr/other/pcu-
bb/agch0392.txtH.
USDA-NRCS (2022) Conversation practice on cultivated croplands: A comparison of CEAP I and
CEAP II survey data and modeling. United States Department of Agriculture, Natural Resources
Conservation Service, https://www.nrcs.usda.gov/sites/default/files/2022-09/CEAP-Croplands-
ConservationPracticesonCultivatedCroplands-Report-March2022.pdf.
USDA-NRCS (2020) Summary Report: 2017 National Resources Inventory. Natural Resources
Conservation Service, Washington, DC, and Center for Survey Statistics and Methodology, Iowa
State University, Ames, Iowa.
https://www.nrcs.usda.gov/wps/portal/nrcs/main/national/technical/nra/nri/results/.
USDA-NRCS (2018) CEAP Cropland Farmer Surveys. USDA Natural Resources Conservation
Service.
https://www.nrcs.usda.gov/wps/portal/nrcs/detail/national/technical/nra/ceap/na/?cid=nrcs1
43 014163.
USDA-NRCS (2012) Assessment of the Effects of Conservation Practices on Cultivated Cropland in
the Upper Mississippi River Basin. U.S. Department of Agriculture, Natural Resources
Conservation Service,
https://www.nrcs.usda.gov/lnternet/FSE DQCUMENTS/stelprdb1042093.pdf.
USFS (2019) Forest Inventory and Analysis Program. United States Department of Agriculture, U.S.
Forest Service, https://www.fia.fs.fed.us/tools-data/default.asp.
Van Buuren, S. (2012) "Flexible imputation of missing data." Chapman & Hall/CRC, Boca Raton, FL.
Wagner-Riddle, C., Congreves, K. A., Abalos, D., Berg, A. A., Brown, S. E., Ambadan, J. T., Gao, X. &
Tenuta, M. (2017) "Globally important nitrous oxide emissions from croplands induced by
freeze-thaw cycles." Nature Geosciences 10(4): 279-283.
10-86 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Wisconsin Department of Natural Resources (1993) Wisconsin Greenhouse Gas Emissions:
Estimates for 1990. Bureau of Air Management, Wisconsin Department of Natural Resources,
Madison, Wl.
Yang, L., Jin, S., Danielson, R, Homer, C., Gass, L., Bender, S. M., Case, A., Costello, C., Dewitz, J.,
Fry, J., Funk, M., Granneman, B., Liknes, G. C., Rigge, M. &Xian, G. (2018) "A new generation of
the United States National Land Cover Database: Requirements, research priorities, design,
and implementation strategies." ISPRS Journal of Photogrammetry and Remote Sensing 146:
108-123.
Liming
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change.
[H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa,
Japan.
Tepordei, V.V. (1994 through 2015) "Crushed Stone," In Minerals Yearbook. U.S. Department of the
Interior/U.S. Geological Survey. Washington, D.C. Available online at:
http://minerals.usgs.gov/minerals/.
Tepordei, V.V. (2003b) Personal communication. Valentin Tepordei, U.S. Geological Survey and ICF
Consulting, August 18, 2003.
USGS (2020a) Mineral Industry Surveys: Crushed Stone and Sand and Gravel in the First Quarter of
2020, U.S. Geological Survey, Reston, VA. Available online at:
http://minerals.usgs.gov/minerals/pubs/commodity/stone crushed/index.html#mis.
USGS (2020b) Mineral Industry Surveys: Crushed Stone and Sand and Gravel in the Fourth Quarter
of 2020, U.S. Geological Survey, Reston, VA. Available online at:
http://minerals.usgs.gov/minerals/pubs/commodity/stone crushed/index.html#mis.
USGS (2021 a) Mineral Industry Surveys: Crushed Stone and Sand and Gravel in the First Quarter of
2021, U.S. Geological Survey, Reston, VA. Available online at:
http://minerals.usgs.gov/minerals/pubs/commodity/stone crushed/index.html#mis.USGS
(2021 b) Mineral Industry Surveys: Crushed Stone and Sand and Gravel in the Fourth Quarter of
2021, U.S. Geological Survey, Reston, VA. Available online at:
http://minerals.usgs.gov/minerals/pubs/commodity/stone crushed/index.html#mis.USGS
(2022a) Mineral Industry Surveys: Crushed Stone and Sand and Gravel in the First Quarter of
2022, U.S. Geological Survey, Reston, VA. Available online at:
http://minerals.usgs.gov/minerals/pubs/commodity/stone crushed/index.html#mis.
USGS (2022b) Mineral Industry Surveys: Crushed Stone and Sand and Gravel in the Fourth Quarter
of 2022, U.S. Geological Survey, Reston, VA. Available online at:
http://minerals.usgs.gov/minerals/pubs/commodity/stone crushed/index.html#mis.
USGS (2023a) Mineral Industry Surveys: Crushed Stone and Sand and Gravel in the First Quarter of
2023, U.S. Geological Survey, Reston, VA. Available online at:
http://minerals.usgs.gov/minerals/pubs/commodity/stone crushed/index.html#mis.
References and Abbreviations 10-87
-------
USGS (2023b) Mineral Industry Surveys: Crushed Stone and Sand and Gravel in the Fourth Quarter
of 2023, U.S. Geological Survey, Reston, VA. Available online at:
http://minerals.usgs.gov/nninerals/pubs/connnnodity/stone crushed/index.html#mis.
USGS (2024a) Mineral Industry Surveys: Crushed Stone and Sand and Gravel in the First Quarter of
2024, U.S. Geological Survey, Reston, VA. Available online at:
http://minerals.usgs.gov/minerals/pubs/commodity/stone crushed/index.html#mis.
USGS (2024b) Mineral Industry Surveys: Crushed Stone and Sand and Gravel in the Second Quarter
of 2024, U.S. Geological Survey, Reston, VA. Available online at:
http://minerals.usgs.gov/minerals/pubs/commodity/stone crushed/index.html#mis.West. T.O.,
and A.C. McBride (2005) "The contribution of agricultural lime to carbon dioxide emissions in
the United States: dissolution, transport, and net emissions," Agricultural Ecosystems &
Environment 108:145-154.
West, T.O. (2008) Email correspondence. Tristram West, Environmental Sciences Division, Oak
Ridge National Laboratory, U.S. Department of Energy and Nikhil Nadkarni, ICF International on
suitability of liming emission factor for the entire United States. June 9, 2008.
Willett, J.C. (2020a) "Crushed Stone," In Minerals Yearbook 2016. U.S. Department of the
Interior/U.S. Geological Survey, Washington, D.C. Available online at:
http://minerals.usgs.gov/minerals/pubs/commodity/stone crushed/index.html#mis. Accessed
November 2020.Willett, J.C. (2022a) "Crushed Stone," In Minerals Yearbook 2018. U.S.
Department of the Interior/U.S. Geological Survey, Washington, D.C. Available online at:
https://www.usgs.gov/centers/national-minerals-information-center/crushed-stone-statistics-
and-information. Accessed October 2022.
Willett, J.C. (2022b) "Crushed Stone," In Minerals Yearbook 2019. U.S. Department of the
Interior/U.S. Geological Survey, Washington, D.C. Available online at:
https://www.usgs.gov/centers/national-minerals-information-center/crushed-stone-statistics-
and-information. Accessed October 2022
Willett, J.C. (2022c) "Crushed Stone," In Minerals Yearbook 2020. U.S. Department of the
Interior/U.S. Geological Survey, Washington, D.C. Available online at:
https://www.usgs.gov/centers/national-minerals-information-center/crushed-stone-statistics-
and-information. Accessed October 2022
Willett, J.C. (2022d) Personal communication. Jason Willett. Preliminary data tables from "Crushed
Stone," In 2021 Minerals Yearbook. U.S. Department of the Interior/U.S. Geological Survey.
Washington, D.C. October, 2022.
Willett, J.C. (2023a) "Crushed Stone," In Minerals Yearbook 2021. U.S. Department of the
Interior/U.S. Geological Survey, Washington, D.C. Available online at:
https://www.usgs.gov/centers/national-minerals-information-center/crushed-stone-statistics-
and-information. Accessed November 2023
Willett, J.C. (2023b) Personal communication. Jason Willett. Preliminary data tables from "Crushed
Stone," In 2022 Minerals Yearbook. U.S. Department of the Interior/U.S. Geological Survey.
Washington, D.C. November, 2023.Willett, J.C. (2024a) "Crushed Stone," In Minerals Yearbook
10-88 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
2021. U.S. Department of the Interior/U.S. Geological Survey, Washington, D.C. Available online
at: https://www.usgs.gov/centers/national-minerals-information-center/crushed-stone-
statistics-and-information. Accessed August 2024
Willett, J.C. (2024b) "Crushed Stone," In Minerals Yearbook 2022. U.S. Department of the
Interior/U.S. Geological Survey, Washington, D.C. Available online at:
https://www.usgs.gov/centers/national-minerals-information-center/crushed-stone-statistics-
and-information. Accessed August 2024
Willett, J.C. (2024c) Personal communication. Jason Willett. Preliminary data tables from "Crushed
Stone," In 2023 Minerals Yearbook. U.S. Department of the Interior/U.S. Geological Survey.
Washington, D.C. November, 2024.
Urea Fertilization
AAPFCO (2008 through 2022) Commercial Fertilizers. Association of American Plant Food Control
Officials. University of Missouri. Columbia, MO.
AAPFCO (1995 through 2000a, 2002 through 2007) Commercial Fertilizers. Association of American
Plant Food Control Officials. University of Kentucky. Lexington, KY.
AAPFCO (2000b) 1999-2000 Commercial Fertilizers Data, ASCII files. Available from David Terry,
Secretary, AAPFCO.
EPA (2000) Preliminary Data Summary: Airport Deicing Operations (Revised). EPA-821-R-00-016.
August 2000.
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change.
[H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa,
Japan.
Itle, C. (2009) Email correspondence. Cortney Itle, ERG and Tom Wirth, U.S. Environmental
Protection Agency on the amount of urea used in aircraft deicing. January 7, 2009.
TVA (1991 through 1994) Commercial Fertilizers. Tennessee Valley Authority, Muscle Shoals, AL.
TVA (1992b) Fertilizer Summary Data 1992. Tennessee Valley Authority, Muscle Shoals, AL.
Field Burning of Agricultural Residues
Akintoye, H.A., Agbeyi, E.O., and Olaniyan, A.B. (2005) "The effects of live mulches on tomato
(Lycopersicon esculentum) yield under tropical conditions." Journal of Sustainable Agriculture
26: 27-37.
Bange, M.P., Milroy, S.P., and Thongbai, P. (2004) "Growth and yield of cotton in response to
waterlogging." Field Crops Research 88:129-142.
Beyaert, R.P. (1996) The effect of cropping and tillage management on the dynamics of soil organic
matter. PhD Thesis. University of Guelph.
References and Abbreviations 10-89
-------
Bouquet, D.J., and Breitenbeck, G.A. (2000) "Nitrogen rate effect on partitioning of nitrogen and dry
matter by cotton." Crop Science 40:1685-1693.
Brockwell, Peter J., and Richard A. Davis (2016) Introduction to time series and forecasting.
Springer. Cantens, G. (2004 through 2005) Personal Communication. Janet Lewis, Assistant to
Gaston Cantens, Vice President of Corporate Relations, Florida Crystals Company and ICF
International.
Brouder, S.M., and Cassman, K.G (1990) "Root development of two cotton cultivars in relation to
potassium uptake and plant growth in a vermiculitic soil." Field Crops Res. 23:187-203.
Costa, L.D., and Gianquinto, G. (2002) "Water stress and watertable depth influence yield, water
use efficiency, and nitrogen recovery in bell pepper: lysimeter studies." Aust. J. Agric. Res. 53:
201-210.
Crafts-Brandner, S.J., Collins, M., Sutton, T.G., and Burton, H.R. (1994) "Effect of leaf maleic
hydrazide concentration on yield and dry matter partitioning in burley tobacco (Nicotiana
tabacum L.)." Field Crops Research 37:121-128.
De Pinheiro Henriques, A.R., and Marcelis, L.F.M. (2000) "Regulation of growth at steady-state
nitrogen nutrition in lettuce (Lactuca sativa L.): Interactive effects of nitrogen and irradiance."
Annals of Botany 86:1073-1080.0.
Diaz-Perez, J.C., Silvoy, J., Phatak,S.C., Ruberson, J., and Morse, R. (2008) Effect of winter cover
crops and co-till on the yield of organically-grown bell pepper (Capsicum annum L.). Acta Hort.
767:243-247.
Dua, K.L., and Sharma, V.K. (1976) "Dry matter production and energy contents of ten varieties of
sugarcane at Muzaffarnagar (Western Uttar Pradesh)." Tropical Ecology 17: 45-49.
Fritschi, F.B., Roberts, B.A., Travis, R.L., Rains, D.W., and Hutmacher, R.B. (2003) "Seasonal nitrogen
concentration, uptake, and partitioning pattern of irrigated Acala and Pima cotton as influenced
by nitrogen fertility level." Crop Science 44:516-527.
Gerik, T.J., K.L. Faver, P.M. Thaxton, and K.M. El-Zik. (1996) "Late season water stress in cotton: I.
Plant growth, water use, and yield." Crop Science 36: 914-921.
Gibberd, M.R., McKay, A.G., Calder, T.C., and Turner, N.C. (2003) "Limitations to carrot (Daucus
carota L.) productivity when grown with reduced rates of frequent irrigation on a free-draining,
sandy soil." Australian Journal of Agricultural Research 54: 499-506.
Giglio, L., I. Csiszar, and C.O. Justice (2006) "Global distribution and seasonality of active fires as
observed with the Terra and Aqua Moderate Resolution Imaging Spectroradiometer (MODIS)
sensors" J. Geophys. Res. 111, G02016, doi:10.1029/2005JG000142.
Giglio, L., Justice, C., Boschetti, L., Roy, D. (2015) "MCD64A1 MODIS/Terra+Aqua Burned Area
Monthly L3 Global 500m SIN Grid V006 [Data set]". NASA EOSDIS Land Processes Distributed
Active Archive Center. Accessed 2023-12-05 from
https://doi.org/10.5067/MQDIS/MCD64A1.006.
10-90 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Halevy, J. (1976) "Growth rate and nutrient uptake of two cotton cultivars grown under irrigation."
Agronomy Journal 68: 701-705.
Halvorson, A.D., Follett, R.F., Bartolo, M.E., and Schweissing, F.C. (2002) "Nitrogen fertilizer use
efficiency of furrow-irrigated onion and corn." Agronomy Journal 94: 442-449.
Heitholt, J.J., Pettigrew, W.T., and Meredith, W.R. (1992) "Light interception and lint yield of narrow-
row cotton." Crop Science 32: 728-733.
Hollifield, C.D., Silvertooth, J.C., and Moser, H. (2000) "Comparison of obsolete and modern cotton
cultivars for irrigated production in Arizona." 2000 Arizona Cotton Report, University of Arizona
College of Agriculture, http://ag.arizona.edu/pubs/crops/az1170/.
Hopkinson, J.M. (1967) "Effects of night temperature on the growth of Nicotiana tabacum."
Australian Journal of Experimental Agriculture and Animal Husbandry 7: 78-82.
Huett, D.O., and Dettman, E.B. (1991) Effect of nitrogen on growth, quality and nutrient uptake of
cabbages grown in sand culture. Australian Journal of Experimental Agriculture 29: 875-81.
Huett, D.O., and Dettman, B. (1989) "Nitrogen response surface models of zucchini squash, head
lettuce and potato." Plant and Soil 134: 243-254.
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change.
[H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa,
Japan.
IPCC/UNEP/OECD/IEA (1997) Revised 1996 IPCC Guidelines for National Greenhouse Gas
Inventories. Intergovernmental Panel on Climate Change, United Nations Environment
Programme, Organization for Economic Co-Operation and Development, International Energy
Agency, Paris, France.
Jacobs, J.L., Ward, G.N., and Kearney, G. (2004) "Effects of irrigation strategies and nitrogen fertilizer
on turnip dry matter yield, water use efficiency, nutritive characteristics and mineral content in
western Victoria." Australian Journal of Experimental Agriculture 44:13-26.
Jacobs, J.L., Ward, G.N., McDowell, A.M., and Kearney, G. (2002) "Effect of seedbed cultivation
techniques, variety, soil type and sowing time, on brassica dry matter yields, water use
efficiency and crop nutritive characteristics in western Victoria." Australian Journal of
Experimental Agriculture 42: 945-952.
Jacobs, J.L., Ward, G.N., McDowell, A.M., and Kearney, G.A. (2001) "A survey on the effect of
establishment techniques, crop management, moisture availability and soil type on turnip dry
matter yields and nutritive characteristics in western Victoria." Australian Journal of
Experimental Agriculture 41: 743-751.
Kage, H., Alt, C., and Stutzel, H. (2003) "Aspects of nitrogen use efficiency of cauliflower II.
Productivity and nitrogen partitioning as influenced by N supply." Journal of Agricultural Science
141:17-29.
References and Abbreviations 10-91
-------
Kumar, A., Singh, D.P., and Singh, P. (1994) "Influence of water stress on photosynthesis,
transpiration, water-use efficiency and yield of Brassica juncea L." Field Crops Research 37: 95-
101.
LANDFIRE (2008) Existing Vegetation Type Layer, LANDFIRE 1.1.0, U.S. Department of the Interior,
Geological Survey. Accessed 28 October 2010 at http://landfire.cr.usgs.gov/viewer/.
MacLeod, L.B., Gupta, U.C., and Cutcliffe, J.A. (1971) "Effect of N, P, and K on root yield and nutrient
levels in the leaves and roots of rutabagas grown in a greenhouse." Plant and Soil 35: 281-288.
Mahrani, A., and Aharonov, B. (1964) "Rate of nitrogen absorption and dry matter production by
upland cotton grown under irrigation." Israel J. Agric. Res. 14: 3-9.
Marcussi, F.F.N., Boas, R.L.V., de Godoy, L.J.G., and Goto, R. (2004) "Macronutrient accumulation
and partitioning in fertigated sweet pepper plants." Sci. Agric. (Piracicaba, Braz.) 61: 62-68.
McCarty, J.L. (2011) "Remote Sensing-Based Estimates of Annual and Seasonal Emissions from
Crop Residue Burning in the Contiguous United States ."Journal of the Air & Waste Management
Association, 61:1, 22-34, DOI: 10.3155/1047-3289.61.1.22.
McCarty, J.L. (2010) Agricultural Residue Burning in the Contiguous United States by Crop Type and
State. Geographic Information Systems (GIS) Data provided to the EPA Climate Change Division
by George Pouliot, Atmospheric Modeling and Analysis Division, EPA. Dr. McCarty's research
was supported by the NRI Air Quality Program of the Cooperative State Research, Education,
and Extension Service, USDA, under Agreement No. 20063511216669 and the NASA Earth
System Science Fellowship.
McCarty, J.L. (2009) Seasonal and Interannual Variability of Emissions from Crop Residue Burning in
the Contiguous United States. Dissertation. University of Maryland, College Park.
McPharlin, I.R., Aylmore, P.M., and Jeffery, R.C. (1992) "Response of carrots (Daucus carota L.) to
applied phosphorus and phosphorus leaching on a Karrakatta sand, under two irrigation
regimes." Australian Journal of Experimental Agriculture 32:225-232.
Mondino, M.H., Peterlin, O.A., and Garay, F. (2004) "Response of late-planted cotton to the
application of growth regulator (chlorocholine chloride, CYCOCEL 75)." Expl Agric. 40: 381-387.
Moustakas, N.K., and Ntzanis, H. (2005) "Dry matter accumulation and nutrient uptake in flue-
cured tobacco (Nicotiana tabacum L.)." Field Crops Research 94:1-13.
Peach, L., Benjamin, L.R., and Mead, A. (2000) "Effects on the growth of carrots (Daucus carota L.),
cabbage (Brassica oleracea var. capitata L.) and onion (Allium cepa L.) of restricting the ability
of the plants to intercept resources ."Journal of Experimental Botany 51: 605-615.
Pettigrew, W.T., and Meredith, W.R., Jr. (1997) "Dry matter production, nutrient uptake, and growth
of cotton as affected by potassium fertilization." J. Plant Nutr. 20:531-548.
Pettigrew, W.T., Meredith, W.R., Jr., and Young, L.D. (2005) "Potassium fertilization effects on cotton
lint yield, yield components, and reniform nematode populations." Agronomy Journal 97:1245-
1251.
10-92 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
PRISM Climate Group (2015) PRISM Climate Data. Oregon State University. July 24, 2015. Available
online at: http://prism.oregonstate.edu.
Reid, J.B., and English, J.M. (2000) "Potentialyield in carrots (Daucus carota L.): Theory, test, and an
application." Annals of Botany 85: 593-605.
Sadras, V.O., and Wilson, L.J. (1997) "Growth analysis of cotton crops infested with spider mites: II.
Partitioning of dry matter." Crop Science 37: 492-497.
Scholberg, J., McNeal, B.L., Jones, J.W., Boote, K.J., Stanley, C.D., and Obreza, T.A. (2000a) "Growth
and canopy characteristics of field-grown tomatoAgronomy Journal 92:152-159.
Scholberg, J., McNeal, B.L., Boote, K.J., Jones, J.W., Locasio, S.J., and Olson, S.M. (2000b) "Nitrogen
stress effects on growth and nitrogen accumulation by field-grown tomatoAgronomy Journal
92:159-167.
Singels, A. and Bezuidenhout, C.N. (2002) "A new method of simulating dry matter partitioning in
the Canegro sugarcane model." Field Crops Research 78:151 -164.
Sitompul, S.M., Hairiah, K., Cadisch, G., and Van Noordwuk, M. (2000) "Dynamics of density
fractions of macro-organic matter after forest conversion to sugarcane and woodlots,
accounted for in a modified Century model." Netherlands Journal of Agricultural Science 48: 61-
73.
Stirling, G.R., Blair, B.L., Whittle, P.J.L., and Garside, A.L. (1999) "Lesion nematode (Pratylenchus
zeae) is a component of the yield decline complex of sugarcane." In: Magarey, R.C. (Ed.),
Proceedings of the First Australasian Soilborne Disease Symposium. Bureau of Sugar
Experiment Stations, Brisbane, pp. 15-16.
Tan, D.K.Y., Wearing, A.H., Rickert, K.G., and Birch, C.J. (1999) "Broccoli yield and quality can be
determined by cultivar and temperature but not photoperiod in south-east Queensland."
Australian Journal of Experimental Agriculture 39: 901-909.
Tadesse, T., Nichols, M.A., and Fisher, K.J. (1999) Nutrient conductivity effects on sweet pepper
plants grown using a nutrient film technique. 1. Yield and fruit quality. New Zealand Journal of
Crop and Horticultural Science, 27: 229-237.
Torbert, H.A., and Reeves, D.W. (1994) "Fertilizer nitrogen requirements for cotton production as
affected by tillage and traffic." Soil Sci. Soc. Am. J. 58:1416-1423.
USDA-NASS (2021) Published crop data layer. Available at
https://nassgeodata.gmu.edu/CropScape/. Accessed July 2021, USDA-NASS, Washington, DC.
USDA-NASS (2019) Quick Stats: U.S. & All States Data; Crops; Production and Area Harvested;
1990-2018. National Agricultural Statistics Service, U.S. Department of Agriculture.
Washington, D.C. U.S. Department of Agriculture, National Agricultural Statistics Service.
Washington, D.C., Available online at: http://quickstats.nass.usda.gov/.
USDA-NRCS (2020) Summary Report: 2017 National Resources Inventory. Natural Resources
Conservation Service, Washington, DC, and Center for Survey Statistics and Methodology, Iowa
References and Abbreviations 10-93
-------
State University, Ames, Iowa.
https://www.nrcs.usda.gov/wps/portal/nrcs/main/national/technical/nra/nri/results/.
USDA-NRCS (2018) Summary Report: 2015 National Resources Inventory, Natural Resources
Conservation Service, Washington, D.C., and Center for Survey Statistics and Methodology,
Iowa State University, Ames, Iowa.
https://www.nrcs.usda.gov/lnternet/FSE DOCUMENTS/nrcseprd1422028.pdf.
Valantin, M., Gary, C., Vaissiere, B.E., and Frossard, J.S. (1999) "Effect of fruit load on partitioning of
dry matter and energy in cantaloupea (Cucumis melo L.)." Annals of Botany 84:173-181.
Wallach, D., Marani, A., and Kletter, E. (1978) "The relation of cotton crop growth and development
to final yield." Field Crops Research 1: 283-294.
Wells, R., and Meredith, W.R., Jr. (1984) "Comparative growth of obsolete and modern cultivars. I.
Vegetative dry matter partitioning." Crop Science 24: 858-872.4.
Wiedenfels, R.R (2000) "Effects of irrigation and N fertilizer application on sugarcane yield and
quality." Field Crops Research 43:101 -108.
Land Use, Land-Use Change and Forestry
Representation of the U.S. Land Base
Alaska Department of Natural Resources (2006) Alaska Infrastructure 1:63,360. Available online at:
http://dnr.alaska.gov/SpatialUtility/SUC?cmd=extract&laverid=75.
Alaska Interagency Fire Management Council (1998) Alaska Interagency Wildland Fire Management
Plan. Available online at: http://agdc.usgs.gov/data/blm/fire/index.html.
Alaska Oil and Gas Conservation Commission (2009) Oil and Gas Information System. Available
online at: http://doa.alaska.gov/ogc/publicdb.html.
Dewitz, J., 2023, National Land Cover Database (NLCD) 2021 Products: U.S. Geological Survey data
release, https://doi.org/10.5066/P9JZ7AQ3.
EIA (2011) Coal Production and Preparation Report Shapefile. Available online at:
http://www.eia.goV/state/notes-sources.cfm#maps.
ESRI (2008) ESRI Data &Maps. Redlands, CA: Environmental Systems Research Institute. [CD-
ROM].
Fry, J., Xian, G., Jin, S., Dewitz, J., Homer, C., Yang, L., Barnes, C., Herold, N., and J. Wickham. (2011)
Completion of the 2006 National Land Cover Database for the Conterminous United States,
PE&RS, Vol. 77(9):858-864.
Homer, C., J. Dewitz, J. Fry, M. Coan, N. Hossain, C. Larson, N. Herold, A. McKerrow, J.N. VanDriel
and J. Wickham. (2007) Completion of the 2001 National Land Cover Database for the
Conterminous United States, Photogrammetric Engineering and Remote Sensing, Vol. 73, No.
4, pp 337-341.
10-94 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Homer, C.G., Dewitz, J.A., Yang, L., Jin, S., Danielson, P., Xian, G., Coulston, J., Herold, N.D.,
Wickham, J.D., and Megown, K. (2015) Completion of the 2011 National Land Cover Database
for the conterminous United States-Representing a decade of land cover change information.
Photogrammetric Engineering and Remote Sensing, v. 81, no. 5, p. 345-354.
IPCC (2019) 207 9 Refinement to the 2006IPCC Guidelines for National Greenhouse Gas
Inventories, Calvo Buendia, E., Tanabe, K., Kranjc, A., Baasansuren, J., Fukuda, M., Ngarize S.,
Osako, A., Pyrozhenko, Y., Shermanau, P. and Federici, S. (eds). Published: IPCC, Switzerland.
IPCC (2014) 2073 Supplement to the 2006 IPCC Guidelines for National Greenhouse Gas
Inventories: Wetlands. Hiraishi, T., Krug, T., Tanabe, K., Srivastava, N., Baasansuren, J., Fukuda,
M. and Troxler, T.G. (eds.). Published: IPCC, Switzerland.
IPCC (2010) Revisiting the use of managed land as a proxy for estimating national anthropogenic
emissions and removals. [Eggleston HS, Srivastava N, Tanabe K, Baasansuren J, (eds.)].
Institute for Global Environmental Studies, Intergovernmental Panel on Climate Change,
Hayama, Kanagawa, Japan.
IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change.
[H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa,
Japan.
Jin, S., L. Yang, P. Danielson, C. Homer, J. Fry, and G. Xian. (2013) A comprehensive change
detection method for updating the National Land Cover Database to circa 2011. Remote
Sensing of Environment, 132:159-175.
U.S. Department of Agriculture, Natural resources Conservation Service. (2023) National resources
Inventory (NRI) Program Technical Manual. Available online at:
www.nrcs.usda.gov/sites/default/files/2023-10/NRI-Tech nicalManual-October2Q23.pdf.
Nelson, M.D., Riitters, K.H., Coulston, J.W., Domke, G.M., Greenfield, E.J., Langner, L.L., Nowak,
D.J., O'Dea, C.B., Oswalt, S.N., Reeves, M.C. and Wear, D.N. (2020) Defining the United States
land base: a technical document supporting the USDA Forest Service 2020 RPA assessment.
Gen. Tech. Rep. NRS-191., 191, pp.1-70.
NOAA Coastal Change Analysis Program (C-CAP) Regional Land Cover Database. Data collected
1995-present Charleston, SC: National Oceanic and Atmospheric Administration (NOAA)
Coastal Services Center. Data accessed at: www.csc.noaa.gov/landcover.
Nusser, S.M. and J.J. Goebel (1997) "The national resources inventory: a long-term multi-resource
monitoring programme." Environmental and Ecological Statistics 4:181-204.
Ogle, S.M., G. Domke, W.A. Kurz, M.T. Rocha, T. Huffman, A. Swan, J.E. Smith, C. Woodall, T. Krug
(2018) Delineating managed land for reporting greenhouse gas emissions and removals to the
United Nations Framework Convention on Climate Change. Carbon Balance and Management
13:9.
References and Abbreviations 10-95
-------
U.S. Census Bureau (2010) Topologically Integrated Geographic Encoding and Referencing (TIGER)
system shapefiles. U.S. Census Bureau, Washington, D.C. Available online at:
http://www.census.gov/geo/www/tiger.
U.S. Department of Agriculture (2015) County Data - Livestock, 1990-2014. U.S. Department of
Agriculture, National Agriculture Statistics Service, Washington, D.C.
U.S. Department of Agriculture, Forest Service. (2023) Timber Product Output (TPO) Reports.
https://www.fia.fs.usda.gov/program-features/tpo/
U.S. Department of Interior (2005) Federal Lands of the United States. National Atlas of the United
States, U.S. Department of the Interior, Washington D.C. Available online at:
http://nationalatlas.gov/atlasftp.html?openChapters=chpbound#chpbound.
United States Geological Survey (USGS), Gap Analysis Program (2012) Protected Areas Database of
the United States (PADUS), version 1.3 Combined Feature Class. November 2012.
USGS (2012) Alaska Resource Data File. Available online at: http://ardf.wr.usgs.gov/.
USGS (2005) Active Mines and Mineral Processing Plants in the United States in 2003. U.S.
Geological Survey, Reston, VA.
Yang, L., Jin, S., Danielson, P., Homer, C., Gass, L., Bender, S. M., Case, A., Costello, C., Dewitz, J.,
Fry, J., Funk, M., Granneman, B., Liknes, G. C., Rigge, M. &Xian, G. (2018) A new generation of
the United States National Land Cover Database: Requirements, research priorities, design,
and implementation strategies. ISPRS Journal of Photogrammetry and Remote Sensing 146:
108-123.
Forest Land Remaining Forest Land: Changes in Forest
Carbon Stocks
AF&PA (2006a and earlier) Statistical roundup. (Monthly). Washington, D.C. American Forest &
Paper Association.
AF&PA (2006b and earlier) Statistics of paper, paperboard and wood pulp. Washington, D.C.
American Forest & Paper Association.
AF&PA (2021) 2020 Statistics - Paper Industry-Annual Summary Data Through 2020. Washington,
D.C.: American Forest and Paper Association, 54 p.
AF&PA (2023) Capacity & Fiber Consumption Survey - Paper Industry-63rd Annual 2022-2023.
Washington, D.C.: American Forest and Paper Association, 44 p.
Amichev, B.Y. and J.M. Galbraith (2004) "A Revised Methodology for Estimation of Forest Soil Carbon
from Spatial Soils and Forest Inventory Data Sets." Environmental Management 33(Suppl.
1 ):S74-S86.
Bechtold, W.A.; Patterson, P.L. (2005) The enhanced forest inventory and analysis program—
national sampling design and estimation procedures. Gen. Tech. Rep. SRS-80. Asheville, NC:
U.S. Department of Agriculture Forest Service, Southern Research Station. 85 p.
10-96 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Birdsey, R. (1996) "Carbon Storage for Major Forest Types and Regions in the Conterminous United
States." In R.N. Sampson and D. Hair, (eds.). Forest and Global Change, Volume 2: Forest
Management Opportunities for Mitigating Carbon Emissions. American Forests. Washington,
D.C., 1-26 and 261-379 (appendices 262 and 263).
Coulston, J.W., Wear, D.N., and Vose, J.M. (2015) Complex forest dynamics indicate potential for
slowing carbon accumulation in the southeastern United States. Scientific Reports. 5: 8002.
Deenik J, McClellan AT (2007) Soils of Hawai'i. Soil and Crop Management, SCM-20. College of
Tropical Agriculture and Human Resources, University of Hawai'i at Manoa, Honolulu.
Domke, G.M., Perry, C.H., Walters, B.F., Woodall, C.W., and Smith, J.E. (2016) A framework for
estimating litter carbon stocks in forests of the United States. Science of the Total Environment
557-558: 469-478.
Domke, G.M., Perry, C.H., Walters, B.F., Woodall, C.W., Nave, L., Swanston, C. (2017) Toward
inventory-based estimates of soil organic carbon in forests of the United States. Ecological
Applications. 27(4), 1223-1235.
Domke, G.M., Walters, B.F., Smith, J.E., Woodall, C.W. (2022) Chapter 6: FIA Carbon Attributes. In
Sampling and estimation documentation for the Enhanced Forest Inventory and Analysis
Program: 2022. Westfall, J.A., Coulston, J.W., Moisen, G.G., Andersen, H.-E., eds. Gen. Tech.
Rep. NRS-GTR-207, Madison, Wl: U.S. Department of Agriculture, Forest Service, Northern
Research Station. 129 p. https://doi.org/10.2737/NRS-GTR-207.
Domke, G.M., Woodall, C.W., Smith, J.E., Westfall, J.A., McRoberts, R.E. (2012) Consequences of
alternative tree-level biomass estimation procedures on U.S. forest carbon stock estimates.
Forest Ecology and Management. 270:108-116.
Domke, G.M., Woodall, C.W., Walters, B.F., Smith, J.E. (2013) From models to measurements:
comparing down dead wood carbon stock estimates in the U.S. forest inventory. PLoS ONE 8(3):
e59949.
EPA (2006) Municipal solid waste in the United States: 2005 Facts and figures. Office of Solid
Waste, U.S. Environmental Protection Agency. Washington, D.C. (5306P) EPA530-R-06-011.
Available online at: http://www.epa.gov/msw/msw99.htm.
FAO. Forestry Production and Trade. License: CC BY-NC-SA 3.0 IGO. Extracted from:
https://www.fao.Org/faostat/en/#data/FQ. Date of Access: 13-09-2023.
Frayer, W.E., and G.M. Furnival (1999) "Forest Survey Sampling Designs: A History." Journal of
Forestry 97(12): 4-10.
Freed, R. (2004) Open-dump and Landfill timeline spreadsheet (unpublished). ICF International.
Washington, D.C.
Hair, D. (1958) "Historical forestry statistics of the United States." Statistical Bull. 228. U.S.
Department of Agriculture Forest Service, Washington, D.C.
Hair. D. and A.H. Ulrich (1963) The Demand and price situation for forest products - 1963. U.S.
Department of Agriculture Forest Service, Misc Publication No. 953. Washington, D.C.
References and Abbreviations 10-97
-------
Harmon, M.E., C.W. Woodall, B. Fasth, J. Sexton, M. Yatkov. (2011) Differences between standing
and downed dead tree wood density reduction factors: A comparison across decay classes and
tree species. Res. Paper. NRS-15. Newtown Square, PA: U.S. Department of Agriculture, Forest
Service, Northern Research Station. 40 p.
Howard, J. L. and Liang, S. (2019) U.S. timber production, trade, consumption, and price statistics
1965 to 2017. Res. Pap. FPL-RP-701. Madison, Wl: USDA, Forest Service, Forest Products
Laboratory.
Howard, J. L. and Jones, K.C. (2016) U.S. timber production, trade, consumption, and price
statistics 1965 to 2013. Res. Pap. FPL-RP-679. Madison, Wl: USDA, Forest Service, Forest
Products Laboratory.
Howard, J. L. (2007) U.S. timber production, trade, consumption, and price statistics 1965 to 2005.
Res. Pap. FPL-RP-637. Madison, Wl: USDA, Forest Service, Forest Products Laboratory.
Howard, J. L. (2003) U.S. timber production, trade, consumption, and price statistics 1965 to 2002.
Res. Pap. FPL-RP-615. Madison, Wl: USDA, Forest Service, Forest Products Laboratory.
Available online at: http://www.fpl.fs.fed.us/documnts/fplrp/fplrp615/fplrp615.pdf.
IPCC (2014) 207 3 Supplement to the 2006IPCC Guidelines for National Greenhouse Gas
Inventories: Wetlands. [Hiraishi, T., Krug, T., Tanabe, K., Srivastava, N., Baasansuren, J., Fukuda,
M., and Troxler, T.G. (eds.)]. Switzerland.
IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to
the Fourth Assessment Report of the Intergovernmental Panel on Climate Change [Solomon, S.,
D. Qin, M. Manning, Z. Chen, M. Marquis, K.B. Averyt, M. Tignorand H.L. Miller (eds.)].
Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, 996 pp.
IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change.
[H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa,
Japan.
Jandl, R., Rodeghiero, M., Martinez, C., Cotrufo, M. F., Bampa, F., van Wesemael, B., Harrison, R.B.,
Guerrini, I.A., deB Richter Jr., D., Rustad, L., Lorenz, K., Chabbi, A., Miglietta, F. (2014) Current
status, uncertainty and future needs in soil organic carbon monitoring. Science of the Total
Environment, 468, 376-383.
Jenkins, J.C., D.C. Chojnacky, L.S. Heath, and R.A. Birdsey (2003) "National-scale biomass
estimators for United States tree species." Forest Science 49(1):12-35.
Johnson, K. Domke, G.M., Russell, M.B., Walters, B.F., Horn, J., Peduzzi, A., Birdsey, R., Dolan, K.,
Huang, W. (2017) Estimating aboveground live understory vegetation carbon in the United
States. Environmental Research Letters.
Johnson, M.G. and Kern, J.S., (2002.) Quantifying the organic carbon held in forested soils of the
United States and Puerto Rico. The potential of US forest soils to sequester and mitigate the
greenhouse effect, Lewis, Boca Raton, pp.47-72.
10-98 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Nelson, M.D., Riitters, K.H., Coulston, J.W., Domke, G.M., Greenfield, E.J., Langner, L.L., Nowak,
D.J., O'Dea, C.B., Oswalt, S.N., Reeves, M.C. and Wear, D.N. (2020) Defining the United States
land base: a technical document supporting the USDA Forest Service 2020 RPA assessment.
Gen. Tech. Rep. NRS-191., 191, pp.1-70.
Ogle, S. M., G. M. Domke, W. A. Kurz, M. T. Rocha, T. Huffman, A. Swan, J. E. Smith, C. W. Woodall,
and T. Krug. (2018) Delineating managed land for reporting national greenhouse gas emissions
and removals to the United Nations framework convention on climate change. Carbon Balance
and Management 13:9.
O'Neill, K.R, Amacher, M.C., Perry, C.H. (2005) Soils as an indicator of forest health: a guide to the
collection, analysis, and interpretation of soil indicator data in the Forest Inventory and Analysis
program. Gen. Tech. Rep. NC-258. St. Paul, MN: U.S. Department of Agriculture, Forest Service,
North Central Research Station. 53 p.
Oswalt, S.N., Brandeis, T.J. and Woodall, C.W., (2008). Contribution of dead wood to biomass and
carbon stocks in the Caribbean: St. John, US Virgin Islands. Biotropica, 40(1), pp.20-27.
Oswalt, S.N., Smith, W.B., Miles, P.D. and Pugh, S.A. (2019) Forest resources of the United States,
2017: A technical document supporting the Forest Service 2020 RPA Assessment. Gen. Tech.
Rep. WO-97. Washington, DC: U.S. Department of Agriculture, Forest Service, Washington
Office., 97.
Perry, C.H., C.W. Woodall, and M. Schoeneberger (2005) Inventorying trees in agricultural
landscapes: towards an accounting of "working trees". In: "Moving Agroforestry into the
Mainstream." Proc. 9th N. Am. Agroforestry Conf., Brooks, K.N. and Folliott, P.F. (eds.). 12-15
June 2005, Rochester, MN [CD-ROM]. Dept. of Forest Resources, Univ. Minnesota, St. Paul, MN,
12 p. Available online at: http://cinram.umn.edu/afta2005/. (verified 23 Sept 2006).
Russell, M.B.; D'Amato, A.W.; Schulz, B.K.; Woodall, C.W.; Domke, G.M.; Bradford, J.B. (2014)
Quantifying understory vegetation in the U.S. Lake States: a proposed frameworkto inform
regional forest carbon stocks. Forestry. 87: 629-638.
Russell, M.B.; Domke, G.M.; Woodall, C.W.; D'Amato, A.W. (2015) Comparisons of allometric and
climate-derived estimates of tree coarse root carbon in forests of the United States. Carbon
Balance and Management. 10: 20.
Selmants, P.C., Giardina, C.P., Jacobi, J.D., and Zhu, Zhiliang, eds. (2017) Baseline and projected
future carbon storage and carbon fluxes in ecosystems of Hawai'i: U.S. Geological Survey
Professional Paper 1834,134 p., https://doi.org/10.3133/pp1834.
Skog, K.E. (2008) Sequestration of carbon in harvested wood products for the United States. Forest
Products Journal 58:56-72.
Smith, J.E., Domke, G.M. and Woodall, C.W. (2022) Predicting downed woody material carbon
stocks in forests of the conterminous United States. Science of the Total Environment, 803,
p.150061.
References and Abbreviations 10-99
-------
Smith, J.E., L.S. Heath, and M.C. Nichols (2010) U.S. Forest Carbon Calculation Tool User's Guide:
Forestland Carbon Stocks and Net Annual Stock Change. General Technical Report NRS-13
revised, U.S. Department of Agriculture Forest Service, Northern Research Station, 34 p.
Smith, J.E.; Heath, L.S.; Skog, K.E.; Birdsey, R.A. (2006) Methods for calculating forest ecosystem
and harvested carbon with standard estimates for forest types of the United States. Gen. Tech.
Rep. NE-343. Newtown Square, PA: U.S. Department of Agriculture, Forest Service,
Northeastern Research Station. 216 p.
Smith, W. B., R D. Miles, C. H. Perry, and S. A. Pugh (2009) Forest Resources of the United States,
2007. General Technical Report WO-78, U.S. Department of Agriculture Forest Service,
Washington Office.
Soil Survey Staff (2020a) Gridded National Soil Survey Geographic (gNATSGO) Database for the
Conterminous United States. United States Department of Agriculture, Natural Resources
Conservation Service. Available online at https://nrcs.app.box.eom/v/soils.
Soil Survey Staff (2020b) Gridded National Soil Survey Geographic (gNATSGO) Database for Alaska.
United States Department of Agriculture, Natural Resources Conservation Service. Available
online at https://nrcs.app.box.eom/v/soils.
Steer, Henry B. (1948) Lumber production in the United States. Misc. Pub. 669, U.S. Department of
Agriculture Forest Service. Washington, D.C.
Ulrich, Alice (1985) U.S. Timber Production, Trade, Consumption, and Price Statistics 1950-1985.
Misc. Pub. 1453, U.S. Department of Agriculture Forest Service. Washington, D.C.
Ulrich, Alice (1989) U.S. Timber Production, Trade, Consumption, and Price Statistics, 1950-1987.
USDA Miscellaneous Publication No. 1471, U.S. Department of Agriculture Forest Service.
Washington, D.C., 77.
United Nations Framework Convention on Climate Change (UNFCC) (2013) Report on the individual
review of the inventory submission of the United States of America submitted in 2012.
FCCC/ARR/2012/USA. 42 p.
USDA Forest Service (2022a) PNW-FIA Hawai'i Inventory Database. Last accessed: 28 September
2023. https://www.fs.usda.gov/research/pnw/products/dataandtools/tools/pnw-fia-hawaii-
inventorv-database
USDA Forest Service (2022b) PNW-FIA Pacific Islands Inventory Database. Last accessed: 28
September 2023. https://www.fs.usda.gov/research/pnw/products/dataandtools/tools/pnw-
fia-pacific-islands-database
https://www.fs.usda.gov/research/pnw/products/dataandtools/tools/pnw-fia-hawaii-inventory-
database
USDA Forest Service (2024a)Forest Inventory and Analysis: Nationwide Forest Inventory (NFI). U.S.
Department of Agriculture Forest Service. Washington, D.C. Available online at:
https://research.fs.usda.gov/programs/nfi. Accessed 19 September 2024.
USDA Forest Service. (2024b) Forest Inventory and Analysis: FIA DataMart. U.S. Department of
Agriculture Forest Service. Washington, D.C. Available online at:
10-100 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
https://research.fs.usda.gov/products/dataandtools/tools/fia-datamart. Accessed on 19
September 2024.
USDA Forest Service. (2024c) Forest Inventory and Analysis: Nationwide Forest Inventory Field
Guide. U.S. Department of Agriculture Forest Service. Washington, D.C. Available online at:
https://research.fs.usda.gov/understory/nationwide-forest-inventory-field-guide. Accessed on
26 September 2024.
USDA Forest Service (2024d) Forest Inventory and Analysis: The Forest Inventory and Analysis
Database User Guide (NFI). U.S. Department of Agriculture, Forest Service, Washington Office.
Available online at: https://research.fs.usda.gov/understory/forest-inventory-and-analysis-
database-user-guide-nfi. Accessed on 26 September 2024.
U.S. Census Bureau (1976) Historical Statistics of the United States, Colonial Times to 1970, Vol. 1.
Washington, D.C.
Wear, D.N., Coulston, J.W. (2015) From sink to source: Regional variation in U.S. forest carbon
futures. Scientific Reports. 5:16518.
Westfall, J.A., Coulston, J.W., Gray, A.N., Shaw, J.D., Radtke, P.J., Walker, D.M., Weiskittel, A.R.,
MacFarlane, D.W., Affleck, D.L.R., Zhao, D., Temesgen, H., Poudel, K.R, Frank, J.M., Prisley, S.P.,
Wang, Y., Sanchez Meador, A.J., Auty, D., and Domke, G.M. (2024) A national-scale tree volume,
biomass, and carbon modeling system for the United States. Gen. Tech. Rep. WO-104.
Washington, DC: U.S. Department of Agriculture, Forest Service. 37 p.
https//doi.org/10.7737/WQ-GTR-104.
Westfall, J.A., Woodall, C.W., Hatfield, M.A. (2013) A statistical power analysis of woody carbon flux
from forest inventory data. Climatic Change. 118: 919-931.
Woodall, C.W., Amacher, M.C., Bechtold, W.A., Coulston, J.W., Jovan, S., Perry, C.H., Randolph,
K.C., Schulz, B.K., Smith, G.C., Tkacz, B., Will-Wolf, S. (2011 b) "Status and future of the forest
health indicators program of the United States." Environmental Monitoring and Assessment.
177:419-436.
Woodall, C.W., and V.J. Monleon (2008) Sampling protocol, estimation, and analysis procedures for
the down woody materials indicator of the FIA program. Gen. Tech. Rep. NRS-22. Newtown
Square, PA: U.S. Department of Agriculture, Forest Service, Northern Research Station. 68 p.
Woodall, C.W., Coulston, J.W., Domke, G.M., Walters, B.F., Wear, D.N., Smith, J.E., Anderson, H.-E.,
Clough, B.J., Cohen, W.B., Griffith, D.M., Hagan, S.C., Hanou, I.S.; Nichols, M.C., Perry, C.H.,
Russell, M.B., Westfall, J.A., Wilson, B.T. (2015a) The U.S. Forest Carbon Accounting Framework:
Stocks and Stock change 1990-2016. Gen. Tech. Rep. NRS-154. Newtown Square, PA: U.S.
Department of Agriculture, Forest Service, Northern Research Station. 49 pp.
Woodall, C.W., L.S. Heath, G.M. Domke, and M.C. Nichols (2011 a) Methods and equations for
estimating aboveground volume, biomass, and carbon for trees in the U.S. forest inventory,
2010. Gen. Tech. Rep. NRS-88. Newtown Square, PA: U.S. Department of Agriculture, Forest
Service, Northern Research Station. 30 p.
References and Abbreviations 10-101
-------
Woodall, C.W., Walters, B.F., Coulston, J.W., D'Amato, A.W., Domke, G.M., Russell, M.B., Sowers,
P.A. (2015b) Monitoring network confirms land use change is a substantial component of the
forest carbon sink in the eastern United States. Scientific Reports. 5:17028.
Woodall, C.W., Walters, B.F., Oswalt, S.N., Domke, G.M., Toney, C., Gray, A.N. (2013) Biomass and
carbon attributes of downed woody materials in forests of the United States. Forest Ecology
and Management 305: 48-59.
Zhu, Zhiliang, and McGuire, A.D., eds. (2016) Baseline and projected future carbon storage and
greenhouse-gas fluxes in ecosystems of Alaska: U.S. Geological Survey Professional Paper
1826, 196 p., Available online at: http://dx.doi.org/10.3133/pp1826.
Forest Land Remaining Forest Land: Non-C02 Emissions
from Forest Fires
Eidenshink, J., Schwind, B., Brewer, K., Zhu, Z.L., Quayle, B. and Howard, S. (2007) A project for
monitoring trends in burn severity. Fire ecology, 3(1), pp.3-21.
French, N.H.F., W.J. de Groot, L.K. Jenkins, B.M. Rogers, E.C. Alvarado, B. Amiro, B. de Jong, S.
Goetz, E. Hoy, E. Hyer, R. Keane, D. McKenzie, S.G. McNulty, B.E Law, R. Ottmar, D.R. Perez-
Salicrup, J. Randerson, K.M. Robertson, and M. Turetsky (2011) "Model comparisons for
estimating carbon emissions from North American wildland fire." Journal of Geophysical
Research 116.10.1029/2010JG001469
French, N.H.F., D. McKenzie, T. Erickson, B. Koziol, M. Billmire, K.A. Endsley, N.K.Y. Scheinerman, L.
Jenkins, M.E. Miller, R. Ottmar, and S. Prichard (2014) "Modeling regional-scale fire emissions
with the Wildland Fire Emissions Information System." Earth Interactions 18, no. 16
Giglio, L., Boschetti, L., Roy, D. P., Humber, M. L., and Justice, C. O. (2018) The Collection 6 MODIS
burned area mapping algorithm and product. Remote Sensing of Environment, 217, 72-85.
IPCC (2019) Chapter 2: Generic Methodologies Applicable to Multiple Land-Use Categories.
Refinement to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories, Calvo
Buendia, E., Tanabe, K., Kranjc, A., Baasansuren, J., Fukuda, M., Ngarize, S., Osako, A.,
Pyrozhenko, Y., Shermanau, P. and Federici, S. (eds). Published: IPCC, Switzerland.
IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to
the Fifth Assessment Report of the Intergovernmental Panel on Climate Change [Stocker, T.F., D.
Qin, G.-K. Plattner, M. Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley
(eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, 1535
pp.
IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change.
[H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa,
Japan.
Larkin, N. K., S. Raffuse, and T. T. Strand (2014) Wildland fire emissions, carbon, and climate: U.S.
emissions inventories. For. Ecol. Manage. 317:61-69. doi:10.1016/j.foreco.2013.09.012.
10-102 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
MTBS Data Access: Fire Level Geospatial Data (August 2023) MTBS Project (USDA Forest
Service/U.S. Geological Survey). Available online at: http://mtbs.gov/direct-download.
Accessed on 15 August 2023.
Ogle, S. M., G. M. Domke, W. A. Kurz, M. T. Rocha, T. Huffman, A. Swan, J. E. Smith, C. W. Woodall,
and T. Krug. (2018) Delineating managed land for reporting national greenhouse gas emissions
and removals to the United Nations framework convention on climate change. Carbon Balance
and Management 13:9.
Smith, James E.; Billmire, Michael; French, Nancy H.F.; Domke, Grant M. 2024. Application of the
wildland fire emissions inventory system to estimate fire emissions on forest lands of the
United States. Carbon Balanceand Management. 19: 26. https://doi.org/10.1186/s13021-024-
00274-0.
WFEIS (2023) Wildland Fire Emissions Inventory System, home page, https://wfeis.mtri.org/.
WFIGS (2023) WFIGS, The Wildland Fire Interagency Geospatial Service Interagency Fire
Perimeters. National Interagency Fire Center. https://data-
nifc.opendata.arcgis.com/datasets/nifc: :wfigs-interagency-fire-
perimeters/explore?location=0.000000%2C0.000000%2C1.82.
Forest Land Remaining Forest Land: N20 Emissions from
Soils
Albaugh, T.J., Allen, H.L., Fox, T.R. (2007) Historical Patterns of Forest Fertilization in the
Southeastern United States from 1969 to 2004. Southern Journal of Applied Forestry, 31,129-
137(9).
Binkley, D. (2004) Email correspondence regarding the 95 percent confidence interval for area
estimates of southern pine plantations receiving N fertilizer (±20%) and the rate applied for
areas receiving N fertilizer (100 to 200 pounds/acre). Dan Binkley, Department of Forest,
Rangeland, and Watershed Stewardship, Colorado State University and Stephen Del Grosso,
Natural Resource Ecology Laboratory, Colorado State University. September 19, 2004.
Binkley, D., R. Carter, and H.L.Allen (1995) Nitrogen Fertilization Practices in Forestry. In: Nitrogen
Fertilization in the Environment, P.E. Bacon (ed.), Marcel Decker, Inc., New York.
Briggs, D. (2007) Management Practices on Pacific Northwest West-Side Industrial Forest Lands,
1991-2005: With Projections to 2010. Stand Management Cooperative, SMC Working Paper
Number 6, College of Forest Resources, University of Washington, Seattle.
Fox, T.R., H. L. Allen, T.J. Albaugh, R. Rubilar, and C.A. Carlson (2007) Tree Nutrition and Forest
Fertilization of Pine Plantations in the Southern United States. Southern Journal of Applied
Forestry, 31,5-11.
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change.
[H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa,
Japan.
References and Abbreviations 10-103
-------
USDA Forest Service (2001) U.S. Forest Facts and Historical Trends. FS-696. U.S. Department of
Agriculture Forest Service, Washington, D.C. Available online at:
http://www.fia.fs.fed.us/librarv/ForestFactsMetric.pdf.
Forest Land Remaining Forest Land: Drained Organic
Soils
IPCC (2014) 207 3 Supplement to the 2006IPCC Guidelines for National Greenhouse Gas
Inventories: Wetlands, Hiraishi, T., Krug, T., Tanabe, K., Srivastava, N., Baasansuren, J., Fukuda,
M. and Troxler, T.G. (eds.). Published: IPCC, Switzerland.
IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to
the Fifth Assessment Report of the Intergovernmental Panel on Climate Change [Stocker, T.F., D.
Qin, G.-K. Plattner, M. Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley
(eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, 1535
pp.
IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change, H.S.
Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Hayama, Kanagawa, Japan.
STATSG02 (2016) Soil Survey Staff, Natural Resources Conservation Service, United States
Department of Agriculture. U.S. General Soil Map (STATSG02). Available online at
https://sdmdataaccess.sc.egov.usda.gov. Accessed 10 November 2016.
USDA Forest Service. (2024b) Forest Inventory and Analysis: FIA DataMart. U.S. Department of
Agriculture Forest Service. Washington, D.C. Available online at:
https://research.fs.usda.gov/products/dataandtools/tools/fia-datamart. Accessed on 19
September 2024.
USDA Forest Service. (2024c) Forest Inventory and Analysis: Nationwide Forest Inventory Field
Guide. U.S. Department of Agriculture Forest Service. Washington, D.C. Available online at:
https://research.fs.usda.gov/understory/nationwide-forest-inventory-field-guide. Accessed on
26 September 2024.
USDA Forest Service (2024d) Forest Inventory and Analysis: The Forest Inventory and Analysis
Database User Guide (NFI). U.S. Department of Agriculture, Forest Service, Washington Office.
Available online at: https://research.fs.usda.gov/understory/forest-inventory-and-analysis-
database-user-guide-nfi. Accessed on 26 September 2024.
Land Converted to Forest Land
Birdsey, R. (1996) "Carbon Storage for Major Forest Types and Regions in the Conterminous United
States." In R.N. Sampson and D. Hair, (eds.). Forest and Global Change, Volume 2: Forest
Management Opportunities for Mitigating Carbon Emissions. American Forests. Washington,
D.C., 1-26 and 261-379 (appendices 262 and 263).
10-104 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Brockwell, Peter J., and Richard A. Davis (2016) Introduction to time series and forecasting.
Springer.
Domke, G.M., Perry, C.H., Walters, B.F., Woodall, C.W., and Smith, J.E. (2016) A framework for
estimating litter carbon stocks in forests of the United States. Science of the Total Environment
557-558: 469-478.
Domke, G.M., Woodall, C.W., Walters, B.F., Smith, J.E. (2013) From models to measurements:
comparing down dead wood carbon stock estimates in the U.S. forest inventory. PLoS ONE 8(3):
e59949.
Harmon, M.E., C.W. Woodall, B. Fasth, J. Sexton, M. Yatkov. (2011) Differences between standing
and downed dead tree wood density reduction factors: A comparison across decay classes and
tree species. Res. Paper. NRS-15. Newtown Square, PA: U.S. Department of Agriculture, Forest
Service, Northern Research Station. 40 p.
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change.
[H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa,
Japan.
Jenkins, J.C., D.C. Chojnacky, L.S. Heath, and R.A. Birdsey (2003) "National-scale biomass
estimators for United States tree species." Forest Science 49(1):12-35.0gle, S.M., M.D. Eve, F.J.
Breidt, and K. Paustian (2003) "Uncertainty in estimating land use and management impacts on
soil organic carbon storage for U.S. agroecosystems between 1982 and 1997." Global Change
Biology 9:1521-1542.
Ogle, S.M., F.J. Breidt, and K. Paustian. (2006) "Bias and variance in model results due to spatial
scaling of measurements for parameterization in regional assessments." Global Change Biology
12:516-523.
Smith, J.E.; Heath, L.S.; Skog, K.E.; Birdsey, R.A. (2006) Methods for calculating forest ecosystem
and harvested carbon with standard estimates for forest types of the United States. Gen. Tech.
Rep. NE-343. Newtown Square, PA: U.S. Department of Agriculture, Forest Service,
Northeastern Research Station. 216 p.
USDA Forest Service (2024b) Forest Inventory and Analysis: FIA DataMart. U.S. Department of
Agriculture Forest Service. Washington, D.C. Available online at:
https://research.fs.usda.gov/products/dataandtools/tools/fia-datamart. Accessed on 19
September 2024.
USDA Forest Service (2024c) Forest Inventory and Analysis:Nationwide Forest Inventory Field
Guide. U.S. Department of Agriculture Forest Service. Washington, D.C. Available online at:
https://research.fs.usda.gov/understory/nationwide-forest-inventory-field-guide. Accessed on
26 September 2024.
USDA-NRCS (2020) Summary Report: 2017 National Resources Inventory. Natural Resources
Conservation Service, Washington, DC, and Center for Survey Statistics and Methodology, Iowa
References and Abbreviations 10-105
-------
State University, Ames, Iowa.
https://www.nrcs.usda.gov/wps/portal/nrcs/main/national/technical/nra/nri/results/.
USDA-NRCS (2018) Summary Report: 2015 National Resources Inventory, Natural Resources
Conservation Service, Washington, D.C., and Center for Survey Statistics and Methodology,
Iowa State University, Ames, Iowa.
https://www.nrcs.usda.gov/lnternet/FSE DOCUMENTS/nrcseprd1422028.pdf.
USDA-NRCS (1997) "National Soil Survey Laboratory Characterization Data," Digital Data, Natural
Resources Conservation Service, U.S. Department of Agriculture. Lincoln, NE.
Westfall, J.A., Coulston, J.W., Gray, A.N., Shaw, J.D., Radtke, P.J., Walker, D.M., Weiskittel, A.R.,
MacFarlane, D.W., Affleck, D.L.R., Zhao, D., Temesgen, H., Poudel, K.R, Frank, J.M., Prisley, S.P.,
Wang, Y., Sanchez Meador, A.J., Auty, D., and Domke, G.M. (2024) A national-scale tree volume,
biomass, and carbon modeling system for the United States. Gen. Tech. Rep. WO-104.
Washington, DC: U.S. Department of Agriculture, Forest Service. 37 p.
https//doi.org/10.2737/WQ-GTR-104.
Woodall, C.W., L.S. Heath, G.M. Domke, and M.C. Nichols (2011 a) Methods and equations for
estimating aboveground volume, biomass, and carbon for trees in the U.S. forest inventory,
2010. Gen. Tech. Rep. NRS-88. Newtown Square, PA: U.S. Department of Agriculture, Forest
Service, Northern Research Station. 30 p.
Woodall, C.W., and V.J. Monleon (2008) Sampling protocol, estimation, and analysis procedures for
the down woody materials indicator of the FIA program. Gen. Tech. Rep. NRS-22. Newtown
Square, PA: U.S. Department of Agriculture, Forest Service, Northern Research Station. 68 p.
Woodall, C.W., Walters, B.F., Coulston, J.W., D'Amato, A.W., Domke, G.M., Russell, M.B., Sowers,
P.A. (2015b) Monitoring network confirms land use change is a substantial component of the
forest carbon sink in the eastern United States. Scientific Reports. 5:17028.
Woodall, C.W., Walters, B.F., Oswalt, S.N., Domke, G.M., Toney, C., Gray, A.N. (2013) Biomass and
carbon attributes of downed woody materials in forests of the United States. Forest Ecology
and Management 305: 48-59.
Yang, L., Jin, S., Danielson, P., Homer, C., Gass, L., Bender, S. M., Case, A., Costello, C., Dewitz, J.,
Fry, J., Funk, M., Granneman, B., Liknes, G. C., Rigge, M. &Xian, G. (2018) A new generation of
the United States National Land Cover Database: Requirements, research priorities, design,
and implementation strategies. ISPRS Journal of Photogrammetry and Remote Sensing 146:
108-123
Cropland Remaining Cropland
Armentano, T. V., and E.S. Menges (1986) Patterns of change in the carbon balance of organic soil-
wetlands of the temperate zone. Journal of Ecology 74: 755-774.
Brady, N.C. and R.R. Weil (1999) The Nature and Properties of Soils. Prentice Hall. Upper Saddle
River, NJ, 881.
10-106 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Brockwell, Peter J., and Richard A. Davis (2016) Introduction to time series and forecasting.
Springer.
Conant, R. T., K. Paustian, and E.T. Elliott (2001) "Grassland management and conversion into
grassland: effects on soil carbon." Ecological Applications 11: 343-355.
CTIC (2004) National Crop Residue Management Survey: 1989-2004. Conservation Technology
Information Center, Purdue University, Available online at: http://www.ctic.purdue.edu/CRM/.
Daly, C., R.P. Neilson, and D.L. Phillips (1994) "A Statistical-Topographic Model for Mapping
Climatological Precipitation Over Mountainous Terrain." Journal of Applied Meteorology 33:140-
158.
Del Grosso, S.J., W.J. Parton, A.R. Mosier, M.D. Hartman, J. Brenner, D.S. Ojima, and D.S. Schimel
(2001) "Simulated Interaction of Carbon Dynamics and Nitrogen Trace Gas Fluxes Using the
DAYCENT Model." In Modeling Carbon and Nitrogen Dynamics for Soil Management, Schaffer,
M., L. Ma, S. Hansen, (eds.). CRC Press, Boca Raton, Florida, pp. 303-332.
Del Grosso, S.J., S.M. Ogle, W.J. Parton (2011) Soil organic matter cycling and greenhouse gas
accounting methodologies, Chapter 1, pp 3-13 DOI: 10.1021/bk-2011-1072.ch001 .In:
Understanding Greenhouse Gas Emissions from Agricultural Management, L. Guo, A.
Gunasekara, L. McConnell (eds.). American Chemical Society, Washington, D.C.
Edmonds, L., R. L. Kellogg, B. Kintzer, L. Knight, C. Lander, J. Lemunyon, D. Meyer, D.C. Moffitt, and
J. Schaefer (2003) "Costs associated with development and implementation of Comprehensive
Nutrient Management Plans." Part I—Nutrient management, land treatment, manure and
wastewater handling and storage, and recordkeeping. Natural Resources Conservation Service,
U.S. Department of Agriculture. Available online at:
http://www.nrcs.usda.gov/technical/land/pubs/cnmp1 .html.
Euliss, N., and R. Gleason (2002) Personal communication regarding wetland restoration factor
estimates and restoration activity data. Ned Euliss and Robert Gleason of the U.S. Geological
Survey, Jamestown, ND, to Stephen Ogle of the National Resource Ecology Laboratory, Fort
Collins, CO. August 2002.
Friedman, J.H. (2001) "Greedy function approximation: A gradient boosting machine." Ann. Statist.
29(5) 1189-1232.
Fry, J., Xian, G., Jin, S., Dewitz, J., Homer, C., Yang, L., Barnes, C., Herold, N., and Wickham, J. (2011)
Completion of the 2006 National Land Cover Database for the Conterminous United States,
PE&RS, Vol. 77(9):858-864.
Griscom, B. W., Adams, J., Ellis, P. W., Houghton, R. A., Lomax, G., Miteva, D. A., Schlesinger, W. H.,
Shoch, D., Siikamaki, J. V., Smith, P., Woodbury, P., Zganjar, C., Blackman, A., Campari, J.,
Conant, R. T., Delgado, C., Elias, P., Gopalakrishna, T., Hamsik, M. R., Herrero, M., Kiesecker, J.,
Landis, E., Laestadius, L., Leavitt, S. M., Minnemeyer, S., Polasky, S., Potapov, P., Putz, F. E.,
Sanderman, J., Silvius, M., Wollenberg, E. & Fargione, J. (2017) "Natural climate solutions."
Proceedings of the National Academy of Sciences of the United States of America 114(44):
11645-11650.
References and Abbreviations 10-107
-------
Hagen, S. C., G. Delgado, P. Ingraham, I. Cooke, R. Emery, J. P. Fisk, L. Melendy, T. Olson, S. Patti, N.
Rubin, B. Ziniti, H. Chen, W. Salas, P. Elias, and D. Gustafson. 2020. Mapping Conservation
Management Practices and Outcomes in the Corn Belt Using the Operational Tillage
Information System (OpTIS) and the Denitrification-Decomposition (DNDC) Model. Land 9:408
Homer, C., Dewitz, J., Fry, J., Coan, M., Hossain, N., Larson, C., Herald, N., McKerrow, A., VanDriel,
J.N., and Wickham, J. (2007) Completion of the 2001 National Land Cover Database for the
Conterminous United States. Photogrammetric Engineering and Remote Sensing, Vol. 73, No.
4, pp 337-341.
Homer, C.G., Dewitz, J.A., Yang, L., Jin, S., Danielson, P., Xian, G., Coulston, J., Herald, N.D.,
Wickham, J.D., and Megown, K. (2015) Completion of the 2011 National Land Cover Database
for the conterminous United States-Representing a decade of land cover change information.
Photogrammetric Engineering and Remote Sensing, v. 81, no. 5, p. 345-354.
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change.
[H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa,
Japan.
IPCC (2003) Good Practice Guidance for Land Use, Land-Use Change, and Forestry. The
Intergovernmental Panel on Climate Change, National Greenhouse Gas Inventories
Programme, J. Penman, et al., eds. August 13, 2004. Available online at: http://www.ipcc-
nggip.iges.or.jp/public/gpglulucf/gpglulucf.htm.
IPCC (2019) 2079 Refinement to the 2006 IPCC Guidelines for National Greenhouse Gas
Inventories. The National Greenhouse Gas Inventories Programme, The Integovernmental Panel
on Climate Change, Calvo Buendia, E., Tanabe, K., Kranjc, A., Baasansuren, J., Fukuda, M.,
Ngarize S., Osako, A., Pyrozhenko, Y.,Shermanau, P. and Federici, S. (eds). Switzerland.
Lai, R., Kimble, J. M., Follett, R. F. & Cole, C. V. (1998) The potential of U.S. cropland to sequester
carbon and mitigate the greenhouse effect. Chelsea, Ml: Sleeping Bear Press, Inc.
Little, R. (1988) "Missing-data adjustments in large surveys." Journal of Business and Economic
Statistics 6: 287-296.
McGill, W.B., and C.V. Cole (1981) Comparative aspects of cycling of organic C, N, S and P through
soil organic matter. Geoderma 26:267-286.
Metherell, A.K., L.A. Harding, C.V. Cole, and W.J. Parton (1993) "CENTURY Soil Organic Matter Model
Environment." Agroecosystem version 4.0. Technical documentation, GPSR Tech. Report No. 4,
USDA/ARS, Ft. Collins, CO.
Mesinger, F., G. DiMego, E. Kalnay, K. Mitchell, P. C. Shafran, W. Ebisuzaki, D. Jovic, J. Woollen, E.
Rogers, E. H. Berbery, M. B. Ek, Y. Fan, R. Grumbine, W. Higgins, H. Li, Y. Lin, G. Manikin, D.
Parrish, and W. Shi (2006) North American regional reanalysis. Bulletin of the American
Meteorological Society 87:343-360.
Nelson, Mark D.; Riitters, Kurt H.; Coulston, John W.; Domke, Grant M.; Greenfield, Eric J.; Langner,
Linda L.; Nowak, David J.; O'Dea, Claire B.; Oswalt, Sonja N.; Reeves, Matthew C.; Wear, David
10-108 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
N. 2020. Defining the United States land base: a technical document supporting the USDA
Forest Service 2020 RPA assessment. Gen. Tech. Rep. NRS-191. Madison, Wl: U.S. Department
of Agriculture, Forest Service, Northern Research Station. 70 p. https://doi.org/10.2737/NRS-
GTR-191.
NRCS (1999) Soil Taxonomy: A basic system of soil classification for making and interpreting soil
surveys, 2nd Edition. Agricultural Handbook Number 436, Natural Resources Conservation
Service, U.S. Department of Agriculture, Washington, D.C.
NRCS (1997) "National Soil Survey Laboratory Characterization Data," Digital Data, Natural
Resources Conservation Service, U.S. Department of Agriculture. Lincoln, NE.
NRCS (1981) Land Resource Regions and Major Land Resource Areas of the United States, USDA
Agriculture Handbook 296, United States Department of Agriculture, Natural Resources
Conservation Service, National Soil Survey Cente., Lincoln, NE, pp. 156.
Ogle, S.M., Breidt, F.J., Del Grosso, S., Gurung, R., Marx, E., Spencer, S., Williams, S., Manning, D.
(2023) "Counterfactual scenarios reveal historical impact of cropland management on soil
organic carbon stocks in the United States." Scientific Reports 13(1):14564.
Ogle, S. M., Alsaker, C., Baldock, J., Bernoux, M., Breidt, F. J., McConkey, B., Regina, K. & Vazquez-
Amabile, G. G. (2019) "Climate and Soil Characteristics Determine Where No-Till Management
Can Store Carbon in Soils and Mitigate Greenhouse Gas Emissions." Scientific Reports 9(1):
11665.
Ogle, S.M., F.J. Breidt, M. Easter, S. Williams, K. Killian, and K. Paustian (2010) "Scale and
uncertainty in modeled soil organic carbon stock changes for U.S. croplands using a process-
based model." Global Change Biology 16:810-820.
Ogle, S.M., F.J. Breidt, M. Easter, S. Williams and K. Paustian (2007) "Empirically-Based Uncertainty
Associated with Modeling Carbon Sequestration Rates in Soils." Ecological Modeling 205:453-
463.
Ogle, S.M., F.J. Breidt, and K. Paustian (2006) "Bias and variance in model results due to spatial
scaling of measurements for parameterization in regional assessments." Global Change Biology
12:516-523.
Ogle, S. M., et al. (2005) "Agricultural management impacts on soil organic carbon storage under
moist and dry climatic conditions of temperate and tropical regions." Biogeochemistry72: 87-
121.
Ogle, S.M., M.D. Eve, F.J. Breidt, and K. Paustian (2003) "Uncertainty in estimating land use and
management impacts on soil organic carbon storage for U.S. agroecosystems between 1982
and 1997." Global Change Biology 9:1521-1542.
Parton, W.J., M.D. Hartman, D.S. Ojima, and D.S. Schimel (1998) "DAYCENT: Its Land Surface
Submodel: Description and Testing". Glob. Planet. Chang. 19: 35-48.
Parton, W.J., D.S. Ojima, C.V. Cole, and D.S. Schimel (1994) "A General Model for Soil Organic
Matter Dynamics: Sensitivity to litter chemistry, texture and management," in Quantitative
References and Abbreviations 10-109
-------
Modeling of Soil Forming Processes. Special Publication 39, Soil Science Society of America,
Madison, Wl, 147-167.
Parton, W.J., D.S. Schimel, C.V. Cole, D.S. Ojima (1987) "Analysis of factors controlling soil organic
matter levels in Great Plains grasslands." Soil Science Society of America Journal 51:1173-1179.
Parton, W.J., J.W.B. Stewart, C.V. Cole. (1988) "Dynamics of C, N, P, and S in grassland soils: a
model." Biogeochemistry 5:109-131.
Paustian, K., et al. (1997a) "Agricultural soils as a sink to mitigate C02 emissions." Soil Use and
Management 13: 230-244.
Paustian, K., et al. (1997b) Management controls on soil carbon. In Soil organic matter in temperate
agroecosystems: long-term experiments in North America (Paul E.A., K. Paustian, and C.V. Cole,
eds.). Boca Raton, CRC Press, pp. 15-49.
Potter, C. S., J.T. Randerson, C.B. Fields, P.A. Matson, P.M. Vitousek, H.A. Mooney, and S.A. Klooster
(1993) "Terrestrial ecosystem production: a process model based on global satellite and
surface data." Global Biogeochemical Cycles 7:811-841.
Potter, C., S. Klooster, A. Huete, and V. Genovese (2007) Terrestrial carbon sinks for the United
States predicted from MODIS satellite data and ecosystem modeling. Earth Interactions 11,
Article No. 13, DO110.1175/EI228.1.
PRISM Climate Group (2022) PRISM Climate Data, Oregon State University,
http://prism.oregonstate.edu. downloaded January 2022.
Pukelsheim, F. (1994) "The 3-Sigma-Rule." American Statistician 48:88-91
Soil Survey Staff (2020) Gridded Soil Survey Geographic (gSSURGO) Database for the
Conterminous United States. United States Department of Agriculture, Natural Resources
Conservation Service, Accessed February 2020 (FY2020 official release), Available online at
https://gdg.sc.egov.usda.gov/.
Spencer, S., S.M. Ogle, F.J. Breidt, J. Goebel, and K. Paustian. (2011) "Designing a national soil
carbon monitoring network to support climate change policy: a case example for US
agricultural lands." Greenhouse Gas Management & Measurement 1:167-178.
Towery, D. (2001) Personal Communication. Dan Towery regarding adjustments to the CTIC (1998)
tillage data to reflect long-term trends, Conservation Technology Information Center, West
Lafayette, IN, and Marlen Eve, National Resource Ecology Laboratory, Fort Collins, CO. February
2001.
USDA-ERS (2020) Agricultural Resource Management Survey (ARMS) Farm Financial and Crop
Production Practices: Tailored Reports. Available online at: https://www.ers.usda.gov/data-
products/arms-farm-financial-and-crop-production-practices/.
USDA-ERS (2018) Agricultural Resource Management Survey (ARMS) Farm Financial and Crop
Production Practices: Tailored Reports. Available online at: https://www.ers.usda.gov/data-
products/arms-farm-financial-and-crop-production-practices/.
10-110 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
USDA-ERS (1997) Cropping Practices Survey Data—1995. Economic Research Service, United
States Department of Agriculture. Available online at:
http://www.ers.usda.gov/data/archive/93018/.
USDA Forest Service. (2022) Forest Inventory and Analysis National Program: FIA Data Mart. U.S.
Department of Agriculture Forest Service. Washington, D.C. Available online at:
https://apps.fs.usda.gov/fia/datamart/datamart.html. Accessed on 07 October 2022.
USDA-FSA (2015) Conservation Reserve Program Monthly Summary-September 2015. U.S.
Department of Agriculture, Farm Service Agency, Washington, D.C. Available online at:
https://www.fsa.usda.gov/Assets/USDA-FSA-
Public/usdafiles/Conservation/PDF/sep2015summarv.pdf.
USDA-NASS (2022) Quick Stats. National Agricultural Statistics Service, United States Department
of Agriculture, Washington, D.C., Accessed October 2022, http://quickstats.nass.usda.gov/.
USDA-NASS (2021) Published crop data layer. Available at
https://nassgeodata.gmu.edu/CropScape/. Accessed July 2021, USDA-NASS, Washington, DC.
USDA-NASS (2017) 2017 Census of Agriculture. USDA National Agricultural Statistics Service,
Complete data available at http://www.nass.usda.gov/AgCensus.
USDA-NASS (2012) 2012 Census of Agriculture. USDA National Agricultural Statistics Service,
Complete data available at http://www.nass.usda.gov/AgCensus.
USDA-NASS (2004) Agricultural Chemical Usage: 2003 Field Crops Summary. Report AgChl (04)a.
National Agricultural Statistics Service, U.S. Department of Agriculture, Washington, D.C.
Available online at: http://usda.mannlib.cornell.edu/reports/nassr/other/pcu-
bb/agcs0504.pdfH.
USDA-NASS (1999) Agricultural Chemical Usage: 1998 Field Crops Summary. Report AgCH1 (99).
National Agricultural Statistics Service, U.S. Department of Agriculture, Washington, DC.
Available online at: http://usda.mannlib.cornell.edu/reports/nassr/other/pcu-bb/agch0599.pdf.
USDA-NASS (1992) Agricultural Chemical Usage: 1991 Field Crops Summary. Report AgChl (92).
National Agricultural Statistics Service, U.S. Department of Agriculture, Washington, D.C.
Available online at: http://usda.mannlib.cornell.edu/reports/nassr/other/pcu-
bb/agch0392.txtH.
USDA-NRCS (2020) Summary Report: 2017 National Resources Inventory. Natural Resources
Conservation Service, Washington, DC, and Center for Survey Statistics and Methodology, Iowa
State University, Ames, Iowa, https://www.nrcs.usda.gov/sites/default/files/2022-
10/2017NRISummary Final.pdf.
USDA-NRCS (2018) CEAP Cropland Farmer Surveys. USDA Natural Resources Conservation
Service.
https://www.nrcs.usda.gov/wps/portal/nrcs/detail/national/technical/nra/ceap/na/?cid=nrcs1
43 014163.
USDA-NRCS (2012) Assessment of the Effects of Conservation Practices on Cultivated Cropland in
the Upper Mississippi River Basin. U.S. Department of Agriculture, Natural Resources
References and Abbreviations 10-111
-------
Conservation Service,
https://www.nrcs.usda.gov/lnternet/FSE DOCUMENTS/stelprdb1042093.pdf.
Yang, L., Jin, S., Danielson, P., Homer, C., Gass, L., Bender, S. M., Case, A., Costello, C., Dewitz, J.,
Fry, J., Funk, M., Granneman, B., Liknes, G. C., Rigge, M. &Xian, G. (2018) "A new generation of
the United States National Land Cover Database: Requirements, research priorities, design,
and implementation strategies." ISPRS Journal of Photogrammetry and Remote Sensing 146:
108-123.
Land Converted to Cropland
Sampson and D. Hair, (eds.). Forest and Global Change, Volume 2: Forest Management
Opportunities for Mitigating Carbon Emissions. American Forests. Washington, D.C., 1 -26 and
261-379 (appendices 262 and 263).
Brockwell, Peter J., and Richard A. Davis (2016) Introduction to time series and forecasting.
Springer.
Del Grosso, S.J., W.J. Parton, A.R. Mosier, M.D. Hartman, J. Brenner, D.S. Ojima, and D.S. Schimel
(2001) "Simulated Interaction of Carbon Dynamics and Nitrogen Trace Gas Fluxes Using the
DAYCENT Model." In Modeling Carbon and Nitrogen Dynamics for Soil Management, Schaffer,
M., L. Ma, S. Hansen, (eds.). CRC Press, Boca Raton, Florida, pp. 303-332.
Del Grosso, S.J., S.M. Ogle, W.J. Parton (2011) "Soil organic matter cycling and greenhouse gas
accounting methodologies." Chapter 1, pp 3-13 DOI: 10.1021/bk-2011-1072.ch001 .In:
Understanding Greenhouse Gas Emissions from Agricultural Management (L. Guo, A.
Gunasekara, L. McConnell. Eds.), American Chemical Society, Washington, D.C.
Del Grosso, S.J., W.J. Parton, A.R. Mosier, M.D. Hartman, J. Brenner, D.S. Ojima, and D.S. Schimel
(2001) "Simulated Interaction of Carbon Dynamics and Nitrogen Trace Gas Fluxes Using the
DAYCENT Model." In Schaffer, M., L. Ma, S. Hansen, (eds.); Modeling Carbon and Nitrogen
Dynamics for Soil Management. CRC Press. Boca Raton, Florida. 303-332.
Domke, G.M., J.E. Smith, and C.W. Woodall. (2011) "Accounting for density reduction and structural
loss in standing dead trees: Implications for forest biomass and carbon stock estimates in the
United States". Carbon Balance and Management 6:14.
Domke, G.M., et al. (2013) "From models to measurements: comparing down dead wood carbon
stock estimates in the U.S. forest inventory." PLoS ONE 8(3): e59949.
Domke, G.M., Perry, C.H., Walters, B.F., Woodall, C.W., and Smith, J.E. (2016) "A framework for
estimating litter carbon stocks in forests of the United States." Science of the Total Environment
557-558: 469-478.
Fry, J., Xian, G., Jin, S., Dewitz, J., Homer, C., Yang, L., Barnes, C., Herold, N., and Wickham, J. (2011)
"Completion of the 2006 National Land Cover Database for the Conterminous United States."
PE&RS, Vol. 77(9):858-864.
Harmon, M.E., C.W. Woodall, B. Fasth, J. Sexton, M. Yatkov. (2011) Differences between standing
and downed dead tree wood density reduction factors: A comparison across decay classes and
10-112 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
tree species. Res. Paper. NRS-15. Newtown Square, PA: U.S. Department of Agriculture, Forest
Service, Northern Research Station. 40 p.
Homer, C., Dewitz, J., Fry, J., Coan, M., Hossain, N., Larson, C., Herald, N., McKerrow, A., VanDriel,
J.N., and Wickham, J. (2007) "Completion of the 2001 National Land Cover Database for the
Conterminous United States." Photogrammetric Engineering and Remote Sensing, Vol. 73, No.
4, pp 337-341.
Homer, C.G., Dewitz, J.A., Yang, L., Jin, S., Danielson, P., Xian, G., Coulston, J., Herald, N.D.,
Wickham, J.D., and Megown, K. (2015) "Completion of the 2011 National Land Cover Database
for the conterminous United States-Representing a decade of land cover change information."
Photogrammetric Engineering and Remote Sensing 81: 345-354.
Houghton, R.A., et al. (1983) "Changes in the carbon content of terrestrial biota and soils between
1860 and 1980: a net release of C02 to the atmosphere." Ecological Monographs 53: 235-262.
Houghton, R. A. and Nassikas, A. A. (2017) "Global and regional fluxes of carbon from land use and
land cover change 1850-2015." Global Biogeochemical Cycles 31 (3): 456-472.
IPCC (2006)2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change.
[H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa,
Japan.
Jenkins, J.C., D.C. Chojnacky, L.S. Heath, and R.A. Birdsey (2003) "National-scale biomass
estimators for United States tree species." Forest Science 49(1): 12-35.
Metherell, A.K., L.A. Harding, C.V. Cole, and W.J. Parton (1993) CENTURY Soil Organic Matter Model
Environment. Agroecosystem version 4.0. Technical documentation, GPSRTech. Report No. 4,
USDA/ARS, Ft. Collins, CO.Ogle, S.M., Breidt, F.J., DelGrosso, S., Gurung, R., Marx, E., Spencer,
S., Williams, S., Manning, D. (2023) "Counterfactual scenarios reveal historical impact of
cropland management on soil organic carbon stocks in the United States." Scientific Reports
13(1):14564.
Ogle, S.M., F.J. Breidt, M. Easter, S. Williams, K. Killian, and K. Paustian (2010) "Scale and
uncertainty in modeled soil organic carbon stock changes for U.S. croplands using a process-
based model." Global Change Biology 16:810-820.
Ogle, S.M., M.D. Eve, F.J. Breidt, and K. Paustian (2003) "Uncertainty in estimating land use and
management impacts on soil organic carbon storage for U.S. agroecosystems between 1982
and 1997." Global Change Biology 9:1521-1542.
Ogle, S.M., F.J. Breidt, and K. Paustian (2006) "Bias and variance in model results due to spatial
scaling of measurements for parameterization in regional assessments." Global Change Biology
12:516-523.
Parton, W.J., M.D. Hartman, D.S. Ojima, and D.S. Schimel (1998) "DAYCENT: Its Land Surface
Submodel: Description and Testing". Glob. Planet. Chang. 19: 35-48.
Parton, W.J., D.S. Ojima, C.V. Cole, and D.S. Schimel (1994) "A General Model for Soil Organic
Matter Dynamics: Sensitivity to litter chemistry, texture and management," in Quantitative
References and Abbreviations 10-113
-------
Modeling of Soil Forming Processes. Special Publication 39, Soil Science Society of America,
Madison, Wl, 147-167.
Parton, W.J., D.S. Schimel, C.V. Cole, D.S. Ojima (1987) "Analysis of factors controlling soil organic
matter levels in Great Plains grasslands." Soil Science Society of America Journal 51:1173-1179.
Parton, W.J., J.W.B. Stewart, C.V. Cole. (1988) "Dynamics of C, N, P, and S in grassland soils: a
model." Biogeochemistry 5:109-131.
PRISM Climate Group (2022) PRISM Climate Data, Oregon State University,
http://prism.oregonstate.edu. downloaded January 2022.
Smith, J.E.; Heath, L.S.; Skog, K.E.; Birdsey, R.A. (2006) Methods for calculating forest ecosystem
and harvested carbon with standard estimates for forest types of the United States. Gen. Tech.
Rep. NE-343. Newtown Square, PA: U.S. Department of Agriculture, Forest Service,
Northeastern Research Station. 216 p.
Tubiello, F. N., et al. (2015) "The Contribution of Agriculture, Forestry and other Land Use activities
to Global Warming, 1990-2012." Global Change Biology 21:2655-2660.
USDA Forest Service. (2024) Forest Inventory and Analysis: FIA DataMart. U.S. Department of
Agriculture Forest Service. Washington, D.C. Available online at:
https://research.fs.usda.gov/products/dataandtools/tools/fia-datamart. Accessed on 19
September 2024.
USDA-NASS (2021) Published crop data layer. Available at
https://nassgeodata.gmu.edu/CropScape/. Accessed July 2021, USDA-NASS, Washington, DC.
USDA-NRCS (2020) Summary Report: 2017 National Resources Inventory. Natural Resources
Conservation Service, Washington, DC, and Center for Survey Statistics and Methodology, Iowa
State University, Ames, Iowa.
https://www.nrcs.usda.gov/wps/portal/nrcs/main/national/technical/nra/nri/results/.
USDA-NRCS (2018) Summary Report: 2015 National Resources Inventory. Natural Resources
Conservation Service, Washington, D.C., and Center for Survey Statistics and Methodology,
Iowa State University, Ames, Iowa.
https://www.nrcs.usda.gov/lnternet/FSE DOCUMENTS/nrcseprd1422028.pdf.
Westfall, J.A., Coulston, J.W., Gray, A.N., Shaw, J.D., Radtke, P.J., Walker, D.M., Weiskittel, A.R.,
MacFarlane, D.W., Affleck, D.L.R., Zhao, D., Temesgen, H., Poudel, K.P., Frank, J.M., Prisley, S.P.,
Wang, Y., Sanchez Meador, A.J., Auty, D., and Domke, G.M. (2024) A national-scale tree volume,
biomass, and carbon modeling system for the United States. Gen. Tech. Rep. WO-104.
Washington, DC: U.S. Department of Agriculture, Forest Service. 37 p.
https//doi.org/10.2737/WQ-GTR-104.
Woodall, C.W., and V.J. Monleon (2008) Sampling protocol, estimation, and analysis procedures for
the down woody materials indicator of the FIA program. Gen. Tech. Rep. NRS-22. Newtown
Square, PA: U.S. Department of Agriculture, Forest Service, Northern Research Station. 68 p.
Woodall, C.W., L.S. Heath, G.M. Domke, and M.C. Nichols (2011) Methods and equations for
estimating aboveground volume, biomass, and carbon for trees in the U.S. forest inventory,
10-114 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
2010. Gen. Tech. Rep. NRS-88. Newtown Square, PA: U.S. Department of Agriculture, Forest
Service, Northern Research Station. 30 p.
Yang, L., Jin, S., Danielson, R, Homer, C., Gass, L., Bender, S. M., Case, A., Costello, C., Dewitz, J.,
Fry, J., Funk, M., Granneman, B., Liknes, G. C., Rigge, M. &Xian, G. (2018) "A new generation of
the United States National Land Cover Database: Requirements, research priorities, design,
and implementation strategies." ISPRS Journal of Photogrammetry and Remote Sensing 146:
108-123.
Grassland Remaining Grassland: Soil Carbon Stock
Changes
Brockwell, Peter J., and Richard A. Davis (2016) Introduction to time series and forecasting.
Springer.
Del Grosso, S.J., S.M. Ogle, W.J. Parton (2011) Soil organic matter cycling and greenhouse gas
accounting methodologies, Chapter 1, pp 3-13 DOI: 10.1021/bk-2011-1072.ch001 .In:
Understanding Greenhouse Gas Emissions from Agricultural Management (L. Guo, A.
Gunasekara, L. McConnell. Eds.), American Chemical Society, Washington, D.C.
Del Grosso, S.J., W.J. Parton, A.R. Mosier, M.D. Hartman, J. Brenner, D.S. Ojima, and D.S. Schimel
(2001) "Simulated Interaction of Carbon Dynamics and Nitrogen Trace Gas Fluxes Using the
DAYCENT Model." In Modeling Carbon and Nitrogen Dynamics for Soil Management, Schaffer,
M., L. Ma, S. Hansen, (eds.). CRC Press, Boca Raton, Florida, pp. 303-332.
Edmonds, L., R. L. Kellogg, B. Kintzer, L. Knight, C. Lander, J. Lemunyon, D. Meyer, D.C. Moffitt, and
J. Schaefer (2003) "Costs associated with development and implementation of Comprehensive
Nutrient Management Plans." Part I—Nutrient management, land treatment, manure and
wastewater handling and storage, and recordkeeping. Natural Resources Conservation Service,
U.S. Department of Agriculture. Available online at:
http://www.nrcs.usda.gov/technical/land/pubs/cnmp1 .html.
EPA (1999) Biosolids Generation, Use and Disposal in the United States. Office of Solid Waste, U.S.
Environmental Protection Agency. Available online at:
http://biosolids.policy.net/relatives/18941 .PDF.
Fry, J., Xian, G., Jin, S., Dewitz, J., Homer, C., Yang, L., Barnes, C., Herold, N., and Wickham, J. (2011)
Completion of the 2006 National Land Cover Database for the Conterminous United States,
PE&RS, Vol. 77(9):858-864.
Homer, C., Dewitz, J., Fry, J., Coan, M., Hossain, N., Larson, C., Herold, N., McKerrow, A., VanDriel,
J.N., and Wickham, J. (2007) Completion of the 2001 National Land Cover Database for the
Conterminous United States. Photogrammetric Engineering and Remote Sensing, Vol. 73, No.
4, pp 337-341.
Homer, C.G., Dewitz, J.A., Yang, L., Jin, S., Danielson, P., Xian, G., Coulston, J., Herold, N.D.,
Wickham, J.D., and Megown, K. (2015) Completion of the 2011 National Land Cover Database
References and Abbreviations 10-115
-------
for the conterminous United States-Representing a decade of land cover change information.
Photogrammetric Engineering and Remote Sensing, v. 81, no. 5, p. 345-354.
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change.
[H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa,
Japan.
Kellogg, R.L., C.H. Lander, D.C. Moffitt, and N. Gollehon (2000) Manure Nutrients Relative to the
Capacity of Cropland and Pastureland to Assimilate Nutrients: Spatial and Temporal Trends for
the United States. U.S. Department of Agriculture, Washington, D.C. Publication number npsOO-
0579.
Metherell, A.K., L.A. Harding, C.V. Cole, and W.J. Parton (1993) "CENTURY Soil Organic Matter Model
Environment." Agroecosystem version 4.0. Technical documentation, GPSR Tech. Report No. 4,
USDA/ARS, Ft. Collins, CO.NEBRA (2007) A National Biosolids Regulation, Quality, End Use &
Disposal Survey. North East Biosolids and Residuals Association. July 21, 2007.
Nusser, S.M. and J.J. Goebel (1997) The national resources inventory: a long-term multi-resource
monitoring programme. Environmental and Ecological Statistics 4:181-204.
Ogle, S.M., F.J. Breidt, M. Easter, S. Williams, K. Killian, and K. Paustian (2010) "Scale and
uncertainty in modeled soil organic carbon stock changes for U.S. croplands using a process-
based model." Global Change Biology 16:810-820.
Ogle, S.M., M.D. Eve, F.J. Breidt, and K. Paustian (2003) "Uncertainty in estimating land use and
management impacts on soil organic carbon storage for U.S. agroecosystems between 1982
and 1997." Global Change Biology 9:1521-1542.
Parton, W.J., D.S. Ojima, C.V. Cole, and D.S. Schimel (1994) "A General Model for Soil Organic
Matter Dynamics: Sensitivity to litter chemistry, texture and management," in Quantitative
Modeling of Soil Forming Processes. Special Publication 39, Soil Science Society of America,
Madison, Wl, 147-167.
Parton, W.J., D.S. Schimel, C.V. Cole, D.S. Ojima (1987) "Analysis of factors controlling soil organic
matter levels in Great Plains grasslands." Soil Science Society of America Journal 51:1173-1179.
Parton, W.J., J.W.B. Stewart, C.V. Cole. (1988) "Dynamics of C, N, P, and S in grassland soils: a
model." Biogeochemistry 5:109-131.
Parton, W.J., M.D. Hartman, D.S. Ojima, and D.S. Schimel (1998) "DAYCENT: Its Land Surface
Submodel: Description and Testing". Glob. Planet. Chang. 19: 35-48.PRISM Climate Group,
Oregon State University, http://prism.oregonstate.edu. created 24 July 2015.
PRISM Climate Group (2022) PRISM Climate Data, Oregon State University,
http://prism.oregonstate.edu. downloaded January 2022.
United States Bureau of Land Management (BLM) (2014) Rangeland Inventory, Monitoring, and
Evaluation Reports. Bureau of Land Management. U.S. Department of the Interior. Available
online at:
http://www.blm.gov/wo/st/en/prog/more/rangeland management/rangeland inventory.html.
10-116 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
USDA-NASS (2021) Published crop data layer. Available at
https://nassgeodata.gmu.edu/CropScape/. Accessed July 2021, USDA-NASS, Washington, DC.
USDA-NRCS (2020) Summary Report: 2017 National Resources Inventory. Natural Resources
Conservation Service, Washington, DC, and Center for Survey Statistics and Methodology, Iowa
State University, Ames, Iowa.
https://www.nrcs.usda.gov/wps/portal/nrcs/main/national/technical/nra/nri/results/.
USDA Forest Service. (2024) Forest Inventory and Analysis: FIA DataMart. U.S. Department of
Agriculture Forest Service. Washington, D.C. Available online at:
https://research.fs.usda.gov/products/dataandtools/tools/fia-datamart. Accessed on 19
September 2024.
Yang, L., Jin, S., Danielson, R, Homer, C., Gass, L., Bender, S. M., Case, A., Costello, C., Dewitz, J.,
Fry, J., Funk, M., Granneman, B., Liknes, G. C., Rigge, M. &Xian, G. (2018) "A new generation of
the United States National Land Cover Database: Requirements, research priorities, design,
and implementation strategies." ISPRS Journal of Photogrammetry and Remote Sensing 146:
108-123.
Grassland Remaining Grassland: Non-C02 Emissions
from Grassland Fires
Anderson, R.C. Evolution and origin of the Central Grassland of North America: climate, fire and
mammalian grazers. Journal of the Torrey Botanical Society 133: 626-647.
Andreae, M.O. and R Merlet (2001) Emission of trace gases and aerosols from biomass burning.
Global Biogeochemical Cycles 15:955-966.
Brockwell, Peter J., and Richard A. Davis (2016) Introduction to time series and forecasting.
Springer.
Chapin, F.S., S.F. Trainor, O. Huntington, A.L. Lovecraft, E. Zavaleta, D.C. Natcher, A.D. McGuire, J.L.
Nelson, L. Ray, M. Calef, N. Fresco, H. Huntington, T.S. Rupp, L. DeWilde, and R.L. Naylor (2008)
Increasing wildfires in Alaska's Boreal Forest: Pathways to potential solutions of a wicked
problem. Bioscience 58:531-540.
Daubenmire, R. (1968) Ecology of fire in grasslands. Advances in Ecological Research 5:209-266.
Fry, J., Xian, G., Jin, S., Dewitz, J., Homer, C., Yang, L., Barnes, C., Herold, N., and Wickham, J. (2011)
Completion of the 2006 National Land Cover Database for the Conterminous United States,
PE&RS, Vol. 77(9):858-864.
Homer, C., Dewitz, J., Fry, J., Coan, M., Hossain, N., Larson, C., Herold, N., McKerrow, A., VanDriel,
J.N., and Wickham, J. (2007) Completion of the 2001 National Land Cover Database for the
Conterminous United States. Photogrammetric Engineering and Remote Sensing, Vol. 73, No.
4, pp 337-341.
Homer, C.G., Dewitz, J.A., Yang, L., Jin, S., Danielson, P., Xian, G., Coulston, J., Herold, N.D.,
Wickham, J.D., and Megown, K. (2015) Completion of the 2011 National Land Cover Database
References and Abbreviations 10-117
-------
for the conterminous United States-Representing a decade of land cover change information.
Photogrammetric Engineering and Remote Sensing, v. 81, no. 5, p. 345-354.
IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to
the Fifth Assessment Report of the Intergovernmental Panel on Climate Change [Stocker, T.F., D.
Qin, G.-K. Plattner, M. Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley
(eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, 1535
pp.
IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change.
[H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa,
Japan.
MTBS (2023) Burned Areas Boundaries Dataset. (2023, August - last revised). MTBS Project (USDA
Forest Service/U.S. Geological Survey). Available online: https://mtbs.gov/direct-download
Accessed: August 9, 2023.
Ogle, S.M., S. Spencer, M. Hartman, L. Buendia, L. Stevens, D. du Toit, J. Witi (2016) "Developing
national baseline GHG emissions and analyzing mitigation potentials for agriculture and
forestry using an advanced national GHG inventory software system." In Advances in
Agricultural Systems Modeling 6, Synthesis and Modeling of Greenhouse Gas Emissions and
Carbon Storage in Agricultural and Forestry Systems to Guide Mitigation and Adaptation, S. Del
Grosso, L.R. Ahuja and W.J. Parton (eds.), American Society of Agriculture, Crop Society of
America and Soil Science Society of America, pp. 129-148.
Nusser, S.M. and J.J. Goebel (1997) The national resources inventory: a long-term multi-resource
monitoring programme. Environmental and Ecological Statistics 4:181-204.
Picotte, J.J., K. Bhattarai, D. Howard, J. Lecker, J. Epting, B. Quayle, N. Benson, and K. Nelson (2020)
"Changes to the Monitoring Trends in Burn Severity program mapping production procedures
and data products." Fire Ecology. 16:16. https://doi.org/10.1186/s42408-020-00076-y.
USDA-NRCS (2015) Summary Report: 2012 National Resources Inventory, Natural Resources
Conservation Service, Washington, D.C., and Center for Survey Statistics and Methodology,
Iowa State University, Ames, Iowa. Available online at:
http://www.nrcs.usda.gov/lnternet/FSE DOCUMENTS/nrcseprd396218.pdf.
Land Converted to Grassland
Asner, G.P., Archer, S., Hughes, R.F., Ansley, R.J. and Wessman, C.A. (2003) "Net changes in
regional woody vegetation cover and carbon storage in Texas drylands, 1937-1999." Global
Change Biology 9(3): 316-335.
Birdsey, R. (1996) "Carbon Storage for Major Forest Types and Regions in the Conterminous United
States." In R.N. Sampson and D. Hair, (eds.). Forest and Global Change, Volume 2: Forest
Management Opportunities for Mitigating Carbon Emissions. American Forests. Washington,
D.C., 1-26 and 261-379 (appendices 262 and 263).
10-118 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Breshears, D.D., Knapp, A.K., Law, D.J., Smith, M.D., Twidwell, D. and Wonkka, C.L., 2016.
Rangeland Responses to Predicted Increases in Drought Extremity. Rangelands, 38(4), pp.191 -
196.
Brockwell, Peter J., and Richard A. Davis (2016) Introduction to time series and forecasting.
Springer.
Del Grosso, S.J., S.M. Ogle, W.J. Parton. (2011) Soil organic matter cycling and greenhouse gas
accounting methodologies, Chapter 1, pp 3-13 DOI: 10.1021/bk-2011-1072.ch001 .In:
Understanding Greenhouse Gas Emissions from Agricultural Management (L. Guo, A.
Gunasekara, L. McConnell. Eds.), American Chemical Society, Washington, D.C.
Del Grosso, S.J., W.J. Parton, A.R. Mosier, M.D. Hartman, J. Brenner, D.S. Ojima, and D.S. Schimel
(2001) "Simulated Interaction of Carbon Dynamics and Nitrogen Trace Gas Fluxes Using the
DAYCENT Model." In Modeling Carbon and Nitrogen Dynamics for Soil Management (Schaffer,
M., L. Ma, S. Hansen, (eds.). CRC Press, Boca Raton, Florida, pp. 303-332.
Domke, G.M., J.E. Smith, and C.W. Woodall. (2011) Accounting for density reduction and structural
loss in standing dead trees: Implications for forest biomass and carbon stock estimates in the
United States. Carbon Balance and Management. 6:14.
Domke, G.M., et al. (2013) From models to measurements: comparing down dead wood carbon
stock estimates in the U.S. forest inventory. PLoS ONE 8(3): e59949.
Domke, G.M., Perry, C.H., Walters, B.F., Woodall, C.W., and Smith, J.E. (2016) A framework for
estimating litter carbon stocks in forests of the United States. Science of the Total Environment
557-558: 469-478.
Domke, G.M., Walters, B.F., Smith, J.E., Woodall, C.W. (2022) Chapter 6: FIA Carbon Attributes. In
Westfall, J.A.; Coulston, J.W.; Moisen, G.G.; Andersen, H.-E., eds. 2022. Sampling and
estimation documentation for the Enhanced Forest Inventory and Analysis Program: 2022. Gen.
Tech. Rep. NRS-GTR-207, Madison, Wl: U.S. Department of Agriculture, Forest Service,
Northern Research Station. 129 p. https://doi.org/10.2737/NRS-GTR-207.
Epstein, H.E., Gill, R.A., Paruelo, J.M., Lauenroth, W.K., Jia, G.J. and Burke, I.C. (2002) The relative
abundance of three plant functional types in temperate grasslands and shrublands of North
and South America: effects of projected climate change. Journal of Biogeography, 29(7),
pp.875-888.
Fry, J., Xian, G., Jin, S., Dewitz, J., Homer, C., Yang, L., Barnes, C., Herold, N., and Wickham, J. (2011)
Completion of the 2006 National Land Cover Database for the Conterminous United States,
PE&RS, Vol. 77(9):858-864.
Harmon, M.E., C.W. Woodall, B. Fasth, J. Sexton, M. Yatkov. (2011) Differences between standing
and downed dead tree wood density reduction factors: A comparison across decay classes and
tree species. Res. Paper. NRS-15. Newtown Square, PA: U.S. Department of Agriculture, Forest
Service, Northern Research Station. 40 p.
Homer, C., Dewitz, J., Fry, J., Coan, M., Hossain, N., Larson, C., Herold, N., McKerrow, A., VanDriel,
J.N., and Wickham, J. (2007) Completion of the 2001 National Land Cover Database for the
References and Abbreviations 10-119
-------
Conterminous United States. Photogrammetric Engineering and Remote Sensing, Vol. 73, No.
4, pp 337-341.
Homer, C.G., Dewitz, J.A., Yang, L., Jin, S., Danielson, P., Xian, G., Coulston, J., Herold, N.D.,
Wickham, J.D., and Megown, K. (2015) Completion of the 2011 National Land Cover Database
for the conterminous United States-Representing a decade of land cover change information.
Photogrammetric Engineering and Remote Sensing, v. 81, no. 5, p. 345-354.
Houghton, R.A., et al. (1983) "Changes in the carbon content of terrestrial biota and soils between
1860 and 1980: a net release of C02 to the atmosphere." Ecological Monographs 53: 235-262.
Houghton, R. A. and Nassikas, A. A. (2017) "Global and regional fluxes of carbon from land use and
land cover change 1850-2015." Global Biogeochemical Cycles 31 (3): 456-472.
Huang, C.Y., Asner, G.R, Martin, R.E., Barger, N.N. and Neff, J.C. (2009) "Multiscale analysis of tree
cover and aboveground carbon stocks in pinyon-juniper woodlands." Ecological Applications
19(3): 668-681.
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change,
[H.S. Eggleston, L. Buendia, K. Miwa, T Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
Jenkins, J.C., D.C. Chojnacky, L.S. Heath, and R.A. Birdsey (2003) "National-scale biomass
estimators for United States tree species." Forest Science 49(1):12-35.
Jurena, P.N. and Archer, S., (2003) Woody plant establishment and spatial heterogeneity in
grasslands. Ecology, 84(4), pp.907-919.
Lenihan, J.M., Drapek, R., Bachelet, D. and Neilson, R.P., (2003) Climate change effects on
vegetation distribution, carbon, and fire in California. Ecological Applications, 13(6), pp.1667-
1681.
Metherell, A.K., L.A. Harding, C.V. Cole, and W.J. Parton (1993) "CENTURY Soil Organic Matter Model
Environment." Agroecosystem version 4.0. Technical documentation, GPSR Tech. Report No. 4,
USDA/ARS, Ft. Collins, CO.
Ogle, S.M., F.J. Breidt, M. Easter, S. Williams, K. Killian, and K. Paustian (2010) "Scale and
uncertainty in modeled soil organic carbon stock changes for U.S. croplands using a process-
based model." Global Change Biology 16:810-820.
Ogle, S.M., M.D. Eve, F.J. Breidt, and K. Paustian (2003) "Uncertainty in estimating land use and
management impacts on soil organic carbon storage for U.S. agroecosystems between 1982
and 1997." Global Change Biology 9:1521-1542.
Parton, W.J., D.S. Ojima, C.V. Cole, and D.S. Schimel (1994) "A General Model for Soil Organic
Matter Dynamics: Sensitivity to litter chemistry, texture and management," in Quantitative
Modeling of Soil Forming Processes. Special Publication 39, Soil Science Society of America,
Madison, Wl, 147-167.
Parton, W.J., D.S. Schimel, C.V. Cole, D.S. Ojima (1987) "Analysis of factors controlling soil organic
matter levels in Great Plains grasslands." Soil Science Society of America Journal 51:1173-1179.
10-120 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Parton, W.J., J.W.B. Stewart, C.V. Cole (1988) "Dynamics of C, N, P, and S in grassland soils: a
model." Biogeochemistry 5:109-131.
Parton, W.J., M.D. Hartman, D.S. Ojima, and D.S. Schimel (1998) "DAYCENT: Its Land Surface
Submodel: Description and Testing". Glob. Planet. Chang. 19: 35-48.
PRISM Climate Group (2022) PRISM Climate Data, Oregon State University,
http://prism.oregonstate.edu. downloaded January 2022.
Scholes, R.J. and Archer, S.R. (1997) Tree-grass interactions in savannas 1. Annual review of
Ecology and Systematics, 28(1), pp.517-544.
Sims, P.L., Singh, J.S. and Lauenroth, W.K. (1978) The structure and function of ten western North
American grasslands: I. Abiotic and vegetational characteristics. The Journal of Ecology,
pp.251-285.
Smith, J.E.; Heath, L.S.; Skog, K.E.; Birdsey, R.A. (2006) Methods for calculating forest ecosystem
and harvested carbon with standard estimates for forest types of the United States. Gen. Tech.
Rep. NE-343. Newtown Square, PA: U.S. Department of Agriculture, Forest Service,
Northeastern Research Station. 216 p.
Tarhouni, M., etal. (2016) Measurement of the aboveground biomass of some rangeland species
using a digital non-destructive technique. Botany Letters 163(3):281-287.
Tubiello, F. N., et al. (2015) "The Contribution of Agriculture, Forestry and other Land Use activities
to Global Warming, 1990-2012." Global Change Biology 21:2655-2660.
United States Bureau of Land Management (BLM) (2014) Rangeland Inventory, Monitoring, and
Evaluation Reports. Bureau of Land Management. U.S. Department of the Interior. Available
online at:
http://www.blm.gov/wo/st/en/prog/more/rangeland management/rangeland inventorv.html.
USDA Forest Service. (2024) Forest Inventory and Analysis: FIA DataMart. U.S. Department of
Agriculture Forest Service. Washington, D.C. Available online at:
https://research.fs.usda.gov/products/dataandtools/tools/fia-datamart. Accessed on 19
September 2024.
USDA-NASS (2021) Published crop data layer. Available at
https://nassgeodata.gmu.edu/CropScape/. Accessed July 2021, USDA-NASS, Washington, DC.
USDA-NRCS (2020) Summary Report: 2017 National Resources Inventory. Natural Resources
Conservation Service, Washington, DC, and Center for Survey Statistics and Methodology, Iowa
State University, Ames, Iowa.
https://www.nrcs.usda.gov/wps/portal/nrcs/main/national/technical/nra/nri/results/.
USDA-NRCS (2018) Summary Report: 2015 National Resources Inventory. Natural Resources
Conservation Service, Washington, D.C., and Center for Survey Statistics and Methodology,
Iowa State University, Ames, Iowa.
https://www.nrcs.usda.gov/lnternet/FSE DOCUMENTS/nrcseprd1422028.pdf.
References and Abbreviations 10-121
-------
Westfall, J.A., Coulston, J.W., Gray, A.N., Shaw, J.D., Radtke, P.J., Walker, D.M., Weiskittel, A.R.,
MacFarlane, D.W., Affleck, D.L.R., Zhao, D., Temesgen, H., Poudel, K.P., Frank, J.M., Prisley, S.P.,
Wang, Y., Sanchez Meador, A.J., Auty, D., and Domke, G.M. (2024) A national-scale tree volume,
biomass, and carbon modeling system for the United States. Gen. Tech. Rep. WO-104.
Washington, DC: U.S. Department of Agriculture, Forest Service. 37 p.
https//doi.org/10.7737/WQ-GTR-104.
Woodall, C.W., and V.J. Monleon (2008) Sampling protocol, estimation, and analysis procedures for
the down woody materials indicator of the FIA program. Gen. Tech. Rep. NRS-22. Newtown
Square, PA: U.S. Department of Agriculture, Forest Service, Northern Research Station. 68 p.
Woodall, C.W., L.S. Heath, G.M. Domke, and M.C. Nichols. (2011) Methods and equations for
estimating aboveground volume, biomass, and carbon for trees in the U.S. forest inventory,
2010. Gen. Tech. Rep. NRS-88. Newtown Square, PA: U.S. Department of Agriculture, Forest
Service, Northern Research Station. 30 p.
Yang, L., Jin, S., Danielson, P., Homer, C., Gass, L., Bender, S. M., Case, A., Costello, C., Dewitz, J.,
Fry, J., Funk, M., Granneman, B., Liknes, G. C., Rigge, M. &Xian, G. (2018) "A new generation of
the United States National Land Cover Database: Requirements, research priorities, design,
and implementation strategies." ISPRS Journal of Photogrammetry and Remote Sensing 146:
108-123.
Wetlands Remaining Wetlands: CO2, CH4, and N20
Emissions from Peatlands Remaining Peatlands
Apodaca, L. (2011) Email correspondence. Lori Apodaca, Peat Commodity Specialist, USGS and
Emily Rowan, ICF International. November.
Apodaca, L. (2008) E-mail correspondence. Lori Apodaca, Peat Commodity Specialist, USGS and
Emily Rowan, ICF International. October and November.
Cleary, J., N. Roulet and T.R. Moore (2005) "Greenhouse gas emissions from Canadian peat
extraction, 1990-2000: A life-cycle analysis." Ambio 34:456-461.
Division of Geological & Geophysical Surveys (DGGS), Alaska Department of Natural Resources
(1997-2015) Alaska's Mineral Industry Report (1997-2014). Alaska Department of Natural
Resources, Fairbanks, AK. Available online at
http://www.dggs.dnr.state.ak.us/pubs/pubs?reqtvpe=minerals.
IPCC (2014) Climate Change 2014: Synthesis Report. Contribution of Working Groups I, II and III to
the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. R.K. Pachauri
and L.A. Meyer (eds.). IPCC, Geneva, Switzerland.
IPCC (2013) 2073 Supplement to the 2006 IPCC Guidelines for National Greenhouse Gas
Inventories: Wetlands. Hiraishi, T., Krug, T., Tanabe, K., Srivastava, N., Baasansuren, J., Fukuda,
M. and Troxler, T.G. (eds.). Published: IPCC, Switzerland.
10-122 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
IPCC (2007) Climate Change 2007: Synthesis Report. Contribution of Working Groups I, II and III to
the Fourth Assessment Report (AR4) of the IPCC. The Intergovernmental Panel on Climate
Change, R.K. Pachauri, A. Resinger (eds.). Geneva, Switzerland.
IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change. H.S.
Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Hayama, Kanagawa, Japan.
Szumigala, D.J. (2011) Phone conversation. Dr. David Szumigala, Division of Geological and
Geophysical Surveys, Alaska Department of Natural Resources and Emily Rowan, ICF
International. January 18, 2011.
Szumigala, D.J. (2008) Phone conversation. Dr. David Szumigala, Division of Geological and
Geophysical Surveys, Alaska Department of Natural Resources and Emily Rowan, ICF
International. October 17, 2008.
USGS (1991-2018) Minerals Yearbook: Peat (1994-2018). United States Geological Survey, Reston,
VA. Available online at http://minerals.usgs.gov/minerals/pubs/commodity/peat/index.html.
USGS (2023) Email correspondence. Amanda Brioche, Mineral Commodity Specialist, USGS.
August 2023.
USGS (2024a) Minerals Yearbook: Peat (2019) Tables-only release. United States Geological Survey,
Reston, VA. Available online at https://www.usgs.gov/centers/nmic/peat-statistics-and-
information.
USGS (2024b) Minerals Yearbook: Peat (2020) Tables-only release. United States Geological Survey,
Reston, VA. Available online at https://www.usgs.gov/centers/nmic/peat-statistics-and-
information.
USGS (2024c) Minerals Yearbook: Peat (2021) Tables-only release. United States Geological Survey,
Reston, VA. Available online at https://www.usgs.gov/centers/nmic/peat-statistics-and-
information.
USGS (2024d) Mineral Commodity Summaries: Peat (1996-2024). United States Geological Survey,
Reston, VA. Available online at https://www.usgs.gov/centers/nmic/peat-statistics-and-
information.
Wetlands Remaining Coastal Wetlands: Emissions and
Removals from Coastal Wetlands Remaining Coastal
Wetlands
Abbott, K. M., Elsey-Quirk, T., and DeLaune, R. D. (2019) Factors influencing blue carbon
accumulation across a 32-year chronosequence of created coastal marshes. Ecosphere, 10(8):
e02828. https://doi.Org/10.1002/ecs2.2828
Allen, J. R., Cornwell, J. C., and Baldwin, A. H. (2021) Contributions of organic and mineral matter to
vertical accretion in tidal wetlands across a Chesapeake Bay subestuary. Journal of Marine
Science and Engineering 9(7): 751. https://doi.Org/10.3390/jmse9070751
References and Abbreviations 10-123
-------
Arias-Ortiz, A., Oikawa, P. Y., Carlin, J., Masque, P., Shahan, J., Kanneg, S.,... and Baldocchi, D. D.
(2021) Tidal and nontidal marsh restoration: a trade-off between carbon sequestration,
methane emissions, and soil accretion. Journal of Geophysical Research:
Biogeosciences, 126(12): e2021 JG006573. https://doi.org/10.1079/7071JG0Q6573.
Arriola, J. M., and Cable, J. E. (2017) Variations in carbon burial and sediment accretion along a tidal
creek in a Florida salt marsh. Limnology and Oceanography 62(S1): S15-S28.
https://doi.Org/10.1002/lno.10652
Baustian, M. M., Stagg, C. L., Perry, C. L., Moss, L. C., and Carruthers, T. J. (2021) Long-term carbon
sinks in marsh soils of coastal Louisiana are at risk to wetland loss. Journal of Geophysical
Research: Biogeosciences 126(3): e2020JG005832. https://doi.Org/10.1029/2020JG005832.
Bianchi, T. S., Allison, M. A., Zhao, J., Li, X., Comeaux, R. S., Feagin, R. A., & Kulawardhana, R. W.
(2013) Historical reconstruction of mangrove expansion in the Gulf of Mexico: linking climate
change with carbon sequestration in coastal wetlands. Estuarine, Coastal and Shelf Science
119:7-16. https://doi.Org/10.1016/j.ecss.7017.17.007.
Blum. M., Rahn, D., Frederick, B., and Polanco, S. (2023). Land loss in the Mississippi River Delta:
Role of subsidence, global sea-level rise, and coupled atmospheric and oceanographic
processes. Global and Planetary Change, 222,104048.
https://doi.Org/10.1016/j.gloplacha.2023.104048.
Boyd, B. (2012) Comparison of sediment accumulation and accretion in impounded and
unimpounded marshes of the Delaware Estuary. Doctoral dissertation, University of Delaware.
https://udspace.udel.edu/items/4580aa31 -636b-41 dd-8346-31188356cc6f.
Boyd, B. M. and Sommerfield, C. K. (2016) Marsh accretion and sediment accumulation in a
managed tidal wetland complex of Delaware Bay. Ecological Engineering, 92: 37-46.
10.1016/j.ecoleng.2016.03.045
Boyd, B. M., Sommerfield, C. K., and Elsey-Quirk, T. (2017) Hydrogeomorphic influences on salt
marsh sediment accumulation and accretion in two estuaries of the US Mid-Atlantic
coast. Marine Geology, 383:132-145. https://doi.Org/10.1016/j.margeo.2016.11.008.
Breithaupt, J. L., Smoak, J. M., Smith III, T. J., and Sanders, C. J. (2014) Temporal variability of carbon
and nutrient burial, sediment accretion, and mass accumulation over the past century in a
carbonate platform mangrove forest of the Florida Everglades. Journal of Geophysical
Research: Biogeosciences, 119(10): 2032-2048. https://doi.org/10.1002/2014JG0Q2715.
Byrd, K. B., Ballanti, L. R., Thomas, N. M., Nguyen, D. K., Holmquist, J. R., Simard, M., Windham-
Myers, L., Schile, L. M., Parker, V. T.,... and Castaneda-Moya, E. (2017) Biomass/Remote
Sensing dataset: 30m resolution tidal marsh biomass samples and remote sensing data for six
regions in the conterminous United States: U.S. Geological Survey data release,
https://doi.org/10.5066/F77943K8.
Byrd, K. B., Ballanti, L., Thomas, N., Nguyen, D., Holmquist, J.R., Simard, M., and Windham-Myers,
L. (2018) A remote sensing-based model of tidal marsh aboveground carbon stocks for the
10-124 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
conterminous United States. ISPRS Journal of Photogrammetry and Remote Sensing 139: 255-
271. https://doi.org/10.1016/j.isprsjprs.2Q18.03.019.
Byrd, K. B., Ballanti, L., Thomas, N., Nguyen, D., Holmquist, J.R., Simard, M., and Windham-Myers,
L. (2020) Corrigendum to "A remote sensing-based model of tidal marsh aboveground carbon
stocks for the conterminous United States". ISPRS Journal of Photogrammetry and Remote
Sensing 166: 63-67. https://doi.org/10.1016/j.isprsjprs.2020.05.005.
Callaway, J. C., Borgnis, E. L., Turner, R. E. & Milan, C. S. (2012a) Carbon sequestration and
sediment accretion in San Francisco Bay tidal wetlands. Estuaries and Coasts 35(5): 1163-
1181. https://www.doi.org/10.1007/s17737-017-9508-910.1007
Callaway, J. C., Borgnis, E. L., Turner, R. E., Milan, C. S., Goodfriend, W., & Richmond, S. (2012b)
"Wetland Sediment Accumulation at Corte Madera Marsh and Muzzi Marsh". San Francisco Bay
Conservation and Development Commission, https://www.adaptingtorisingtides.org/wp-
content/uploads/2015/04/Ma rsh-Sediment-Accumulation-Rates-Callaway-et-al-2012.pdf
Church, T. M., Sommerfield, C. K., Velinsky, D. J., Point, D., Benoit, C., Amouroux, D. & Donard, O. F.
X. (2006) Marsh sediments as records of sedimentation, eutrophication and metal pollution in
the urban Delaware Estuary. Marine Chemistry 102(1-2): 72-95.
https://doi.org/10.1016/j.marchem.2005.10.026.
Couvillion, B. R., Barras, J. A., Steyer, G. D., Sleavin, W., Fischer, M., Beck, H., & Heckman, D. (2011)
Land area change in coastal Louisiana (1932 to 2010) (pp. 1-12). U.S. Department of the Interior,
U.S. Geological Survey, https://pubs.usgs.gov/sim/3164/.
Couvillion, B. R., Fischer, M. R., Beck, H. J. and Sleavin, W. J. (2016) Spatial Configuration Trends in
Coastal Louisiana from 1986 to 2010. Wetlands 1-13. https://doi.org/10.1007/s13157-016-
0744-9.
Craft, C. B., & Richardson, C. J. (1998) Recent and long-term organic soil accretion and nutrient
accumulation in the Everglades. Soil Science Society of America Journal 62(3): 834-843.
https://doi.org/10.7136/sssaj1998.03615995006700030Q47x.
Creamer, C.A., Waldrop, M. P., Stagg, C. L., Manies, K. L., Baustian, M. M., Laurenzano, C., et al.
(2024). Vegetation loss following vertical drowning of Mississippi River Deltaic wetlands leads to
faster microbial decomposition and decreases in soil carbon. Journal of Geophysical Research:
Biogeosciences, 129, e2023JG007832. https://doi.org/10.1029/2023JG007832.
Crooks, S., Findsen, J., Igusky, K., Orr, M. K. and Brew, D. (2009) Greenhouse Gas Mitigation
Typology Issues Paper: Tidal Wetlands Restoration. Report by PWAand SAIC to the California
Climate Action Reserve, https://www.climateactionreserve.org/wp-
content/uploads/2009/03/future-protocol-development tidal-wetlands.pdf
Crooks, S., Rybczyk, J., O'Connell, K., Devier, D. L., Poppe, K., Emmett-Mattox, S. (2014) Coastal
Blue Carbon Opportunity Assessment for the Snohomish Estuary: The Climate Benefits of
Estuary Restoration. Report by Environmental Science Associates, Western Washington
University, EarthCorps, and Restore America's Estuaries, https://estuaries.org/wp-
References and Abbreviations 10-125
-------
content/uploads/2020/11 /Crooks.-Coastal-Blue-Ca rbon-Opportunity-Assessment-for-the-
Snohomish-Estuary-ilovepdf-compressed.pdf
DeLaune, R. D., & White, J. R. (2012) Will coastal wetlands continue to sequester carbon in
response to an increase in global sea level?: A case study of the rapidly subsiding Mississippi
river deltaic plain. Climatic Change, 110(1), 297-314.
https://link.springer.com/article/10.1007/s10584-Q11-0089-6
Drexler, J. Z., de Fontaine, C. S., and Brown, T. A. (2009) Peat accretion histories during the past
6,000 years in marshes of the Sacramento-San Joaquin Delta, CA, USA. Estuaries and Coasts
32: 871-892. https://www.doi.org/10.1007/s17737-009-97Q7-8
Drexler, J. Z., Krauss, K. W., Sasser, M. C., Fuller, C. C., Swarzenski, C. M., Powell, A.,... and
Orlando, J. (2013) A long-term comparison of carbon sequestration rates in impounded and
naturally tidal freshwater marshes along the lower Waccamaw River, South Carolina. Wetlands
33: 965-974. https://www.doi.org/10.1007/s13157-013-0456-3
Drexler, J. Z., Woo, I., Fuller, C. C., and Nakai, G. (2019) Carbon accumulation and vertical accretion
in a restored versus historic salt marsh in southern Puget Sound, Washington, United States.
Restoration Ecology 27(5): 1117-1127. https://doi.org/10.1111 /rec.12941
Ensign, S. H., Noe, G. B., Hupp, C. R., and Skalak, K. J. (2015) Head-of-tide bottleneck of particulate
material transport from watersheds to estuaries. Geophysical Research Letters 42(24): 10-671.
https://doi.org/10.1007/7015GI 066830
Gerlach, M. J., Engelhart, S. E., Kemp, A. C., Moyer, R. P., Smoak, J. M., Bernhardt, C. E., and Cahill,
N. (2017) Reconstructing Common Era relative sea-level change on the Gulf Coast of Florida.
Marine Geology 390: 254-269. https://doi.Org/10.1016/j.margeo.2017.07.001
Giblin, A., Forbrich, I., & LTER, P. I. E. (2018) PIE LTER high marsh sediment chemistry and activity
measurements, Nelson Island Creek marsh, Rowley, MA.
https://portal.edirepository.org/nis/mapbrowse?scope=knb-lter-pie&identifier=427
Holmquist, J. R., Windham-Myers, L., Bliss, N., Crooks, S., Morris, J. T., Megonigal, J. P. & Woodrey,
M. (2018) Accuracy and Precision of Tidal Wetland Soil Carbon Mapping in the Conterminous
United States. Scientific reports 8(1): 9478. https://www.doi.org/10.1038/s41598-018-26948-7
Hu, Z., Lee, J. W., Chandran, K., Kim, S. and Khanal, S. K. (2012) N20 Emissions from Aquaculture: A
Review. Environmental Science & Technology 46(12): 6470-6480.
https://doi.org/10.1021/es300110x
Hussein, A. H., Rabenhorst, M. C. &Tucker, M. L. (2004) Modeling of carbon sequestration in
coastal marsh soils. Soil Science Society of America Journal 68(5): 1786-1795.
https://doi.org/10.2136/sssaj2004.1786
IPCC (2014) 207 3 Supplement to the 2006IPCC Guidelines for National Greenhouse Gas
Inventories: Wetlands. Hiraishi, T., Krug, T., Tanabe, K., Srivastava, N., Baasansuren, J., Fukuda,
M. and Troxler, T.G. (eds.). Published: IPCC, Switzerland.
IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to
the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. Stocker, T.,
10-126 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Qin, D., Plattner, G.-K., Tignor, M. Allen, S.K., Boschung, J., Nauels, A., Xia, Y., Bex, V. and
Midgley, P.M. (eds.). Cambridge University Press, Cambridge, United Kingdom and New York,
NY, USA.
IPCC (2006) IPCC Guidelines for National Greenhouse Gas Inventories. Prepared by the National
Greenhouse Gas Inventories Programme, Eggleston H.S., Buendia L., Miwa K., Ngara T. and
Tanabe K. (eds). IGES, Japan.
IPCC (2003) Good Practice Guidance for Land Use, Land-Use Change and Forestry. LUCF Sector
Good Practice Guidance, Chapter 3. Penman, J., Gytarsky, M., Hiraishi, T., Krug, T., Kruger, D.,
Pipatti, R., Buendia, L., Miwa, K., Ngara, T., Tanabe, K. and Wagner, F. (eds). Institute of Global
Environmental Strategies (IGES), on behalf of the Intergovernmental Panel on Climate Change
(IPCC): Hayama, Japan.
IPCC (2000) Good Practice Guidance and Uncertainty Management in National Greenhouse Gas
Inventories. Quantifying Uncertainties in Practice, Chapter 6. Penman, J., Kruger, D., Galbally, I.,
Hiraishi, T., Nyenzi, B., Emmanuel, S., Buendia, L., Hoppaus, R., Martinsen, T., Meijer, J., Miwa,
K. and Tanabe, K. (eds). Institute of Global Environmental Strategies (IGES), on behalf of the
Intergovernmental Panel on Climate Change (IPCC): Hayama, Japan.
Kearney, M. S. &Stevenson, J. C. (1991) Island land loss and marsh vertical accretion rate evidence
for historical sea-level changes in Chesapeake Bay. Journal of Coastal Research 7(2): 403-415.
Kemp, A. C., Sommerfield, C. K., Vane, C. H., Horton, B. P., Chenery, S., Anisfeld, S., and Nikitina, D.
(2012) Use of lead isotopes for developing chronologies in recent salt-marsh sediments.
Quaternary Geochronology 12: 40-49. https://doi.Org/10.1016/j.quageo.2012.05.004
Koster, D., Lichter, J., Lea, P. D., & Nurse, A. (2007) Historical eutrophication in a river-estuary
complex in mid-coast Maine. Ecological Applications 17(3): 765-778.
Krauss, K. W., Noe, G. B., Duberstein, J. A., Conner, W. H., Stagg, C. L., and Jones, M. C. (2018)
Carbon budget assessment of tidal freshwater forested wetland and oligohaline marsh
ecosystems along the Waccamaw and Savannah Rivers, USA (2005-2016). US Geological
Survey Data Release. https://doi.org/10.5066/F7TM7930.
Lagomasino, D., Corbett, D. R., and Walsh, J. P. (2013) Influence of wind-driven inundation and
coastal geomorphology on sedimentation in two microtidal marshes, Pamlico River Estuary,
NC. Estuaries and Coasts 36:1165-1180. http://dx.doi.Org/10.1007/s12237-013-9625-0
Lu, M & Megonigal, J. P. (2017) Final Report for RAE Baseline Assessment Project. Memo to
Silvestrum Climate Associates by Smithsonian Environmental Research Center, Maryland.
Lynch, J. C. (1989) Sedimentation and nutrient accumulation in mangrove ecosystems of the Gulf of
Mexico. M.S. thesis, Univ. of Southwestern Louisiana, Lafayette, LA.
Luk, S. Y., Todd-Brown, K., Eagle, M., McNichol, A. P., Sanderman, J., Gosselin, K., and Spivak, A. C.
(2021) Soil organic carbon development and turnover in natural and disturbed salt marsh
environments. Geophysical Research Letters 48(2): e2020GL090287.
https://doi.org/10.1029/2020GL090287
References and Abbreviations 10-127
-------
Marchio, D. A., Savarese, M., Bovard, B., & Mitsch, W. J. (2016) Carbon sequestration and
sedimentation in mangrove swamps influenced by hydrogeomorphic conditions and
urbanization in Southwest Florida. Forests 7:116-135. https://doi.org/10.3390/f7060116
McCombs, J. W., Herold, N. D., Burkhalter, S. G. and Robinson C. J. (2016) Accuracy Assessment of
NOAA Coastal Change Analysis Program 2006-2010 Land Cover and Land Cover Change Data.
Photogrammetric Engineering & Remote Sensing. 82:711-718.
McTigue, N., Davis, J., Rodriguez, A. B., McKee, B., Atencio, A., and Currin, C. (2019) Sea level rise
explains changing carbon accumulation rates in a salt marsh over the past two millennia. JGR
Biogeosciences. https://doi.org/10.1029/2019JG005207
Merrill, J. Z. (1999) Tidal Freshwater Marshes as Nutrient Sinks: particulate Nutrient Burial and
Denitrification. Ph.D. Dissertation, University of Maryland, College Park, MD, 342 pp.
Miller, C. B., Rodriguez, A. B., Bost, M. C., McKee, B. A., and McTigue, N. D. (2022) Carbon
accumulation rates are highest at young and expanding salt marsh edges. Communications
Earth & Environment 3(1): 173. https://doi.org/10.1038/s43247-022-00501 -x
National Marine Fisheries Service (2022). Fisheries of the United States, 2020. U.S. Department of
Commerce, NOAA Current Fishery Statistics No. 2020. Available at:
https://www.fisheries.noaa.gov/national/sustainable-fisheries/fisheries-united-states.
National Oceanic and Atmospheric Administration, Office for Coastal Management (2020) Coastal
Change Analysis Program (C-CAP) Regional Land Cover. Charleston, SC: NOAA Office for
Coastal Management. Accessed October 2024 at
www.coast.noaa.gov/htdata/raster1/landcover/bulkdownload/30m lc/.
Noe, G. B., Hupp, C. R., Bernhardt, C. E., & Krauss, K. W. (2016) Contemporary deposition and long-
term accumulation of sediment and nutrients by tidal freshwater forested wetlands impacted
by sea level rise. Estuaries and Coasts 39(4): 1006-1019. https://doi.org/10.1007/s12237-016-
0066-4
Orson, R. A., Simpson, R. L., & Good, R. E. (1990) Rates of sediment accumulation in a tidal
freshwater marsh. Journal of Sedimentary Research 60(6): 859-869.
Orson, R., Warren, R. & Niering, W. (1998) Interpreting sea level rise and rates of vertical marsh
accretion in a southern New England tidal salt marsh. Estuarine, Coastal and Shelf Science
47(4): 419-429.
Peck, E. K., Wheatcroft, R. A., and Brophy, L. S. (2020) Controls on sediment accretion and blue
carbon burial in tidal saline wetlands: insights from the Oregon Coast, USA. Journal of
Geophysical Research: Biogeosciences 125(2): e2019JG005464.
https://doi.org/10.1029/2019JG005464
Poppe, K. L., and Rybczyk, J. M. (2021) Tidal marsh restoration enhances sediment accretion and
carbon accumulation in the Stillaguamish River estuary, Washington. PloS one 16(9): e0257244.
https://doi.org/10.75573/data.10005748.v1
10-128 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Roman, C., Peck, J., Allen, J., King, J. & Appleby, P. (1997) Accretion of a New England (USA) salt
marsh in response to inlet migration, storms, and sea-level rise. Estuarine, Coastal and Shelf
Science 45(6): 717-727.
Smith, K. E., Flocks, J. G., Steyer, G. D., and Piazza, S. C. (2015) Wetland Paleoecological Study of
Southwest Coastal Louisiana: Sediment Cores and Diatom Calibration Dataset. US Department
of the Interior, US Geological Survey, https://doi.org/10.3133/ds877
Thom, R. M. (1992) Accretion rates of low intertidal salt marshes in the Pacific Northwest. Wetlands
12:147-156. https://doi.org/10.1 QQ7/BFQ3160603
Vaughn, D. R., Bianchi, T. S., Shields, M. R., Kenney, W. F., and Osborne, T. Z. (2020) Increased
organic carbon burial in northern Florida mangrove-salt marsh transition zones. Global
Biogeochemical Cycles 34(5): e2019GB006334.
Villa, J. A. & Mitsch W. J. (2015) Carbon sequestration in different wetland plant communities of
Southwest Florida. International Journal for Biodiversity Science, Ecosystems Services and
Management 11:17-28
Watson, E. B., and Byrne, R. (2013) Late Holocene Marsh Expansion in Southern San Francisco Bay,
California. Estuaries and Coasts 36: 643-653. http://dx.doi.org/10.1007/s12237-013-9598-z
Weis, D. A., Callaway, J. C., and Gersberg, R. M. (2001) Vertical accretion rates and heavy metal
chronologies in wetland sediments of the Tijuana Estuary. Estuaries 24: 840-850.
https://doi.org/10.7307/1353175
Weston, N. B., Neubauer, S. C., Velinsky, D. J., & Vile, M. A. (2014) Net ecosystem carbon exchange
and the greenhouse gas balance of tidal marshes along an estuarine salinity gradient.
Biogeochemistry 120: 163-189. http://dx.doi.org/10.1007/s10533-014-9989-7
Weston, N. B., Rodriguez, E., Donnelly, B., Solohin, E., Jezycki, K., Demberger, S.,... and Craft, C. B.
(2023) Recent acceleration of wetland accretion and carbon accumulation along the US East
Coast. Earth's Future 11 (3): e2022EF003037. https://doi.org/10.10P9/P0PPFF003Q37
Wetlands Remaining Wetlands: Flooded Land Remaining
Flooded Land
Abril, G., Gu 'erin, F., Richard, S., Delmas, R., Galy-Lacaux, C., Gosse, P., et al. (2005) Carbon
dioxide and methane emissions and the carbon budget of a 10-year old tropical reservoir (Petit
Saut, French Guiana). Global Biogeochem. Cycles 19 (GB4007), 1-16.
https://doi.org/10.1029/20Q5GB0Q2457.
Barros, N., Cole, J.J., Tranvik, L.J., Prairie, Y.T., Bastviken, D., Huszar, V.L.M., et al. (2011) Carbon
emission from hydroelectric reservoirs linked to reservoir age and latitude. Nat. Geosci. 4 (9),
593-596. https://doi.org/10.1038/ngeo1 P11.
Davis, D. W. (1973) Louisiana Canals and Their Influence on Wetland Development. Louisiana State
University and Agricultural & Mechanical College. LSU Historical Dissertations and Theses.
2386., Louisiana State University.
References and Abbreviations 10-129
-------
IPCC (2019) 2019 Refinement to the 2006 Guidelines for National Greenhouse Gas Inventories.
Wetlands, Chapter 7. Lovelock, C. E., Evans, C., Barros, N., Prairie, Y. T., Aim, J., Bastviken, D.,
Beaulieu, J. J., Garneau, M., Harby, A., Harrison, J. A., Pare, David, Raadal, Hanne Lerche,
Sherman, B., Zhang, Chengyi, Ogle, S. M.IPCC (2013) 2013 Supplement to the 2006 IPCC
Guidelines for National Greenhouse Gas Inventories: Wetlands. Hiraishi, T., Krug, T., Tanabe, K.,
Srivastava, N., Baasansuren, J., Fukuda, M. and Troxler, T.G. (eds). In: IPCC, Switzerland.
IPCC (2006) IPCC Guidelines for National Greenhouse Gas Inventories. Prepared by the National
Greenhouse Gas Inventories Programme, Eggleston H.S., Buendia L., Miwa K., Ngara T. and
Tanabe K. (eds). IGES, Japan.
IPCC (2003) Good Practice Guidance for Land Use, Land-Use Change and Forestry. LUCF Sector
Good Practice Guidance, Chapter 3. Penman, J., Gytarsky, M., Hiraishi, T., Krug, T., Kruger, D.,
Pipatti, R., Buendia, L., Miwa, K., Ngara, T., Tanabe, K. and Wagner, F. (eds). Institute of Global
Environmental Strategies (IGES), on behalf of the Intergovernmental Panel on Climate Change
(IPCC): Hayama, Japan.
Lehner B, Reidy Liermann C, Revenga C, Vorosmarty C, Fekete B, Crouzet P, Doll P, et al. (2011 b)
Global Reservoir and Dam Database, Version 1 (GRanDvl): Dams, Revision 01. In: Palisades,
NY: NASA Socioeconomic Data and Applications Center (SEDAC).
Prairie, Y. T., etal. (2017) TheGHG Reservoir Tool (G-res) User guide. UNESCO/IHA research project
on the GHG status of freshwater reservoirs. Joint publication of the UNESCO Chair in Global
Environmental Change and the International Hydropower Association: 38.
Teodoru, C.R., Bastien, J., Bonneville, M.C., Del Giorgio, P.a., Demarty, M., Garneau, M., et al., 2012.
The net carbon footprint of a newly created boreal hydroelectric reservoir. Global Biogeochem.
Cycles 26 (GB2016), 1-14. https://doi.org/10.1029/2011GB0Q4187.
Land Converted to Wetlands: Emissions and Removals
from Land Converted to Vegetated Coastal Wetlands
Abbott, K. M., Elsey-Quirk, T., and DeLaune, R. D. (2019) Factors influencing blue carbon
accumulation across a 32-year chronosequence of created coastal marshes. Ecosphere, 10(8):
e02828.
Allen, J. R., Cornwell, J. C., and Baldwin, A. H. (2021) Contributions of organic and mineral matter to
vertical accretion in tidal wetlands across a Chesapeake Bay subestuary. Journal of Marine
Science and Engineering 9(7): 751.
Arias-Ortiz, A., Oikawa, P. Y., Carlin, J., Masque, P., Shahan, J., Kanneg, S.,... and Baldocchi, D. D.
(2021) Tidal and nontidal marsh restoration: a trade-off between carbon sequestration,
methane emissions, and soil accretion. Journal of Geophysical Research:
Biogeosciences, 126(12): e2021JG006573.
Arriola, J. M., and Cable, J. E. (2017) Variations in carbon burial and sediment accretion along a tidal
creek in a Florida salt marsh. Limnology and Oceanography 62(S1): S15-S28.
10-130 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Baustian, M. M., Stagg, C. L., Perry, C. L., Moss, L. C., and Carruthers, T. J. (2021) Long-term carbon
sinks in marsh soils of coastal Louisiana are at risk to wetland loss. Journal of Geophysical
Research: Biogeosciences 126(3): e2020JG005832.
Bianchi, T. S., Allison, M. A., Zhao, J., Li, X., Comeaux, R. S., Feagin, R. A., & Kulawardhana, R. W.
(2013) Historical reconstruction of mangrove expansion in the Gulf of Mexico: linking climate
change with carbon sequestration in coastal wetlands. Estuarine, Coastal and Shelf Science
119:7-16.
Boyd, B. (2012) Comparison of sediment accumulation and accretion in impounded and
unimpounded marshes of the Delaware Estuary. Doctoral dissertation, University of Delaware.
Boyd, B. M. and Sommerfield, C. K. (2016) Marsh accretion and sediment accumulation in a
managed tidal wetland complex of Delaware Bay. Ecological Engineering, 92: 37-46.
Boyd, B. M., Sommerfield, C. K., and Elsey-Quirk, T. (2017) Hydrogeomorphic influences on salt
marsh sediment accumulation and accretion in two estuaries of the US Mid-Atlantic
coast. Marine Geology, 383:132-145.
Breithaupt, J. L., Smoak, J. M., Smith III, T. J., and Sanders, C. J. (2014) Temporal variability of carbon
and nutrient burial, sediment accretion, and mass accumulation over the past century in a
carbonate platform mangrove forest of the Florida Everglades. Journal of Geophysical
Research: Biogeosciences, 119(10): 2032-2048.
Byrd, K. B., Ballanti, L. R., Thomas, N. M., Nguyen, D. K., Holmquist, J. R., Simard, M., Windham-
Myers, L., Schile, L. M., Parker, V. T.,... and Castaneda-Moya, E. (2017) Biomass/Remote
Sensing dataset: 30m resolution tidal marsh biomass samples and remote sensing data for six
regions in the conterminous United States: U.S. Geological Survey data release,
https://doi.org/10.5066/F77943K8.
Byrd, K. B., Ballanti, L., Thomas, N., Nguyen, D., Holmquist, J.R., Simard, M., and Windham-Myers,
L. (2018) A remote sensing-based model of tidal marsh aboveground carbon stocks for the
conterminous United States. ISPRS Journal of Photogrammetry and Remote Sensing 139: 255-
271.
Byrd, K. B., Ballanti, L., Thomas, N., Nguyen, D., Holmquist, J.R., Simard, M., and Windham-Myers,
L. (2020) Corrigendum to "A remote sensing-based model of tidal marsh aboveground carbon
stocks for the conterminous United States". ISPRS Journal of Photogrammetry and Remote
Sensing 166: 63-67.
Callaway, J. C., Borgnis, E. L., Turner, R. E. & Milan, C. S. (2012a) Carbon sequestration and
sediment accretion in San Francisco Bay tidal wetlands. Estuaries and Coasts 35(5): 1163-
1181.
Callaway, J. C., Borgnis, E. L., Turner, R. E., Milan, C. S., Goodfriend, W., & Richmond, S. (2012b)
"Wetland Sediment Accumulation at Corte Madera Marsh and Muzzi Marsh". San Francisco Bay
Conservation and Development Commission.
References and Abbreviations 10-131
-------
Church, T. M., Sommerfield, C. K., Velinsky, D. J., Point, D., Benoit, C., Amouroux, D. & Donard, O. F.
X. (2006) Marsh sediments as records of sedimentation, eutrophication and metal pollution in
the urban Delaware Estuary. Marine Chemistry 102(1-2): 72-95.
Craft, C. B., & Richardson, C. J. (1998) Recent and long-term organic soil accretion and nutrient
accumulation in the Everglades. Soil Science Society of America Journal 62(3): 834-843.
Crooks, S., Rybczyk, J., O'Connell, K., Devier, D.L., Poppe, K., Emmett-Mattox, S. (2014) Coastal
Blue Carbon Opportunity Assessment for the Snohomish Estuary: The Climate Benefits of
Estuary Restoration. Report by Environmental Science Associates, Western Washington
University, EarthCorps, and Restore America's Estuaries.
Drexler, J. Z., de Fontaine, C. S., and Brown, T. A. (2009) Peat accretion histories during the past
6,000 years in marshes of the Sacramento-San Joaquin Delta, CA, USA. Estuaries and Coasts
32: 871-892.
Drexler, J. Z., Krauss, K. W., Sasser, M. C., Fuller, C. C., Swarzenski, C. M., Powell, A.,... and
Orlando, J. (2013) A long-term comparison of carbon sequestration rates in impounded and
naturally tidal freshwater marshes along the lower Waccamaw River, South Carolina. Wetlands
33: 965-974.
Drexler, J. Z., Woo, I., Fuller, C. C., and Nakai, G. (2019) Carbon accumulation and vertical accretion
in a restored versus historic salt marsh in southern Puget Sound, Washington, United States.
Restoration Ecology 27(5): 1117-1127.
Ensign, S. H., Noe, G. B., Hupp, C. R., and Skalak, K. J. (2015) Head-of-tide bottleneck of particulate
material transport from watersheds to estuaries. Geophysical Research Letters 42(24): 10-671.
Gerlach, M. J., Engelhart, S. E., Kemp, A. C., Moyer, R. P., Smoak, J. M., Bernhardt, C. E., and Cahill,
N. (2017) Reconstructing Common Era relative sea-level change on the Gulf Coast of Florida.
Marine Geology 390: 254-269.
Giblin, A., Forbrich, I., & LTER, P. I. E. (2018) PIE LTER high marsh sediment chemistry and activity
measurements, Nelson Island Creek marsh, Rowley, MA.
Hussein, A. H., Rabenhorst, M. C. &Tucker, M. L. (2004) Modeling of carbon sequestration in
coastal marsh soils. Soil Science Society of America Journal 68(5): 1786-1795.
IPCC (2019) Refinement to the 2006IPCC Guidelines for National Greenhouse Gas Inventories,
Volume 4: Agriculture, Forestry, and Other Land Use. Calvo Buendia, E., Tanabe K., Kranjc, A.,
Baasansuren, J., Fukuda, M., Ngarize, S., Osako, A., Pyrozhenko, Y., Shermanau, P., & Federici,
S. (eds). IPCC, Switzerland.
IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. Prepared by the
National Greenhouse Gas Inventories Programme, H.S.Eggleston, L. Buendia, K. Miwa, T. Ngara
& K. Tanabe (eds). IGES, Japan.
IPCC (2014) 2073 Supplement to the 2006 IPCC Guidelines for National Greenhouse Gas
Inventories: Wetlands. Hiraishi, T., Krug, T., Tanabe, K., Srivastava, N., Baasansuren, J., Fukuda,
M. and Troxler, T.G. (eds.). Published: IPCC, Switzerland.
10-132 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to
the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. Stocker, T.,
Qin, D., Plattner, G.-K., Tignor, M. Allen, S.K., Boschung, J., Nauels, A., Xia, Y., Bex, V. and
Midgley, P.M. (eds.). Cambridge University Press, Cambridge, United Kingdom and New York,
NY, USA.
IPCC (2003) Good Practice Guidance for Land Use, Land-Use Change and Forestry. LUCF Sector
Good Practice Guidance, Chapter 3. Penman, J., Gytarsky, M., Hiraishi, T., Krug, T., Kruger, D.,
Pipatti, R., Buendia, L., Miwa, K., Ngara, T., Tanabe, K. & F. Wagner (eds). Institute of Global
Environmental Strategies (IGES), on behalf of the Intergovernmental Panel on Climate Change
(IPCC): Hayama, Japan.
IPCC (2000) Good Practice Guidance and Uncertainty Management in National Greenhouse Gas
Inventories. Quantifying Uncertainties in Practice, Chapter 6. Penman, J and Kruger, D and
Galbally, I and Hiraishi, T and Nyenzi, B and Emmanuel, S and Buendia, L and Hoppaus, R and
Martinsen, T and Meijer, J and Miwa, Kand Tanabe, K (eds). Institute of Global Environmental
Strategies (IGES), on behalf of the Intergovernmental Panel on Climate Change (IPCC): Hayama,
Japan.
Kearney, M. S. &Stevenson, J. C. (1991) Island land loss and marsh vertical accretion rate evidence
for historical sea-level changes in Chesapeake Bay. Journal of Coastal Research 7(2): 403-415.
Kemp, A. C., Sommerfield, C. K., Vane, C. H., Horton, B. P., Chenery, S., Anisfeld, S., and Nikitina, D.
(2012) Use of lead isotopes for developing chronologies in recent salt-marsh sediments.
Quaternary Geochronology 12: 40-49.
Koster, D., Lichter, J., Lea, P. D., & Nurse, A. (2007) Historical eutrophication in a river-estuary
complex in mid-coast Maine. Ecological Applications 17(3): 765-778.
Krauss, K. W., Noe, G. B., Duberstein, J. A., Conner, W. H., Stagg, C. L., and Jones, M. C. (2018)
Carbon budget assessment of tidal freshwater forested wetland and oligohaline marsh
ecosystems along the Waccamaw and Savannah Rivers, USA (2005-2016). US Geological
Survey Data Release. https://doi.org/10.5066/F7TM7930.
Lagomasino, D., Corbett, D. R., and Walsh, J. P. (2013) Influence of wind-driven inundation and
coastal geomorphology on sedimentation in two microtidal marshes, Pamlico River Estuary,
NC. Estuaries and Coasts 36:1165-1180.
Lu, M & Megonigal, J.P. (2017) Final Report for RAE Baseline Assessment Project. Memo to
Silvestrum Climate Associates by Smithsonian Environmental Research Center, Maryland.
Lynch, J. C. (1989) Sedimentation and nutrient accumulation in mangrove ecosystems of the Gulf of
Mexico, M.S. thesis, Univ. of Southwestern Louisiana, Lafayette, La.
Luk, S. Y., Todd-Brown, K., Eagle, M., McNichol, A. P., Sanderman, J., Gosselin, K., and Spivak, A. C.
(2021) Soil organic carbon development and turnover in natural and disturbed salt marsh
environments. Geophysical Research Letters 48(2): e2020GL090287.
References and Abbreviations 10-133
-------
Marchio, D.A., Savarese, M., Bovard, B., & Mitsch, W.J. (2016) Carbon sequestration and
sedimentation in mangrove swamps influenced by hydrogeomorphic conditions and
urbanization in Southwest Florida. Forests 7:116-135.
McCombs, J.W., Herald, N.D., Burkhalter, S.G. and Robinson C.J., (2016) Accuracy Assessment of
NOAA Coastal Change Analysis Program 2006-2010 Land Cover and Land Cover Change Data.
Photogrammetric Engineering & Remote Sensing. 82:711-718.
McTigue, N., Davis, J., Rodriguez, A. B., McKee, B., Atencio, A., and Currin, C. (2019) Sea level rise
explains changing carbon accumulation rates in a salt marsh over the past two millennia. JGR
Biogeosciences.
Merrill, J. Z. (1999) Tidal Freshwater Marshes as Nutrient Sinks: particulate Nutrient Burial and
Denitrification. Ph.D. Dissertation, University of Maryland, College Park, MD, 342pp.
Miller, C. B., Rodriguez, A. B., Bost, M. C., McKee, B. A., and McTigue, N. D. (2022) Carbon
accumulation rates are highest at young and expanding salt marsh edges. Communications
Earth & Environment 3(1): 173.
National Oceanic and Atmospheric Administration, Office for Coastal Management (2020) Coastal
Change Analysis Program (C-CAP) Regional Land Cover. Charleston, SC: NOAA Office for
Coastal Management. Accessed October 2020 at
www.coast.noaa.gov/htdata/raster1/landcover/bulkdownload/30m lc/.[°Bj
Noe, G. B., Hupp, C. R., Bernhardt, C. E., & Krauss, K. W. (2016) Contemporary deposition and long-
term accumulation of sediment and nutrients by tidal freshwater forested wetlands impacted
by sea level rise. Estuaries and Coasts 39(4): 1006-1019.
Orson, R. A., Simpson, R. L., & Good, R. E. (1990) Rates of sediment accumulation in a tidal
freshwater marsh. Journal of Sedimentary Research 60(6): 859-869.
Orson, R., Warren, R. & Niering, W. (1998) Interpreting sea level rise and rates of vertical marsh
accretion in a southern New England tidal salt marsh. Estuarine, Coastal and Shelf Science
47(4): 419-429.
Peck, E. K., Wheatcroft, R. A., and Brophy, L. S. (2020) Controls on sediment accretion and blue
carbon burial in tidal saline wetlands: insights from the Oregon Coast, USA. Journal of
Geophysical Research: Biogeosciences 125(2): e2019JG005464.
Poppe, K. L., and Rybczyk, J. M. (2021) Tidal marsh restoration enhances sediment accretion and
carbon accumulation in the Stillaguamish River estuary, Washington. PloS one 16(9): e0257244.
Roman, C., Peck, J., Allen, J., King, J. & Appleby, P. (1997) Accretion of a New England (USA) salt
marsh in response to inlet migration, storms, and sea-level rise. Estuarine, Coastal and Shelf
Science 45(6): 717-727.
Smith, K. E., Flocks, J. G., Steyer, G. D., and Piazza, S. C. (2015) Wetland Paleoecological Study of
Southwest Coastal Louisiana: Sediment Cores and Diatom Calibration Dataset. US Department
of the Interior, US Geological Survey.
10-134 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Thom, R. M. (1992 Accretion rates of low intertidal salt marshes in the Pacific Northwest. Wetlands
12: 147-156.
Vaughn, D. R., Bianchi, T. S., Shields, M. R., Kenney, W. F., and Osborne, T. Z. (2020) Increased
organic carbon burial in northern Florida mangrove-salt marsh transition zones. Global
Biogeochemical Cycles 34(5): e2019GB006334.
Villa, J. A. & Mitsch W. J. (2015) "Carbon sequestration in different wetland plant communities of
Southwest Florida". International Journal for Biodiversity Science, Ecosystems Services and
Management 11:17-28.
Watson, E. B., and Byrne, R. (2013) Late Holocene Marsh Expansion in Southern San Francisco Bay,
California. Estuaries and Coasts 36: 643-653.
Weis, D. A., Callaway, J. C., and Gersberg, R. M. (2001) Vertical accretion rates and heavy metal
chronologies in wetland sediments of the Tijuana Estuary. Estuaries 24: 840-850.
Weston, N. B., Neubauer, S. C., Velinsky, D. J., & Vile, M. A. (2014) Net ecosystem carbon exchange
and the greenhouse gas balance of tidal marshes along an estuarine salinity gradient.
Biogeochemistry 120: 163-189.
Weston, N. B., Rodriguez, E., Donnelly, B., Solohin, E., Jezycki, K., Demberger, S.,... and Craft, C. B.
(2023) Recent acceleration of wetland accretion and carbon accumulation along the US East
Coast. Earth's Future 11 (3): e2022EF003037.
Land Converted to Wetlands: Land Converted to Flooded
Land
Abril, G., Gu 'erin, F., Richard, S., Delmas, R., Galy-Lacaux, C., Gosse, R, et al. (2005) Carbon
dioxide and methane emissions and the carbon budget of a 10-year old tropical reservoir (Petit
Saut, French Guiana). Global Biogeochem. Cycles 19 (GB4007), 1-16.
https://doi.org/10.1079/7005GB007457.
Barros, N., Cole, J.J., Tranvik, L.J., Prairie, Y.T., Bastviken, D., Huszar, V.L.M., et al. (2011). Carbon
emission from hydroelectric reservoirs linked to reservoir age and latitude. Nat. Geosci. 4 (9),
593-596. https://doi.org/10.1038/ngeo1211.
IPCC (2019) 207 9 Refinement to the 2006 Guidelines for National Greenhouse Gas Inventories.
Wetlands, Chapter 7. Lovelock, C. E., Evans, C., Barros, N., Prairie, Y. T., Aim, J., Bastviken, D.,
Beaulieu, J. J., Garneau, M., Harby, A., Harrison, J. A., Pare, David, Raadal, Hanne Lerche,
Sherman, B., Zhang, Chengyi, Ogle, S. M.
IPCC (2013) 2073 Supplement to the 2006 IPCC Guidelines for National Greenhouse Gas
Inventories: Wetlands. Hiraishi, T., Krug, T., Tanabe, K., Srivastava, N., Baasansuren, J., Fukuda,
M. and Troxler, T.G. (eds). In: IPCC, Switzerland.
IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. Prepared by the
National Greenhouse Gas Inventories Programme, H.S.Eggleston, L. Buendia, K. Miwa, T. Ngara
& K. Tanabe (eds). IGES, Japan.
References and Abbreviations 10-135
-------
IPCC (2003) Good Practice Guidance for Land Use, Land-Use Change and Forestry. LUCF Sector
Good Practice Guidance, Chapter 3. Penman, J., Gytarsky, M., Hiraishi, T., Krug, T., Kruger, D.,
Pipatti, R., Buendia, L., Miwa, K., Ngara, T., Tanabe, K. and Wagner, F. (eds). Institute of Global
Environmental Strategies (IGES), on behalf of the Intergovernmental Panel on Climate Change
(IPCC): Hayama, Japan.
Lehner B, Reidy Liermann C, Revenga C, Vorosmarty C, Fekete B, Crouzet P, Doll P, et al. (2011 b)
Global Reservoir and Dam Database, Version 1 (GRanDvl): Dams, Revision 01. In: Palisades,
NY: NASA Socioeconomic Data and Applications Center (SEDAC).
Prairie, Y. T., etal. (2017) TheGHG Reservoir Tool (G-res) User guide. UNESCO/IHA research project
on the GHG status of freshwater reservoirs. Joint publication of the UNESCO Chair in Global
Environmental Change and the International Hydropower Association: 38.
Teodoru, C.R., Bastien, J., Bonneville, M.C., Del Giorgio, P.a., Demarty, M., Garneau, M., et al.
(2012). The net carbon footprint of a newly created boreal hydroelectric reservoir. Global
Biogeochem. Cycles 26 (GB2016), 1-14. https://doi.org/10.1029/2011GB004187.
Settlements Remaining Settlements: Soil Carbon Stock
Changes
AAPFCO (2016 through 2022) Commercial Fertilizers: 2013-2017. Association of American Plant
Food Control Officials. University of Missouri. Columbia, MO.
Armentano, T. V., and E.S. Menges (1986) Patterns of change in the carbon balance of organic soil-
wetlands of the temperate zone. Journal of Ecology 74: 755-774.
Brady, N.C. and R.R. Weil (1999) The Nature and Properties of Soils. Prentice Hall. Upper Saddle
River, NJ, 881.
Brockwell, Peter J., and Richard A. Davis (2016) Introduction to time series and forecasting.
Springer.
Fry, J., Xian, G., Jin, S., Dewitz, J., Homer, C., Yang, L., Barnes, C., Herold, N., and J. Wickham. (2011)
Completion of the 2006 National Land Cover Database for the Conterminous United States,
PE&RS 77(9):858-864.
Homer, C., J. Dewitz, J. Fry, M. Coan, N. Hossain, C. Larson, N. Herold, A. McKerrow, J.N. VanDriel
and J. Wickham. (2007) Completion of the 2001 National Land Cover Database for the
Conterminous United States. Photogrammetric Engineering and Remote Sensing 73(4): 337-
341.
Homer, C.G., Dewitz, J.A., Yang, L., Jin, S., Danielson, P., Xian, G., Coulston, J., Herold, N.D.,
Wickham, J.D., and Megown, K. (2015) Completion of the 2011 National Land Cover Database
for the conterminous United States-Representing a decade of land cover change information.
Photogrammetric Engineering and Remote Sensing 81 (5):345-354.
10-136 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change. H.S.
Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Hayama, Kanagawa, Japan.
Nelson, Mark D.; Riitters, Kurt H.; Coulston, John W.; Domke, Grant M.; Greenfield, Eric J.; Langner,
Linda L.; Nowak, David J.; O' Dea, Claire B.; Oswalt, Sonja N.; Reeves, Matthew C.; Wear,
David N. 2020. Defining the United States land base: a technical document supporting the
USDA Forest Service 2020 RPA assessment. Gen. Tech. Rep. NRS-191. Madison, Wl: U.S.
Department of Agriculture, Forest Service, Northern Research Station. 70 p.
https://doi.org/10.2737/NRS-GTR-191.
NRCS (1999) Soil Taxonomy: A basic system of soil classification for making and interpreting soil
surveys, 2nd Edition. Agricultural Handbook Number 436, Natural Resources Conservation
Service, U.S. Department of Agriculture, Washington, D.C.
Nusser, S.M. and J.J. Goebel (1997) The national resources inventory: a long-term multi-resource
monitoring programme. Environmental and Ecological Statistics 4:181-204.
Ogle, S.M., M.D. Eve, F.J. Breidt, and K. Paustian (2003) Uncertainty in estimating land use and
management impacts on soil organic carbon storage for U.S. agroecosystems between 1982
and 1997. GlobalChange Biology 9:1521-1542.
Ogle, S.M., F.J. Breidt, M. Easter, S. Williams, K. Killian, and K. Paustian (2010) Scale and uncertainty
in modeled soil organic carbon stock changes for U.S. croplands using a process-based model.
Global Change Biology 16:810-822.
Sarndal C-E, Swensson B, Wretman, J (1992). Model Assisted Survey Sampling. Springer, New York.
Soil Survey Staff (2020) Gridded Soil Survey Geographic (gSSURGO) Database for the
Conterminous United States. United States Department of Agriculture, Natural Resources
Conservation Service, Accessed February 2020 (FY2020 official release), Available online at
https://gdg.sc.egov.usda.gov/.
USDA-NRCS (2020) Summary Report: 2017 National Resources Inventory. Natural Resources
Conservation Service, Washington, DC, and Center for Survey Statistics and Methodology, Iowa
State University, Ames, Iowa.
https://www.nrcs.usda.gov/wps/portal/nrcs/main/national/technical/nra/nri/results/.
Yang, L., Jin, S., Danielson, P., Homer, C., Gass, L., Bender, S. M., Case, A., Costello, C., Dewitz, J.,
Fry, J., Funk, M., Granneman, B., Liknes, G. C., Rigge, M. &Xian, G. (2018) A new generation of
the United States National Land Cover Database: Requirements, research priorities, design,
and implementation strategies. ISPRS Journal of Photogrammetry and Remote Sensing 146:
108-123.
Settlements Remaining Settlements: Changes in Carbon
Stocks in Settlement Trees
deVries R.E. (1987) A Preliminary Investigation of the Growth and Longevity of Trees in Central Park.
M.S. thesis, Rutgers University, New Brunswick, NJ.
References and Abbreviations 10-137
-------
Fleming, L.E. (1988) Growth Estimation of Street Trees in Central New Jersey. M.S. thesis, Rutgers
University, New Brunswick, NJ.
Frelich, L.E. (1992) Predicting Dimensional Relationships for Twin Cities Shade Trees. University of
Minnesota, Department of Forest Resources, St. Paul, MN, p. 33.
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change. H.S.
Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Hayama, Kanagawa, Japan.
MRLC (2013) National Land Cover Database 2001 (NLCD2001). Available online at:
http://www.mrlc.gov/nlcd2001 .php. Accessed August 2013.
Nowak, D.J. (1986) Silvics of an Urban Tree Species: Norway maple (Acer platanoides L.). M.S.
thesis, College of Environmental Science and Forestry, State University of New York, Syracuse,
NY.
Nowak, D.J. (1994) Atmospheric carbon dioxide reduction by Chicago's urban forest. In: Chicago's
Urban Forest Ecosystem: Results of the Chicago Urban Forest Climate Project. E.G. McPherson,
D.J. Nowak, and R.A. Rowntree (eds.). General Technical Report NE-186. U.S. Department of
Agriculture Forest Service, Radnor, PA. pp. 83-94.
Nowak, D.J. (2012) Contrasting natural regeneration and tree planting in 14 North American cities.
Urban Forestry and Urban Greening. 11: 374- 382.
Nowak, D.J. and D.E. Crane (2002) Carbon storage and sequestration by urban trees in the United
States. Environmental Pollution 116(3):381-389.
Nowak, D.J. and E. Greenfield (2010) Evaluating the National Land Cover Database tree canopy and
impervious cover estimates across the conterminous United States: A comparison with photo-
interpreted estimates. Environmental Management. 46: 378-390.
Nowak, D.J. and E.J. Greenfield (2018a) U.S. urban forest statistics, values and projections. Journal
of Forestry. 116(2): 164-177
Nowak, D.J. and E.J. Greenfield (2018b) Declining urban and community tree cover in the United
States. Urban Forestry and Urban Greening. 32:32-55.
Nowak, D.J., D.E. Crane, J.C. Stevens, and M. Ibarra (2002) Brooklyn's Urban Forest. General
Technical Report NE-290. U.S. Department of Agriculture Forest Service, Newtown Square, PA.
Nowak, D.J., R.E. Hoehn, D.E. Crane, J.C. Stevens, J.T. Walton, and J. Bond (2008) A ground-based
method of assessing urban forest structure and ecosystem services. Arboric. Urb. For. 34(6):
347-358.
Nowak, D.J., E.J. Greenfield, R.E. Hoehn, and E. Lapoint (2013) Carbon storage and sequestration
by trees in urban and community areas of the United States." Environmental Pollution 178: 229-
236.
Nowak, D.J.A.R. Bodine, R.E. Hoehn, C.B. Edgar, D.R. Hartel, T.W. Lister, T.J. Brandeis (2016)
Austin's Urban Forest, 2014. USDA Forest Service, Northern Research Station Resources
Bulletin. NRS-100. Newtown Square, PA. 55 p.
10-138 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Nowak, D.J. A.R. Bodine, R.E. Hoehn, C.B. Edgar, G. Riley, D.R. Hartel, K.J. Dooley, S.M. Stanton,
M.A. Hatfield, T.J. Brandeis, T.W. Lister (2017) Houston's Urban Forest, 2015. USDA Forest
Service, Southern Research Station Resources Bulletin. SRS-211. Newtown Square, PA. 91 p.
Smith, W.B. and S.R. Shifley (1984) Diameter Growth, Survival, and Volume Estimates for Trees in
Indiana and Illinois. Research Paper NC-257. North Central Forest Experiment Station, U.S.
Department of Agriculture Forest Service, St. Paul, MN.
U.S. Department of Interior (2018) National Land Cover Database 2011 (NLCD2011). Accessed
online August 16, 2018. Available online at: https://www.mrlc.gov/nlcd11 leg.php.
Settlements Remaining Settlements: N20 Emissions
from Soils
AAPFCO (2016 through 2022) Commercial Fertilizers: 2013-2017. Association of American Plant
Food Control Officials. University of Missouri. Columbia, MO.
Brakebill, J.W. and Gronberg, J.M. (2017) County-Level Estimates of Nitrogen and Phosphorus from
Commercial Fertilizer for the Conterminous United States, 1987-2012. U.S. Geological Survey,
https://doi.org/10.5066/F7H41 PKX.
Brockwell, Peter J., and Richard A. Davis (2016) Introduction to time series and forecasting.
Springer.
IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to
the Fifth Assessment Report of the Intergovernmental Panel on Climate Change [Stocker, T.F., D.
Qin, G.-K. Plattner, M. Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley
(eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, 1535
pp.
IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to
the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. [S. Solomon,
D. Qin, M. Manning, Z. Chen, M. Marquis, K.B. Averyt, M. Tignorand H.L. Miller (eds.)].
Cambridge University Press. Cambridge, United Kingdom 996 pp.
IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change. H.S.
Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Hayama, Kanagawa, Japan.
Nelson, Mark D.; Riitters, Kurt H.; Coulston, John W.; Domke, Grant M.; Greenfield, Eric J.; Langner,
Linda L.; Nowak, David J.; O' Dea, Claire B.; Oswalt, Sonja N.; Reeves, Matthew C.; Wear,
David N. 2020. Defining the United States land base: a technical document supporting the
USDA Forest Service 2020 RPA assessment. Gen. Tech. Rep. NRS-191. Madison, Wl: U.S.
Department of Agriculture, Forest Service, Northern Research Station. 70 p.
https://doi.org/10.2737/NRS-GTR-191.
Ogle, S.M., F.J. Breidt, M. Easter, S. Williams, K. Killian, and K. Paustian (2010) Scale and uncertainty
in modeled soil organic carbon stock changes for U.S. croplands using a process-based model.
Global Change Biology 16:810-822.
References and Abbreviations 10-139
-------
Sarndal C-E, Swensson B, Wretman, J (1992). Model Assisted Survey Sampling. Springer, New York.
Soil Survey Staff (2020) Gridded Soil Survey Geographic (gSSURGO) Database for the
Conterminous United States. United States Department of Agriculture, Natural Resources
Conservation Service, Accessed February 2020 (FY2020 official release), Available online at
https://gdg.sc.egov.usda.gov/.
USDA-NRCS (2020) Summary Report: 2017 National Resources Inventory. Natural Resources
Conservation Service, Washington, DC, and Center for Survey Statistics and Methodology, Iowa
State University, Ames, Iowa.
https://www.nrcs.usda.gov/wps/portal/nrcs/main/national/technical/nra/nri/results/
Yang, L., et al. (2018). "A new generation of the United States National Land Cover Database:
Requirements, research priorities, design, and implementation strategies." ISPRS Journal of
Photogrammetry and Remote Sensing 146:108-123.
Settlements Remaining Settlements: Changes in Yard
Trimmings and Food Scrap Carbon Stocks in Landfills
Barlaz, M.A. (2008) "Re: Corrections to Previously Published Carbon Storage Factors."
Memorandum to Randall Freed, ICF International. February 28, 2008.
Barlaz, M.A. (2005) "Decomposition of Leaves in Simulated Landfill." Letter report to Randall Freed,
ICF Consulting. June 29, 2005.
Barlaz, M.A. (1998) "Carbon Storage during Biodegradation of Municipal Solid Waste Components
in Laboratory-Scale Landfills." Global Biogeochemical Cycles 12:373-380.
De la Cruz, F.B. and M.A. Barlaz (2010) "Estimation of Waste Component Specific Landfill Decay
Rates Using Laboratory-Scale Decomposition Data" Environmental Science & Technology
44:4722- 4728.
Eleazer, W.E., W.S. Odle, Y. Wang, and M.A. Barlaz (1997) "Biodegradability of Municipal Solid
Waste Components in Laboratory-Scale Landfills." Environmental Science & Technology
31:911-917.
EPA (2020) Advancing Sustainable Materials Management: Facts and Figures 2018. U.S.
Environmental Protection Agency, Office of Solid Waste and Emergency Response, Washington,
D.C. Available online at https://www.epa.gov/smm/advancing-sustainable-materials-
management-facts-and-figures-report.
EPA (2019) Advancing Sustainable Materials Management: Facts and Figures. U.S. Environmental
Protection Agency, Office of Solid Waste and Emergency Response, Washington, D.C. Available
online at https://www.epa.gov/smm/advancing-sustainable-materials-management-facts-and-
figu res-report.
EPA (2016) Municipal Solid Waste Generation, Recycling, and Disposal in the United States: Facts
and Figures. U.S. Environmental Protection Agency, Office of Solid Waste and Emergency
10-140 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Response, Washington, D.C. Available online at
https://archive.epa.gov/epawaste/nonhaz/municipal/web/htnnl/nnsw99.htnnl.
EPA (1995) Compilation of Air Pollutant Emission Factors. U.S. Environmental Protection Agency,
Office of Air Quality Planning and Standards, Research Triangle Park, NC. AP-42 Fifth Edition.
Available online at http://www3.epa.gov/ttnchie1/ap42/.
EPA (1991) Characterization of Municipal Solid Waste in the United States: 1990 Update. U.S.
Environmental Protection Agency, Office of Solid Waste and Emergency Response, Washington,
D.C. EPA/530-SW-90-042.
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change. H.S.
Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Hayama, Kanagawa, Japan.
IPCC (2003) Good Practice Guidance for Land Use, Land-Use Change, and Forestry. The
Intergovernmental Panel on Climate Change, National Greenhouse Gas Inventories
Programme, J. Penman etal. (eds.). Available online at http://www.ipcc-
nggip.iges.or.jp/public/gpglulucf/gpglulucf.htm.
Oshins, C. and D. Block (2000) "Feedstock Composition at Composting Sites." Biocycle 41 (9):31-
34.
Tchobanoglous, G., H. Theisen, and S.A. Vigil (1993) Integrated Solid Waste Management, 1st
edition. McGraw-Hill, NY. Cited by Barlaz (1998) "Carbon Storage during Biodegradation of
Municipal Solid Waste Components in Laboratory-Scale Landfills." Global Biogeochemical
Cycles 12:373-380.
Land Converted to Settlements
Birdsey, R. (1996) "Carbon Storage for Major Forest Types and Regions in the Conterminous United
States." In R.N. Sampson and D. Hair, (eds.). Forest and Global Change, Volume 2: Forest
Management Opportunities for Mitigating Carbon Emissions. American Forests. Washington,
D.C., 1-26 and 261-379 (appendices 262 and 263).
Brockwell, Peter J., and Richard A. Davis (2016) Introduction to time series and forecasting.
Springer. Domke, G.M., Perry, C.H., Walters, B.F., Woodall, C.W., and Smith, J.E. (2016) A
framework for estimating litter carbon stocks in forests of the United States. Science of the
Total Environment 557-558: 469-478.
Domke, G.M., J.E. Smith, and C.W. Woodall. (2011) Accounting for density reduction and structural
loss in standing dead trees: Implications for forest biomass and carbon stock estimates in the
United States. Carbon Balance and Management. 6:14.
Domke, G.M., Woodall, C.W., Walters, B.F., Smith, J.E. (2013) From models to measurements:
comparing down dead wood carbon stock estimates in the U.S. forest inventory. PLoS ONE 8(3):
e59949.
References and Abbreviations 10-141
-------
Domke, G.M., Perry, C.H., Walters, B.F., Woodall, C.W., and Smith, J.E. (2016) A framework for
estimating litter carbon stocks in forests of the United States. Science of the Total Environment
557-558: 469-478.
Fry, J., Xian, G., Jin, S., Dewitz, J., Homer, C., Yang, L., Barnes, C., Herold, N., and Wickham, J. (2011)
Completion of the 2006 National Land Cover Database for the Conterminous United States,
PE&RS, Vol. 77(9):858-864.
Harmon, M.E., C.W. Woodall, B. Fasth, J. Sexton, M. Yatkov. (2011) Differences between standing
and downed dead tree wood density reduction factors: A comparison across decay classes and
tree species. Res. Paper. NRS-15. Newtown Square, PA: U.S. Department of Agriculture, Forest
Service, Northern Research Station. 40 p.
Homer, C., Dewitz, J., Fry, J., Coan, M., Hossain, N., Larson, C., Herold, N., McKerrow, A., VanDriel,
J.N., and Wickham, J. (2007) Completion of the 2001 National Land Cover Database for the
Conterminous United States. Photogrammetric Engineering and Remote Sensing, Vol. 73, No.
4, pp 337-341.
Homer, C.G., Dewitz, J.A., Yang, L., Jin, S., Danielson, P., Xian, G., Coulston, J., Herold, N.D.,
Wickham, J.D., and Megown, K. (2015) Completion of the 2011 National Land Cover Database
for the conterminous United States-Representing a decade of land cover change information.
Photogrammetric Engineering and Remote Sensing, v. 81, no. 5, p. 345-354.
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change, H.S.
Eggleston, L. Buendia, K. Miwa, T Ngara, and K. Tanabe (eds.). Hayama, Kanagawa, Japan.
IPCC (2019) 2079 Refinement to the 2006 IPCC Guidelines for National Greenhouse Gas
Inventories. The National Greenhouse Gas Inventories Programme, The Integovernmental Panel
on Climate Change, Calvo Buendia, E., Tanabe, K., Kranjc, A., Baasansuren, J., Fukuda, M.,
Ngarize S., Osako, A., Pyrozhenko, Y.,Shermanau, P. and Federici, S. (eds). Switzerland.
Jenkins, J.C., D.C. Chojnacky, L.S. Heath, and R.A. Birdsey (2003) "National-scale biomass
estimators for United States tree species." Forest Science 49(1):12-35.
Ogle, S.M., M.D. Eve, F.J. Breidt, and K. Paustian (2003) "Uncertainty in estimating land use and
management impacts on soil organic carbon storage for U.S. agroecosystems between 1982
and 1997." Global Change Biology 9:1521 -1542.
Ogle, S.M., F.J. Breidt, and K. Paustian (2006) "Bias and variance in model results due to spatial
scaling of measurements for parameterization in regional assessments." Global Change Biology
12:516-523.
Schimel, D.S. (1995) "Terrestrial ecosystems and the carbon cycle." Global Change Biology 1: 77-
91.
Smith, J.E.; Heath, L.S.; Skog, K.E.; Birdsey, R.A. (2006) Methods for calculating forest ecosystem
and harvested carbon with standard estimates for forest types of the United States. Gen. Tech.
Rep. NE-343. Newtown Square, PA: U.S. Department of Agriculture, Forest Service,
Northeastern Research Station. 216 p.
10-142 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Tubiello, F. N., et al. (2015) "The Contribution of Agriculture, Forestry and other Land Use activities
to Global Warming, 1990-2012." Global Change Biology 21:2655-2660.
USDA Forest Service. (2022) Forest Inventory and Analysis National Program: FIA Data Mart. U.S.
Department of Agriculture Forest Service. Washington, D.C. Available online at:
https://apps.fs.usda.gov/fia/datamart/datamart.html. Accessed on 07 October 2022.
USDA-NRCS (2020) Summary Report: 2017 National Resources Inventory. Natural Resources
Conservation Service, Washington, DC, and Center for Survey Statistics and Methodology, Iowa
State University, Ames, Iowa.
https://www.nrcs.usda.gov/wps/portal/nrcs/main/national/technical/nra/nri/results/.
USDA-NRCS (1997) "National Soil Survey Laboratory Characterization Data," Digital Data, Natural
Resources Conservation Service, U.S. Department of Agriculture. Lincoln, NE.
Westfall, J.A., Coulston, J.W., Gray, A.N., Shaw, J.D., Radtke, P.J., Walker, D.M., Weiskittel, A.R.,
MacFarlane, D.W., Affleck, D.L.R., Zhao, D., Temesgen, H., Poudel, K.P., Frank, J.M., Prisley, S.P.,
Wang, Y., Sanchez Meador, A.J., Auty, D., and Domke, G.M. (2024) A national-scale tree volume,
biomass, and carbon modeling system for the United States. Gen. Tech. Rep. WO-104.
Washington, DC: U.S. Department of Agriculture, Forest Service. 37 p.
https//doi.org/10.7737/WQ-GTR-104.
Woodall, C.W., L.S. Heath, G.M. Domke, and M.C. Nichols. (2011) Methods and equations for
estimating aboveground volume, biomass, and carbon for trees in the U.S. forest inventory,
2010. Gen. Tech. Rep. NRS-88. Newtown Square, PA: U.S. Department of Agriculture, Forest
Service, Northern Research Station. 30 p.
Woodall, C.W., and V.J. Monleon (2008) Sampling protocol, estimation, and analysis procedures for
the down woody materials indicator of the FIA program. Gen. Tech. Rep. NRS-22. Newtown
Square, PA: U.S. Department of Agriculture, Forest Service, Northern Research Station. 68 p.
Yang, L., Jin, S., Danielson, P., Homer, C., Gass, L., Bender, S. M., Case, A., Costello, C., Dewitz, J.,
Fry, J., Funk, M., Granneman, B., Liknes, G. C., Rigge, M. &Xian, G. (2018) A new generation of
the United States National Land Cover Database: Requirements, research priorities, design,
and implementation strategies. ISPRS Journal of Photogrammetry and Remote Sensing 146:
108-123.
Waste
Landfills
40 CFR Parts 9 and 98 (2024) Revisions and Confidentiality Determinations for Date Elements
Under the Greenhouse Gas Reporting Rule. Code of Federal Regulations, Title 40. Available
online at: https://www.federalregister.gov/documents/2024/04/25/2024-Q7413/revisions-and-
confidentiality-determinations-for-data-elements-under-the-greenhouse-gas-reporting.
References and Abbreviations 10-143
-------
40 CFR Part 60, Subpart WWW (2005) Standards of Performance for Municipal Solid Waste
Landfills, 60.750--60.759, Code of Federal Regulations, Title 40. Available online at:
https://www.ecfr.gov/current/title-40/chapter-l/subchapter-C/part-60/subpart-WWW.
40 CFR Part 258, Subtitle D of RCRA (2012) Criteria for Municipal Solid Waste Landfills, 258.1 —
258.75, Code of Federal Regulations, Title 40. Available online at: https://www.ecfr.gov/cgi-
bin/text-idx?node=pt40.25.258..
BioCycle (2010) "The State of Garbage in America" By L. Arsova, R. Van Haaren, N. Goldstein, S.
Kaufman, and N. Themelis. BioCycle. December 2010. Available online at:
https://www.biocycle.net/2010/10/26/the-state-of-garbage-in-america-4/.
Bronstein, K., Coburn, J., and R. Schmeltz (2012) "Understanding the EPA's Inventory of U.S.
Greenhouse Gas Emissions and Sinks and Mandatory GHG Reporting Program for Landfills:
Methodologies, Uncertainties, Improvements and Deferrals." Prepared for the U.S. EPA
International Emissions Inventory Conference, August 2012, Tampa, Florida. Available online
at: https://www3.epa.gov/ttnchie1/conference/ei20/session3/kbronstein.pdf.
Cusworth D., Duren R., Ayasse A., Jiorle R., Howell K., Aubrey A., Green R., Eastwood M., Chapman
J., Thorpe A., Heckler J., Asner G., Smith M., Thoma E., Krause M., Heins D., Thorneloe S. (2024)
Quantifying Methane Emissions from United States Landfills. Science. 2024 Mar 29;383(6690):
1499-1504. doi: 10.1126/science.adi7735.
Czepiel, P., B. Mosher, P. Crill, and R. Harriss (1996) "Quantifying the Effect of Oxidation on Landfill
Methane Emissions." Journal of Geophysical Research, 101 (D11):16721-16730.Dou, Z.;
Ferguson, J. D.; Galligan, D. T.; Kelly, A. M.; Finn, S. T.; Giegengack, R. (2016) "Assessing U.S.
food wastage and opportunities for reduction." Global Food Security Volume 8, March 2016,
Pages 19-26. https://doi.Org/10.1016/i.gfs.2016.02.001.
EIA (2007) Voluntary Greenhouse Gas Reports for EIA Form 1605B (Reporting Year 2006). Available
online at: https://www.eia.gov/environment/pdfpages/0608s(2009)index.php.
EPA (2024a) Landfill Methane Outreach Program (LMOP). 2024 Landfill and Project Level Data.
September 2024. Available online at: https://www.epa.gov/lmop/landfill-gas-energy-project-
data.
EPA (2024b) Greenhouse Gas Reporting Program Data Published 2024.Envirofacts data download
https://enviro.epa.gov/envirofacts/ghg/search. Subpart HH: Municipal Solid Waste Landfills and
Subpart TT: Industrial Waste Landfills. Accessed on January 10, 2025.
EPA (2020a) Wasted Food Measurement Methodology Scoping Memo. July 2020. Available online at
https://www.epa.gov/sites/production/files/2020-
06/documents/food measurement methodology scoping memo-6-18-20.pdf.
EPA (2020b) Advancing Sustainable Materials Management: 2018 Tables and Figures. December
2020. Available online at: https://www.epa.gov/sites/default/files/2021-
01/documents/2018 tables and figures dec 2020 fnl 5Q8.pdf.
10-144 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
EPA (2019) Advancing Sustainable Materials Management: 2016 and 2017 Tables and Figures.
November 2019. Available online at: https://www.epa.gov/sites/default/files/2019-
11/documents/2016 and 2017 facts and figures data tables O.pdf.
EPA (2018) Advancing Sustainable Materials Management: Facts and Figures 2015. July 2018.
Available online at: https://www.epa.gov/sites/production/files/2018-
07/documents/smm 2015 tables and figures 07252018 fnl 508 O.pdf.
EPA (2016a) Industrial and Construction and Demolition Landfills. Available online at:
https://www.epa.gov/landfills/industrial-and-construction-and-demolition-cd-landfills.
EPA (2016b) Advancing Sustainable Materials Management: Facts and Figures 2014. December
2016. Available online at: https://www.epa.gov/sites/production/files/2016-
11/documents/2014 smm tablesfigures 508.pdf.
EPA (2014) Advancing Sustainable Materials Management: Facts and Figures 2014. February 2014.
Available online at: https://www.epa.gov/sites/production/files/2015-
09/documents/2012 msw dat tbls.pdf.
EPA (2008) Compilation of Air Pollution Emission Factors, Publication AP-42, Draft Section 2.4
Municipal Solid Waste Landfills. October 2008.
EPA (1993) Anthropogenic Methane Emissions in the United States, Estimates for 1990: Report to
Congress, U.S. Environmental Protection Agency, Office of Air and Radiation. Washington, D.C.
EPA/430-R-93-003. April 1993.
EPA (1988) National Survey of Solid Waste (Municipal) Landfill Facilities, U.S. Environmental
Protection Agency. Washington, D.C. EPA/530-SW-88-011. September 1988.
EREF (The Environmental Research & Education Foundation) (2016) Municipal Solid Waste
Management in the United States: 2010 & 2013.
ERG (2023) Production Data Supplied by ERG for 1990-2022 for Pulp and Paper, Fruits and
Vegetables, and Meat. September 7, 2021.
Food Waste Reduction Alliance (FWRA) (2016) Analysis of U.S. Food Waste Among Food
Manufacturers, Retailers, and Restaurants. A joint project by the Food Marketing Institute, the
Grocery Manufacturers Association, and the National Restaurant Association. Available online
at: https://foodwastealliance.org/wp-content/uploads/2020/Q5/FWRA-Food-Waste-Survey-
2016-Report Final.pdf.
IPCC (2019) 207 9 Refinement to the 2006IPCC Guidelines for National Greenhouse Gas
Inventories. Calvo Buendia, E., Tanabe, K., Kranjc, A., Baasansuren, J., Fukuda, M., Ngarize, S.,
Osako, A., Pyrozhenko, Y., Shermanau, P. and Federici, S. (eds). Published: IPCC, Switzerland.
IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change.
[H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa,
Japan.
References and Abbreviations 10-145
-------
Mancinelli, R. and C. McKay (1985) "Methane-Oxidizing Bacteria in Sanitary Landfills." Proc. First
Symposium on Biotechnical Advances in Processing Municipal Wastes for Fuels and
Chemicals, Minneapolis, MN, 437-450. August.
Nesser, H., Jacob, D. J., Maasakkers, J. D., Lorente, A., Chen, Z., Lu, X., Shen, L., Qu, Z., Sulprizio, M.
P., Winter, M., Ma, S., Bloom, A. A., Worden, J. R., Stavins, R. N., and Randies, C. A. (2024) High-
resolution US methane emissions inferred from an inversion of 2019 TROPOMI satellite data:
contributions from individual states, urban areas, and landfills, Atmos. Chem. Phys., 24, 5069-
5091, https://doi.org/10.5194/acp-24-5069-2024.
RTI (2018) Methodological changes to the scale-up factor used to estimate emissions from
municipal solid waste landfills in the Inventory. Memorandum prepared by K. Bronstein and M.
McGrath for R. Schmeltz (EPA). March 22, 2018.
RTI (2017) Methodological changes to the methane emissions from municipal solid waste landfills
as reflected in the public review draft of the 1990-2015 Inventory. Memorandum prepared by K.
Bronstein and M. McGrath for R. Schmeltz (EPA). March 31, 2017.
RTI (2011) Updated Research on Methane Oxidation in Landfills. Memorandum prepared by K. Weitz
(RTI) for R. Schmeltz (EPA). January 14, 2011.
Waste Business Journal (WBJ) (2021) Directory of Waste Processing & Disposal Sites 2021.
WBJ (2016) Directory of Waste Processing & Disposal Sites 2016.
WBJ (2010) Directory of Waste Processing & Disposal Sites 2010.
Wastewater Treatment and Discharge
AF&PA (2022) "AF&PA Members Achieve Progress on Water Stewardship Goal for 2020." American
Forest & Paper Association. Available online at: https://www.afandpa.org/statistics-
resources/afpa-members-achieve-progress-water-stewardship-goal-2020. Accessed July 2022.
AF&PA (2020) "2020 AF&PA Sustainability Report: Advancing the sustainability of an essential
industry." American Forest & Paper Association. Available online at:
https://www.afandpa.org/sites/default/files/2021-07/2020 AF-PA-Sustainability-Report.pdf.
Accessed June 2021.
AF&PA (2018) "2018 AF&PA Sustainability Report: Advancing U.S. Paper and Wood Products
Industry Sustainability Performance." American Forest & Paper Association. Available online at:
http://sustainability.afandpa.org/wp-
content/uploads/2018/06/2018SustainabilityReport PAGES.pdf. Accessed July 2019.
AF&PA (2016) "2016 AF&PA Sustainability Report: Advancing U.S. Paper and Wood Products
Industry Sustainability Performance." American Forest & Paper Association.
AF&PA (2014) "2014 AF&PA Sustainability Report." American Forest & Paper Association.
Babanova et al. (2022) Bioelectrochemical Treatment Technology—The New Practical Approach for
Wastewater Management and GHG Emissions Reduction. Frontiers in Chem Eng 4 (2022).
https://doi.org/10.3389/fceng.7077.837505.
10-146 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
Beecher et al. (2007) "A National Biosolids Regulation, Quality, End Use & Disposal Survey,
Preliminary Report." Northeast Biosolids and Residuals Association, April 14, 2007. Available
online at:
https://static1 .squarespace.com/static/54806478e4b0dc44e1698e88/t/5480c7a2e4b0787f2c
73ad81/1417725858575/NtlBiosldsRpt-AppD-FINAL.pdf. Accessed August 2021.
Beer Institute (2011) Brewers Almanac. Available online at:
http://www.beerinstitute.org/multimedia/brewers-almanac.
Benyahia, F., M. Abdulkarim, A. Embaby, and M. Rao. (2006) Refinery Wastewater Treatment: Atrue
Technological Challenge. Presented at the Seventh Annual U.A.E. University Research
Conference.
BIER (2024) 2023 Benchmarking Study Trends & Observations. Available online at
https://www.bieroundtable.com/publication/2023-water-and-energy-use-benchmarking-
study/. Accessed September 2024.
Brewers Association (2021) Statistics: Number of Breweries. Available online at:
https://www.brewersassociation.org/statistics-and-data/national-beer-stats/. Accessed
August 2021.
Brewers Association (2016a) 2015 Sustainability Benchmarking Report. Available online at:
https://www.brewersassociation.org/best-practices/sustainability/sustainabilitv-
benchmarking-tools. Accessed March 2018.
Brewers Association (2016b) Wastewater Management Guidance Manual. Available online at:
https://www.brewersassociation.org/educational-publications/wastewater-management-
guidance-manual. Accessed September 2017.
Cabrera (2017) "Pulp Mill Wastewater: Characteristics and Treatment." Biological Wastewater
Treatment and Resource Recovery. InTech. pp. 119-139.
CAST (1995) Council for Agricultural Science and Technology. Waste Management and Utilization in
Food Production and Processing. U.S.A. October 1995. ISBN 1-887383-02-6. Available online
at: https://www.cast-science.org/publication/waste-management-and-utilization-in-food-
production-and-processing/.
Chong et al. (2024) "More applicable quantification of non-C02 greenhouse gas emissions from
wastewater treatment plants by on-site plant-integrated measurements." Science of The Total
Environment 929 (2024): 172598. https://doi.Org/10.1016/j.scitotenv.2024.172598
Climate Action Reserve (CAR) (2011) Landfill Project Protocol V4.0, June 2011. Available online at:
http://www.climateactionreserve.org/how/protocols/us-landfill/.
Cooper (2018) Email correspondence. Geoff Cooper, Renewable Fuels Association to Kara Edquist,
ERG. "Wet Mill vs. Dry Mill Ethanol Production." May 18, 2018.
DOE (2013) U.S. Department of Energy Bioenergy Technologies Office. Biofuels Basics. Available
online at: http://energy.gov/eere/bioenergy/biofuels-basics. Accessed September 2013.
References and Abbreviations 10-147
-------
Donovan (1996) Siting an Ethanol Plant in the Northeast. C.T. Donovan Associates, Inc. Report
presented to Northeast Regional Biomass Program (NRBP). (April). Available online at:
https://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.614.856&rep=rep1&type=pdf.
Accessed October 2006.
EIA (2024) Energy Information Administration. U.S. Refinery and Blender Net Production of Crude
Oil and Petroleum Products (Thousand Barrels). Available online at:
https://www.eia.gov/dnav/pet/pet pnp refp dc nus mbbl m.htm. Accessed September 2024.
EPA (2019a) Preliminary Effluent Guidelines Program Plan 14. EPA-821-R-19-005. Office ofWater,
U.S. Environmental Protection Agency. Washington, DC. October 2019. Available online at:
https://www.epa.gov/sites/production/files/2019-10/documents/prelim-eg-plan-14 oct-
2019.pdf. Accessed July 2020.
EPA (2019b) The EPA's Review of Nutrients in Industrial Wastewater Discharge. EPA-830-R-19-001.
Office ofWater, U.S. Environmental Protection Agency. Washington, DC. October 2019.
Available online at: https://www.regulations.gov/document?D=EPA-HO-OW-2Q18-0618-0569.
Accessed March 2020.
EPA (2013) U.S. Environmental Protection Agency. Report on the Performance of Secondary
Treatment Technology. EPA-821-R-13-001. Office ofWater, U.S. Environmental Protection
Agency. Washington, D.C. March 2013. Available online at:
https://www.epa.gov/sites/production/files/2015-
11/documents/npdes secondary treatment report march2013.pdf.
EPA (2012) U.S. Environmental Protection Agency. Clean Watersheds Needs Survey 2012 - Report
to Congress. U.S. Environmental Protection Agency, Office of Wastewater Management.
Washington, D.C. Available online at: https://www.epa.gov/cwns/clean-watersheds-needs-
survey-cwns-2012-report-and-data#access. Accessed February 2016.
EPA (2010) U.S. Environmental Protection Agency. Nutrient Control Design Manual. U.S.
Environmental Protection Agency, Office of Research and Development. Washington, D.C.
EPA600-R-10-100. Available online at: https://www.epa.gov/sites/default/files/2019-
08/documents/nutrient control design manual.pdf. Accessed December 2023.
EPA (2008) U.S. Environmental Protection Agency. Clean Watersheds Needs Survey 2008 - Report
to Congress. U.S. Environmental Protection Agency, Office of Wastewater Management.
Washington, D.C. Available online at: https://www.epa.gov/cwns/clean-watersheds-needs-
survey-cwns-2008-report-and-data. Accessed December 2015.
EPA (2004) U.S. Environmental Protection Agency. Clean Watersheds Needs Survey 2004 - Report
to Congress. U.S. Environmental Protection Agency, Office of Wastewater Management.
Washington, D.C. Available online at: https://www.epa.gov/cwns/clean-watersheds-needs-
survey-cwns-report-congress-2004.
EPA (2000) U.S. Environmental Protection Agency. Clean Watersheds Needs Survey 2000 - Report
to Congress. Office of Wastewater Management, U.S. Environmental Protection Agency.
Washington, D.C. Available online at: https://www.epa.gov/cwns/clean-watersheds-needs-
survey-cwns-2000-report-and-data. Accessed July 2007.
10-148 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
EPA (1999) U.S. Environmental Protection Agency. Biosolids Generation, Use and Disposal in the
United States. Office of Solid Waste and Emergency Response, U.S. Environmental Protection
Agency. Washington, D.C. EPA530-R-99-009. September 1999.
EPA (1998) U.S. Environmental Protection Agency. "AP-42 Compilation of Air Pollutant Emission
Factors." Chapter 2.4, Table 2.4-3, page 2.4-13. Available online at:
https://www.epa.gov/sites/default/files/2Q2Q-10/documents/c02s04.pdf.
EPA (1997a) U.S. Environmental Protection Agency. Estimates of Global Greenhouse Gas Emissions
from Industrial and Domestic Wastewater Treatment. EPA-600/R-97-091. Office of Policy,
Planning, and Evaluation, U.S. Environmental Protection Agency. Washington, D.C. September
1997.
EPA (1997b) U.S. Environmental Protection Agency. Supplemental Technical Development
Document for Effluent Guidelines and Standards (Subparts B & E). EPA-821-R-97-011. Office of
Water, U.S. Environmental Protection Agency. Washington, D.C. October 1997.
EPA (1996) U.S. Environmental Protection Agency. 1996 Clean Water Needs Survey Report to
Congress. Assessment of Needs for Publicly Owned Wastewater Treatment Facilities,
Correction of Combined Sewer Overflows, and Management of Storm Water and Nonpoint
Source Pollution in the United States. Office of Wastewater Management, U.S. Environmental
Protection Agency. Washington, D.C.
EPA (1993a) U.S. Environmental Protection Agency, "Anthropogenic Methane Emissions in the U.S.:
Estimates for 1990, Report to Congress." Office of Air and Radiation, Washington, DC. April
1993.
EPA (1993b) U.S. Environmental Protection Agency. Development Document for the Proposed
Effluent Limitations Guidelines and Standards for the Pulp, Paper and Paperboard Point Source
Category. EPA-821-R-93-019. Office of Water, U.S. Environmental Protection Agency.
Washington, D.C. October 1993.
EPA (1993c) Standards for the Use and Disposal of Sewage Sludge. 40 CFR Part 503.
EPA (1992) U.S. Environmental Protection Agency. Clean Watersheds Needs Survey 1992 - Report
to Congress. Office of Wastewater Management, U.S. Environmental Protection Agency.
Washington, D.C.
EPA (1982) U.S. Environmental Protection Agency. Development Document for Effluent Limitations
Guidelines and standards for the Petroleum Refining. United States Environmental Protection
Agency, Office of Water. EPA-440/1-82-014. Washington D.C. October 1982.
EPA (1975) U.S. Environmental Protection Agency. Development Document for Interim Final and
Proposed Effluent Limitations Guidelines and New Source Performance Standards for the
Fruits, Vegetables, and Specialties Segment of the Canned and Preserved Fruits and Vegetables
Point Source Category. United States Environmental Protection Agency, Office of Water. EPA-
440/1-75-046. Washington D.C. October 1975.
EPA (1974) U.S. Environmental Protection Agency. Development Document for Effluent Limitations
Guidelines and New Source Performance Standards for the Apple, Citrus, and Potato
References and Abbreviations 10-149
-------
Processing Segment of the Canned and Preserved Fruits and Vegetables Point Source Category.
Office of Water, U.S. Environmental Protection Agency, Washington, D.C. EPA-440/1-74-027-a.
March 1974.
ERG (2023) Memorandum: Improvements to the 1990-2021 Wastewater Treatment and Discharge
Greenhouse Gas Inventory. February 2023.
ERG (2021a) Revised Memorandum: Improvements to the 1990-2019 Wastewater Treatment and
Discharge Greenhouse Gas Inventory. March 2021.
ERG (2021 b) Draft Memorandum: Improvements to the 1990-2020 Wastewater Treatment and
Discharge Greenhouse Gas Inventory. July 2021.
ERG (2021c) Draft Memorandum: Expert Judgement Documentation for the Wastewater Treatment
and Discharge Greenhouse Gas Inventory Uncertainties, August 2021.
ERG (2019) Memorandum: Recommended Improvements to the 1990-2018 Wastewater
Greenhouse Gas Inventory. August 2019.
ERG (2018a) Memorandum: Updates to Domestic Wastewater BOD Generation per Capita. August
2018.
ERG (2018b) Memorandum: Inclusion of Wastewater Treatment Emissions from Breweries. July
2018.
ERG (2016) Revised Memorandum: Recommended Improvements to the 1990-2015 Wastewater
Greenhouse Gas Inventory. November 2016.
ERG (2013a) Memorandum: Revisions to Pulp and Paper Wastewater Inventory. October 2013.
ERG (2013b) Memorandum: Revisions to the Petroleum Refinery Wastewater Inventory. October
2013.
ERG (2008a) Memorandum: Planned Revisions of the Industrial Wastewater Inventory Emission
Estimates for the 1990-2007 Inventory. 10 August 2008.
ERG (2008b) Memorandum: Estimation of Onsite Treatment at Industrial Facilities and Review of
Wastewater Characterization Data. 15 April 2008.
ERG (2006a) Memorandum: Recommended Improvements to EPA's Wastewater Inventory for
Industrial Wastewater. Prepared for Melissa Weitz, EPA. 11 August 2006.
ERG (2006b) Memorandum: Assessment of Greenhouse Gas Emissions from Wastewater
Treatment of U.S. Ethanol Production Wastewaters. Prepared for Melissa Weitz, EPA. 10
October 2006.
ERG (2008c) Memorandum: Evaluation of Domestic Wastewater Nitrous Oxide Emissions
Calculation Methodology. 15 April 2008.
Esmaeeli et al. (2023). Reducing freshwater consumption in pulp and paper industries using pinch
analysis and mathematical optimization. Journal of Water Process Eng 53 (2023): 103646.
https://doi.org/10.3389/fceng.2022.832505
10-150 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
FAO (2024a) FAOSTAT-Forestry Database. Available online at:
http://www.fao.Org/faostat/en/#data/FQ. Accessed September 2023.
FAO (2024b) FAOSTAT-Food Balance Sheets. Available online at:
http://www.fao.org/faostat/en/#data/FBS. Accessed September 2024.
FAO (2023) "Pulp and Paper Capacities Report." United States. From 1998 - 2003, 2000 - 2005,
2001 - 2006, 2002 - 2007, 2003 - 2008, 2010 - 2015, 2011 - 2016, 2012 - 2017, 2013 - 2018,
2014 - 2019, 2015 - 2020, 2016 - 2021, 2017 - 2022, 2018 - 2023, 2019 - 2024, 2020-2025,
2021-2026, 2022-2024 reports. Available online at:
https://openknowledge.fao.org/items/2747073c-2c3c-4ab8-9da7-1 b757577e027. Accessed
September 2024.
Foley et a I. (2015) N20 and CH4 Emission from Wastewater Collection and Treatment Systems:
State of the Science Report and Technical Report. GWRC Report Series. IWA Publishing,
London, UK.
Great Lakes-Upper Mississippi River Board of State and Provincial Public Health and Environmental
Managers. (2004) Recommended Standards for Wastewater Facilities (Ten-State Standards).
Guisasola etal. (2008) Methane formation in sewer systems. Water Research 42(6-7): 1421-1430.
Instituto de Estadisticas de Puerto Rico (2021) Population of Puerto Rico from 1990-1999 from
"Estimados anuales poblacionales de los municipios desde 1950." Accessed February 2021.
Available online at: https://censo.estadisticas.pr/EstimadosPoblacionales.
IPCC (2022) Emission factor database: Emission Factor Detail (ID:625621). The Intergovernmental
Panel on Climate Change. Available online at: https://www.ipcc-
nggip.iges.or.jp/EFDB/ef detail.php.
IPCC (2019) 2079 Refinement to the 2006 IPCC Guidelines for National Greenhouse Gas
Inventories. The National Greenhouse Gas Inventories Programme, The Intergovernmental
Panel on Climate Change. [CalvoBuendia, E., Tanabe, K., Kranjc, A., Baasansuren, J., Fukuda,
M., Ngarize S., Osako, A., Pyrozhenko, Y., Shermanau, P. and Federici, S. (eds)]. Switzerland.
IPCC (2014) 2073 Supplement to the 2006 IPCC Guidelines for National Greenhouse Gas
Inventories: Wetlands. [Hiraishi, T., Krug, T., Tanabe, K., Srivastava, N., Baasansuren, J., Fukuda,
M. and Troxler, T.G. (eds.)]. Published: IPCC, Switzerland.
IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to
the Fifth Assessment Report of the Intergovernmental Panel on Climate Change [Stocker, T.F., D.
Qin, G.-K. Plattner, M. Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley
(eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, 1535
pp.
IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to
the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. [S. Solomon,
D. Qin, M. Manning, Z. Chen, M. Marquis, K.B. Averyt, M. Tignorand H.L. Miller (eds.)].
Cambridge University Press. Cambridge, United Kingdom 996 pp.
References and Abbreviations 10-151
-------
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change.
[H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa,
Japan.
Kenari et. al(2010) An Investigation on the Nitrogen Content of a Petroleum Refinery Wastewater
and its Removal by Biological Treatment. Journal of Environmental Health, Sciences, and
Engineering. 7(1): 391-394.
Leverenz, H.L., G. Tchobanoglous, and J.L. Darby (2010) "Evaluation of Greenhouse Gas Emissions
from Septic Systems." Water Environment Research Foundation. Alexandria, VA.
Lewis, A. (2019) Email correspondence. Ann Lewis, RFAto Kara Edquist, ERG. "Wet Mill vs Dry Mill
Ethanol Production." August 20, 2019.
Malmberg, B. (2018) Draft Pulp and Paper Information for Revision of EPA Inventory of U.S.
Greenhouse Gas Emissions and Sinks, Waste Chapter. National Council for Air and Stream
Improvement, Inc. Prepared for Rachel Schmeltz, EPA. June 13, 2018.
Merrick (1998) Wastewater Treatment Options for the Biomass-to-Ethanol Process. Report
presented to National Renewable Energy Laboratory (NREL). Merrick & Company. Subcontract
No. AXE-8-18020-01. October 22,1998.
Metcalf & Eddy, Inc. (2014) Wastewater Engineering: Treatment and Resource Recovery, 5th ed.
McGraw Hill Publishing.
Metcalf & Eddy, Inc. (2003) Wastewater Engineering: Treatment, Disposal and Reuse, 4th ed.
McGraw Hill Publishing.
Moore et al. (2023) "Underestimation of sector-wide methane emissions from United States
wastewater treatment." Environmental Science & Technology 57.10 (2023): 4082-4090.
https://doi.org/10.1021/acs.est.2c05373.
NEBRA (2022) "U.S. National Biosolids Data." Northeast Biosolids and Residuals Associations.
Available online at:
https://static1 .squarespace.com/static/601837d1c67bcc4e1 b11862f/t/62f4f5fbae32804dd9f5
1ef6/1660220925356/National BiosolidsDataSummary NBDP 20220811.pdf.
Nemerow, N.L. and A. Dasgupta (1991) Industrial and Hazardous Waste Treatment. Van Nostrand
Reinhold. NY. ISBN 0-442-31934-7.
NRBP (2001) Northeast Regional Biomass Program. An Ethanol Production Guidebook for
Northeast States. Washington, D.C. (May 3).
Rendleman, C.M. and Shapouri, H. (2007) New Technologies in Ethanol Production. USDA
Agricultural Economic Report Number 842.
RFA (2024a) Renewable Fuels Association. Annual U.S. Fuel Ethanol Production. Available online
at: https://ethanolrfa.org/statistics/annual-ethanol-production. Accessed September 2024.
10-152 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
RFA (2024b) Renewable Fuels Association. Monthly Grain Use for U.S. Ethanol Production Report.
Available online at: https://ethanolrfa.org/statistics/feedstock-use-co-product-output.
Accessed September 2024.
Ruocco (2006a) Email correspondence. Dr. Joe Ruocco, Phoenix Bio-Systems to Sarah Holman,
ERG. "Capacity of Bio-Methanators (Dry Milling)." October 6, 2006.
Ruocco (2006b) Email correspondence. Dr. Joe Ruocco, Phoenix Bio-Systems to Sarah Holman,
ERG. "Capacity of Bio-Methanators (Wet Milling)." October 16, 2006.
Short et al. (2017) Dissolved Methane in the Influent of Three Australian Wastewater Treatment
Plants Fed by Gravity Sewers. Sci Total Environ 599-600: 85-93.
Short et al. (2014) Municipal Gravity Sewers: an Unrecognised Source of Nitrous Oxide. Sci Total
Environ 468-469: 211-218.
Sieranen et al. (2024) "Seasonality of nitrous oxide emissions at six full-scale wastewater treatment
plants." Water Science &Technology 89.3 (2024): 603-612.
https://doi.org/10.2166/wst.2023.420
Song et al. (2023). Methane Emissions from Municipal Wastewater Collection and Treatment
Systems. Environ Sci Technol 57-6: 2248-2261.
Song et al. (2024). "Oversimplification and misestimation of nitrous oxide emissions from
wastewater treatment plants." Nature Sustainability (2024): 1-11.
https://doi.org/10.1038/s41893-024-01420-9
Stier, J. (2018) Personal communications between John Stier, Brewers Association Sustainability
Mentor and Amie Aguiar, ERG. Multiple dates.
Sullivan (SCS Engineers) (2010) The Importance of Landfill Gas Capture and Utilization in the U.S.
Presented to SWICS, April 6, 2010. Available online at: https://www.scsengineers.com/scs-
white-papers/the-importance-of-landfill-gas-capture-and-utilization-in-the-u-s/.
Sullivan (SCS Engineers) (2007) Current MSW Industry Position and State of the Practice on
Methane Destruction Efficiency in Flares, Turbines, and Engines. Presented to Solid Waste
Industry for Climate Solutions (SWICS). July 2007. Available online at:
https://www.scsengineers.com/wp-
content/uploads/2015/03/Sullivan LFG Destruction Efficiency White Paper.pdf.
TTB (2024) Alcohol and Tobacco Tax and Trade Bureau. Beer Statistics. Available online at:
https://www.ttb.gov/beer/beer-stats.shtml. Accessed September 2024.
UNFCCC (2012) CDM Methodological tool, Project emissions from flaring (Version 02.0.0). EB68
Report. Annex 15. Available online at:
http://cdm.unfccc.int/methodologies/PAmethodologies/tools/am-tool-06-v1 .pdf/history view.
U.S. Census Bureau (2024) International Database. Available online at:
https://www.census.gov/data-
tools/demo/idb/#/trends?YR ANIM=7070&dashPages=DASH&FIPS SINGI F=US&COUNTRY YF
AR=2022&menu=trendsViz&TREND RANGE=1990.2023&TREND STEP=5&TREND ADD YRS=&
References and Abbreviations 10-153
-------
FIPS=AO.GO.CO.RO.VO&measures=POP&CCODE=AS.GU.MP.PR.US.VI&CCODE SINGLE=US&
COUNTRY YR ANIM=7077. Accessed September 2024.
U.S. Census Bureau (2023a) "American Housing Survey." Table 1A-4: Selected Equipment and
Plumbing-All Housing Units. From 1989, 1991,1993, 1995,1997, 1999, 2001, 2003, 2005,
2007, and 2009 reports. Table C-04-AO Plumbing, Water, and Sewage Disposal-All Occupied
Units. From 2011, 2013, 2015, 2017, 2019, 2021 and 2023 reports. Available online at
http://www.census.gov/programs-surveys/ahs/data.html. Accessed September 2024.
U.S. Census Bureau (2023b) Annual Estimates of the Resident Population for the United States,
Regions, States, District of Columbia, and Puerto Rico: April 1, 2020 to July 1, 2023. Available
online at: https://www.census.gov/data/tables/time-series/demo/popest/2020s-national-
total.html. Accessed September 2024.
U.S. Census Bureau (2021) Annual Estimates of the Resident Population for the United States,
Regions, States, and Puerto Rico: April 1, 2010 to July 1, 2020. Available online at:
https://www.census.gov/data/tables/time- series/demo/popest/2010s-state-total.html.
U.S. Census Bureau (2013) "American Housing Survey." Table 1A-4: Selected Equipmentand
Plumbing-All Housing Units. From 1989, 1991,1993, 1995,1997, 1999, 2001, 2003, 2005,
2007, and 2009 reports. Table C-04-AO Plumbing, Water, and Sewage Disposal-All Occupied
Units. From 2011, and 2013 reports. Available online at http://www.census.gov/programs-
survevs/ahs/data.html. Accessed May 2020.
U.S. Census Bureau, Population Division (2011) Table 1. Intercensal Estimates of the Resident
Population for the United States, Regions, States, and Puerto Rico: April 1, 2000 to July 1, 2010
(ST-EST00INT-01), Release Date: September 2011. Available online at:
https://www2.census.gov/programs-surveys/popest/datasets/2000-201Q/intercensal/state/st-
estOOint- alldata.csv.
U.S. Census Bureau, Population Division (2002) Table CO-EST2001-12-00 - Time Series of
Intercensal State Population Estimates: April 1,1990 to April 1, 2000.Available online at:
https://www2.census.gov/programs-surveys/popest/tables/1990-200Q/intercensal/st- co/co-
est7001-17-00.pdf.
USDA (U.S. Department of Agriculture) (2024a) Livestock Slaughter 2023 Summary. Available online
at: https://downloads.usda.library.cornell.edu/usda-
esmis/files/r207tp32d/wh248d422/p5549g65c/lsan0424.pdf.
USDA (U.S. Department of Agriculture) (2024b) Poultry Slaughter 2023 Summary. Available online
at: https://downloads.usda.library.cornell.edu/usda-
esmis/files/pg15bd88s/q524m975v/zs25zx570/pslaan24.pdf.
USDA (U.S. Department of Agriculture) (2024c) Vegetables 2023 Summary. Available online at:
https://downloads.usda.library.cornell.edu/usda-
esmis/files/02870v86p/qz20vd735/ht24z584t/vegean24.pdf.
10-154 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
USDA (U.S. Department of Agriculture) (2024d) Noncitrus Fruits and Nuts 2023 Summary. Available
online at: https://downloads.usda.library.cornell.edu/usda-
esmis/files/zs25x846c/6682zt197/qf85q241d/ncit0524.pdf.
USDA (U.S. Department of Agriculture) (2024e) Potato Annual 2023 Summary. Available online at:
https://downloads.usda.library.cornell.edu/usda-
esmis/files/fx719m44h/6q184c88z/v118t7605/pots0924.pdf.
USDA (U.S. Department of Agriculture) (2023) Citrus Fruits 2023 Summary. Available online at:
https://downloads.usda.library.cornell.edu/usda-
esmis/files/j9607060k/4747bs71j/3n705h50s/cf rt0973.pdf.
USDA (2015) U.S. Department of Agriculture. Economic Research Service. Nutrient Availability
(food energy, nutrients, and dietary components). Washington D.C. Available online at:
https://www.ers.usda.gov/data-products/food-availability-per-capita-data-system/food-
availability-per-capita-data-system. Accessed September 2024.
USDA (2014) U.S. Department of Agriculture. Economic Research Service. Kilocalories and
Macronutrients per Capita per Day in the U.S. Food Supply, 1909-2010. Washington D.C.
Available online at: https://www.fns.usda.gov/cnpp/us-foodsupply/nutrient-content-1909-
2010. Accessed June 2021.
USDA (1997) U.S. Department of Agriculture. Economic Research Service. Estimating and
Addressing Americans Food Losses. Washington D.C. Available online at:
https://endhunger.org/docs hunger/USDA-Jan97a.pdf.
U.S. Poultry (2006) Email correspondence. John Starkey, USPOULTRY to D. Bartram, ERG. 30 August
2006.
Wang et al. (2021). Life-cycle assessment of treating slaughterhouse waste using anaerobic
digestion systems. Journal of Cleaner Prod 292 (2021): 126038.
https://doi.org/10.1016/j.jclepro.2021.126038
White and Johnson (2003) White, P.J. and Johnson, L.A. Editors. Corn: Chemistry and Technology.
2nd ed. AACC Monograph Series. American Association of Cereal Chemists. St. Paul, MN.
World Bank (1999) Pollution Prevention and Abatement Handbook 1998, Toward Cleaner
Production. The International Bank for Reconstruction and Development/The WORLDBANK.
1818 H Street, N.W. Washington, DC. 20433, USA. ISBN 0-8213-3638-X.
Yin etal. (2024) "Quantifying Methane Influx from Sewer into Wastewater Treatment Processes."
Environmental Science &Technology (2024). https://doi.org/10.1021/acs.est.4c00820
Composting
BioCycle (2023) BioCycle Nationwide Survey: Full-Scale Food Waste Composting Infrastructure in
the U.S. Prepared by N. Goldstein, P. Luu, and S. Motta. Available online at:
https://www.biocycle.net/us-food-waste-composting-infrastructure/.
References and Abbreviations 10-155
-------
BioCycle (2018a) Organic Waste Bans and Recycling Laws to Tackle Food Waste. Prepared by E.
Broad Lieb, K. Sandson, L. Macaluso, and C. Mansell. Available online at:
https://www.biocycle.net/2018/09/11/organic-waste-bans-recycling-laws-tackle-food-waste/.
BioCycle (2018b) State Food Waste Recycling Data Collection, Reporting Analysis. Prepared by
Nora Goldstein. Available online at: http://compostcolab.wpengine.com/wp-
content/uploads/2018/11/State-Food-Waste-Recycling-Data-Collection-Reporting-
Analysis.pdf.
BioCycle (2017) The State of Organics Recycling in the U.S. Prepared by Nora Goldstein. Available
online at http://www.biocycle.net/17 10 06 1/0001/BioCycle StateOfOrganicsUS.pdf.
BioCycle (2010) The State of Garbage in America. Prepared by Rob van Haaren, Nickolas Themelis
and Nora Goldstein. Available online at
http://www.biocycle.net/images/art/1010/bc101016 s.pdf.
Cornell Composting (1996) Monitoring Compost Moisture. Cornell Waste Management Institute.
Available online at: http://compost.css.cornell.edu/monitor/monitormoisture.html.
Cornell Waste Management Institute (2007) The Science of Composting. Available online at
http://cwmi.css.cornell.edu/chapter1 .pdf.
EPA (2020) Advancing Sustainable Materials Management: 2018 Tables and Figures. Office of Solid
Waste and Emergency Response, U.S. Environmental Protection Agency, Washington, D.C.
Available online at: https://www.epa.gov/sites/default/files/2021-
01/documents/2018 tables and figures dec 2020 fnl 508.pdf.
EPA (2018) Advancing Sustainable Materials Management: 2015 Tables and Figures. Office of Solid
Waste and Emergency Response, U.S. Environmental Protection Agency, Washington, D.C.
Available online at https://www.epa.gov/sites/production/files/2018-
07/documents/smm 2015 tables and figures 07252018 fnl 508 Q.pdf.
EPA (2016) Advancing Sustainable Materials Management: Facts and Figures 2014. Office of Solid
Waste and Emergency Response, U.S. Environmental Protection Agency, Washington, D.C.
Available online at https://www.epa.gov/sites/production/files/2016-
11/documents/2014 smm tablesfigures 508.pdf.
EPA (2014) Municipal Solid Waste in the United States: 2012 Facts and Figures. Office of Solid
Waste and Emergency Response, U.S. Environmental Protection Agency, Washington, D.C.
Available online at https://www.epa.gov/sites/default/files/2015-
09/documents/2012 msw fs.pdf.
Harvard Law School and Center for EcoTechnology (CET) (2019) Bans and Beyond: Designing and
Implementing Organic Waste Bans and Mandatory Organics Recycling Laws. Prepared by Katie
Sandson and Emily Broad Leib, Harvard Law School Food Law and Policy Clinic, with input from
Lorenzo Macaluso and Coryanne Mansell, Center for EcoTechnology (CET). Available online at
https://wastedfood.cetonline.org/wp-content/uploads/2019/07/Harvard-Law-School-FLPC-
Center-for-EcoTechnology-CET-Organic-Waste-Bans-Toolkit.pdf.
10-156 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. Volume 5: Waste,
Chapter 4: Biological Treatment of Solid Waste, Table 4.1. The National Greenhouse Gas
Inventories Programme, The Intergovernmental Panel on Climate Change, H.S. Eggleston, L.
Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Hayama, Kanagawa, Japan. Available online at
https://www.ipcc-nggip.iges.or.jp/public/2006gl/pdf/5 Volume5/V5 4 Ch4 Bio Treat.pdf.
Kijanka (2020) Email correspondence. Kenin Kijanka, EPA Region 2 to Rachel Schmeltz, EPA HQ.
"Puerto Rico Composting Operations." November 13, 2020.
University of Maine (2016) Compost Report Interpretation Guide. Soil Testing Lab. Available online
at: https://umaine.edu/soiltestinglab/wp-content/uploads/sites/227/2016/07/Compost-
Report-lnterpretation-Guide.pdf.
U.S. Census Bureau, Population Division (2024) Table 1. Annual Estimates of the Resident
Population for the United States, Regions, States, the District of Columbia, and Puerto Rico:
April 1, 2020 to July 1, 2023 (NST-EST2023-POP). Available online at
https://www.census.gov/data/tables/time-series/demo/popest/2020s-state-total.html.
U.S. Census Bureau (2021) Table 1. Annual Estimates of the Resident Population for the United
States, Regions, States, the District of Columbia, and Puerto Rico: April 1, 2010 to July 1, 2019;
April 1, 2020; and July 1, 2020 (NST-EST2020). Available online at
https://www.census.gov/programs-surveys/popest/technical-
documentation/research/evaluation-estimates/2020-evaluation-estimates/2010s-totals-
national.html.
U.S. Composting Council (2024) State and City Organics Bans, as of 2023. Accessed on September
19, 2024. Available at https://www.compostingcouncil.org/page/organicsbans.
U.S. Composting Council (2010) Yard Trimmings Bans: Impact and Support. Prepared by Stuart
Buckner, Executive Director, U.S., Composting Council. Available online at
https://cdn.ymaws.com/www.compostingcouncil.org/resource/resmgr/images/advocacy/Yard-
Trimmings-Ban-lmpacts-a.pdf.
Anaerobic Digestion at Biogas Facilities
Bronstein, Kate (2021) Expert Judgement Uncertainty of quantity of materials digested. RTI
International, Solid Waste Management GHG Expert.
EPA (2024) Anaerobic Digestion Facilities Processing Food Waste in the U.S. (2020 & 2021).
Accessed September 19, 2024. Available online at Anaerobic Digestion Facilities Processing
Food Waste in the U.S. (2020 & 2021) | US EPA.
EPA (2023) Anaerobic Digestion Facilities Processing Food Waste in the United States (2019):
Survey Results. April 2023 EPA 530-R-23-003. April 2023. Available online at
https://www.epa.gov/system/files/documents/2023-
04/Anaerobic Digestion Facilities Processing Food Waste in the United States 2019 2023
0404 508.pdf.
References and Abbreviations 10-157
-------
EPA (2021) Anaerobic Digestion Facilities Processing Food Waste in the United States (2017 &
2018): Survey Results. January 2021 EPA/903/S-21/001. Available online at
https://www.epa.gov/sites/default/files/2021-
02/documents/2021 final ad report feb 2 with links.pdf.
EPA (2020) Types of Anaerobic Digesters: Common Ways to Describe Digesters. Available online at
https://www.epa.gov/anaerobic-digestion/types-anaerobic-digesters.
EPA (2019) Anaerobic Digestion Facilities Processing Food Waste in the United States in 2016:
Survey Results. September 2019 EPA/903/S-19/001. Available online at
https://www.epa.gov/sites/production/files/2018-
08/documents/ad data report final 508 compliant no password.pdf.
EPA (2018) Anaerobic Digestion Facilities Processing Food Waste in the United States in 2015:
Survey Results. May 2018 EPA/903/S-18/001. Available online at
https://www.epa.gov/sites/production/files/2019-09/documents/ad data report v10 -
508 comp v1 .pdf.
EPA (2016) Frequently Asked Questions About Anaerobic Digestion. Available online at
https://www.epa.gov/anaerobic-digestion/frequent-questions-about-anaerobic-
digestion#codigestion.
EPA (1993) Anthropogenic Methane Emissions in the U.S.: Estimates for 1990, Report to Congress.
Office of Air and Radiation, Washington, DC. April 1993.
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. Volume 5: Waste,
Chapter 4: Biological Treatment of Solid Waste, Table 4.7. The National Greenhouse Gas
Inventories Programme, The Intergovernmental Panel on Climate Change, H.S. Eggleston, L.
Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Hayama, Kanagawa, Japan. Available online at
https://www.ipcc-nggip.iges.or.jp/public/2006gl/pdf/5 Volume5/V5 4 Ch4 Bio Treat.pdf.
Water Environment Federation (WEF) (2012) What Every Operator Should Know about Anaerobic
Digestion. Available online at https://www.wef.org/globalassets/assets-wef/direct-download-
librarv/public/operator-essentials/wet-operator-essentials—anaerobic-digestion—dec12.pdf.
Waste Incineration
RTI (2009) Hospital/Medical/Infectious Waste Incinerators: Summary of Requirements for Revised
or New Section 111 (d)/129 State Plans Following Amendments to the Emission Guidelines.
Available online at
https://nepis.epa.gov/Exe/ZyPDF.cgi/P1009ZW6.PDF?Dockev=P1009ZW6.PDF.
Waste Sources of Precursor Greenhouse Gas Emissions
EPA (2024) "Criteria pollutants National Tier 1 for 1970 - 2023." National Emissions Inventory (NEI)
Air Pollutant Emissions Trends Data. Office of Air Quality Planning and Standards, February
2024. Available online at: https://www.epa.gov/air-emissions-inventories/air-pollutant-
emissions-trends-data.
10-158 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
EPA (2023) "2020 National Emissions Inventory Technical Support Document: Introduction." Office
of Air Quality Planning and Standards, March 2023. Available online at:
https://www.epa.gov/system/files/documents/2023-
01/NEI2020 TSD Sectionl lntroduction.pdf.
Recalculations and Improvements
IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to
the Fifth Assessment Report of the Intergovernmental Panel on Climate Change [Stocker, T.F.,
D. Qin, G.-K. Plattner, M. Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M.
Midgley (eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY,
USA, 1535 pp.
IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change.
[H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa,
Japan.
References and Abbreviations 10-159
-------
Abbreviations
ABS Acrylonitrile butadiene styrene
AC Air conditioner
ACC American Chemistry Council
AEDT FAA Aviation Environmental Design Tool
AEO Annual Energy Outlook
AER All-electric range
AF&PA American Forest and Paper Association
AFEAS Alternative Fluorocarbon Environmental
Acceptability Study
AFOLU Agriculture, Forestry, and Other Land Use
AFV Alternative fuel vehicle
AGA American Gas Association
AGR Acid gas removal
AHEF Atmospheric and Health Effects Framework
AHRI Air-Conditioning, Heating, and Refrigeration
Institute
AIM Act American Innovation and Manufacturing Act
AISI American Iron and Steel Institute
ALU Agriculture and Land Use
ANGA American Natural Gas Alliance
ANL Argonne National Laboratory
APC American Plastics Council
API American Petroleum Institute
APTA American Public Transportation Association
AR5 IPCC Fifth Assessment Report
AR6 IPCC Sixth Assessment Report
ARI Advanced Resources International
ARMA Autoregressive moving-average
ARMS Agricultural Resource Management Surveys
ASAE American Society of Agricultural Engineers
ASLRRA American Short-line and Regional Railroad
Association
ASR Annual Statistical Report
ASTM American Society for Testing and Materials
AZR American Zinc Recycling
BCEF Biomass conversion and expansion factors
BEA Bureau of Economic Analysis, U.S. Department
of Commerce
BIER Beverage Industry Environmental Roundtable
BLM Bureau of Land Management
BoC Bureau of Census
BOD Biological oxygen demand
BOD5 Biochemical oxygen demand over a 5-day
period
BOEM Bureau of Ocean Energy Management
BOEMRE Bureau of Ocean Energy Management,
Regulation and Enforcement
BOF Basic oxygen furnace
BSEE Bureau of Safety and Environmental
Enforcement
BTS Bureau of Transportation Statistics, U.S.
Department of Transportation
Btu British thermal unit
C Carbon
C&D Construction and demolition waste
C&EN Chemical and Engineering News
CAAA Clean Air Act Amendments of 1990
CAFOS Concentrated Animal Feeding Operations
CAGR Compound Annual Growth Rate
CaO Calcium oxide
CAPP Canadian Association of Petroleum Producers
CARB California Air Resources Board
CBI Confidential business information
C-CAP Coastal Change Analysis Program
CDAT Chemical Data Access Tool
CEAP USDA-NRCS Conservation Effects Assessment
Program
CEFM Cattle Enteric Fermentation Model
CEMS Continuous emission monitoring system
CFC Chlorofluorocarbon
CFR Code of Federal Regulations
CGA Compressed Gas Association
CH4 Methane
CHAPA California Health and Productivity Audit
CHP Combined heat and power
CI Confidence interval
CIGRE International Council on Large Electric Systems
CKD Cement kiln dust
CLE Crown Light Exposure
CMA Chemical Manufacturer's Association
CMM Coal mine methane
CMOP Coalbed Methane Outreach Program
CMR Chemical Market Reporter
CNG Compressed natural gas
CO Carbon monoxide
C02 Carbon dioxide
COD Chemical oxygen demand
COGCC Colorado Oil and Gas Conservation Commission
CONUS Continental United States
CRM Component ratio method
CRP Conservation Reserve Program
CRT Common Reporting Tables
CSRA Carbon Sequestration Rural Appraisals
CTIC Conservation Technology Information Center
CVD Chemical vapor deposition
CWNS Clean Watershed Needs Survey
d.b.h Diameter breast height
DE Digestible energy
DESC Defense Energy Support Center-DoD's Defense
Logistics Agency
DFAMS Defense Fuels Automated Management System
10-160 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
DGGS Division of Geological & Geophysical Surveys GDP
DHS Department of Homeland Security GEI
DLA DoD's Defense Logistics Agency GHG
DM Dry matter GHGRP
DOC Degradable organic carbon GIS
DOC U.S. Department of Commerce GJ
DoD U.S. Department of Defense GOADS
DOE U.S. Department of Energy GOM
DOI U.S. Department of the Interior GPG
DOM Dead organic matter GRI
DOT U.S. Department of Transportation GSAM
DRE Destruction or removal efficiencies GTI
DRI Direct Reduced Iron GWP
EAF Electric arc furnace ha
EDB Aircraft Engine Emissions Databank HBFC
EDF Environmental Defense Fund HC
EER Energy economy ratio HCFC
EF Emission factor HCFO
EFMA European Fertilizer Manufacturers Association HDDV
EJ Exajoule HDGV
EGR Exhaust gas recirculation HDPE
EGU Electric generating unit HF
EIA Energy Information Administration, U.S. HFC
Department of Energy HFO
EOR Enhanced oil recovery HFE
EPA U.S. Environmental Protection Agency HHV
EPRI Electric Power Research Institute HMA
EREF Environment Research & Education Foundation HMIWI
ERS Economic Research Service HTF
ETMS Enhanced Traffic Management System HTS
EV Electric vehicle HVAE
EVI Enhanced Vegetation Index HWP
FAA Federal Aviation Administration IBF
FAO Food and Agricultural Organization IC
FAOSTAT Food and Agricultural Organization database ICAO
FAS Fuels Automated System ICBA
FCCC Framework Convention on Climate Change ICE
FEB Fiber Economics Bureau ICR
FEMA Federal Emergency Management Agency IEA
FERC Federal Energy Regulatory Commission IFO
FGD Flue gas desulfurization IISRP
FHWA Federal Highway Administration
FIA Forest Inventory and Analysis ILENR
FIADB Forest Inventory and Analysis Database
FIPR Florida Institute of Phosphate Research IMO
FOD First order decay IPAA
FOEN Federal Office for the Environment IPCC
FOKS Fuel Oil and Kerosene Sales IPPU
FQSV First-quarter of silicon volume ISO
FSA Farm Service Agency ITC
FTA Federal Transit Authority ITRS
FTP Federal Test Procedure
g Gram JWR
G&B Gathering and boosting KCA
GaAs Gallium arsenide kg
GCV Gross calorific value kt
Gross domestic product
Gulfwide Emissions Inventory
Greenhouse gas
EPA's Greenhouse Gas Reporting Program
Geographic Information Systems
Gigajoule
Gulf Offshore Activity Data System
Gulf of Mexico
Good Practice Guidance
Gas Research Institute
Gas Systems Analysis Model
Gas Technology Institute
Global warming potential
Hectare
Hydrobromofluorocarbon
Hydrocarbon
Hydrochlorofluorocarbon
Hydrochlorofluoroolefin
Heavy duty diesel vehicle
Heavy duty gas vehicle
High density polyethylene
Hydraulically fractured
Hydrofluorocarbon
Hydrofluoroolefin
Hydrofluoroether
Higher Heating Value
Hot Mix Asphalt
Hospital/medical/infectious waste incinerator
Heat Transfer Fluid
Harmonized Tariff Schedule
High Voltage Anode Effects
Harvested wood product
International bunker fuels
Integrated Circuit
International Civil Aviation Organization
International Carbon Black Association
Internal combustion engine
Information Collection Request
International Energy Agency
Intermediate Fuel Oil
International Institute of Synthetic Rubber
Products
Illinois Department of Energy and Natural
Resources
International Maritime Organization
Independent Petroleum Association of America
Intergovernmental Panel on Climate Change
Industrial Processes and Product Use
International Organization for Standardization
U.S. International Trade Commission
International Technology Roadmap for
Semiconductors
Jim Walters Resources
Key category analysis
Kilogram
Kiloton
References and Abbreviations 10-161
-------
kWh
Kilowatt hour
MY
Model year
LDPE
Low density polyethylene
N20
Nitrous oxide
LDT
Light-duty truck
NA
Not applicable; Not available
LDV
Light-duty vehicle
NACWA
National Association of Clean Water Agencies
LEV
Low emission vehicles
NAFTA
North American Free Trade Agreement
LFG
Landfill gas
NAHMS
National Animal Health Monitoring System
LFGTE
Landfill gas-to-energy
NAICS
North American Industry Classification System
LHV
Lower Heating Value
NAPAP
National Acid Precipitation and Assessment
LKD
Lime kiln dust
Program
LLDPE
Linear low density polyethylene
NARR
North American Regional Reanalysis Product
LMOP
EPA's Landfill Methane Outreach Program
NAS
National Academies of Sciences, Engineering,
LNG
Liquefied natural gas
and Medicine
LPG
Liquefied petroleum gas(es)
NASA
National Aeronautics and Space Administration
LTO
Landing and take-off
NASF
National Association of State Foresters
LULUCF
Land Use, Land-Use Change, and Forestry
NASS
USDA's National Agriculture Statistics Service
LVAE
Low Voltage Anode Effects
NC
No change
M&R
Metering and regulating
NCASI
National Council of Air and Stream
MARPOL
International Convention for the Prevention of
Improvement
Pollution from Ships
NCV
Net calorific value
MC
Motorcycle
NE
Not estimated
MCF
Methane conversion factor
NEH
National Engineering Handbook
MCL
Maximum Contaminant Levels
NEI
National Emissions Inventory
MCFD
Thousand cubic feet per day
NEMA
National Electrical Manufacturers Association
MDI
Metered dose inhalers
NEMS
National Energy Modeling System
MDP
Management and design practices
NESHAP
National Emission Standards for Hazardous Air
MECS
EIA Manufacturer's Energy Consumption Survey
Pollutants
MEMS
Micro-electromechanical systems
NEU
Non-Energy Use
MER
Monthly Energy Review
NEV
Neighborhood Electric Vehicle
MGO
Marine gas oil
NFs
Nitrogen trifluoride
MgO
Magnesium oxide
NFI
National forest inventory
MJ
Megajoule
NGL
Natural gas liquids
MLRA
Major Land Resource Area
NGO
Non-Governmental Organization
mm
Millimeter
NID
National inventory of Dams
MMBtu
Million British thermal units
NIR
National Inventory Report
MMCF
Million cubic feet
NLA
National Lime Association
MMCFD
Million cubic feet per day
NLCD
National Land Cover Dataset
MMS
Minerals Management Service
NMOC
Non-methane organic compounds
MMT
Million metric tons
NMVOC
Non-methane volatile organic compound
MMTCE
Million metric tons carbon equivalent
NMOG
Non-methane organic gas
MMTCO2
Million metric tons carbon dioxide equivalent
NO
Not occurring
Eq.
N02
Nitrogen dioxide
MODIS
Moderate Resolution Imaging
NOx
Nitrogen oxides
Spectroradiometer
NOAA
National Oceanic and Atmospheric
MoU
Memorandum of Understanding
Administration
MOVES
U.S. EPA's Motor Vehicle Emission Simulator
NOF
Not on feed
model
NPDES
National Pollutant Discharge Elimination System
MPG
Miles per gallon
NPP
Net primary productivity
MRLC
Multi-Resolution Land Characteristics
NPRA
National Petroleum and Refiners Association
Consortium
NRBP
Northeast Regional Biomass Program
MRV
Monitoring, reporting, and verification
NRC
National Research Council
MSHA
Mine Safety and Health Administration
NRCS
Natural Resources Conservation Service
MSW
Municipal solid waste
NREL
National Renewable Energy Laboratory
MT
Metric ton
NRI
National Resources Inventory
MTBE
Methyl Tertiary Butyl Ether
NSCEP
National Service Center for Environmental
MTBS
Monitoring Trends in Burn Severity
Publications
MVAC
Motor vehicle air conditioning
NSCR
Non-selective catalytic reduction
10-162 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
-------
NSPS New source performance standards RMA
NWS National Weather Service RPA
OAG Official Airline Guide RTO
OAP EPA Office of Atmospheric Programs SAE
OAQPS EPA Office of Air Quality Planning and Standards SAGE
ODP Ozone depleting potential SAIC
ODS Ozone depleting substances SAN
OECD Organization of Economic Co-operation and SAR
Development SCR
OEM Original equipment manufacturers SCSE
OGJ Oil & Gas Journal SDR
OGOR Oil and Gas Operations Reports SEC
OH Hydroxyl radical SEMI
OPEC Organization of Petroleum-Exporting Countries
OMS EPA Office of Mobile Sources SF6
ORNL Oak Ridge National Laboratory SIA
OSHA Occupational Safety and Health Administration SiC
OTA Office of Technology Assessment SICAS
OTAQ EPA Office of Transportation and Air Quality SNAP
OVS Offset verification statement SNG
PADUS Protected Areas Database of the United States S02
PAH Polycyclic aromatic hydrocarbons SOC
PCA Portland Cement Association SOG
PCC Precipitate calcium carbonate SOHIO
PDF Probability Density Function SSURGO
PECVD Plasma enhanced chemical vapor deposition STMC
PET Polyethylene terephthalate SULEV
PET Potential evapotranspiration SWANA
PEVM PFC Emissions Vintage Model SWDS
PFC Perfluorocarbon SWICS
PFPE Perfluoropolyether TA
PHEV Plug-in hybrid vehicles TAM
PHMSA Pipeline and Hazardous Materials Safety TAME
Administration TAR
PI Productivity index TBtu
PLS Pregnant liquor solution TDN
PM Particulate matter TEDB
POTW Publicly Owned Treatment Works TFI
ppbv Parts per billion (109) by volume TIGER
ppm Parts per million
ppmv Parts per million (106) by volume TJ
pptv Parts per trillion (1012) by volume TLEV
PRCI Pipeline Research Council International TMLA
PRP Pasture/Range/Paddock TOW
PS Polystyrene TPO
PSU Primary Sample Unit TRI
PU Polyurethane TSDF
PVC Polyvinyl chloride
PV Photovoltaic TTB
QA/QC Quality Assurance and Quality Control TVA
QBtu Quadrillion Btu UAN
R&D Research and Development UDI
RECs Reduced Emissions Completions UFORE
RCRA Resource Conservation and Recovery Act UG
RFA Renewable Fuels Association U.S.
RFS Renewable Fuel Standard U.S. ITC
Rubber Manufacturers' Association
Resources Planning Act
Regression-through-the-origin
Society of Automotive Engineers
System for assessing Aviation's Global Emissions
Science Applications International Corporation
Styrene Acrylonitrile
IPCC Second Assessment Report
Selective catalytic reduction
South central and southeastern coastal
Steel dust recycling
Securities and Exchange Commission
Semiconductor Equipment and Materials
Industry
Sulfur hexafluoride
Semiconductor Industry Association
Silicon carbide
Semiconductor International Capacity Statistics
Significant New Alternative Policy Program
Synthetic natural gas
Sulfur dioxide
Soil Organic Carbon
State of Garbage survey
Standard Oil Company of Ohio
Soil Survey Geographic Database
Scrap Tire Management Council
Super Ultra Low Emissions Vehicle
Solid Waste Association of North America
Solid waste disposal sites
Solid Waste Industry for Climate Solutions
Treated anaerobically (wastewater)
Typical animal mass
Tertiary amyl methyl ether
IPCC Third Assessment Report
Trillion Btu
Total digestible nutrients
Transportation Energy Data Book
The Fertilizer Institute
Topological^ Integrated Geographic Encoding
and Referencing survey
Terajoule
Traditional low emissions vehicle
Total Manufactured Layer Area
Total organics in wastewater
Timber Product Output
Toxic Release Inventory
Hazardous waste treatment, storage, and
disposal facility
Tax and Trade Bureau
Tennessee Valley Authority
Urea ammonium nitrate
Utility Data Institute
U.S. Forest Service's Urban Forest Effects model
Underground (coal mining)
United States
United States International Trade Commission
References and Abbreviations 10-163
-------
UEP United Egg Producers VOCs
ULEV Ultra low emission vehicle VS
UNEP United Nations Environmental Programme WBJ
UNFCCC United Nations Framework Convention on WEF
Climate Change WERF
USAA U.S. Aluminum Association WFF
USAF United States Air Force
USDA United States Department of Agriculture WGC
USFS United States Forest Service WIP
USGS United States Geological Survey WMO
USITC U.S. International Trade Commission WMS
VAIP EPA's Voluntary Aluminum Industrial WRRF
Partnership WTE
VAM Ventilation air methane WW
VKT Vehicle kilometers traveled WWTP
VMT Vehicle miles traveled ZEVs
Volatile organic compounds
Volatile solids
Waste Business Journal
Water Environment Federation
Water Environment Research Federation
World Fab Forecast (previously WFW, World
Fab Watch)
World Gas Conference
Waste-in-place
World Meteorological Organization
Waste management systems
Water resource recovery facilities
Waste-to-energy
Wastewater
Wastewater treatment plant
Zero emissions vehicles
10-164 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2023
------- |