&EPA
United a*K
Envirainwnlal Protection
Agency
Welfare Risk and Exposure Assessment for
Ozone
Final
-------
This page intentionally left blank.
-------
EPA-452/R-14-005a
August 2014
Welfare Risk and Exposure Assessment for Ozone
Final
U.S. Environmental Protection Agency
Office of Air and Radiation
Office of Air Quality Planning and Standards
Health and Environmental Impacts Division
Risk and Benefits Group
Research Triangle Park, North Carolina 27711
-------
DISCLAIMER
This document has been prepared by staff from the Risk and Benefits Group, Health and
Environmental Impacts Division, Office of Air Quality Planning and Standards, U.S. Environmental
Protection Agency. Any findings and conclusions are those of the authors and do not necessarily
reflect the views of the Agency.
Questions on this document should be addressed to Dr. Bryan Hubbell at the U.S.
Environmental Protection Agency, Office of Air Quality Planning and Standards, 109 TW Alexander
Drive, C504-02, Research Triangle Park, North Carolina 27711 or email — hubbell.bryan@epa.gov.
-------
TABLE OF CONTENTS
1 INTRODUCTION 1-1
1.1 HISTORY 1-2
1.2 CURRENT RISK AND EXPOSURE ASSESSMENTS: GOALS AND PLANNED
APPROACH 1-4
1.3 ORGANIZATION OF DOCUMENT 1-5
2 FRAMEWORK 2-1
2.1 O3 CHEMISTRY 2-2
2.2 SOURCES OF O3 AND O3 PRECURSORS 2-4
2.3 ECOLOGICAL EFFECTS 2-5
2.4 ECOSYSTEM SERVICES 2-8
3 SCOPE 3-1
3.1 OVERVIEW OF RISK AND EXPOSURE ASSESSMENTS FROM PREVIOUS
REVIEW 3-2
3.1.1 Exposure Characterization 3-3
3.1.2 Assessment of Risks to Vegetation 3-4
3.2 OVERVIEW OF CURRENT ASSESSMENT PLAN 3-5
3.2.1 Air Quality Considerations 3-6
3.2.2 Relative TreeBiomass Loss and Crop Yield Loss 3-8
3.2.3 Visible Foliar Injury 3-11
3.3 UNCERTAINTY AND VARIABILITY 3-13
4 AIR QUALITY CONSIDERATIONS 4-1
4.1 INTRODUCTION 4-1
4.2 OVERVIEW OF O3 MONITORING AND AIR QUALITY 4-1
4.3 OVERVIEW OF AIR QUALITY INPUTS TO RISK AND EXPOSURE
ASSESSMENTS 4-4
4.3.1 Air Quality Metrics 4-5
4.3.2 Ambient Air Quality Measurements 4-7
4.3.3 National-scale Air Quality Surfaces for Recent Air Quality 4-8
4.3.4 Air Quality Adjustments to Meet Existing Primary and Potential Alternative
Secondary O3 Standards 4-10
4.4 ASSESSMENT OF UNCERTAINTY 4-29
4.5 SUMMARY OF AIR QUALITY RESULTS 4-43
4.6 REFERENCES 4-43
v
-------
5 O3 RISK TO ECOSYSTEM SERVICES 5-1
5.1 INTRODUCTION 5-1
5.2 REGULATING SERVICES 5-3
5.2.1 Hydrologic Cycle 5-3
5.2.2 Pollination 5-4
5.2.3 Fire Regulation 5-4
5.3 SUPPORTING SERVICES 5-8
5.3.1 Net Primary Productivity 5-8
5.3.2 Community Composition and Habitat Provision 5-9
5.4 PROVISIONING SERVICES 5-10
5.5 CULTURAL SERVICES 5-14
5.5.1 Non-Use Services 5-15
5.6 QUALITATIVE ASSESSMENT OF UNCERTAINTY 5-17
5.7 KEY OBSERVATIONS 5-21
6 BIOMASS LOSS 6-1
6.1 INTRODUCTION 6-1
6.2 RELATIVE BIOMASS LOSS 6-2
6.2.1 Species-Level Analyses 6-7
6.3 COMMERCIAL TIMBER EFFECTS 6-23
6.4 NON-TIMBER FOREST PRODUCTS 6-33
6.4.1 Commercial Non-Timber Forest Products 6-36
6.4.2 Informal Economy or Subsistence Use of Non-Timber Forest Products 6-38
6.5 AGRICULTURE 6-40
6.5.1 Commercial Agriculture 6-40
6.6 CLIMATE REGULATION 6-48
6.6.1 National Scale Forest Carbon Sequestration 6-48
6.6.2 Urban Case Study Carbon Storage 6-50
6.7 URBAN CASE STUDY AIR POLLUTION REMOVAL 6-55
6.8 ECOSYSTEM-LEVEL EFFECTS 6-59
6.8.1 Potential Biomass Loss in Federally Designated Areas 6-63
6.9 QUALITATIVE ASSESSMENT OF UNCERTAINTY 6-64
6.10 KEY OBSERVATIONS 6-72
7 VISIBLE FOLIAR INJURY 7-1
7.1 INTRODUCTION 7-1
7.1.1 Ecosystem Services 7-4
7.2 NATIONAL-SCALE ANALYSIS OF FOLIAR INJURY 7-10
VI
-------
7.2.1 Forest Health Monitoring Network 7-10
7.2.2 NOAA Palmer Z Drought Index 7-12
7.2.3 Results of National-Scale Analysis 7-13
7.3 SCREENING-LEVEL ASSESSMENT OF VISIBLE FOLIAR INJURY IN 214
NATIONAL PARKS 7-19
7.3.1 Screening-Level Assessment Methods 7-20
7.3.2 Screening-Level Assessment Results and Discussion 7-25
7.3.3 Sensitivity Analyses for Screening-Level Assessment 7-31
7.4 NATIONAL PARK CASE STUDY AREAS 7-34
7.4.1 Great Smoky Mountains National Park 7-37
7.4.2 Rocky Mountain National Park 7-45
7.4.3 Sequoia and Kings Canyon National Parks 7-52
7.5 QUALITATIVE ASSESSMENT OF UNCERTAINTY 7-58
7.6 KEY OBSERVATIONS 7-63
8 SUMMARY OF ANALYSES AND SYNTHESIS OF RESULTS 8-1
8.1 INTRODUCTION 8-1
8.2 SUMMARY OF ANALYSES AND KEY RESULTS 8-2
8.2.1 National-Scale Analyses 8-3
8.2.2 Case Study-Scale Analyses 8-13
8.3 PATTERNS OF RISK 8-16
8.3.1 Risk Patterns Across or Between Geographic Areas 8-17
8.3.2 Risk Patterns Across Years 8-20
8.3.3 Risk Patterns Across Alternative W126 Standard Levels 8-22
8.4 REPRESENTATIVENESS 8-26
8.4.1 Species Representativeness 8-26
8.4.2 Geographic Representativeness 8-27
8.4.3 Temporal Representativeness 8-28
8.5 OVERALL CONFIDENCE IN WELFARE EXPOSURE AND RISK RESULTS 8-29
8.5.1 Confidence and Key Uncertainties in Air Quality Analyses 8-29
8.5.2 Confidence and Key Uncertainties in Biomass Loss Analyses 8-30
8.5.3 Confidence and Key Uncertainties in Visible Foliar Injury Analyses 8-30
8.6 CONCLUSIONS 8-31
9 REFERENCES 9-1
vn
-------
LIST OF TABLES
Table 4-1 Percent Reductions in U.S. Anthropogenic NOx Emissions Applied to
Independently Reach Existing and Alternative Secondary Standards in the Nine
Climate Regions 4-13
Table 4-2 Percent Reductions in U.S. Anthropogenic NOx Emissions Applied to Create the
W126 Surfaces Representing Just Meeting Existing and Alternative Standards in
the Nine Climate Regions 4-15
Table 4-3 Highest 2006-2008 Average W126 Concentrations in the Observed and Existing
Standard Air Quality Adjustment Scenarios; Highest 2006-2008 8-hour Os Design
Values in the Observed and Potential Alternative Standard Air Quality
Adjustment Scenarios 4-16
Table 4-4 Comparison of Projected NOx Emissions Reductions to those Applied to Meet
Various Standard Levels in the WREA Analysis 4-38
Table 4-5 Summary of Qualitative Uncertainty Analysis of Key Air Quality Elements in the
O3 NAAQS Risk Assessment 4-39
Table 5-1 Area of Moderate to High-Fire Threat, Mixed Conifer Forest for Existing and
Alternative Standard Levels (in km2) 5-7
Table 5-2 Responses to NSRE Wildlife Value Questions 5-9
Table 5-3 Area (km2) 'At Risk' of High Pine Beetle Loss at Various
W126 Index Values 5-13
Table 5-4 Tree Basal Area Considered 'At Risk' of High Pine Beetle Loss By W126 Index
Values after Just Meeting the Existing and Alternative Standard Levels
(in millions of square feet) 5-14
Table 5-5 NSRE Responses to Non-Use Value Questions For Forests 5-16
Table 5-6 Value Components for WTP for Extensive Protection Program for Southern
Appalachian Spruce-Fir Forests 5-17
Table 5-7 Summary of Qualitative Uncertainty Analysis in Semi-Quantitative Ecosystem
Services Assessments 5-19
Table 6-1 Relative Biomass Loss Functions for Tree Species 6-4
Table 6-2 Relative Biomass Loss Functions for Crop Species 6-4
Table 6-3 Comparison of Adult to Seedling Biomass Loss 6-8
Table 6-4 Comparison of Seedling Biomass Loss to Adult Circumference 6-9
Table 6-5 Summary of Uncertainty in Seedling to Adult Tree Biomass Loss
Comparisons 6-10
vin
-------
Table 6-6 Individual Species Relative Biomass Loss Values - Median, 75th Percentile,
Maximum Percentages 6-17
Table 6-7 Number of Counties w/Tree Species Exceeding 2 Percent Relative
Biomass Loss 6-19
Table 6-8 Mapping Os Impacts to FASOMGHG Forest Types 6-25
Table 6-9 Percent Relative Yield Loss for Forest Types by Region for Modeled
Scenarios 6-26
Table 6-10 Percent Relative Yield Gain for Forest Types by Region with Respect to the
Existing Standard 6-27
Table 6-11 Percentage Changes in National Timber Prices 6-31
Table 6-12 Consumer and Producer Surplus in Forestry, Million $2010 6-33
Table 6-13 Os Sensitive Trees and Their Uses 6-34
Table 6-14 Quantity of NTFP Harvested on U.S. Forest Service and Bureau of Land
Management Land 6-37
Table 6-15 Definition of FASOMGHG Production Regions and Market Regions 6-41
Table 6-16 Mapping of Os Impacts on Crops to FASOMGHG Crops 6-42
Table 6-17 Consumer and Producer Surplus in Agriculture (Million 2010$) 6-47
Table 6-18 Annualized Changes in Consumer and Producer Surplus in Agriculture and
Forestry, 2010-2040, Million 2010$ (4% Discount Rate) 6-48
Table 6-19 Increase in Carbon Storage, MMtCChe, Cumulative over 30 years 6-50
Table 6-20 Tree Species with Available E-R Functions in Selected Urban Study Areas.... 6-52
Table 6-21 Os Effects on Carbon Storage for Five Urban Areas over 25 Years (in millions of
metric tons) 6-54
Table 6-22 Comparison of Pollutant Removal Between an Unadjusted Scenario and
Alternatives and Gains Between the Existing Standard and Alternatives
(metric tons) 6-58
Table 6-23 Percent of Total Basal Area Covered by 12 Assessed Tree Species 6-59
Table 6-24 Grid Cells With No Data That Exceed W126 Index Values under Recent
Conditions 6-60
Table 6-25 Percent of Area Exceeding 2 Percent Weighted Biomass 6-62
Table 6-26 Weighted RBL and Percent Cover in Class I Areas 6-64
Table 6-27 Summary of Qualitative Uncertainty Analysis in Relative Biomass Loss
Assessments 6-66
Table 7-1 Percent of Cover Category Exceeding W126 Index Values 7-4
Table 7-2 National Outdoor Activity Participation 7-8
IX
-------
Table 7-3 National Expenditures for Wildlife Watching, Trail, and Camp-Related
Recreation (in billions of 2010$) 7-10
Table 7-4 Summary of Biosite Index Values for 2006 to 2010 Os Biomonitoring Sites... 7-14
Table 7-5 Censored Regression Results 7-16
Table 7-6 W126 Benchmarks by Relative Soil Moisture Category in Five Scenarios 7-25
Table 7-7 Parks Exceeding W126 Benchmarks in Five Scenarios from 2006 to 2010
(Cumulative) 7-26
Table 7-8 Parks Exceeding W126 Benchmarks in Five Scenarios in Individual Years from
2006 to 2010 7-27
Table 7-9 Screening-level Foliar Injury Results in 42 Parks with an Os Monitor using 3
Methods for Assigning Cb Exposure to Each Park in Base Scenario 7-29
Table 7-10 Foliar Injury Sensitivity Analyses for 214 Parks 7-32
Table 7-11 Soil Moisture Sensitivity Analyses in 57 Os Monitors in Parks 7-33
Table 7-12 Value of Most Frequent Visitor Activities at Great Smoky Mountains National
Park 7-38
Table 7-13 Visitor Spending and Local Area Economic Impact of Great Smoky Mountains
National Park 7-38
Table 7-14 Median Travel Cost for Great Smoky Mountains National Park Visitors 7-39
Table 7-15 Geographic Area of Great Smoky Mountains National Park after Just Meeting
Existing and Alternative Standard Levels (km2) 7-44
Table 7-16 Value of Most Frequent Visitor Activities at Rocky Mountain National Park.. 7-45
Table 7-17 Visitor Spending and Local Area Economic Impact of Rocky Mountain National
Park 7-46
Table 7-18 Median Travel Cost for Rocky Mountain National Park Visitors 7-46
Table 7-19 Geographic Area of Rocky Mountain National Park after Just Meeting Existing
and Alternative Standard Levels (km2) 7-48
Table 7-20 Value of Most Frequent Visitor Activities at Sequoia and Kings Canyon National
Parks 7-53
Table 7-21 Visitor Spending and Local Area Economic Impact of Sequoia and Kings Canyon
National Parks 7-53
Table 7-22 Median Travel Cost for Sequoia and Kings Canyon National Parks Visitors... 7-54
Table 7-23 Geographic Area of Sequoia and Kings Canyon National Parks after Just Meeting
Existing and Alternative Standard Levels (km2) 7-55
Table 7-24 Summary of Qualitative Uncertainty Analysis in Visible Foliar Injury
Assessments 7-59
-------
Table 8-1 Summary of Os-Exposure Risk Across Alternative W126 Standards Relative to
Just Meeting Existing Standard - National-Scale Analyses 8-23
Table 8-2 Summary of Os-Exposure Risk Across Alternative Standards Relative to Just
Meeting Existing Standard - Case-Study Scale Analyses 8-25
XI
-------
LIST OF FIGURES
Figure 2-1 W126 Sigmoidal Weighting Function 2-3
Figure 2-2 Causal Determinations for Os Welfare Effects 2-6
Figure 2-3 Linkages Between Ecosystem Services Categories and Components of Human
Weil-Being 2-10
Figure 4-1 Map of U.S. Ambient Os Monitoring Sites in Operation during 2006-2010 4-3
Figure 4-2 Map of Monitored 8-hour O3 Design Values for the 2006-2008 Period 4-4
Figure 4-3 Flowchart of Air Quality Data Processing for Different Parts of the Welfare Risk
and Exposure Assessments 4-5
Figure 4-4 Monitored 2006-2008 Average W126 Concentrations in ppm-hrs 4-8
Figure 4-5 National Surface of Observed 2006-2008 Average W126 Concentrations,
in ppm-hrs 4-9
Figure 4-6 Map of the 9 NOAA Climate Regions Used in the National-scale Air Quality
Adjustments 4-13
Figure 4-7 National Surface of 2006-2008 Average W126 Concentrations (in ppm-hrs)
Adjusted to Just Meet the Existing Os Standard of 75 ppb 4-19
Figure 4-8 Difference in ppm-hrs between the National Surface of Observed 2006-2008
Average W126 Concentrations and the National Surface of 2006-2008
Average W126 Concentrations Adjusted to Just Meet the Existing Os Standard of
75 ppb 4-20
Figure 4-9 National Surface of 2006-2008 Average W126 concentrations (in ppm-hrs)
Adjusted to Just Meet the Potential Alternative Standard of 15 ppm-hrs 4-21
Figure 4-10 Difference in ppm-hrs Between the National Surface of 2006-2008 Average
W126 Concentrations Adjusted to Just Meet the Existing Os Standard of 75 ppb
and the National Surface of 2006-2008 Average W126 Concentrations Adjusted
to Just Meet the Potential Alternative Standard of 15 ppm-hrs 4-22
Figure 4-11 National Surface of 2006-2008 Average W126 concentrations (in ppm-hrs)
Adjusted to Just Meet the Potential Alternative Standard of 11 ppm-hrs 4-23
Figure 4-12 Difference in ppm-hrs Between the National Surface of 2006-2008 Average
W126 Concentrations Adjusted to Just Meet the Existing Os Standard of 75 ppb
and the National Surface of 2006-2008 Average W126 Concentrations Adjusted
to Just Meet the Potential Alternative Standard of 11 ppm-hrs 4-24
Figure 4-13 National Surface of 2006-2008 Average W126 concentrations (in ppm-hrs)
Adjusted to Just Meet the Potential Alternative Standard of 7 ppm-hrs 4-25
xn
-------
Figure 4-14 Difference in ppm-hrs Between the National Surface of 2006-2008 Average
W126 Concentrations Adjusted to Just Meet the Existing Os Standard of 75 ppb
and the National Surface of 2006-2008 Average W126 Concentrations Adjusted
to Just Meet the Potential Alternative Standard of 7 ppm-hrs 4-26
Figure 4-15 Empirical Frequency Distribution (top) and Cumulative Distribution (bottom)
Functions for the Monitored 2006-2008 8-hour Os Design Values, and the 2006-
2008 8-hour Os Design Values after Adjusting to Just Meet the Existing and
Potential Alternative Standards 4-27
Figure 4-16 Empirical Frequency Distribution (top) and Cumulative Distribution (bottom)
Functions for the Monitored 2006-2008 Average W126 Concentrations, and the
2006-2008 Average W126 Concentrations after Adjusting to Just Meet the
Existing and Potential Alternative Standards 4-28
Figure 4-17 Boxplots of Standard Errors for 2006-2008 Average W126 Values Adjusted to
Meet the Existing and Alternative Standards 4-32
Figure 4-18 Map of Standard Errors for 2006-2008 Average W126 Values (in ppm-hrs)
Adjusted to Meet the Existing Standard of 75 ppb 4-32
Figure 4-19 Map of Standard Errors for 2006-2008 Average W126 Values (in ppm-hrs)
Adjusted to Meet the Alternative Standard of 15 ppm-hrs 4-33
Figure 4-20 Map of Standard Errors for 2006-2008 Average W126 Values (in ppm-hrs)
Adjusted to Meet the Alternative Standard of 11 ppm-hrs 4-33
Figure 4-21 Map of Standard Errors for 2006-2008 Average W126 Values (in ppm-hrs)
Adjusted to Meet the Alternative Standard of 7 ppm-hrs 4-34
Figure 4-22 Percent Reduction in State NOx Emissions Projected to Occur Between
2007 and 2020 4-37
Figure 4-23 Percent NOx Reductions Projected to Occur from 2007 to 2020 Aggregated by
Climate Region for Counties Designated in Attainment with the 2008 Os NAAQS
and Counties Designated Nonattainment for the 2008 Os NAAQS 4-37
Figure 5-1 Conceptual Diagram of the Major Pathway through which Os Enters Plants and
the Major Endpoints that Os May Affect in Plants and Ecosystems 5-1
Figure 5-2 Relationship between Ecological Effects of O3 Exposure and
Ecosystem Services 5-2
Figure 5-3 Overlap of W126 Index Values for the Existing Standard and Alternative W126
Standard Levels, Fire Threat > 2, and Mixed Conifer Forest 5-6
Figure 5-4 Location of Fires in 2010 in Mixed Conifer Forest Areas (under Recent Os
Conditions) 5-8
Figure 5-5 Southern Pine Beetle Damage 5-10
xin
-------
Figure 5-6 Southern Pine Beetle Damage 5-11
Figure 5-7 W126 Index Values for Just Meeting the Existing and Alternative Standard
Levels in Areas Considered 'At Risk' of High Basal Area
Loss (>25% Loss) 5-13
Figure 5-8 Percent of Forest Land in the US by Ownership Category, 2007 5-15
Figure 6-1 Conceptual Diagram of Relationship of Relative Biomass Loss to Ecosystem
Services 6-1
Figure 6-2 Relative Biomass Loss Functions for 12 Tree Species 6-5
Figure 6-3 Relative Yield Loss Functions for 10 Crop Species 6-6
Figure 6-4 Elasticity in Relative Biomass Loss Compared to Changes in W126 6-7
Figure 6-5 W126 Index Values for Alternative Percent Biomass Loss for Tree Species ... 6-12
Figure 6-6 W126 Index Values for Alternative Percent Biomass Loss for Crop Species... 6-12
Figure 6-7 Relative Biomass Loss (RBL) of Ponderosa Pine (Pinusponder osd) Seedlings
under Recent Ambient W126 Index Values (2006 - 2008) 6-14
Figure 6-8 Relative Biomass Loss (RBL) of Ponderosa Pine with Ch Exposure After
Adjusted to Meet the Existing (8-hr) Primary Standard (75 ppb) 6-14
Figure 6-9 Relative Biomass Loss (RBL) of Ponderosa Pine with Ch Exposure After
Adjusted to Meet an Alternative Secondary Standard of 15 ppm-hrs (after
Meeting Existing Os Standard) 6-15
Figure 6-10 Relative Biomass Loss (RBL) of Ponderosa Pine with Ch Exposure After
Adjusted to Meet an Alternative Secondary Standard of 11 ppm-hrs (after
Meeting Existing Os Standard) 6-15
Figure 6-11 Relative Biomass Loss (RBL) of Ponderosa Pine with Ch Exposure After
Adjusted to Meet an Alternative Secondary Standard of 7 ppm-hrs (after Meeting
Existing Cb Standard) 6-16
Figure 6-12 Relative Biomass Loss of Ponderosa Pine at the Existing Primary and Alternative
Secondary Standards 6-20
Figure 6-13 Proportion of Current Standard, Ponderosa Pine - Recent Conditions and
Alternative Secondary Standards 6-21
Figure 6-14 Three-Year Compounded Relative Biomass Loss, by Region 6-23
Figure 6-15 RYG for Softwoods by Region 6-29
Figure 6-16 RYG for Hardwoods by Region 6-30
Figure 6-17 Percentage Changes in Corn RYG with Respect to 75 ppb 6-44
Figure 6-18 Percentage Changes in Soybean RYG with Respect to 75 ppb 6-45
Figure 7-1 Relationship between Visible Foliar Injury and Ecosystem Services 7-2
Figure 7-2 Tree Species Sensitive to Foliar Injury 7-3
xiv
-------
Figure 7-3 Examples of Foliar Injury from Os Exposure 7-6
Figure 7-4 Examples of Southern Bark Beetle Damage 7-7
Figure 7-5 Os Biomonitoring Sampling Sites ("Biosites") 7-12
Figure 7-6 344 Climate Divisions with Palmer Z Soil Moisture Data 7-13
Figure 7-7 General Relationship of Os (ppm-hrs) andBiosite Index 7-14
Figure 7-8 General Relationship of Average Palmer Z (April to August) and
Biosite Index 7-15
Figure 7-9 Cumulative Proportion of Sites with Foliar Injury Present, by Year 7-17
Figure 7-10 Cumulative Proportion of Sites with Foliar Injury Present, by Moisture
Category 7-18
Figure 7-11 Cumulative Proportion of Sites with Foliar Injury Present,
by Climate Region 7-19
Figure 7-12 Distribution of Os and Soil Moisture in 214 Parks by Year 7-23
Figure 7-13 Foliar Injury Results Maps for the Base Scenario in 214 Parks 7-27
Figure 7-14 Identification of W126 Index Value where 10 Percent of Biosites Show Any
Foliar Injury 7-36
Figure 7-15 Cover of Sensitive Species in Great Smoky Mountains National Park 7-41
Figure 7-16 Percent of Sensitive Species Near Trails in Great Smoky Mountains National Park
7-42
Figure 7-17 Trail Kilometers of Sensitive Species by Cover Category in Great Smoky
Mountains National Park 7-43
Figure 7-18 Sensitive Vegetation Cover in Great Smoky Mountains National Park Scenic
Overlooks (3km) 7-44
Figure 7-19 Sensitive Species Cover in Rocky Mountain National Park 7-49
Figure 7-20 Percent Cover of Sensitive Species Near Trails in Rocky Mountain National Park
7-50
Figure 7-21 Trail Kilometers of Sensitive Species by Cover Category in Rocky Mountain
National Park 7-51
Figure 7-22 Sensitive Species Cover in Sequoia and Kings Canyon National Parks 7-56
Figure 7-23 Percent Cover of Sensitive Species Near Trails in Sequoia and Kings Canyon
National Parks 7-57
Figure 7-24 Trail Kilometers of Sensitive Species by Cover Category in Sequoia and Kings
Canyon National Parks 7-58
Figure 8-1 Map of the 9 NOAA Climate Regions (Karl and Koss, 1984) used in the Welfare
Risk and Exposure Assessment 8-5
xv
-------
Figure 8-2 Cumulative Proportion of Biosites with Visible Foliar Injury Present, by Moisture
Category 8-12
Figure 8-3 National Surface of 2006-2008 Average W126 Index Values Adjusted to Just
Meet the Alternative Standard Level of 15 ppm-hrs 8-18
Figure 8-4 National Surface of 2006-2008 Average W126 Index Values Adjusted to Just
Meet the Alternative Standard Level of 11 ppm-hrs 8-19
Figure 8-5 Cumulative Proportion of Sites with Foliar Injury Present, by Year 8-21
xvi
-------
APPENDICES
4A: SPATIAL FIELDS FOR THE W126 METRIC
5A: LARGER MAPS OF FIRE THREAT AND BASAL AREA LOSS
6A: MAPS OF INDIVIDUAL TREE SPECIES
6B: ASSESSMENT OF THE IMPACTS OF ALTERNATIVE OZONE
CONCENTRATIONS ON THE U.S. FOREST AND AGRICULTURE SECTORS
6C: SUPPLY CURVE SHIFTS
6D: iTREE MODEL
6E: CLASS I AREAS AND WEIGHTED RBL AT CURRENT STANDARD
AND ALTERNATIVE W126 STANDARD LEVELS
6F: RELATIVE BIOMASS LOSS AND CROP YIELD LOSS ESTIMATES
7A: ADDITIONAL INFORMATION FOR SCREENING-LEVEL ASSESSMENT OF
VISIBLE FOLIAR INJURY IN NATIONAL PARKS
7B: NATIONAL PARKS CASE STUDY LARGE SCALE MAPS
xvn
-------
LIST OF ACRONYMS/ABBREVIATIONS
AGSIM
AQCD
AQS
BLM
CAA
CALFIRE
CASAC
CASTNET
C.F.R.
CH4
CMAQ
CO
C-R
CSTR
EGU
EPA
FACE
FASOMGHG
FHM
FHTET
FIA
FR
GIS
GRSM
HDDM
HNC-3
HO2
IMPLAN®
IRP
ISA
i-Tree
MEA
MSA
Agriculture Simulation Model
Air Quality Criteria Document
Air Quality System
Bureau of Land Management
Clean Air Act
California Department of Forestry and Fire Protection
Clean Air Scientific Advisory Committee
Clean Air Status and Trends Network
Code of Federal Regulations
Methane
Community Multi-Scale Air Quality
Carbon Monoxide
Concentration-Response
Continuous Stirred Tank Reactors
Electric Generating Unit
Environmental Protection Agency
Free- Air Carbon Dioxide/Ozone Enrichment
Forest and Agriculture Sector Optimization Model with Greenhouse Gases
Forest Health Monitoring
Forest Health Technology Enterprise Team
Forest Inventory and Analysis
Federal Register
Geographic Information System
Great Smoky Mountains National Park
Higher-Order Decoupled Direct Method
Nitric Acid
Hydro-Peroxy Radical
Impact Analysis for Planning Model
Integrated Review Plan
Integrated Science Assessment
Urban Forestry Analysis and Benefits Assessment Tool
Millennium Ecosystem Assessment
Metropolitan Statistical Area
xvin
-------
NAAQS
NCDC
NCLAN
NCore
NEI
FHWAR
NHEERL-WED
NOAA
NOx
NPP
NFS
NSRE
NTFP
O3
OAQPS
OBP
OH
OIF
OTC
PA
PAMS
ppb
ppm-hrs
POMS
RBL
REA
ROMO
RYG
RYL
SEKI
SLAMS
SOx
SPMS
STE
TREGRO
National Ambient Air Quality Standards
National Climatic Data Center
National Crop Loss Assessment Network
National Core
National Emissions Inventory
National Survey of Fishing, Hunting, and Wildlife Associated Recreation
National Health and Environmental Effects Laboratory - Western Ecology
Division
National Oceanographic and Atmospheric Administration
Oxides of Nitrogen
Net Primary Productivity
National Park Service
National Survey on Recreation and the Environment
Non-Timber Forest Products
Ozone
Office of Air Quality Planning and Standards
Ozone Biomonitoring Program
Hydroxl Radical
Outdoor Industry Foundation
Open-Top Camber
Policy Assessment
Photochemical Assessment Monitoring Stations
Parts per Billion
Parts per Million Hours
Portable O3 Monitoring System
Relative Biomass Loss
Risk and Exposure Assessment
Rocky Mountains National Park
Relative Yield Gain
Relative Yield Loss
Sequoia/Kings Canyon National Parks
State and Local Monitoring Stations
Oxides of Sulfur
Special Purpose Monitoring Stations
Stratosphere-Troposphere Exchange
Tree Growth Model
xix
-------
UNESCO
U.S.
USDA
U.S. EPA
USFS
USGS
VegBank
VNA
voc
WHO
W126
WTP
ZELIG
United National Education, Scientific, and Cultural Organization
United States
United States Department of Agriculture
United States Environmental Protection Agency
United States Forest Service
United States Geological Society
Vegetation Plot Database
Voronoi Neighbor Averaging
Volatile Organic Compound
World Health Organization
Cumulative Integrated Exposure Index with a Sigmoidal Weighting
Function
Willingness-to-Pay
A Forest Succession Simulation Model
xx
-------
1 INTRODUCTION
The U.S. Environmental Protection Agency (EPA) is presently conducting a review of
the national ambient air quality standard (NAAQS) for ozone (Os) and related photochemical
oxidants. The NAAQS review process includes four key phases: planning, science assessment,
risk/exposure assessment, and policy assessment/rulemaking.1 This process and the overall plan
for this review of the Os NAAQS are presented in the Integrated Review Plan for the Ozone
National Ambient Air Quality Standards (IRP, US EPA, 201 la). The IRP additionally presents
the schedule for the review; identifies key policy-relevant issues; and discusses the key scientific,
technical, and policy documents. These documents include an Integrated Science Assessment
(ISA), Risk and Exposure Assessments (REAs), and a Policy Assessment (PA). This final
Welfare REA (WREA) is one of the two quantitative REAs developed for the review by EPA's
Office of Air Quality Planning and Standards (OAQPS); the second is a Health REA (HREA).
This WREA focuses on assessments to inform consideration of the review of the secondary
(welfare-based) NAAQS for O3.
The existing secondary standard for Os is set identical to the primary standard at a level
of 0.075 ppm, based on the annual fourth-highest daily maximum 8-hour average concentration,
averaged over three years (73 FR 16436). The EPA initiated the current review of the Os
NAAQS on September 29, 2008 with an announcement of the development of an Os ISA and a
public workshop to discuss policy-relevant science to inform EPA's integrated plan for the
review of the Os NAAQS (73 FR 56581). Discussions at the workshop, held on October 29-30,
2008, informed identification of key policy issues and questions to frame the review of the Os
NAAQS. Drawing from the workshop discussions, EPA developed a draft and then final IRP
(U.S. EPA, 201 la).2 In early 2013, EPA completed the Integrated Science Assessment for Ozone
and Related Photochemical Oxidants (ISA, U.S. EPA, 2013). The Os ISA provides a concise
review, synthesis, and evaluation of the most policy-relevant science to serve as a scientific
1 For more information on the NAAQS review process, see http://www.epa.gov/ttn/naaqs/review.html.
2 On March 30, 2009, EPA held a public consultation with the CASAC O3 Panel on the draft IRP. The final IRP
took into consideration comments received from CASAC and the public on the draft plan, as well as input from
senior Agency managers.
1-1
-------
foundation for the review of the NAAQS. The scientific and technical information in the Os
ISA, including that newly available since the previous review on the welfare effects of Cb,
includes information on exposure, physiological mechanisms by which Os might adversely
impact vegetation, and an evaluation of the ecological evidence, including information on
reported exposure-response (E-R) relationships for Os-related changes in plant biomass.
The REA is a concise presentation of the conceptual model, scope, methods, key results,
observations, and related uncertainties associated with the quantitative analyses performed. This
WREA builds upon the welfare effects evidence presented and assessed in the Os ISA, as well as
CAS AC advice (Samet, 2011) and public comments on a scope and methods planning document
for the REAs (here after, "Scope and Methods Plan", U.S. EPA, 201 Ib). Preparation of this
WREA draws upon the final Os ISA and reflects consideration of CAS AC and public comments
on the first and second drafts of the WREAs (Frey and Samet, 2012a, Frey, 2014). This WREA
is being released, concurrently with the UREA and PA to inform the proposed NAAQS
rulemaking.
The PA presents a staff evaluation and conclusions of the policy implications of the key
scientific and technical information in the Os ISA and final REAs. The PA is intended to help
"bridge the gap" between the Agency's scientific assessments presented in the ISA and REAs
and the judgments required of the EPA Administrator in determining whether it is appropriate to
retain or revise the NAAQS. The PA integrates and interprets the information from the ISA and
REAs to frame policy options for consideration by the Administrator. In so doing, the PA
recognizes that the selection of a specific approach to reaching final decisions on primary and
secondary NAAQS will reflect the judgments of the Administrator. The development of the
various scientific, technical and policy documents and their roles in informing this NAAQS
review are described in more detail in the PA.
1.1 HISTORY
As part of the previous Os NAAQS review completed in 2008, EPA's OAQPS conducted
quantitative risk and exposure assessments to estimate risks to human welfare based on
ecological effects associated with exposure to ambient Os (U.S. EPA 2007a, U.S. EPA 2007b).
The assessment scope and methodology were developed with considerable input from CASAC
1-2
-------
and the public, with CASAC generally concluding that the exposure assessment reflected
generally-accepted modeling approaches, and that the risk assessments were well done, balanced
and reasonably communicated (Henderson, 2006a). The final quantitative risk and exposure
assessments took into consideration CASAC advice (Henderson, 2006a; Henderson, 2006b) and
public comments on two drafts of the risk and exposure assessments.
The assessments conducted as part of the previous review focused on national-level Os-
related impacts to sensitive vegetation and their associated ecosystems. The vegetation exposure
assessment was performed using an interpolation approach that included information from
ambient monitoring networks and results from air quality modeling. The vegetation risk
assessment included both tree and crop analyses. The tree risk analysis included three distinct
lines of evidence: (1) observations of visible foliar injury in the field linked to monitored Os air
quality for the years 2001 - 2004; (2) estimates of seedling growth loss under then-current and
alternative Os exposure conditions; and (3) simulated mature tree growth reductions using the
TREGRO model to simulate the effect of meeting alternative air quality standards on the
predicted annual growth of mature trees from three different species. The crop risk analysis
included estimates of crop yields under current and alternative Os exposure conditions. The
assessments also analyzed the associated changes in economic value upon meeting the levels of
various alternative standards using an agricultural sector economic model.3
Based on the 2006 Air Quality Criteria for Ozone (U.S. EPA, 2006), the 2007 Staff Paper
(U.S. EPA, 2007) and related technical support documents (including the risk and exposure
assessments), EPA published a proposed decision in the Federal Register on July 11, 2007 (72
FR 37818). The EPA proposed to revise the level of the primary standard to a level within the
range of 0.075 to 0.070 ppm. Two options were proposed for the secondary standard: (1)
replacing the then existing standard with a cumulative, seasonal standard, expressed as an index
of the annual sum of weighted hourly concentrations cumulated over 12 daylight hours during
the consecutive 3-month period within the Os season with the maximum index value (W126), set
3 We addressed key observations and insights from the O3 risk assessment, in addition to important caveats and
limitations, in Section II.B of the Final Rule notice (73 FR 16440 to 16443, March 27, 2008).
1-3
-------
at a level within the range of 7 to 21 ppm-hours4, and (2) setting the secondary standard identical
to the revised primary standard. EPA completed the review with publication of a final decision
on March 27, 2008 (73 FR 16436), revising the level of the 8-hour primary Os standard from
0.08 ppm to 0.075 ppm, as the 3-year average of the fourth highest daily maximum 8-hour
average concentration, and revising the secondary standard to be identical to the revised primary
standard.
In May 2008, state, public health, environmental, and industry petitioners filed suit
against EPA regarding the 2008 decision. At EPA's request, the consolidated cases were held in
abeyance pending EPA's reconsideration of the 2008 decision. The Administrator issued a
notice of proposed rulemaking to reconsider the 2008 final decision on January 6, 2010. EPA
held three public hearings. The Agency solicited CAS AC review of the proposed rule on January
25, 2010 and additional CASAC advice on January 26, 2011. On September 2, 2011, the Office
of Management and Budget returned the draft final rule on reconsideration to EPA for further
consideration. EPA decided to coordinate further proceedings on its voluntary rulemaking on
reconsideration with the ongoing periodic review, by deferring the completion of its voluntary
rulemaking on reconsideration until it completes its statutorily-required periodic review. In light
of that, the litigation on the 2008 final decision proceeded. On July 23, 2013, the Court ruled on
the litigation of the 2008 decision, denying the petitioners suit except with respect to the
secondary standard, which was remanded to the Agency for reconsideration. The PA provides
additional description of the court ruling with regard to the secondary standard.
1 2 CURRENT RISK AND EXPOSURE ASSESSMENTS: GOALS AND PLANNED
APPROACH
This final WREA provides an assessment of exposure and risk associated with recent
ambient concentrations of Os and Os air quality adjusted to just meet the existing secondary Os
standard and to just meet potential alternative Os standards based on recommendations provided
in the first and second drafts of the PA. To inform the PA regarding the adequacy of existing
standards and the potential for reductions in adverse effects associated with alternative standards
4 See Chapter 2, Section 2.1 for additional discussion on the W126 metric.
1-4
-------
that might be considered, the goals of this quantitative WREA are to (1) provide estimates of the
ecological effects of Os exposure across a range of environments; (2) provide estimates of
ecological effects within selected case study areas; (3) provide estimates of the effects of Os
exposure on specific urban and non-urban ecosystem services based on the causal ecological
effects; and (4) develop a better understanding of the response of ecological systems and
ecosystem services to changing Os exposure. This quantitative risk and exposure assessment
builds on the approach used and lessons learned in the previous Os risk assessments and focuses
on improving the characterization of the overall confidence in the risk estimates, including
related uncertainties, by improving the methods and data used in the analyses; this risk and
exposure assessment also incorporates the range of ecosystem effects and expands the
characterization of adversity to include consideration of impacts to ecosystem services. This
assessment considers a variety of welfare endpoints for which, in our judgment, there is adequate
information to develop quantitative risk estimates that can meaningfully inform the review of the
secondary Os NAAQS.
1 3 ORGANIZATION OF DOCUMENT
The remainder of this document is organized into chapters. Chapter 2 provides a
conceptual framework for the risk and exposure assessment, including discussions of Cb
chemistry, sources of Os precursors, ecological exposure pathways and uptake into plants,
ecological effects, and ecosystem services endpoints associated with Os. This conceptual
framework sets the stage for the scope of the risk and exposure assessments. Chapter 3 provides
an overview of the scope of the quantitative risk and exposure assessments, including a summary
of the previous risk and exposure assessments and an overview of the current risk and exposure
assessments. Chapter 4 discusses air quality considerations relevant to the exposure and risk
assessments, including available Os monitoring data and important air quality inputs to the risk
and exposure assessments. Chapter 5 describes the ecological effects of Os exposure and the
associated ecosystem services, including the ecosystem services for which data and methods for
incremental analysis of direct Os are not yet available. Chapter 6 provides quantitative analysis
of the biomass loss effects of Os and the ecosystem services affected by this loss, such as
provision of food and fiber, carbon sequestration and storage, and pollution removal. Chapter 7
provides quantitative assessments of the effects of Os on foliar injury and associated ecosystem
1-5
-------
services, particularly cultural services related to recreation and the three selected National Park
case studies. Chapter 8 provides a summary of these analyses and an integrated discussion of the
risk estimates generated in these analyses, drawing on the results of the quantitative analyses and
incorporating considerations from the qualitative discussion of ecosystem services.
1-6
-------
2 FRAMEWORK
In this chapter, we summarize the conceptual framework for assessing exposures of
ecosystems to ozone (Os) and the associated risks to public welfare. This conceptual framework
includes elements related to characterizing: (1) Os chemistry (Section 2.1); (2) important sources
of Os precursors, including oxides of nitrogen (NOx) and volatile organic compounds (VOC)
(Section 2.2); (3) Os-induced effects occurring on Os-sensitive species and in their associated
ecosystems (Section 2.3); and (4) ecosystem services that are likely to be negatively impacted by
changes in ecological functions resulting from Os exposures (Section 2.4). We conclude the
chapter with key observations relevant for developing the scope of the quantitative risk and
exposure assessments.
In the previous review of the secondary standards, we focused the ecological risk
assessment on estimating changes in biomass loss in forest tree species and yield loss in
agricultural crops, as well as qualitatively considering effects on ecosystem services. In this
review, EPA expanded the analysis to consider the broader array of impacts on ecosystem
services resulting from known effects of Os exposure on ecosystem functions. This expanded
scope is addressed in the risk assessment by quantifying the risks not just to ecosystems, but also
to the aspects of public welfare dependent on those ecosystems, i.e., services. EPA has started
using an ecosystem services framework to help inform determinations of the adversity to public
welfare associated with changes in ecosystem functions (Rea et al., 2012). The Risk and
Exposure Assessment conducted as part of the Review of the Secondary National Ambient Air
Quality Standards for Oxides of Nitrogen and Oxides of Sulfur (U.S. EPA, 2009) presented
detailed discussions of how ecosystem services and public welfare are related and how an
ecosystem services framework may be employed to evaluate effects on welfare. In this risk
assessment, we identify the ecosystem services associated with the ecological effects caused by
Os exposure for both the national-scale assessment and the more refined case study areas. These
services may be characterized as: supporting services that are necessary for all other services
(e.g., primary production); cultural services including existence and bequest values, aesthetic
values, and recreation values, among others; provisioning services (e.g., food and timber); and
regulating services such as climate regulation or hydrologic cycle (Millenium Ecosystem
Assessment, 2005).
2-1
-------
2.1 O3 CHEMISTRY
Ozone occurs naturally in the stratosphere where it provides protection against harmful
solar ultraviolet radiation; Os is also formed closer to the Earth's surface in the troposphere by
both natural and anthropogenic sources. Ozone is not emitted directly into the air, but is created
when its two primary precursors, VOC and NOx, combine in the presence of sunlight. Volatile
organic compounds and NOx are, for the most part, emitted directly into the atmosphere. Carbon
monoxide (CO) and methane (CH/t) are also important for Os formation (U.S. EPA, 2013, section
3.2.2).
Rather than varying directly with emissions of its precursors, Os changes in a nonlinear
fashion with the concentrations of its precursors. Nitrogen oxide emissions lead to both the
formation and destruction of Os, depending on the local quantities of NOx, VOC, and radicals
such as the hydroxyl (OH) and hydro-peroxy (IKh) radicals. In areas dominated by fresh NOx
emissions, these radicals are removed via the production of nitric acid (HNOs), which lowers the
Os formation rate. The reduction in, or scavenging of, Os by this reaction is called "titration"
and is often found in downtown metropolitan areas, especially near busy streets and roads, and in
power plant plumes. Titration is usually short-lived and confined to areas close to strong NOx
sources; titration results in localized valleys in which Os concentrations are low compared to
surrounding areas. Consequently, Os response to reductions in NOx emissions is complex and
may include Os decreases at some times and locations and Os increases to fill in the local valleys
of low Os. In contrast, in areas with low NOx concentrations, such as remote continental areas
and rural and suburban areas downwind of urban centers, the net production of Os varies directly
with NOx concentrations and typically increases with increasing NOx emissions.
In general, the rate of Os production is limited by the concentration of VOC or NOx, and
Os formation based on these two precursors depends on the relative sources of OH and NOx.
When OH radicals are abundant and are not depleted by reaction with NOx and/or other species,
Os production is "NOx-limited" (U.S. EPA, 2013, section 3.2.4). In this NOx-limited
circumstance, Os concentrations are most effectively reduced by lowering NOx emissions rather
than by lowering VOC emissions. When OH and other radicals are not abundant, either through
low production or reactions with NOx and other species, Os production is referred to as "VOC-
limited", "radical-limited", or "NOx-saturated" (Jaegle et al., 2001), and Os is most effectively
reduced by lowering VOC emissions. However, even in NOx-saturated conditions, very large
2-2
-------
decreases in NOX emissions can cause the Os formation regime to become NOx-limited.
Consequently, large reductions in NOx emissions can make further emissions reductions more
effective at reducing Os. Between the NOx-limited and NOx-saturated extremes there is a range
where Os is relatively insensitive to marginal changes in both NOx and VOC emissions.
In rural areas and downwind of urban areas, Os production is generally NOx-limited.
This is particularly true in rural areas such as national parks, national forests, and state parks
where VOC emissions from vegetation are high and anthropogenic NOx emissions are relatively
low. Due to lower chemical scavenging in non-urban areas, Os tends to persist longer in rural
than in urban areas and tends to lead to higher cumulative exposures in rural areas than in urban
areas (U.S. EPA, 2013, Section 3.6.2.2).
We focused the analyses in the welfare risk and exposure assessments on the W126 Os
exposure metric. The W126 metric is a seasonal sum of hourly Os concentrations, designed to
measure the cumulative effects of Os exposure on vulnerable plant and tree species. The W126
metric uses a sigmoidal weighting function to place less emphasis on exposure to low
concentrations and more emphasis on exposure to high concentrations (see Figure 2-1).
Figure 2-1 W126 Sigmoidal Weighting Function
2-3
-------
2.2 SOURCES OF O3 AND O3 PRECURSORS
Ozone precursor emissions can be divided into anthropogenic and natural source
categories, with natural sources further divided into biogenic emissions (from vegetation,
microbes, and animals) and abiotic emissions (from biomass burning, lightning, and geogenic
sources). The anthropogenic precursors of Os originate from a wide variety of stationary and
mobile sources.
In urban areas, both biogenic and anthropogenic VOC emissions are relevant to Os
formation. Hundreds of VOC are emitted by evaporation and combustion processes from a large
number of anthropogenic sources. Based on the 2005 national emissions inventory (NEI),
solvent use and highway vehicles are the two main sources of VOC emissions, with roughly
equal contributions to total emissions (U.S. EPA, 2013, Figure 3-2). The emissions inventory
categories of "miscellaneous" (which includes agriculture and forestry, wildfires, prescribed
burns, and structural fires) and off-highway mobile sources are the next two largest contributing
emissions categories, with a combined total of over 5.5 million metric tons of VOC emissions a
year (MT/year).
In rural areas and at the global scale, VOC emissions from vegetation are much larger
than those from anthropogenic sources. In the 2005 NEI, U.S. rural emissions from biogenic
sources were 29 MT/year, and emissions of VOC from anthropogenic sources were
approximately 17 MT/year (wildfires constitute -1/6 of that total). Vegetation emits substantial
quantities of VOC, such as isoprene and other terpenoid and sesqui-terpenoid compounds. Most
biogenic emissions occur during the summer because they depend on temperature and incident
sunlight. Biogenic emissions are also higher in southern and eastern states than in northern and
western states for these reasons and because of species variations.
Anthropogenic NOX emissions are associated with combustion processes. Based on the
2005 NEI, the three largest sources of NOX emissions in the U.S. are on-road and off-road mobile
sources (e.g., construction and agricultural equipment) and electric power generation plants
(electric generating units, or EGUs) (U.S. EPA, 2013, Figure 3-2). Emissions of NOx are highest
in areas with a high density of power plants and in urban regions with high traffic density.
However, it is not possible to make an overall statement about their relative impacts on Os in all
2-4
-------
local areas because there are fewer EGUs than mobile sources, particularly in the west and south,
and because of the nonlinear chemistry discussed in Section 2.1.
Major natural sources of NOx in the U.S. include lightning, soils, and wildfires. Biogenic
NOx emissions are generally highest during the summer and occur across the entire country,
including areas where anthropogenic emissions are low. It should be noted that uncertainties in
estimating natural NOx emissions are much larger than uncertainties in estimating anthropogenic
NOx emissions.
Ozone concentrations in a region are affected both by local formation and by transport
from surrounding areas. Ozone transport occurs on many spatial scales, including local transport
between cities, regional transport over large regions of the U.S., and international/long-range
transport. In addition, Os is also transferred from the stratosphere into the troposphere, which is
rich in Os, through stratosphere-troposphere exchange (STE). These inversions or "foldings"
usually occur behind cold fronts, bringing stratospheric air with them (U.S. EPA, 2013, section
3.4.1.1). Contribution to Os concentrations in an area from STE are defined as being part of
background Os (U.S. EPA, 2013, section 3.4).
Rural areas, such as national parks, national forests, and state parks, tend to be less
directly affected by anthropogenic pollution sources than urban sites. However, they can be
regularly affected by transport of Os or Os precursors from upwind urban areas. In addition,
biogenic VOC emissions tend to be higher in rural areas, and major anthropogenic sources of Os
precursor emissions such as highways, power plants, biomass combustion, and oil and gas
operations are commonly found in rural areas, adding to the Os produced in these areas. Areas at
higher elevations, such as many of the national parks in the western U.S., can also be affected
more significantly by international transport of Os or stratospheric intrusions that transport Os
into the area (U.S. EPA, 2013, section 3.7.3).
2.3 ECOLOGICAL EFFECTS
Recent studies reviewed in the Os ISA support and strengthen the findings reported in the
2006 Os Air Quality Criteria Document (AQCD) (U.S. EPA, 2006). The most significant new
body of evidence since the 2006 Os AQCD comes from research on molecular mechanisms of
the biochemical and physiological changes observed in many plant species in response to Os
exposure. These newer molecular studies not only provide very important information regarding
2-5
-------
the many mechanisms of plant responses to Os, they also allow for the analysis of interactions
between various biochemical pathways that are induced in response to Os. However, many of
these studies have been conducted in artificial conditions with model plants, which are typically
exposed to very high, short doses of Os and are not quantifiable as part of this risk assessment.
Chapter 9 of the Os ISA (U.S. EPA, 2013) provides a detailed review of the effects of Os
on vegetation including the major pathways of exposure and known ecological and ecosystem
effects. In general, Os is taken up through the stomata into the leaves. Once inside the leaves, Os
affects a number of biological and physiological processes, including photosynthesis. This leads,
in some cases, to visible foliar injury as well as reduced plant growth, which are the main
ecological effects assessed in this review. Visible foliar injury and reduced growth can lead to a
reduction in ecosystem services, including crop and timber yield loss, decreased carbon
sequestration, alteration in community composition, and loss of recreational or cultural value.
Overall causal determinations are made based on the full range of evidence including
controlled exposure studies and field-based ecological studies. Figure 2-2 shows the Os welfare
effects that have been categorized by strength of evidence for causality in the Os ISA (U.S. EPA,
2013, Chapter 2). These determinations support causal or likely causal relationships between
exposure to Os and ecological and ecosystem-level effects.
Reduced carbon
sequestration in
terrestrial ecosystems
Alteration of
terrestrial ecosystem
water cycling
Alteration of
terrestrial community
composition
Not likely
Inadequate
to infer
Suggestive
Likely
Visible foliar injury
effects
Reduced vegetation growth
Reduced productivity in
terrestrial ecosystems
Reduced yield and
quality of agricultural
crops
Alteration of below
ground biogeochemical
cycles
Causal
Figure 2-2 Causal Determinations for Os Welfare Effects
2-6
-------
The adequate characterization of the effects of Os on plants for the purpose of setting air
quality standards depends not only on the choice of the index used (e.g., W126) to summarize Os
concentrations (Section 9.5 of the Os ISA), but also on quantifying the response of the plant
variables of interest at specific values of the selected index. The factors that determine the
response of plants to Os exposure include species, genotype and other genetic characteristics,
biochemical and physiological status, previous and current exposure to other stressors, and
characteristics of the exposure.
Quantitative characterization of exposure-response in the 2006 Os AQCD was based on
experimental data generated for projects conducted by the National Crop Loss Assessment
Network (NCLAN) and the EPA's National Health and Environmental Effects Research
Laboratory, Western Ecology Division (NHEERL-WED) that used open-top chambers (OTCs)
to expose crops and trees seedling to Os. In recent years, additional yield and growth results for
soybean and aspen, respectively, (two of the species that provided extensive exposure-response
information in those projects) have become available from studies that used free-air carbon
dioxide/ozone enrichment (FACE) technology, which is intended to provide conditions much
closer to natural environments (Pregitzer et al., 2008; Morgan et al., 2006; Morgan et al., 2004;
Dickson et al., 2000). The results of these FACE studies provided support for the earlier
findings reported in the OTC studies.
The quantitative exposure-response relationships described in the 2006 Os AQCD have
not changed in the Os ISA, with the exception of the addition of one new species. The
exposure-response models are summarized in the final Os ISA (U.S. EPA, 2013) and are
computed using the W126 metric, cumulated over 90 days. These response functions provide an
adequate basis for quantifying biomass loss damages.
Visible foliar injury resulting from exposure to Os has also been well characterized and
documented over several decades of research on many tree, shrub, herbaceous, and crop species
(U.S. EPA, 2006, 1996a, 1984, 1978). Ozone-induced visible foliar injury symptoms on certain
bioindicator plant species are considered diagnostic as they have been verified experimentally in
exposure-response studies, using exposure methodologies such as continuous stirred tank
reactors (CSTRs), OTCs, and free-air fumigation. Experimental evidence has clearly established
a consistent association of visible injury with Os exposure, with greater exposure often resulting
2-7
-------
in greater and more prevalent injury. This welfare risk and exposure assessment assesses the risk
of visible foliar injury at differing concentrations of Os using U.S. Forest Service biomonitoring
data along with soil moisture information.
2.4 ECOSYSTEM SERVICES
The Risk and Exposure Assessment conducted as part of the Review of the Secondary
National Ambient Air Quality Standards for Oxides of Nitrogen and Oxides of Sulfur evaluates
the benefits received from the resources and processes that are supplied by ecosystems.
Collectively, these benefits are known as ecosystem services and include products or provisions,
such as food and fiber; processes that regulate ecosystems, such as carbon sequestration; cultural
enrichment; and supportive processes for services, such as nutrient cycling. Ecosystem services
are distinct from other ecosystem products and functions because there is human demand for
these services. In the Millennium Ecosystem Assessment (MEA), ecosystem services are
classified into four main categories:
• Provisioning — includes products obtained from ecosystems, such as the production
of food and water.
• Regulating — includes benefits obtained from the regulation of ecosystem processes,
such as the control of climate and disease.
• Cultural — includes the nonmaterial benefits that people obtain from ecosystems
through spiritual enrichment, cognitive development, reflection, recreation, and
aesthetic experiences.
Supporting — includes those services necessary for the production of all other
ecosystem services, such as nutrient cycles and crop pollination (MEA, 2005).
The concept of ecosystem services can be used to help define adverse effects as they
pertain to NAAQS reviews. The most recent secondary NAAQS reviews have characterized
known or anticipated adverse effects to public welfare by assessing changes in ecosystem
structure or processes using a weight-of-evidence approach that includes both quantitative and
qualitative data. For example, the previous Os NAAQS review evaluated changes in foliar
injury, tree and crop growth loss, and biomass reduction in trees beyond the seedling stage. The
2-8
-------
presence or absence of foliar damage in counties meeting the existing standard has been used as
a way to evaluate the adequacy of the secondary NAAQS. Characterizing a known or
anticipated adverse effect to public welfare is an important component of developing any
secondary NAAQS. According to the Clean Air Act (CAA), welfare effects include the
following:
"Effects on soils, water, crops, vegetation, manmade materials, animals, wildlife,
weather, visibility, and climate, damage to and deterioration of property, and hazards to
transportation, as well as effect on economic values and on personal comfort and well-being,
whether caused by transformation, conversion, or combination with other air pollutants."
(Section 302(h))
In other words, welfare effects are those effects that are important to individuals and/or
society in general. Ecosystem services can be generally defined as the benefits that individuals
and organizations obtain from ecosystems. The EPA has defined ecological goods and services
as the "outputs of ecological functions or processes that directly or indirectly contribute to social
welfare or have the potential to do so in the future. Some outputs may be bought and sold, but
most are not marketed" (U.S. EPA, 2006). Conceptually, changes in ecosystem services may be
used to aid in characterizing a known or anticipated adverse effect to public welfare. In the
context of this review, ecosystem services may also aid in assessing the magnitude and
significance of a resource and in assessing how Os concentrations may impact that resource.
Figure 2-3 provides the World Resources Institute's schematic demonstrating the
connections between the categories of ecosystem services and human well-being (MEA, 2005).
The interrelatedness of these categories means that any one ecosystem may provide multiple
services. Changes in these services can impact human well-being by affecting security, health,
social relationships, and access to basic material goods (MEA, 2005). The strength of the
linkages, as indicated by arrow width, and the potential for mediation, as indicated by arrow
color, differ in different ecosystems and regions.
2-9
-------
ECOSYSTEM SERVICES
Provisioning
FOOD
FRESH WATER
WOOD AND FIBER
FUEL
Supporting
NUTRIENT CYCLING
SOIL FORMATION
PRIMARY PRODUCTION
Regulating
CLIMATE REGULATION
FLOOD REGULATION
DISEASE REGULATION
WATER PURIFICATION
Cultural
AESTHETIC
SPIRITUAL
EDUCATIONAL
RECREATIONAL
LIFE ON EARTH - BIODIVERSITY
CONSTITUENTS OF WELL-BEING
Security
PERSONAL SAFETY
SECURE RESOURCE ACCESS
SECURITY FROM DISASTERS
Basic material
for good life
ADEQUATE LIVELIHOODS
SUFFICIENT NUTRITIOUS FOOD
SHELTER
ACCESS TO GOODS
Health
STRENGTH
FEELING WELL
ACCESS TO CLEAN AIR
AND WATER
Good social relations
SOCIAL COHESION
MUTUAL RESPECT
ABILITY TO HELP OTHERS
Freedom
of choice
and action
OPPORTUNITY TO BE
ABLE TO ACHIEVE
WHAT AN INDIVIDUAL
VALUES DOING
AND BEING
Source: Millennium Ecosystem Assessment
Figure 2-3 Linkages Between Ecosystem Services Categories and Components of Human
Weil-Being
The ecosystems of interest in this welfare risk and exposure assessment are impacted by
the effects of anthropogenic air pollution, which may alter the services provided by the
ecosystems in question. For example, changes in forest conditions as a result of Os exposure
may affect supporting services such as net primary productivity; provisioning services such as
timber production; regulating services such as climate regulation; provisioning services such as
food; and cultural services such as recreation and ecotourism.
Where possible, we developed linkages to ecosystem services from indicators of each
effect identified in the Os ISA (U.S. EPA, 2013). These linkages were based on existing
literature and models, focus on the services identified in the peer-reviewed literature, and are
essential to any attempt to evaluate Os-induced changes on the quantity and/or quality of
ecosystem services provided. According to the EPA's Science Advisory Board Committee on
Valuing the Protection of Ecological Systems and Services, these linkages are critical elements
for determining the valuation of benefits of EPA-regulated air pollutants (SAB CVPESS, 2009).
2-10
-------
We have identified the primary ecosystem service(s) potentially impacted by Os for
major ecosystem types and components (i.e., terrestrial ecosystems, productivity) under
consideration in this risk and exposure assessment. The impacts associated with various
ecosystem services for each targeted effect are assessed in Chapters 5, 6, and 7 of this document
at a national scale and in the more refined case studies.
2-11
-------
3 SCOPE
This chapter provides an overview of the scope and key design elements of the welfare
risk and exposure assessment. The design of this assessment began with a review of the risk and
exposure assessments completed during the previous review of the National Ambient Air Quality
Standard for Ozone (Os NAAQS) (U.S. EPA, 2007), with an emphasis on considering key
limitations and sources of uncertainty recognized in that analysis.
In October 2008, as an initial step in the current Os NAAQS review, the Environmental
Protection Agency (EPA) invited outside experts, representing a broad range of expertise, to
participate in a workshop with EPA staff to help inform the EPA's plan for the review. The
participants discussed key policy-relevant issues that would frame the review, as well as the most
relevant new science that would be available to inform our understanding of these issues. One
workshop session focused on planning for quantitative risk and exposure assessments, taking
into consideration what new research and/or improved methodologies would be available to
inform the design of a quantitative welfare risk and exposure risk assessment. Based in part on
the workshop discussions, the EPA developed a draft Integrated Re view Plan for the Ozone
National Ambient Air Quality Standards (IRP) (U.S. EPA, 2009) outlining the schedule, process,
and key policy-relevant questions that would frame this review. On November 13, 2009, the
EPA held a consultation with the Clean Air Scientific Advisory Committee (CASAC) on the
draft IRP (74 FR 54562, October 22, 2009), which included opportunity for public comment.
The final IRP incorporated comments from CASAC (Samet, 2009) and the public on the draft
plan, as well as input from senior Agency managers. The final IRP included initial plans for the
quantitative risk and exposure assessments for both human health and welfare (U.S. EPA, 201 la,
chapters 5 and 6).
As a next step in the design of these quantitative assessments, the Office of Air Quality
Planning and Standards (OAQPS) staff developed more detailed planning documents, including
the following: Os National Ambient Air Quality Standards: Scope and Methods Plan for Health
Risk and Exposure Assessment (Health Scope and Methods Plan; U.S. EPA, 201 Ib) and Os
National Ambient Air Quality Standards: Scope and Methods Plan for Welfare Risk and
Exposure Assessment (Welfare Scope and Methods Plan, U.S. EPA, 201 Ic). These plans were
the subject of a May 19-20, 2011, consultation with CASAC (76 FR 23809, April 28, 2011).
3-1
-------
Based on consideration of CASAC (Samet, 2011) and public comments on these plans and
information in the second draft Integrated Science Assessment (ISA), we modified the scope and
design of the risk and exposure assessment and drafted a memo with updates to the information
presented in these plans (Wegman, 2012). We further modified the scope in response to
comments from CASAC on the first draft assessment (Frey and Samet, 2012a). These plans,
together with the update memo and comments from CASAC and the public, provide the basis for
the discussion of the scope of the risk and exposure assessment provided in this chapter.
Section 3.1 of this chapter provides a brief overview of the risk and exposure assessment
completed for the previous Os NAAQS review, including key limitations and uncertainties
associated with that analysis. Section 3.2 provides a summary of the design of the current
exposure assessment, including the ecosystem services framework, assessments for biomass loss
and visible foliar injury. Section 3.3 provides an overview of the uncertainty and variability
assessments.
3.1 OVERVIEW OF RISK AND EXPOSURE ASSESSMENTS FROM PREVIOUS
REVIEW
The assessments conducted as part of the previous review focused on national-level Os-
related impacts to sensitive vegetation and their associated ecosystems. The vegetation exposure
assessment was performed using an interpolation approach that included information from
ambient monitoring networks and results from air quality modeling. The vegetation risk
assessment included both tree and crop analyses. The tree risk analysis included three distinct
lines of evidence: (1) observations of visible foliar injury in the field linked to monitored Os air
quality for the years 2001 - 2004; (2) estimates of tree seedling growth loss under then current
and alternative Os exposure conditions; and (3) simulated mature tree growth reductions of
meeting alternative air quality standards on the predicted annual growth of mature trees from
three different species. The crop risk analysis included estimates of crop yields under current and
alternative Os exposure conditions. The EPA analyzed the associated changes in economic value
upon meeting the levels of various alternative standards using an agricultural sector economic
model. Key elements and observations from these risk and exposure assessments are outlined in
the following sections.
3-2
-------
3.1.1 Exposure Characterization
In many rural and remote areas where sensitive species of vegetation can occur,
monitoring coverage is limited. Thus, the 2007 Staff Paper (U.S. EPA, 2007) concluded that it
was necessary to use an interpolation method to better characterize Os concentrations over broad
geographic areas and at the national scale. Based on the significant difference in monitoring
network density between the eastern and western U.S., the 2007 Staff Paper further concluded
that it was appropriate to use separate interpolation techniques in these two regions. The EPA
used monitoring data for the eastern interpolation, and in the western U.S., where rural
monitoring is sparser, the EPA used the Community Multi-scale Air Quality (CMAQ) model
(http://www.epa.gov/asmdnerl/CMAQ, Byun and Ching, 1999; Byun and Schere, 2006) to
develop scaling factors to augment the monitor interpolation.
To evaluate changing vegetation exposures under selected air quality scenarios, the EPA
conducted a number of analyses. One analysis adjusted 2001 base year Os concentration
distributions using a rollback method (Rizzo, 2005, 2006) to reflect meeting the current and
alternative secondary standard options. For the "just meet" and alternative 8-hour average
standard scenarios, the EPA generated the associated maps of estimated 3-month, 12-hour, W126
exposures.1
A second analysis in the 2007 Staff Paper identified the overlap between different
alternative forms of the secondary standard. The analysis was designed to evaluate the extent to
which county-level Os concentrations measured in terms of various concentrations of the then
current 8-hour average form overlapped with concentrations measured in terms of the 3-month,
12-hour W126 cumulative, seasonal form. This analysis found that the number of counties
meeting either one or both of the standard forms depended greatly on the level of the forms
selected as well as the air quality pattern that exists in a particular year or set of years. Thus, the
2007 Staff Paper indicated that it remained uncertain as to the extent to which air quality
improvements designed to reduce 8-hour average Os concentrations would also reduce Os
exposures measured by a seasonal, cumulative W126 index. The 2007 Staff Paper stated this was
an important consideration because: (1) the biological database stresses the importance of
cumulative, seasonal exposures in determining plant response; (2) plants have not been
1 See Section 4.3.1 for more information regarding the W126 Cb exposure metric.
3-3
-------
specifically tested for the importance of daily maximum 8-hour Os concentrations in relation to
plant response; and (3) the effects of attainment of an 8-hour standard in upwind urban areas on
rural air quality distributions cannot be characterized with confidence because of the lack of
monitoring data in rural and remote areas.
3.1.2 Assessment of Risks to Vegetation
The risk assessments in the previous review reflected the availability of several lines of
evidence that provided a picture of the scope of Os-related vegetation risks for seedling, sapling
and mature tree species growing in field settings and, indirectly, for forested ecosystems. To
assess visible foliar injury, the 2007 Staff Paper presented an assessment that combined U.S.
Forest Service (USFS) Forest Inventory and Analysis (FIA) biomonitoring site data with the
county-level air quality data for those counties containing the FIA biomonitoring sites.
The EPA conducted separate assessments for seedlings and mature trees. To estimate
growth reductions in seedlings, the EPA used exposure-response (E-R) functions developed from
open- top chamber (OTC) studies for biomass loss for available seedling tree species and from
information on tree growing regions derived from the U.S. Department of Agriculture's (USD A)
Atlas of United States Trees. The E-R functions were then combined with projections of air
quality based on 2001 interpolated exposures. To estimate growth reductions in mature trees, the
EPA used a tree growth model (TREGRO) to evaluate the effect of changing Os concentration
scenarios from just meeting alternative Os standards on the growth of mature trees. TREGRO is
a process-based, individual tree growth simulation model (Weinstein et al, 1991) that is linked
with concurrent climate data to account for Os and climate/meteorology interactions on tree
growth. The model was run for a single western species (ponderosa pine) and two eastern species
(red maple and tulip poplar). These three species were chosen based on the availability of
species-specific parameterization in the model, their relative abundance in their respective
regions, and the importance of their associated ecosystem services.
To estimate yield loss in agricultural commodity, fruit and vegetable crops, the EPA
applied information from the National Crop Loss Assessment Network (NCLAN) program and a
1996 California fruit and vegetable analysis to develop E-R functions. The crop risk assessment,
like the tree seedling assessment, combined E-R information on nine commodity crops and six
3-4
-------
fruit and vegetable species with crop growing regions, and interpolated exposures during each
crop growing season.
The 2007 Staff Paper also presented estimates of economic valuation for crops associated
with the then current and alternative standards. The Agriculture Simulation Model (AGSIM)
(Taylor, 1993) was used to calculate annual average changes in total undiscounted economic
surplus for commodity crops and fruits and vegetables when then current and alternative
standard levels were met. The 2007 Staff Paper recognized that the modeled economic impacts
from AGSIM had many associated uncertainties, which limited the usefulness of these estimates.
3.2 OVERVIEW OF CURRENT ASSESSMENT PLAN
Since the 2008 Os NAAQS review, new scientific information on the direct and indirect
effects of Os on vegetation and ecosystems, respectively, has become available. With respect to
mature trees and forests, the information regarding Os impacts to forest ecosystems has
continued to expand, including limited new evidence that implicates Os as an indirect contributor
to decreases in stream flow resulting from direct impacts on whole tree-level water use. Recently
published results from the long-term Free-Air Carbon Dioxide Enrichment (FACE) studies
provide additional evidence regarding chronic Os exposures in forests, including decreased tree
heights, stem volumes (Kubiske et al., 2006), seed weight and seed germination (Darbah et
al., 2008, 2007); and changes in tree community structure (Kubiske et al., 2007). In addition, a
comparison, presented in the Os ISA (Section 9.6.3), using recent data from Aspen FACE found
that Os effects on biomass accumulation in aspen during the first seven years of the experiment
closely agreed with the E-R function based on data from earlier OTC experiments. In addition,
recent available data from annual field surveys conducted by the USFS to assess visible foliar
injury to selected tree species is available. In light of this more recent information, we are
updating the analysis that combines the USFS data with recent air quality data to determine the
incidence of visible foliar injury occurring across the U.S. at recent air quality concentrations and
have included new assessments that combine foliar injury information with soil moisture data.
One of the objectives of the risk assessment for a secondary NAAQS is to quantify the
risks to public welfare, including ecosystem services. For example, the Risk and Exposure
Assessment for Review of the Secondary National Ambient Air Quality Standards for Oxides of
3-5
-------
Nitrogen and Oxides of Sulfur (U.S. EPA, 2009) includes detailed discussions of how ecosystem
services and public welfare are related and how an ecosystem services framework may be
employed to evaluate effects on welfare. To the extent applicable, we provide qualitative and/or
quantitative assessments of ecosystem services impacted by Os to inform the current review. In
Chapter 5 of this assessment, we identify and describe the ecosystem services associated with the
ecological effects for which data and methods for incremental analysis of direct Os are not yet
available. For example, we overlay data on fire incidence, risk, and expenditures related to fires
in California with Os data to better characterize areas where Os may result in increased risks of
fires. Similarly, we also overlay data on bark beetle infestation with Os data. In chapters 6 and 7,
we identify and describe the ecosystem services associated with the ecological effects for
biomass loss and foliar injury, respectively, including national-scale assessments and more
refined case study areas.
3.2.1 Air Quality Considerations
Air quality information and analyses are used to inform and support welfare-related
assessments. The air quality information and analyses for this review build upon those in the Cb
ISA and include: (1) summaries of recent ambient air quality data; (2) application of a
methodology to extrapolate measured Os concentrations to areas without monitors, including
natural areas important to a welfare effects assessment such as national parks; and (3)
adjustments of air quality to just meet the existing standard and potential alternative W126
secondary standards. In this assessment, we use W126 as a shorthand for the maximum
consecutive 3-month, 12-hour daylight W126 index value. Consistent with the 2007 Staff Paper
(U.S. EPA, 2007) and CASAC recommendation (Henderson et al., 2007), the air quality analyses
in this assessment focus on the W126 metric. We provide more information regarding the air
quality analyses in Chapter 4.
3.2.1.1 Recent Ambient Data
In addition to updating air quality summaries from the previous review, these air quality
analyses include summaries of the recent ambient measurements for 2006 to 2010 for the form of
the existing standard (ppb) and a potential alternative form of secondary standard (W126). The
ambient measurements are from monitor data from the EPA's Air Quality System (AQS)
database (which includes National Park Service monitors) and the EPA's Clean Air Status and
3-6
-------
Trends Network (CASTNET) network. Since the previous review, the extent of monitoring
coverage in non-urban areas has not significantly changed, and the monitoring network in some
locations of the Western U.S. is sparse. We provide more information regarding the air quality
analyses for recent ambient data in section 4.3.2.
3.2.1.2 National Os Exposure Surfaces for Recent Conditions
National-scale Os surfaces are used as inputs to the vegetation exposure and risk
assessments described in subsequent sections. To estimate Os exposure in areas without
monitors, particularly those gaps left by a sparse rural monitoring network in the western U.S.,
we used a spatial interpolation technique, called Voronoi Neighbor Averaging (VNA), (Gold,
1997; Chen et al., 2004) to create an air quality surface for the contiguous U.S. at a 12 kilometer
grid resolution. We created annual W126 surfaces for each year from 2006 to 2010 and for a 3-
year average for 2006-2008. We provide more information regarding the recent W126 exposure
surfaces in section 4.3.1.
3.2.1.3 Adjustments to Just Meet Existing and Alternative Standards
The vegetation exposure assessments also rely on recent Os concentrations adjusted to
just meet the existing standard and potential alternative secondary standards. All adjustments
were made to monitored values. New VNA surfaces were then created from the adjusted
monitored values. These surfaces are used in several vegetation assessments, including the
geographic analysis for fire risk and bark beetle damage, the national- and case-study scale
biomass loss assessments, and the national park case studies for foliar injury.
First, we adjusted hourly Os concentrations for recent conditions (2006-2008) to just
meet the existing standard at 75 ppb. These hourly Os concentrations at monitor locations were
then aggregated to the 3-year average of the W126 metric and compared against three potential
alternative secondary standard levels of 15, 11, and 7 ppm-hrs. We selected these potential
alternative standard levels for analysis in this WREA because CASAC recommended and
supported a range of potential alternative W126 standard levels from 7 to 15 ppm-hrs during the
previous review. In regions of the country for which the 75 ppb adjustment case left some
monitors above the secondary standard level being evaluated, hourly Os was further adjusted to
meet alternative W126 standard levels. In other words, these surfaces assume that the existing
3-7
-------
standard is met prior to adjustments to meet alternative standards. We describe the adjustment
process in detail in section 4.3.2.
3.2.2 Relative Tree Biomass Loss and Crop Yield Loss
3.2.2.1 National-Scale Assessment: Exposure-Response Functions for
Tree Seedlings and Crops
In the 2007 Staff Paper, the EPA derived information on tree species growing regions
from the USDA Atlas of United States Trees (Little, 1971). In this assessment, we use more
recent information (2006-2008) from the USFS Forest Health Technology Enterprise Team
(FHTET) to update growing ranges for the 12 tree species studied by National Health and
Environmental Effects Research Laboratory, Western Ecology Division (NHEERL-WED). We
combine the national O3 surface with seedling E-R functions for each of the tree species and
information on each tree species growing region to produce estimates of OS-induced seedling
biomass loss for each of the 12 tree species. From this information, we generate GIS maps
depicting seedling biomass loss for each species for each air quality scenario. For crops, we
estimate yield loss for each of the 10 crop species from NCLAN. This analysis enabled direct
evaluation of estimated seedling biomass loss for trees and yield loss for crops expected to occur
under air quality exposure scenarios expressed in terms of recent air quality and, after adjusting
to just meet the existing standard and potential alternative secondary standards. In addition, this
assessment can be used to determine the W126 benchmark values associated with 1 to 2 percent
seedling biomass loss for trees and 5 percent yield loss for crops. For biomass loss, CASAC
recommended that the EPA should consider options for W126 standard levels based on factors
including a predicted 1 to 2 percent biomass loss for trees and a predicted 5 percent loss of crop
yield. Small losses for trees on a yearly basis compound over time and can result in substantial
biomass losses over the decades-long lifespan of a tree (Frey and Samet, 2012b).
3.2.2.2 National-Scale Assessment: National Weighted RBL and Class I
Areas
To assess overall ecosystem-level effects from biomass loss, we used FHTET data for
modeled predictions of stand density and basal area. The resolution of the FHTET data is 1,000
3-S
-------
square meter grids, and we summed these data into the larger CMAQ grid cells (12 km x 12 km).
For the individual species analyses, these data were used only as a predictor of presence or
absence. In the ecosystem-level analysis, these data were used to scale the biomass loss by the
proportion of total basal area for each species. We combined the RBL values for 12 tree species
into a weighted RBL rate and considered the weighted value in relation to proportion of basal
area covered (as measured by proportion of geographic area with available data on species). A
weighted RBL value is a relatively straightforward metric to attempt to understand the potential
ecological effect on some ecosystem services. We provide more information regarding the
individual species analysis in section 6.2.1.3 and the combined analysis in 6.8.
We also analyzed federally designated Class I areas in relation to the W126 surface and
the weighted RBL values in the same manner as the analyses across the entire range of data. Out
of 156 Class I areas nation-wide, 145 Class I areas had tree data available for this analysis. This
analysis was conducted for air quality exposure scenarios expressed in terms of recent air quality
(2006-2008) and after adjusting to just meeting the existing standard and potential alternative
secondary standards. We provide more information regarding this analysis in section 6.8.1.1.
3.2.2.3 National-Scale Assessment: Ecosystem Services
The national-level ecosystem services quantified in this review associated with biomass
and yield loss include provisioning services (e.g., timber and crops) and regulating services (e.g.,
carbon sequestration). Where information is available, we describe the impacts on other
ecosystem services such as impacts on biodiversity, biological community composition, health of
forest ecosystems, aesthetic values of trees and plants, and the nutritive quality of forage crops.
We also describe the cultural ecosystem services associated with non-timber forest products. In
addition, there is new preliminary evidence that Cb adversely affects the ability of pollinators to
find their targets, which could have broad implications for agriculture, horticulture, and forestry.
We use the Forest and Agricultural Sector Optimization Model with Greenhouse Gases
(FASOMGHG) model (Adams et al., 2005) to estimate Os impacts on the agriculture and
forestry sectors and quantify how Os exposure to vegetation affects the provision of timber and
crops and carbon sequestration. FASOM, including the GHG version, has been used recently in
many evaluations of effects of climate change on the timber and agriculture market sectors, in
part because it accounts for the tradeoffs between land use for forestry and agriculture.
3-9
-------
Specifically, FASOMGHG is a dynamic, non-linear programming model of the U.S. forest and
agricultural sectors. The EPA uses this model to evaluate welfare benefits and market effects of
Os-induced biomass loss in trees and of carbon sequestration in trees, understory, forest floor,
wood products and landfills that would occur under different agricultural and forestry scenarios.
Using this model, we calculate the economic impacts of yield changes between recent ambient
Os conditions and after adjusting to just meet the existing 75 ppb standard and alternative W126
standards.
3.2.2.4 Case Study Areas: Five Urban Areas
In selecting urban case study areas for more in-depth analysis of the ecosystem services
associated with urban tree biomass loss, we relied on several criteria:
• Areas expected to have elevated W126 index values where ecological effects might
be expected to occur.
• Occurrence of Os-sensitive tree species and/or species for which Os E-R curves have
been generated.
• Availability of vegetation information in the case study area.
• Geographic coverage representing a cross section of the nation, including urban and
natural settings.
We use the i-Tree model to assess effects on regulating ecosystem services provided by
urban forests, including pollution removal and carbon storage and sequestration for the case
study areas. The i-Tree model is a publicly available, peer-reviewed software suite developed by
the USFS and its partners to assess the ecosystem service impacts of urban forestry (available
here: http://www.itreetools.org/). We collaborated with the USFS to vary the tree growth metric
in the model, which allows us to assess the effects of Os exposure on the ability of the forests in
the selected case study area to provide the services enumerated by the model. Specifically, we
estimate impacts on vegetation in Atlanta, Baltimore, Syracuse, the Chicago region, and the
urban areas of Tennessee. We present results for model runs representing recent ambient Os
conditions, just meeting the existing 75 ppb standard, and just meeting alternative W126
standards.
5-10
-------
3.2.3 Visible Foliar Injury
3.2.3.1 National Analysis of Visible Foliar Injury
To assess visible foliar injury (hereafter referred to as foliar injury) at a national scale, we
compared data from the USFS Forest Health Monitoring Network (USFS, 2011) with Os
exposure estimates and soil moisture data for 2006-2010. For estimates of short-term soil
moisture in the contiguous U.S., we use NOAA's Palmer Z drought index (NCDC, 2012b).
Foliar injury sampling data were not available for several western states (Montana, Idaho,
Wyoming, Nevada, Utah, Colorado, Arizona, New Mexico, Oklahoma, and portions of Texas).
This analysis provides estimates of the presence and absence of foliar injury for each of the 5
years by soil moisture category, which provides insight into the degree of protection that drought
provides from foliar injury. Using this analysis, we derived multiple W126 benchmarks for
evaluating foliar injury at national parks in a screening-level assessment and three case studies.
3.2.3.2 National-Scale Screening-level Assessment of Visible Foliar
Injury in 214 National Parks
A study by Kohut (2007) assessed the risk of Os-induced foliar injury on Os-sensitive
vegetation in 244 parks managed by the National Park Service (NPS). We modified this
screening-level assessment to use more recent Os exposure and soil moisture data and to
incorporate benchmarks derived from the national-scale foliar injury analysis (described above in
section 3.2.3.1). Specifically, we use Os monitoring data to create spatial surfaces of Os exposure
and short-term soil moisture data (Palmer Z) (NCDC, 2012b) for each year from 2006 to 2010.
These data reflect the contiguous U.S. only, which is a key reason why this assessment includes
fewer parks than Kohut (2007). Overall, the screening-level assessment includes 42 parks with
Os monitors and 214 parks with Os exposure estimated from the interpolated Os surface. We
combine these data with lists from the NPS of the parks containing Os-sensitive vegetation
species (NPS, 2003, 2006). Consistent with Kohut (2007), we consider the results for these parks
without identified species as potential until sensitive species are identified in field surveys at
these parks.
Using the results of the national-scale foliar injury analysis, we derived five W126
benchmark scenarios for evaluating foliar injury risk at parks in this screening-level assessment.
One scenario reflects Os exposure only, and four scenarios reflect Os exposure and soil moisture
3-11
-------
jointly for different percentages of biosites with injury. For each of these scenarios, we identify
the number of parks that exceed the benchmark criteria in each year.
3.2.3.3 National-Scale Assessment: Ecosystem Services
We use GIS mapping developed for the ecological effects analysis to illustrate where
foliar injury may be occurring, and we cross reference those areas to national statistics for
recreational use available through the National Survey of Fishing, Hunting, and Wildlife-
Associated Recreation (U.S. DOT, 2011) and the National Survey on Recreation and the
Environment (USDA, 2002). We also scale the resulting estimates of cultural service provision
to the current population and values assigned using existing meta-data on willingness-to-pay
from the Recreation Values Database.2 We understand that these estimates are limited to current
levels of service provision and provide a snapshot of the overall magnitude of services
potentially affected by Os exposure. Currently, estimates of service loss from recent Os exposure
are beyond the available data and resources, as is the calculation of changes in ecosystem
services that might result from meeting existing and alternative Os standards. However, the
current losses in service from Os exposure are embedded in estimates of the current level of
services.
3.2.3.4 Case Study Analysis: Three National Parks
In selecting case study areas for more in-depth analysis of the ecosystem services
associated with foliar injury, we relied on several criteria:
• Areas expected to have elevated W126 index values where ecological effects might
be expected to occur.
• Availability of vegetation mapping, including estimates of species cover.
• Geographic coverage representing a cross section of the nation, including urban and
natural settings.
• Occurrence of Os sensitive species and/or species for which Os E-R curves have been
generated.
: Available at: http://recvaluation.forestry.oregonstate.edu/.
3-12
-------
We selected Great Smoky Mountains National Park, Rocky Mountain National Park, and
Sequoia/Kings National Park. All three of these park units are in areas with elevated ambient
W126 index values, have vegetation maps, and have species that are considered Os sensitive. We
considered including Acadia National Park, but we determined it did not fit our selection criteria.
Specifically, Acadia did not have detailed vegetation mapping comparable to the selected parks,
and the W126 index values were all well below 15 ppm-hrs. Using GIS, we compare the NFS
vegetation maps to the national Os surface to illustrate where foliar injury may be occurring,
particularly with respect to park amenities such as trails. Ecological metrics quantified for each
park include:
• Percent of vegetation cover affected by foliar injury.
• Percent of trail length affected by foliar injury.
In national parks, foliar injury affects primarily cultural values that include existence,
bequest and recreational values. In addition, we describe the other non-use values associated
with national parks including existence and bequest values. We also provide park-specific
statistics for recreational use available and estimates of service provision values using existing
meta-data on willingness-to-pay from Kaval and Loomis (2003). We understand that these
estimates are limited to current levels of service provision. Estimates of service loss due to Os
exposure are beyond the available data and/or resources for many if not all ecosystem services
listed above.
3.3 UNCERTAINTY AND VARIABILITY
An important issue associated with any ecological risk assessment is the characterization
of uncertainty and variability. Variability refers to the heterogeneity in a variable of interest that
is inherent and cannot be reduced through further research. For example, there may be variability
among E-R functions describing the relationship between Os and vegetation injury across
selected study areas. This variability may be due to differences in ecosystems (e.g., species
diversity, habitat heterogeneity, and rainfall), concentrations and distributions of Os and/or co-
pollutants, and/or other factors that vary either within or across ecosystems.
Uncertainty refers to the lack of knowledge regarding both the actual values of model
input variables (parameter uncertainty) and the physical systems or relationships (model
3-13
-------
uncertainty - e.g., the shapes of E-R functions). In any risk assessment, uncertainty is, ideally,
reduced by the maximum extent practical, through improved measurement of key parameters and
ongoing model refinement. However, significant uncertainty often remains, and emphasis is then
placed on characterizing the nature of that uncertainty and its impact on risk estimates. The
characterization of uncertainty can include both qualitative and quantitative analyses, the latter
requiring more detailed information and, often, the application of sophisticated analytical
techniques. Sources of variability that are not fully reflected in the risk assessment can
consequently introduce uncertainty into the analysis.
The goal in designing a quantitative risk assessment is to reduce uncertainty to the extent
practical and to incorporate the sources of variability into the analysis approach to insure that the
risk estimates are representative of the actual response of an ecosystem (including the
distribution of that adverse response across the ecosystem). An additional aspect of variability
that is pertinent to this risk assessment is the degree to which the set of selected case study areas
provide coverage for the range of Os-related ecological risk across the U.S.
Recent guidance from the World Health Organization (WHO, 2008) presents a four-
tiered approach for characterizing uncertainty. With this four-tiered approach, the WHO
framework provides a means for systematically linking the characterization of uncertainty to the
sophistication of the underlying risk assessment, where the decision to proceed to the next tier is
based on the outcome of the previous tier's assessment. Ultimately, the decision as to which tier
of uncertainty characterization to include in a risk assessment will depend both on the overall
sophistication of the risk assessment and the availability of information for characterizing the
various sources of uncertainty. We used the WHO guidance as a framework for developing the
approach used for characterizing uncertainly in this assessment. The four tiers described in the
WHO guidance include:
• Tier 0: recommended for routine screening assessments, uses default uncertainty
factors (rather than developing site-specific uncertainty characterizations);
• Tier 1: the lowest level of site-specific uncertainty characterization, involves
qualitative characterization of sources of uncertainty (e.g., a qualitative assessment of
the general magnitude and direction of the effect on risk results);
5-14
-------
• Tier 2: site-specific deterministic quantitative analysis involving sensitivity analysis,
interval-based assessment, and possibly probability bounded (high-and low-end)
assessment; and
• Tier 3: uses probabilistic methods to characterize the effects on risk estimates of
sources of uncertainty, individually and combined.
In this assessment, we applied a variety of quantitative (WHO Tier 2) and qualitative
(WHO Tierl) analyses to address uncertainty and variability in this assessment of Os-related
ecological risks. In general, we attempted to quantify uncertainty and variability where we had
sufficient data to do so and addressed these aspects qualitatively where we did not have data.
Several analyses in this assessment include quantitative assessments of uncertainty and
variability. In the air quality analyses, we quantified the standard errors associated with using
regressions to relate modeled Os responses to Os concentrations at various locations and times of
day, as well as during different seasons. For the analysis of the alternative percentages of
biomass and yield loss, we plotted the E-R relationship for 54 crop studies and 52 tree seedling
studies to estimate the differences in within-species variability. We also qualitatively compared
the uncertainty in the relationship between E-R functions for tree seedlings and the effects on
adult trees. For the screening-level assessment of foliar injury, we conducted several quantitative
sensitivity analyses, including five scenarios reflecting consideration of soil moisture, three
approaches for estimating Os exposure at monitored parks, three durations for soil moisture data,
and two time periods evaluating different years of analysis. We provide detailed tables
characterizing the uncertainty inherent in the various risk and exposure analyses at the end of
Chapters 4, 5, 6, and 7.
5-15
-------
4 AIR QUALITY CONSIDERATIONS
4.1 INTRODUCTION
Air quality information is used to assess exposures and ecological risks for national-scale
air quality surfaces generated to estimate 2006-2008l average concentrations based on the W126
exposure metric, which is defined later in this chapter. These national-scale air quality surfaces
are generated for five air quality scenarios by the methodology summarized in Sections 4.3.1 and
4.3.4 below. The five scenarios are for recent air quality, air quality adjusted to just meet the
current standard, and air quality further adjusted to just meet three different W126 index values:
15 ppm-hrs, 11 ppm-hrs, and 7 ppm-hrs. Additional national-scale air quality surfaces are
generated using observed W126 concentrations for individual years from 2006-2010. This chapter
describes the air quality information used in these analyses, providing an overview of monitoring
data and air quality (Section 4.2), and an overview of air quality inputs to the welfare risk and
exposure assessments (Section 4.3).
4.2 OVERVIEW OF O3 MONITORING AND AIR QUALITY
To monitor compliance with the NAAQS, state and local environmental agencies operate
Os monitoring sites at various locations, depending on the population of the area and typical peak
Os concentrations. In 2010, there were over 1,300 state, local, and tribal Os monitors reporting
concentrations to EPA. The minimum number of Os monitors required in a Metropolitan
Statistical Area (MSA) ranges from zero, for areas with a population under 350,000 and with no
recent history of an Os design value greater than 85 percent of the NAAQS, to four, for areas with
a population greater than 10 million and an Os design value greater than 85percent of the NAAQS.2
In areas for which Os monitors are required, at least one site must be designed to record the
maximum concentration for that particular metropolitan area (US EPA, 2013, Sections 3.5.6.1 and
3.7.4). Since Os concentrations are usually significantly lower in the colder months of the year,
1 The focus was placed on the years of 2006-2008 based on availability of data during that time period.
2The existing monitoring network requirements (40 CFR Part 58, Appendix D) have an urban focus and do not
address siting in non-urban (rural) areas. States may operate ozone monitors in non-urban (rural) areas to meet other
objectives (e.g., support for research studies of atmospheric chemistry or ecosystem impacts).
4-1
-------
Os is required to be monitored only during the required Ch monitoring season, which varies by
state (US EPA, 2013, Figure 3-24).3
While the existing U.S. Cb monitoring network has a largely urban focus, to address
ecosystem impacts of Os such as biomass loss and foliar injury, it is equally important to focus on
Os monitoring in rural areas. Figure 4-1 shows the location of all U.S. Cb monitors operating
during the 2006-2010 period. The gray dots which make up over 80 percent of the Os monitoring
network are "State and Local Monitoring Stations" (SLAMS) monitors which are largely operated
by state and local governments to meet regulatory requirements and provide air quality information
to public health agencies, and thus are largely focused on urban areas. The blue dots highlight two
important subsets of the SLAMS network: "National Core" (NCore) multipollutant monitoring
sites, and the "Photochemical Assessment Monitoring Stations" (PAMS) network.
The green dots represent the Clean Air Status and Trends Network (CASTNET) monitors
which are focused on rural areas. There were about 80 CASTNET sites operating in 2010, with
sites in the Eastern U.S. being operated by EPA and sites in the Western U.S. being operated by
the National Park Service (NPS). Finally, the black dots represent "Special Purpose Monitoring
Stations" (SPMS), which include about 20 rural monitors as part of the "Portable Os Monitoring
System" (POMS) network operated by the NPS. Between the CASTNET, NCore, and POMS
networks, there were about 120 rural Os monitoring sites in the U.S. in 2010.
3 Some States and Territories are required to operate ozone monitors year-round, including Arizona, California,
Hawaii, Louisiana, Nevada, New Mexico, Puerto Rico, Texas, American Samoa, Guam and the Virgin Islands.
4-2
-------
* SLAMS
• CASTNET
• NCORE/PAMS
• SPMS/OTHER
Figure 4-1 Map of U.S. ambient Os monitoring sites in operation during 2006-2010
To determine whether or not the NAAQS have been met at an ambient Os monitoring site,
a statistic commonly referred to as a "design value" must be calculated based on 3 consecutive
years of data collected from that site. The form of the existing Os NAAQS design value statistic
is the 3-year average of the annual 4th highest daily maximum 8-hour average Os concentration in
parts per billion (ppb), with decimal digits truncated. The existing primary and secondary Os
NAAQS are met at an ambient monitoring site when the design value is less than or equal to 75
ppb.4 Figure 4-2 shows the design values for the existing 8-hour Os NAAQS for all regulatory
monitoring sites in the U.S. for the 2006-2008 period. Monitors shown as red dots had design
values above the existing Os NAAQS of 75 ppb in 2006-2008.
4For more details on the data handling procedures used to calculate design values for the existing O3 NAAQS, see
40 CFR Part 50, Appendix P.
4-3
-------
61-65 ppb (65 Sites)
O 66-70 ppb (140 Sites)
O 71-75 ppb (279 Sites)
• 76-120 ppb (518 Sites)
Figure 4-2 Map of monitored 8-hour Os design values for the 2006-2008 period
4.3 OVERVIEW OF AIR QUALITY INPUTS TO RISK AND EXPOSURE
ASSESSMENTS
In this section, we summarize the air quality inputs for the welfare risk and exposure
assessments, and discuss the methodology used to adjust air quality to meet the existing standard
and potential alternative standards. These steps are summarized in the flowchart in Figure 4-3 and
discussed in more detail in this section.
Section 4.3.1 describes the W126 metric upon which the potential alternative standards are
based. Section 4.3.2 describes the ambient air quality monitoring data used in the welfare risk and
exposure assessments. Section 4.3.3 describes the procedure used to generate the national-scale
4-4
-------
air quality surfaces upon which several of the welfare risk and exposure analyses are based, with
further details in Appendix 4a. Finally, Section 4.4.4 summarizes the method used to adjust
observed air quality concentrations to just meet the existing standard and potential alternative
standards, and discusses the resulting distributions of adjusted W126 concentrations.
Check if a
Monitors meet
targeted
W126 alternative
standard
Roll back to targeted
W126alternative
standard
1
Aggregate hourlyOS
toW126ateach
monitor
Hourly rolled backOS
Values at all monitors
In region
Figure 4-3
Flowchart of air quality data processing for different parts of the welfare
risk and exposure assessments.
4.3.1 Air Quality Metrics
EPA focused the analyses in the welfare risk and exposure assessments on the W126 Os
exposure metric. The W126 metric is a seasonal aggregate of hourly Os concentrations, designed
to measure the cumulative effects of Os exposure on vulnerable plant and tree species, with units
in parts per million-hours (ppm-hrs). The metric uses a logistic weighting function to place less
emphasis on exposure to low hourly Os concentrations and more emphasis on exposure to high
hourly Os concentrations (Lefohn et al, 1988).
4-5
-------
The first step in calculating W126 concentrations was to sum the weighted hourly Os
concentrations within each calendar month, resulting in monthly index values. Since plant and
tree species are not photosynthetically active during nighttime hours, only Os concentrations
observed during daytime hours (defined as 8:00 AM to 8:00 PM local time) were included in the
summations. The monthly W126 index values were calculated from the hourly Os concentration
data as follows:
N 19
Monthly W126 = Y Y —,
y /-l/-l\+ 4403 * exp(-126 * Cdh)
d=lh=8 ^V atlJ
where TV is the number of days in the month,
d is the day of the month (d = 1, 2, ..., N),
h is the hour of the day (h = 0, 1, ..., 23),
Cdh is the hourly Os concentration observed on day J, hour /z, in parts per million.
Next, the monthly W126 index values were adjusted for missing data. IfNm is defined as
the number of daytime Os concentrations observed during month m (i.e. the number of terms in
the monthly index summation), then the monthly data completeness rate is Vm = Nm /12 * N. The
monthly index values were adjusted by dividing them by their respective Vm. Monthly index values
were not computed if the monthly data completeness rate was less than 75 percent (Vm < 0.75).
Finally, the annual W126 index values were computed as the maximum sum of their
respective adjusted monthly index values occurring in three consecutive months (i.e., January-
March, February-April, etc.). Three-month periods spanning across two years (i.e., November-
January, December-February) were not considered, because the seasonal nature of Os makes it
unlikely for the maximum values to occur at that time of year. The annual W126 concentrations
were considered valid if the data met the annual data completeness requirements for the existing
standard. Three-year W126 index values are calculated by taking the average of annual W126
index values in the same three-month period in three consecutive years.5
5 W126 calculations are slightly modified in the case of the model adjustment scenarios described in Section 4.3.4.
When calculating W126 for the model adjustment cases, we first found the three-year average of each three-month
period, and then selected the three-month period with the highest three-year average using the same three-month
period for each of the three years.
4-6
-------
4.3.2 Ambient Air Quality Measurements
Air quality monitoring data from 1,468 U.S. ambient Os monitoring sites were retrieved
for use in the risk and exposure assessments. The initial dataset was the same as the one used for
the health REA (HREA), which consisted of hourly Os concentrations in ppb collected between
1/1/2006 and 12/31/2010 from these monitors. Data for nearly 1,400 of these monitors were
extracted from EPA's Air Quality System (AQS) database6, while the remaining data came from
EPA's Clean Air Status and Trends Network (CASTNET) database which consists of primarily
rural monitoring sites.
Observations flagged in AQS as having been affected by exceptional events were included
the initial dataset, but were not used in design value calculations in accordance with EPA's
exceptional events policy. Missing data intervals of 1 or 2 hours in the initial dataset were filled
in using linear interpolation. These short gaps often occur at regular intervals in the ambient data
due to an EPA requirement for monitoring agencies to perform routine quality control checks on
their Os monitors. Quality control checks are typically performed between midnight and 6:00 AM
when Os concentrations are low. Missing data intervals of 3 hours or more were not replaced, and
interpolated data values were not used in design values calculations.
Annual W126 concentrations were calculated from the ambient data for each year in the
2006-2010 period, as well as 3-year averages of the 2006-2008 annual W126 concentrations.
Figure 4-4 shows the 2006-2008 average W126 concentrations in ppm-hrs at all monitoring sites
in the contiguous U.S. Monitors outside of the contiguous U.S. were not included in the welfare
analyses since they fell outside of the CMAQ 12 km modeling domain, and were already well
below the existing and potential alternative standards.
6 EPA's Air Quality System (AQS) database is a national repository for many types of air quality and related
monitoring data. AQS contains monitoring data for the six criteria pollutants dating back to the 1970's, as well as
more recent additions such as PM2 5 speciation, air toxics, and meteorology data. At present, AQS receives hourly O3
monitoring data collected from nearly 1,400 monitors operated by over 100 state, local, and tribal air quality
monitoring agencies.
4-7
-------
Figure 4-4
5 10 15 20 25
Monitored 2006-2008 average W126 concentrations in ppm-hrs
30+
4.3.3 National-scale Air Quality Surfaces for Recent Air Quality
In addition to ambient monitoring data, the welfare risk and exposure assessments analyzed
national-scale air quality surfaces. For the biomass loss analyses presented in Chapter 6, a
national-scale surface was generated from the monitored 2006-2008 average W126 concentrations
using the Voronoi Neighbor Averaging (VNA) technique (Gold, 1997; Chen et al, 2004) (Figure
4-5). For the foliar injury analysis presented in Chapter 7, national-scale surfaces were generated
from the monitored annual W126 concentrations for individual years 2006-2010, also using VNA.
Maps of the annual W126 air quality surfaces for 2006-2010 are included in Appendix 4-A.
4-8
-------
Figure 4-5 National surface of observed 2006-2008 average W126 concentrations, in
ppm-hrs
In the 1st draft of the welfare REA (WREA), the national-scale air quality surfaces were
created by "fusing" monitored 2006-2008 average W126 concentrations with annual W126
concentrations from a 2007 CMAQ model simulation, using the enhanced Voronoi Neighbor
Averaging (eVNA) technique (Timin et al., 2010). The resulting surfaces contained estimates of
the 2006-2008 average annual W126 concentrations at a 12km grid cell resolution in the
contiguous U.S. modeling domain. Here, the air quality surfaces of the 2006-2008 average W126
concentrations are based solely on monitored W126 concentrations and do not include CMAQ
model predictions. The reason for this change from the first draft WREA is discussed below.
In addition to the VNA methodology, two alternative methods for creating the national-
scale air quality surfaces were also considered: eVNA and Downscaler (Berrocal et al, 2012; used
in the HREA). Both the eVNA and Downscaler methods were tested using updated 2007 12km
4-9
-------
CMAQ modeling7 that is described in detail in Appendix 4-B of the HREA. While each of the
three methods had its own advantages and disadvantages, the VNA method was ultimately selected
because large differences between the modeled W126 surface and the monitored W126
concentrations8 made the two "data fusion" methods more uncertain in some instances, whereas
VNA did not suffer from this problem since it is based solely on monitored values. Technical
justification for the change from eVNA to VNA, including a cross-validation analysis, and
comparisons between the resulting air quality surfaces for these three methods, can be found in
Appendix 4-A.
4.3.4 Air Quality Adjustments to Meet Existing Primary and Potential
Alternative Secondary Os Standards
In addition to observed W126 levels, the risk and exposure assessments also consider the
relative change in risk and exposure after adjusting air quality to just meet the existing Os standard
of 75 ppb, and further adjusting air quality to just meet possible alternative standards with forms
based on the W126 metric and levels of 15 ppm-hrs, 11 ppm-hrs, and 7 ppm-hrs. The sections
below summarize the methodology used to adjust observed air quality concentrations to just meet
the existing standard and potential alternative standards, and discuss the resulting adjusted
distributions of W126 concentrations. More details on these inputs are provided in Appendix 4A.
4.3.4.1 Adjustment Methods
The model-based HDDM Os adjustment approach used for this analysis is the same general
methodology developed for evaluating air quality distributions that could occur if meeting various
alternate levels of the primary standard. This methodology is described in detail in Chapter 4 and
Appendix 4D of the HREA and additional details on HDDM itself are also provided in Appendix
4D of the HREA. A brief description of the HDDM adjustment process and key differences from
the HREA are provided in this section.
The first step of the HDDM adjustment technique is to obtain modeled sensitivities
(responses) of Os concentrations to perturbations in U.S. anthropogenic NOx emissions. Monitor,
7 The updated CMAQ modeling used wildfire emissions based on a multi-year average instead of 2007-specific
wildfires.
8 The 2007 CMAQ simulation over-predicted W126 values by an average of 4 ppm-hrs in monitored locations. A
more in depth model evaluation of CMAQ W126 values is provided in Appendix 4A.
4-10
-------
season and hour-of-day specific relationships between sensitivities and modeled ozone
concentrations were developed based on the modeling data. These responses are then applied to
ambient data to create a 3-year time-series of hourly Os concentrations at each monitor location
which is consistent with meeting various potential levels of the ozone NAAQS for 2006-2008.
There are a few key differences between the adjustments made in the HREA and those
performed here. First, the adjustments in HREA focused on 15 urban study areas while those used
in the WREA cover all monitoring sites across the US. In the HREA, a uniform reduction of U.S.
anthropogenic emissions was applied to all sites within an urban area. By applying equal
proportional decreases in emissions throughout the contiguous U.S., we were able to estimate how
hourly Os concentrations would respond to changes in ambient NOx concentrations without
simulating a specific control strategy. Note that the HDDM-adjustment approach was not designed
to produce an optimal control scenario but instead aimed to characterize a potential distribution of
air quality across a region when all monitors are meeting the existing standard and potential
alternative standards. In contrast to the 15 study area analyses performed for the HREA, both the
ecosystem services analysis (Chapter 5) and biomass loss analysis (Chapter 6) require nationally
consistent surfaces of W126 values as inputs. To create these surfaces, we balanced the need for
nationally consistent surfaces and the regional nature of W126 values with the potential scale over
which the secondary standard might be evaluated. If considering two potential bounding scenarios,
d using a single emissions reduction scenario for the entire U.S. could overstate the amount of
NOx reductions needed to just meet alternative W126 levels in different regions of the U.S., while
creating numerous distinct locally adjusted areas could lead to a patchwork of disjointed W126
surfaces. Consequently in this analysis, we determined the level of U.S. NOx emissions reduction
that would result in Os just meeting the potential alternative W126 levels independently for nine
distinct regions of the contiguous U.S. (Figure 4-6) based on the National Oceanic and
Atmospheric Administration (NOAA) climate regions (Karl and Koss, 1984). NOAA
characterizes each region as being "climatically consistent" and routinely uses these regions to
describe regional climate trends. These regions were deemed an appropriate delineation for this
analysis since geographic patterns of both Os and plant species are driven by climatic features such
as temperature and precipitation and because they broadly align with distinct emissions regions
(for instance the central region contains a greater density of the nation's coal-fired power plants
while the northeast region contains the highly urbanized northeast corridor). The regions are large
4-11
-------
enough to account for pollutant transport but not so large as to unrealistically include the impacts
of emissions reductions at locations far from the monitors in the region with the highest W126
values.
Analogous to the procedure used in the HREA for the urban study areas, a single NOx
emissions adjustment was used to adjust ambient air quality data at all Os monitoring sites within
each region and for each air quality standard scenario considered. The magnitude of this emissions
adjustment was determined independently for each region and standard by determining the
smallest adjustment necessary to ensure all sites within a region would meet the existing standard
or the potential alternative standards (Table 4-1). In a few cases, all monitors in the region met
one or more of the alternative standards based on 2006-2008 observations, and thus there was no
need for model-based adjustments. These cases are represented by values of "0" in Table 4-1. By
evaluating the effect of U.S. anthropogenic emissions reductions on all monitoring sites within a
region, our analysis incorporates the effects of emissions reductions on both local Os production
and regional transport. Since each region is treated independently, the effects of the emissions
reductions required to bring a particular region down to the targeted standard levels do not affect
other regions which require less drastic emissions reductions. In portions of the country with lower
W126 values than nearby locations, the emissions adjustment determined by the controlling or
design monitor in the region may be larger than the emissions reductions that would be required if
the nine climate regions were replaced by many smaller localized areas. However, by considering
larger regions, we are able to account for the fact that nearby emissions reductions will affect Os
monitors already meeting the targeted standard level.9
9 Another rationale for the use of large regions is that the air quality adjustments are computationally intensive, and
focusing on a small number of large regions, rather than many localized areas, greatly reduces the problem size.
4-12
-------
Legend
Central
East North Central
• Northeast
Northwest
• South
• Southeast
• Southwest
West
• West North Central
Figure 4-6 Map of the 9 NOAA climate regions (Karl and Koss, 1984) used in the
national-scale air quality adjustments
Table 4-1 Percent reductions in U.S. anthropogenic NOx emissions applied to
independently reach existing and alternative secondary standards in the nine
climate regions
Region
Central
East North Central
Northeast
Northwest
75 ppb
48
65
96
51
15 ppm-hrs
14
0
36
0
11 ppm-hrs
58
23
51
0
7 ppm-hrs
70
61
81
0
4-13
-------
Region
South
Southeast
Southwest
West
West North Central
75 ppb
54
64
55
90
23
15 ppm-hrs
44
14
67
91
0
11 ppm-hrs
56
38
85
93
6
7 ppm-hrs
66
58
90
95
39
A second distinction between the welfare air quality adjustments and those in the HREA
is that only U.S. anthropogenic NOx emissions reductions were applied in the HDDM adjustment
methodology for the welfare assessment (i.e., changes in U.S. anthropogenic VOC emissions
changes were not considered). NOx emissions reductions are believed to be the most effective
method for reducing Os regionally, since most areas outside of urban population centers tend to be
NOx limited in terms of Os formation. Uncertainties introduced by this assumption are discussed
further in Section 4.4.
Finally, it should be noted that this analysis includes adjustment to four standard levels: 1)
the existing standard of 75 ppb based on the 3-year average of the 4th highest daily maximum 8-
hour average Os concentration, 2) a W126-based standard with a level of 15 ppm-hrs, 3) a W126-
based standard with a level of 11 ppm-hrs, and 4) a W126-based standard with a level of 7 ppm-
hrs. The 2006-2008 average W126 concentrations and 4th highest daily maximum 8-hour average
Os concentrations were calculated for every monitor in each adjusted air quality scenario. For the
analysis of each of the W126 standards, we started with W126 air quality values resulting from
emission reductions required to just meet the existing standard at all monitors in the region, and
only applied the HDDM adjustments to those regions where all sites were not already below the
targeted W126 standard. In some cases, the emissions reductions necessary to meet the existing
standard resulted in W126 values below the level of one or more potential alternative standards at
all monitors within the region. In those cases, there is no change in air quality between the scenario
meeting the existing standard and the scenario meeting the potential alternative standard (compare
Table 4-1 and Table 4-2). For instance, Table 4-1 shows that a 6 percent NOx cut was applied to
4-14
-------
adjust all monitors in the East North Central region down to the current standard of 75 ppb. Since
adjusting to 7 ppm-hrs only required 61 percent cut in US anthropogenic NOx, the primary
standard was determined to be "controlling" and the 65 percent NOx cut was applied to the East
North Central region to create all four W126 surfaces (75 ppb, 15 ppm-hrs, 11 ppm-hrs, 7 ppm-
hrs). Table 4-2 shows the actual NOx reductions that were applied to create the W126 surfaces
for each standard when first adjusting to the 75 ppb standard.
Table 4-2 Percent reductions in U.S. anthropogenic NOx emissions applied to create
the W126 surfaces representing just meeting existing and alternative
standards in the nine climate regions
Region
Central
East North Central
Northeast
Northwest
South
Southeast
Southwest
West
West North Central
75 ppb
48
65
96
51
54
64
55
90
23
15 ppm-hrs
48
65
96
51
54
64
67
91
23
11 ppm-hrs
58
65
96
51
56
64
85
93
23
7 ppm-hrs
70
65
96
51
66
64
90
95
39
National-scale spatial surfaces that represent 2006-2008 W126 concentrations when just
meeting the existing standard and the potential alternate standards (at the highest monitor in the
region) were then created using the monitor values from the appropriate adjustment scenario and
the Voronoi Neighbor Averaging (VNA) spatial interpolation technique. Additional details on the
VNA technique can be found in Appendix 4-A. Note that since each region was adjusted
independently, in some cases distinct boundaries may be visible in the adjusted surfaces. These
boundaries may be obscured to some degree due to the VNA interpolation procedure.
4-15
-------
4.3.4.2
Results
Table 4-3 shows the highest monitored 2006-2008 average W126 concentration in each
region for observed air quality and air quality adjusted to meet the existing Os standard of 75 ppb,
and the highest monitored 2006-2008 8-hour Os design value in each region for observed air
quality and air quality adjusted to meet alternative standards based on the W126 metric with levels
of 15 ppm-hrs, 11 ppm-hrs, and 7 ppm-hrs. Recall that the adjusted air quality surfaces used in
the welfare risk and exposure analyses adjusted each region down to the existing Os standard
before applying additional reductions to meet the alternative standards. So effectively, Table 4-1
shows which standard was the "controlling" standard in each region. For example, when all
monitors in the Central region were adjusted to meet the existing standard, the highest resulting
W126 value was 14 ppm-hrs. Thus, in the Central region, no further adjustments were necessary
to meet the alternative standard of 15 ppm-hrs, but further adjustments were necessary to meet the
alternative standards of 11 ppm-hrs and 7-ppm-hrs.
Table 4-3 Highest 2006-2008 average W126 concentrations in the observed and existing
standard air quality adjustment scenarios; highest 2006-2008 8-hour Os
design values in the observed and potential alternative standard air quality
adjustment scenarios
Region
Central
East North Central
Northeast
Northwest
Southeast
South
Southwest
West
West North Central
Highest W126 value
(ppm-hrs)
Observed
18.3
13.8
17.9
6.6
22.2
18.1
24.3
48.6
12.2
75 ppb
adjustment
14.0
6.4
2.6
3.8
11.9
6.4
17.7
18.9
9.3
Highest 8-hour maximum-based design value (ppb)
Observed
88
86
92
76
95
91
86
119
80
15 ppm-hr
adjustment
83
86
94
76
81
89
71
71
80
11 ppm-hr
adjustment
72
83
89
76
74
91
65
66
79
7 ppm-hr
adjustment
66
76
76
76
67
79
62
61
72
4-16
-------
From Table 4-3, it can be inferred that while each of the 9 regions had at least one monitor
with 2006-2008 air quality data not meeting the existing Os standard, there were 3 regions (East
North Central, Northwest, West North Central) with all monitors meeting the potential alternative
standard with a W126 level of 15 ppm-hrs based on 2006-2008 air quality data. Furthermore, all
monitors in the Northwest region met the alternative standards of 11 ppm-hrs and 7-ppm-hrs based
on 2006-2008 ambient data. When the air quality was adjusted to meet the existing standard, only
two regions (West and Southwest) had monitors with W126 concentrations remaining above 15
ppm-hrs. In addition, there were 4 regions (East North Central, Northeast, Northwest, and South)
that already met 7 ppm-hrs when air quality was adjusted to meet the existing standard.
Figure 4-7 shows the national-scale 2006-2008 average W126 surface adjusted to just meet
the existing Os standard of 75 ppb using the UDDM adjustment procedure described in Section
4.3.2.1, and Figure 4-8 shows the difference between the recent air quality surface (Figure 4-5)
and Figure 4-7. Figure 4-9, Figure 4-11, and Figure 4-13 show the 2006-2008 average W126
surfaces further adjusted to just meet 15 ppm-hrs, 11 ppm-hrs, and 7 ppm-hrs, respectively, while
Figure 4-10, Figure 4-12, and Figure 4-14 show the differences between the surface adjusted to
just meet the existing Os standard of 75 ppb, and the surfaces further adjusted to just meet the
potential alternative standards based on the W126 metric with levels of 15 ppm-hrs, 11 ppm-hrs,
and 7 ppm-hrs. It is immediately apparent from these figures that the reductions in W126 between
recent air quality and air quality just meeting the existing standard (Figure 4-8) are much larger
than the additional reductions in W126 between air quality just meeting the existing standard and
air quality meeting the alternative standards (Figure 4-10, Figure 4-12, Figure 4-14).
This is further exemplified in Figure 4-15 and Figure 4-16, which show empirical
probability density and cumulative distribution functions based on the monitored 8-hour Os design
values (Figure 4-15) and W126 concentrations (Figure 4-16) for each of the air quality scenarios.
Both sets of density functions show a large shift leftward going from observed air quality to just
meeting the existing standard, followed by much smaller leftward shifts from air quality just
meeting the existing standard to air quality just meeting the potential alternative standards. The
shift between air quality just meeting the existing standard and air quality just meeting the potential
alternative standard based on the W126 metric with a level of 15 ppm-hrs is especially small, since
4-17
-------
only a few monitors in the Southwest and West regions did not meet a W126 level of 15 ppm-hrs
when air quality was adjusted to meet the existing standard.
4-18
-------
Figure 4-7 National surface of 2006-2008 average W126 concentrations (in ppm-hrs)
adjusted to just meet the existing Os standard of 75 ppb
4-19
-------
Figure 4-8 Difference in ppm-hrs between the national surface of observed 2006-2008
average W126 concentrations and the national surface of 2006-2008 average
W126 concentrations adjusted to just meet the existing Os standard of 75
ppb
4-20
-------
Figure 4-9 National surface of 2006-2008 average W126 concentrations (in ppm-hrs)
adjusted to just meet the potential alternative standard of 15 ppm-hrs
4-21
-------
Figure 4-10 Difference in ppm-hrs between the national surface of 2006-2008 average
W126 concentrations adjusted to just meet the existing Os standard of 75
ppb and the national surface of 2006-2008 average W126 concentrations
adjusted to just meet the potential alternative standard of 15 ppm-hrs.
White areas indicate no difference.
4-22
-------
Figure 4-11 National surface of 2006-2008 average W126 concentrations (in ppm-hrs)
adjusted to just meet the potential alternative standard of 11 ppm-hrs
4-23
-------
Figure 4-12 Difference in ppm-hrs between the national surface of 2006-2008 average
W126 concentrations adjusted to just meet the existing Os standard of 75
ppb and the national surface of 2006-2008 average W126 concentrations
adjusted to just meet the potential alternative standard of 11 ppm-hrs.
White areas indicate no difference.
4-24
-------
Figure 4-13 National surface of 2006-2008 average W126 concentrations (in ppm-hrs)
adjusted to just meet the potential alternative standard of 7 ppm-hrs
4-25
-------
Figure 4-14 Difference in ppm-hrs between the national surface of 2006-2008 average
W126 concentrations adjusted to just meet the existing Os standard of 75
ppb and the national surface of 2006-2008 average W126 concentrations
adjusted to just meet the potential alternative standard of 7 ppm-hrs. White
areas indicate no difference.
4-26
-------
:i
o **
1= O
p
o
Observed
75 ppb
15 ppm-hrs
11 ppm-hrs
7 ppm-hrs
i
40
i I
60 80
03 Design Value (ppb)
100
120
o
01
o
CD
m
2<=,
03 CO
'CO
01
Observed
75 ppb
15 ppm-hrs
11 ppm-hrs
7 ppm-hrs
i
20
i
40
I
60
i
30
100
Percentile
Figure 4-15
Empirical frequency distribution (top) and cumulative distribution (bottom)
functions for the monitored 2006-2008 8-hour Os design values, and the
2006-2008 8-hour Os design values after adjusting to just meet the existing
and potential alternative standards
4-27
-------
CO
o
IS
(D
Observed
75ppb
15 ppm-hrs
11 ppm-hrs
7 ppm-hrs
7 15
W1 26 (ppm-hrs)
i
50
Q. '
•-_-•
CO
Observed
75ppb
15 ppm-hrs
11 ppm-hrs
7 ppm-hrs
i
20
i
40
I
60
i
80
100
Percentile
Figure 4-16 Empirical frequency distribution (top) and cumulative distribution (bottom)
functions for the monitored 2006-2008 average W126 concentrations, and the
2006-2008 average W126 concentrations after adjusting to just meet the
existing and potential alternative standards. Note W126 concentrations are
displayed using a square root scale.
4-28
-------
4.4 ASSESSMENT OF UNCERTAINTY
As noted in Chapter 3, we have based the design of the uncertainty analysis for this
assessment on the framework outlined in the WHO guidance (WHO, 2008). In this section, we
provide quantitative assessments where possible in addition to an overall qualitative uncertainty
analysis in which we described each key source of uncertainty and qualitatively assessed its
potential impact (including both the magnitude and direction of the impact) on risk results, as
specified in the WHO guidance. In general, this assessment includes qualitative discussions of
the potential impact of uncertainty on the results (WHO Tierl) and quantitative sensitivity
analyses where we have sufficient data (WHO Tier 2).
Table 4-5 includes a summary the key sources of uncertainty identified for the Os REA.
For each source of uncertainty, we have (a) provided a description, (b) estimated the direction of
influence (over, under, both, or unknown) and magnitude (low, medium, high) of the potential
impact of each source of uncertainty on the risk estimates, (c) assessed the degree of uncertainty
(low, medium, or high) associated with the knowledge-base (i.e., assessed how well we
understand each source of uncertainty), and (d) provided comments further clarifying the
qualitative assessment presented. The categories used in describing the potential magnitude of
impact for specific sources of uncertainty on risk estimates (i.e., low, medium, or high) reflect
our consensus on the degree to which a particular source could produce a sufficient impact on
risk estimates to influence the interpretation of those estimates in the context of the secondary Os
NAAQS review. Where appropriate, we have included references to specific sources of
information considered in arriving at a ranking and classification for a particular source of
uncertainty. Discussion of elements in Table 4-5 is provided below.
There is inherent uncertainty in all deterministic air quality models, such as CMAQ, the
photochemical grid model which was used to develop the model-based Os adjustment
methodology. Evaluations of air quality models against observed pollutant concentrations build
confidence that the model performs with reasonable accuracy despite both structural and
parametric uncertainties. A comprehensive model performance evaluation provided in Appendix
4-B of the HREA shows generally acceptable model performance which is equivalent to or better
than typical state-of-the science regional modeling simulations as summarized in Simon et al.
(2012). Dynamic evaluations of CMAQ in the literature have evaluated the ability of the
4-29
-------
modeling system to predict ozone response to emissions changes. As described in more detail in
Appendix 4-B of the HREA, these analyses generally conclude that the predicted model ozone
response is conservative meaning that this analysis may overestimate the required emissions
reductions needed to meet any standard level. The use of the Higher Order Decoupled Direct
Method (HDDM) within CMAQ to estimate Os response to emissions adjustments adds
uncertainty to that inherent in the model itself. HDDM allows for the approximation of Os
concentrations under alternate emission scenarios without re-running the model simulation with
different inputs. This approximation becomes less accurate for larger emissions adjustments. To
accommodate increasing uncertainty at larger emissions adjustments, the HDDM modeling was
performed at three distinct emissions levels to allow for a better characterization of Os response
over the entire range of emissions levels. The accuracy of the HDDM estimates can be quantified
at distinct emissions levels by re-running the model with modified emissions inputs and
comparing the results. This method was applied to quantify the accuracy of 3-step HDDM Cb
estimates for 50 percent and 90 percent NOx cut conditions for each urban study areas (as shown
in Appendix 4-D of the HREA). At 50 percent NOx cut conditions, HDDM using information
from these multiple simulations predicted hourly Os concentrations with a mean bias and a mean
error less than +/- 1 ppb in all study areas compared to brute force model simulations. At 90
percent NOx cut conditions, HDDM using information from these multiple simulations predicted
hourly Os concentrations with a mean bias less than +/- 3ppb and a mean error less than +/- 4
ppb in all study areas. These small bias and error estimates show that uncertainty due to the
HDDM approximation method is small up to 90 percent emissions cuts.
In order to apply modeled Os response to ambient measurements, simple linear regression
relationships were developed which relate Os response to emissions adjustments with ambient Os
concentrations for every season, hour-of-the-day, and monitor location. Applying modeled Os
responses to ambient data based on these relationships adds uncertainty because the variability in
the individual responses is collapsed into a central tendency estimate (i.e., the regression line).
Preliminary work showed that the relationships developed with these regressions were generally
statistically significant for most season, hour-of-the-day, and monitor location combinations for
2005 modeling in Detroit and Charlotte (Simon et al, 2013). Statistical significance was not
evaluated for each regression in this analysis here since there were over 280,000 regressions
created (1,300 monitors x 2 sensitivity coefficients x 3 emissions levels x 3 seasons x 12 hours =
4-30
-------
280,800 regressions). Statistical inferences can quantify the goodness of fit for the modeled
relationships and can quantify the uncertainty in the central tendency estimate at any given Os
concentration. The simple linear regression model provided both a central tendency estimate and
a standard error estimate for Os response at each measured hourly Os concentration.
The propagation of hourly standard error estimates to W126 is not a straightforward
calculation due to the nonlinear weighting function which is applied to the hourly concentrations.
Thus, a bootstrapping approach was employed to estimate the uncertainty in the 3-year average
W126 values adjusted to meet the current and alternative standards due to the use of a central
tendency estimate to represent Os response. Starting with 3 years of hourly Os concentrations
and standard errors for a given monitor adjusted to meet a given standard level, 1,000 random
hourly time-series were generated using Equation 1:
Equation 1: O3boot(h, i) = O3obs(h) + R(d, i) * SE(K)
where O3boot(h,i) is the ith random hourly value for hour h
O3obs(h) is the adjusted hourly concentration value for hour h
R(d,i) is the ith random value sampled from a Normal(0,l) distribution for day d
SE(h) is the standard error estimate for hour h
Note that a single random value was drawn for each day and applied to all hourly concentrations
within that day in order to account for any temporal correlation between the hourly values.
Three-year average W126 values were then calculated from each of the 1,000 random
hourly time-series, and the resulting standard error of the adjusted 3-year average W126 value at
the monitor was the standard deviation of these 1,000 values. This process was repeated for all
1,300 monitors in the contiguous U.S. for the existing standard of 75 ppb, and the alternative
standards of 15, 11, and 7 ppm-hrs. Figure 4-17 shows boxplots of the standard errors in ppm-
hrs at each monitor for the various standards, and Figure 4-18 to Figure 4-21 show maps of the
standard errors in ppm-hrs at each standard level.
4-31
-------
Standard Errors for Adjusted W126 Values
0
1X5
CM _
O
"ft
£ o
E o
Q.
Q.
*- UO
O ,— _
LJJ °
o
ro o
o
o
o
D
O
0 o
0 0
0
0 0
o
0 0
8
1
8
0 °
o
x x —
~ i
0
0
8
o S
o 0
0
0 0
8
i 8
I
x 1 ^ 1
1
75 ppb
15 ppm-hr 11 ppm-hr
Standard Level
7 ppm-hr
Figure 4-17 Boxplots of standard errors for 2006-2008 average W126 values adjusted to
meet the existing and alternative standards. Boxes represent the median and
quartiles, x's represent mean values, whiskers extend up to 1.5x the inter-
quartile range from the boxes, and circles represent data points outside this
range.
Standard Errors for Adjusted W126 Values: 75 ppb
o
0.05
0.2
0.25
0.1 0.15
Standard Error (ppm-hrs)
Figure 4-18 Map of standard errors for 2006-2008 average W126 values (in ppm-hrs)
adjusted to meet the existing standard of 75 ppb
4-32
-------
Standard Errors for Adjusted W126 Values: 15 ppm-hrs
0.05
0.2
0.1 0.15
Standard Error (ppm-hrs)
Figure 4-19 Map of standard errors for 2006-2008 average W126 values (in ppm-hrs)
adjusted to meet the alternative standard of 15 ppm-hrs
Standard Errors for Adjusted W126 Values: 11 ppm-hrs
0.25
0
0.05
02
0.1 0.15
Standard Error (ppm-hrs)
Figure 4-20 Map of standard errors for 2006-2008 average W126 values (in ppm-hrs)
adjusted to meet the alternative standard of 11 ppm-hrs
4-33
0.25
-------
Standard Errors for Adjusted W126 Values: 7 ppm-hrs
0.05
0.2
0.25
0.1 0.15
Standard Error (ppm-hrs)
Figure 4-21 Map of standard errors for 2006-2008 average W126 values (in ppm-hrs)
adjusted to meet the alternative standard of 7 ppm-hrs
The resulting standard error values were generally quite small: all monitors had standard
errors of less than 0.3, and about 98 percent of monitors had standard errors of less than 0.1 ppm-
hr. The standard errors tended to decrease slightly with lower standard levels. In general, the
hourly standard errors increased with larger reductions associated with meeting lower standard
level. Simultaneously, the largest decreases in the peak Os concentrations which have the
greatest impact upon W126 levels also occurred at lower standard levels. These two factors
tended to offset one another, resulting in only a slight decrease in standard error values when
adjusting to meet lower standard levels. Finally, the largest standard errors tended to occur in
urban core areas, which is expected for two reasons. First, the simple linear regression models
tended to fit more poorly in urban core areas due to non-linearities in the ozone chemistry in
those locations, resulting in higher hourly standard error values. Second, the highest standard
errors tended to occur at monitors with the highest adjusted W126 values, which tended to be
located in urban areas under the various adjustment scenarios.
Relationships between Os response and hourly Os concentration were developed based on
8 months of modeling: April-October 2007. These relationships were applied to ambient data
4-34
-------
from 2006-2008 leading to an additional source of uncertainty. The eight months that were
modeled capture a variety of meteorological conditions. However, in cases where other years
have drastically different meteorological conditions, there is uncertainty in how well the
regression central tendency estimates would represent Os responses in those years. In addition, if
emissions were to change drastically between the modeled period and the time of the ambient
data measurements this could also change the relationship between Os response and Os
concentrations. The regressions derived from the 2007 modeling period are only applied to
measurements made within one year of the modeled time period. Although some emissions
changes did occur over this time period, we believe it is still reasonable to apply 2007 modeling
to this relatively small window of measurements which occurs before and after the modeling.
Ozone responses were modeled for "across-the-board" reductions in U.S. anthropogenic
NOx and were applied independently for nine climate regions, e.g. for each region, we looked at
how W126 responded to NOx emissions reductions across the entire U.S. We recognize that this
means that when considered together, the adjusted W126 values will not reflect any single
specific NOx emissions reductions across the U.S. These reductions were chosen for illustrative
purposes and were not meant to reflect actual emissions control strategies. The "across-the-
board" reductions do not optimize the lowest cost or least total emissions combinations that state
and local agencies will likely attempt to achieve to bring NOx emissions levels down uniformly
across time and space within each region. So the reductions do not reflect spatial and temporal
heterogeneity that may occur in local and regional emissions reductions.
To further investigate the implications of the regionally-derived "across-the-board" NOx
reduction scenarios that were used here we evaluate past emissions reductions and EPA
projected future changes to emissions. An EPA analysis (EPA, 2006) has shown that some past
efforts to meet ozone NAAQS have resulted in regional emissions reductions. Specifically, the
NOx SIP Call program implemented to help areas meet the 1997 ozone standard resulted in
substantial reductions in power plant NOx emissions from states across the eastern U.S. We
further evaluated EPA projected emissions changes between 2007 and 2020 (EPA, 2012). These
emissions projections take into account "on the books" controls from state and federal
regulations that were in place at the time of the analysis as well as population growth and do not
consider any actions that would be undertaken to meet a new Os NAAQS level. Nationally, NOx
4-35
-------
is projected to decrease by 45 percent between 2007 and 2020. Two-thirds of these NOx
reductions are projected to come from on-road vehicles as a result of tighter emissions standards
and fleet turnover. Smaller but still substantial reductions are predicted to occur from power
plants, off-road equipment, railroad and marine sources. Figure 4-22 and Figure 4-23 show the
magnitude of NOx reductions that are projected to occur. These figures broadly show large state
and regional reductions in anthropogenic NOx on the order of 40-60 percent. These reductions
are not limited to densely populated states or to current nonattainment areas. In fact, in several
regions larger reductions in NOx emissions are projected to occur in attainment counties than in
nonattainment counties. Table 4-4 compares these projected emissions changes to those applied
to meet the various standard levels in the WREA analysis. In all regions except the West and
Southwest, the projected emissions changes, which are expected to be fairly regional in nature,
are greater than what would be required to meet a 15 ppm-hrs standard based on the HDDM
methodology. In addition, in all regions the projected emissions changes make up at least 40
percent of what would be required to just meet a W126 standard of 7 ppm-hrs and in many cases
they make up a substantially larger portion. These comparisons build confidence that the
regionally-based NOx control scenarios applied in this analysis are not unrealistic since
substantial regional NOx reductions are projected to occur regardless of whether the Os NAAQS
is changed. The comparisons also build confidence in the finding that most areas would have
W126 values below 15 ppm-hrs after meeting the existing 75 ppb standard. In some potential
future scenarios, a portion of the controls applied to just meet a W126-based NAAQS might
come from local controls. While the scenarios implemented in this analysis show that by
bringing down the highest monitor in a region would lead to reductions below the targeted level
through the rest of the region, to the extent that the regional reductions from on-the-books
controls are supplemented with more local controls the additional benefit may be overestimated.
In addition, the assumption of aNOx-only control scenario adds uncertainty. The 2020
projections predict a 20 percent reduction in US anthropogenic VOC from 2007 levels and some
locations may undertake additional VOC emissions reductions. Since ozone in urban areas is
more responsive to VOC emissions reductions than ozone in rural areas, these VOC emissions
would result in lower required NOx reductions and would tend to reduce the W126 benefits in
rural areas.
4-36
-------
rm*^ mam fcimi •HUH nir cri
Figure 4-22 Percent reduction in state NOx emissions projected to occur between 2007
and 2020
Attainment Counties
Nlonattainment Counties
region
Figure 4-23 Percent NOx reductions projected to occur from 2007 to 2020 aggregated by
climate region for counties designated in attainment with the 2008 Os
NAAQS and counties designated nonattainment for the 2008 Os NAAQS
4-37
-------
Table 4-4 Comparison of projected NOx emissions reductions to those applied to meet
various standard levels in the WREA analysis
Region
Percent NOx Emissions Reductions
Applied to
meet 75 ppb
Applied to
meet 15
ppm-hrs
Applied to meet
7 ppm-hrs
Projected from 2007-
2020 (Regional)
Range of projected
2007-2020
(State-level)
Central
ENC
Northeast
Northwest
Southeast
South
Southwest
West
WNC
48
65
96
51
54
64
55
90
23
14
0
36
0
44
14
67
91
0
70
61
81
0
66
58
90
95
39
51.3
44.1
45.6
41.6
37.6
52.6
35.9
44.0
32.9
46.4-59.7
34.6-47.8
37.8-56.6
29.8-45.2
46.5-58.0
31.1-46.1
17.6-50.2
41.1-44.3
19.6-42.2
4-38
-------
Table 4-5 Summary of Qualitative Uncertainty Analysis of Key Air Quality Elements in the Os NAAQS Risk Assessment
Source
Description
Potential influence of
uncertainty on risk estimates
Direction
Magnitude
Knowledge-
Base
uncertainty*
Comments
(KB: knowledge base, INF: influence of uncertainty on risk
estimates)
A. Ambient air quality
measurement data
Os concentrations measured by
ambient monitoring instruments
have inherent uncertainties
associated with them. Additional
uncertainties due to other factors
may include:
- monitoring network design
- required Os monitoring seasons
- monitor malfunctions
- wildfire and smoke impacts
- interpolation of missing data
Both
Low
Low
KB: Os measurements are assumed to be accurate to within '/2 of the
instrument's Method Detection Limit (MDL), which is 2.5 ppb for
most instruments. EPA requires that routine quality assurance checks
are performed on all regulatory instruments, and that all data reported
to AQS are certified by both the monitoring agency and the
corresponding EPA regional office. See 40 CFR Part 58, Appendix A
for details. The CASTNET monitoring data were subject to their own
quality assurance requirements.
KB: Monitor malfunctions sometimes occur causing periods of
missing data or poor data quality. Monitoring data affected by
malfunctions are usually flagged by the monitoring agency and
removed from AQS. In addition, the AQS database managers run
several routines to identify suspicious data for potential removal.
KB: There is a known tendency for smoke produced from wildfires to
cause interference in Os instruments. Measurements collected by Os
analyzers were reported to be biased high by 5.1-6.6 ppb per 100
|ig/m3 of PM2.5 from wildfire smoke (Payton, 2007). However,
smoke concentrations high enough to cause significant interferences
are infrequent and the overall impact is believed to be minimal.
KB: Missing intervals of 1 or 2 hours in the measurement data were
interpolated, which may cause some additional uncertainty.
However, due to the short length of the interpolation periods, and the
tendency for these periods to occur at night when Os concentrations
are low, the overall impact is believed to be minimal.
INF: EPA's current Os monitoring network requirements (40 CFR
Part 58, Appendix D) are primarily focused on urban areas. Rural
areas where Os concentrations are lower tend to be under-represented
by the current monitoring network. The network requirements also
state that at least one monitor within each urban area must be sited to
capture the highest Os concentrations in that area, which may cause
some bias toward higher measured concentrations.
INF: Each state has a required Os monitoring season which varies in
length from May - September to year-round. Some states turn their
Os monitors off during months outside of the required season, while
others leave them on. This can cause differences in the amount of
4-39
-------
Source
Description
Potential influence of
uncertainty on risk estimates
Direction
Magnitude
Knowledge-
Base
uncertainty*
Comments
(KB: knowledge base, INF: influence of uncertainty on risk
estimates)
data available throughout the year across states, especially in months
outside of the required Os monitoring season.
B. Veronoi Neighbor
Averaging (VNA)
spatial fields
VNA is a spatial interpolation
technique used to estimate W126
concentrations in unmonitored
areas, which has inherent
uncertainty
Both
Low-
Medium
Low-Medium
KB: VNA interpolates 2006-2008 average W126 values estimated
from hourly ambient air quality measurements at each CMAQ grid
cell in each of the 9 NOAA climate regions. The VNA estimates are
weighted based on distance from neighboring monitoring sites, thus
the uncertainty tends to increase with distance from the monitoring
sites becomes greater. As a result, there is less uncertainty in the
VNA estimates near urban areas where the monitoring networks are
dense, and more uncertainty in sparsely populated areas where
monitors are further apart, particularly in the Western U.S.
C.CMAQ modeling
Model predictions from CMAQ,
like all deterministic photochemical
models, have both parametric and
structural uncertainty associated
with them
Both
Medium
Medium
KB: Structural uncertainties are uncertainties in the representation of
physical and chemical processes in the model. These include: choice
of chemical mechanism used to characterize reactions in the
atmosphere, choice of land surface model and choice of planetary
boundary layer model.
KB: Parametric uncertainties include uncertainties in model inputs
(hourly meteorological fields, hourly 3-D gridded emissions, initial
conditions, and boundary conditions)
KB: Uncertainties due to initial conditions are minimized by using a
10 day ramp-up period from which model results are not used.
KB: Evaluations of models against observed pollutant concentrations
build confidence that the model performs with reasonable accuracy
despite the uncertainties listed above. A comprehensive model
evaluation provided in Appendix 4-B of the hREA shows generally
acceptable model performance which is equivalent or better than
typical state-of-the science regional modeling simulations as
summarized in Simon et al (2012). However, both under-estimations
and over-estimations do occur at some times and locations. Generally
the largest mean biases occur on low ozone days during the summer
season. In addition, the model did not fully capture rare wintertime
high ozone events occurring in the Western U.S. Both of these types
of biases are not likely to substantially affect W126 performance
since low ozone days are not heavily weighted in the W126
calculation and since the highest 3-month W126 values were only
calculated for April-October in this analysis.
4-40
-------
Source
Description
Potential influence of
uncertainty on risk estimates
Direction
Magnitude
Knowledge-
Base
uncertainty*
Comments
(KB: knowledge base, INF: influence of uncertainty on risk
estimates)
D. Higher Order
Decoupled Direct
Method (HDDM)
HDDM allows for the
approximation of ozone
concentrations under alternate
emissions scenarios without re-
running the model simulation
multiple times using different
emissions inputs. This
approximation becomes less
accurate for larger emissions
perturbations especially under
nonlinear chemistry conditions.
Both
Medium
Medium
KB: To accommodate increasing uncertainty at larger emissions
perturbations, the HDDM modeling was performed at three distinct
emissions levels to allow for a better characterization of ozone
response over the entire range of emissions levels. The replication of
brute force10 hourly ozone concentration model results by the HDDM
approximation was quantified for 50% and 90% NOx cut conditions
for each urban study areas (as shown in Appendix 4-D of the hREA).
At 50% NOx cut conditions, HDDM using information from these
multiple simulations predicted hourly ozone concentrations with a
mean bias and a mean error less than +/- 1 ppb in all urban study
areas compared to brute force model simulations. At 90% NOx cut
conditions, HDDM using information from these multiple simulations
predicted hourly ozone concentrations with a mean bias less than +/-
3ppb and a mean error less than +/- 4 ppb in all urban study areas.
E. Application of
HDDM sensitivities to
ambient data
In order to apply modeled
sensitivities to ambient
measurements, regressions were
developed which relate ozone
response to emissions perturbations
with ambient ozone concentrations
for every season, hour-of-the-day
and monitor location. Applying
ozone responses based on this
relationship adds uncertainty.
Both
Low-
Medium
Low-Medium
KB: Preliminary work showed that the relationships developed with
these regressions were generally statistically significant for most
season, hour-of-the-day, and monitor location combinations for 2005
modeling in Detroit and Charlotte. Statistical significance was not
evaluated for each regression in this analysis since there were over
280,000 regressions created (1300 monitors x 2 sensitivity
coefficients x 3 emissions levels x 3 seasons x 12 hours = 280,800
regressions). Statistical inferences can quantify the goodness of fit
for the modeled relationships. However it is not possible to quantify
the applicability of this modeled relationship to the actual
atmosphere.
KB: The regression model provided both a central tendency estimate
and a standard error estimate for ozone response at each measured
hourly ozone concentration. The base analysis used the central
tendency which will inherently dampen some of the variability in
ozone response. A bootstrapping analysis was used to estimate the
uncertainty in 3-year average W126 concentrations due to use of the
central tendency prediction for ozone response. This analysis showed
that the uncertainty was small: the standard errors were less than 0.3
10 Brute force model concentrations refer to model results obtained by changing the emissions inputs and re-running the CMAQ model. HDDM concentration
estimates are an approximation of the model results that would be obtained by re-running the simulation with different inputs.
4-41
-------
Source
Description
Potential influence of
uncertainty on risk estimates
Direction
Magnitude
Knowledge-
Base
uncertainty*
Comments
(KB: knowledge base, INF: influence of uncertainty on risk
estimates)
ppm-hrs at all monitors, and less than 0.1 ppm-hrs at about 98% of
monitors.
F. Applying modeled
sensitivities to un-
modeled time periods
Relationships between ozone
response and hourly ozone
concentration were developed based
on 7 months of modeling: April-
October 2007. These relationships
were applied to ambient data from
2006-2008.
Both
Medium
Medium
KB: The seven months that were modeled capture a variety of
meteorological and emissions conditions. Applying these 2007
sensitivities to other years with potentially different meteorology and
emissions adds uncertainty to the relationship between ozone
response and ozone concentrations. The regressions derived from the
2007 modeling period are only applied to measurements made within
one year of the modeled time period. Although some emissions
changes did occur over this time period, we believe it is still
reasonable to apply 2007 modeling to this relatively small window of
measurements which occurs before and after the modeling.
G. Assumptions of
regionally-determined
across-the-board
emissions reductions
Ozone response is modeled for
across-the-board reductions11 in
U.S. anthropogenic NOx. These
across-the-board cuts do not reflect
actual emissions control strategies.
Overestimates
W126 benefits
Medium
Medium
KB: The form, locations, and timing of emissions reductions that
would be undertaken to meet various levels of the ozone standard are
unknown. The across-the-board emissions reductions bring levels
down uniformly across time and space to show how ozone would
respond to changes in ambient levels of precursor species but do not
reflect spatial and temporal heterogeneity that may occur in local and
regional emissions reductions.
* Refers to the degree of uncertainty associated with our understanding of the phenomenon, in the context of assessing and characterizing its uncertainty. Sources
classified as having a "low" impact would not be expected to impact the interpretation of risk estimates in the context of the O3 NAAQS review; sources
classified as having a "medium" impact have the potential to change the interpretation; and sources classified as "high" are likely to influence the interpretation
of risk in the context of the Os NAAQS review.
11 "Across the board" emission reductions refer to equal percentage NOx emissions cuts in all source categories and all locations at all times.
4-42
-------
4.5 SUMMARY OF AIR QUALITY RESULTS
Observed W126 levels in 2006-2008 were highest in the Western US (maximum monitored
value was 48.6 ppm-hrs) followed by the Southwest, Southeast, Central and Northeast (24.3, 22.2,
18.3, and 17.9 ppm-hrs respectively). All monitored W126 values in other regions of the US were
below 15 ppm-hrs with the lowest values in the Northwestern U.S. all falling below 7 ppm-hrs.
The air quality adjustments to meet the current 75 ppb standard brought all areas except the West
and Southwest (18.9 and 17.7 ppm-hrs) below 15 ppm-hrs. The air quality adjustments to meet
the current 75 ppb standard additionally resulted in four regions being below 7 ppm-hrs (East
North Central, Northeast, Northwest, and South). The reductions in W126 between recent air
quality and air quality just meeting the existing standard are much larger than the additional
reductions in W126 between air quality just meeting the existing standard and air quality meeting
the alternative standards. The shift between air quality just meeting the existing standard and air
quality just meeting the potential alternative standard based on the W126 metric with a level of 15
ppm-hrs is especially small, since only a few monitors in the Southwest and West regions did not
meet a W126 level of 15 ppm-hrs when air quality was adjusted to meet the existing standard.
4.6 REFERENCES
Berrocal, V.J.; A. E. Gelfand and D.M. Holland. 2012. "Space-Time Data Fusion Under Error in
Computer Model Output: An Application to Modeling Air Quality." Biometrics, 68(3),
837-848.
Chen, J. R.; Zhao; Z. Li. 2004. "Voronoi-based k-order Neighbor Relations for Spatial Analysis."
ISPRS J Photogrammetry Remote Sensing, 59(1-2), 60-72.
Gold, C. 1997. "Voronoi Methods in GIS," Vol. 1340. In Algorithmic Foundation of Geographic
Information Systems (Kereveld M., J. Nievergelt, T. Roos, P. Widmayer eds). Lecture
notes in Computer Science, Berlin: Springer-Verlag, 21-35.
Karl, T. R.; Koss, W. J. 1984. "Regional and National Monthly, Seasonal, and Annual
Temperature Weighted by Area, 1895-1983." Historical Climatology Series, 4-3,
National Climatic Data Center, Asheville, NC, 38 pp.
Lefohn, A. S.; Laurence, J. A.; Kohut, R. J. 1988. "A Comparison of Indices that Describe the
Relationship between Exposure to Ozone and Reduction in the Yield of Agricultural
Crops". Atmospheric Environment, 22(6), 1229-1240.
Simon, H.; Baker, K.R.; Phillips, S. 2012. "Compilation and Interpretation of Photochemical
Model Performance Statistics Published Between 2006 and 2012." Atmospheric
Environment, 61, 124-139.
4-43
-------
Simon, H.; K. R. Baker; F. Akhtar; S.L. Napelenok; N. Possiel; B. Wells and B. Timin. 2013. "A
Direct Sensitivity Approach to Predict Hourly Ozone Resulting from Compliance with
the National Ambient Air Quality Standard" Environmental Science and Technology,
Vol. 47, 2304-2313.
Timin, B.; K. Wesson and J. Thurman. 2010. "Application of Model and Ambient Data Fusion
Techniques to Predict Current and Future Year PM2.5 Concentrations in Unmonitored
Areas, " in D.G. Steyn and St Rao (eds), Air Pollution Modeling and Its Application XX,
Netherlands: Springer, pp. 175-179.
U.S. EPA, 2006. Nox Budget Trading Program 2005: Program Compliance and Environmental
Results. Research Triang Park, NC: EPA Office of Air Quality Planning and Standards
(EPA document number EPA-430-R-06-013)
U.S. EPA, 2012. Technical Support Document (TSD): Preparation of Emissions Inventories for
the Version 5.0, 2007 Emissions Modeling Platform. Research Triangle Park, NC: EPA
Office of Air Quality Planning and Standards. Accessed at:
http://www.epa.gov/ttn/chief/emch/index.htmltfpmnaaqs
U. S. EPA, 2013. Integrated Science Assessment for Ozone and Related Photochemical
Oxidants. Research Triangle Park, NC: EPA Office of Air Quality Planning and
Standards. (EPA document number EPA/600/R-10/076F).
WHO, 2008. World Health Organization Harmonization Project Document No. 6. Part 1:
Guidance Document on Characterizing and Communicating Uncertainty in Exposure
Assessment, .
4-44
-------
5 O3 RISK TO ECOSYSTEM SERVICES
5.1 INTRODUCTION
The EPA is using an ecosystem services framework as described in Chapter 2 to help
define how the damage to ecosystems informs determinations of the adversity to public welfare
associated with changes in ecosystem functions. Figure 9-1 of the Cb ISA (U.S. EPA, 2013) is
reproduced below (Figure 5-1) as a summary of exposure and effects that lead to potential loss of
ecosystem services. Figure numbers in this figure refer to Chapter 9 of the Os ISA.
O3 exposure
O3 uptake & physiology (Fig 9-2)
• *Antioxidant metabolism up-regulated
•Decreased photosynthesis
-Decreased stomatal conductance
or sluggish stomataI response
Effects on leaves
-Visible leaf injury
-Altered leaf production
•Altered leaf chemical composition
Plant growth (Fig 9.8)
•Decreased biomass accumulation
•Altered reproduction
•Altered carbon allocation
•Altered crop quality
1
Belowground processes (Fig 9,8)
-Altered litter production and decomposition
-Altered soil carbon and nutrient cycling
-Altered soil fauna and microbial communities
D
I
w
CD
(ft
2;
Affected ecosystem services
•Decreased productivity
•Decreased C sequestration
•Altered water cycling (Fig 9-7)
•Altered community composition
(i.e., plant, insects microbe)
Figure 5-1 Conceptual Diagram of the Major Pathway through which
and the Major Endpoints that Os May Affect in Plants and Ecosystems
Enters Plants
5-1
-------
This chapter focuses primarily on those ecosystem services anticipated to be at risk from
Os exposure that we were only able to assess qualitatively, due to a lack of sufficient data,
methods, or resources to allow quantification of the incremental effects of Os. It also includes
semi-qualitative, GIS-driven, correlational assessments of the potential impacts of Os on risks of
fire and bark beetle damage and identifies additional anticipated adverse effects associated with
Os exposure that we are not able to assess, even qualitatively. In contrast, Chapters 6 and 7
provide quantitative assessments for risks related to tree biomass loss, timber and crop yield loss
and visible foliar injury. Figure 5-2 illustrates the relationships between the ecological effects of
Os and the anticipated ecosystem services impacts that will be discussed in the following
sections.
Ecological Effects
» Biomass Loss (Chapter 6)
• Foliar Injury (Chapter 7)
Supporting Services
• Primary Productivity
Regulating Services
• Carbon Sequestration
• Pollution Removal
Cultural Services
• Recreational Use
Provisioning Services
• Agricultural Harvest
• Timber Production
Additional Assessments
Support ing Services
• Soil Formation
• Community Structure
• Primary Productivity
Regulating Services
• Nutrient Cycling
• Water Regulation
* Pollination
• Fire Regulation
Cultural Services
• Aesthetic Services
• Non-Use Values
Provisioning Services
• Insect Damage
• Non-Timber Uses
Figure 5-2
Services
Relationship between Ecological Effects of O3 Exposure and Ecosystem
5-2
-------
While most of the impacts of Os on these services cannot be specifically quantified, it is
important to provide an understanding of the magnitude and significance of the services that are
anticipated to be negatively impacted by Os exposures. For many services, we can estimate the
current total magnitude and, for some, we can estimate the current value of the services in
question. The estimates of current service provision will reflect the loss of services potentially
occurring from historical and current Os exposure and provide context for the importance of any
potential impacts of Os on those services, e.g., if the total value of a service is small, the likely
impact of Os exposure will also be small. Likewise, if the total value is large, there is a higher
potential for significant damage, even if the relative contribution of Os as a stressor is small.
Also, in some cases we can provide information on locations where high Os exposures occur in
conjunction with significant ecosystem service impairment. Specifically, we can provide
information on areas where high W126 index values may have the greatest contribution to the
service impairment caused by fires in California and bark beetle damage in forests. This
assessment will address Os impacts on ecosystem services following the framework of the
Millennium Ecosystem Assessment (MEA, 2005). In line with the framework, the subsequent
sections are divided into regulating, supporting, provisioning, and cultural ecosystem services.
5.2 REGULATING SERVICES
Regulating services as defined by the MEA (2005) are those services that regulate
ecosystem processes. Services such as air quality, water, climate, erosion, and pollination
regulation fit within this category. The next sections describe potential impacts of Os on some of
these services.
5.2.1 Hydrologic Cycle
Regulation of the water cycle is another ecosystem service that can be adversely affected
by the effects of Os on plants. Studies of Os-impacted forests in eastern Tennessee in or near the
Great Smoky Mountains have shown that ambient Os exposures resulted in increased water use
in Cb-sensitive species, which led to decreased modeled late-season stream flow in those
watersheds. The increased water use resulted from a sluggish stomatal response that increases
water loss, which in turn increases water requirements (U.S. EPA, 2013). Ecosystem services
potentially affected by such a loss in stream flow could include habitat for species (e.g., trout)
5-3
-------
that depend on an optimum stream flow or temperature. Additional downstream effects could
potentially include a reduction in the quantity and/or quality of water available for irrigation or
drinking and for recreational use. Conversely, one model study reported in the Os ISA (U.S.
EPA, 2013) associated reduced stomatal aperture from Os exposure combined with nitrogen
limitation with decreased water loss, which in turn increased runoff, potentially increasing water
availability. Regardless of the response, water cycling in forests is affected by Os exposure and
potentially impacts ecosystem services associated with both water quality and quantity.
The National Survey on Recreation and the Environment (NSRE) is an ongoing survey of
a random sample of adults over the age of 16 on their interactions with the environment that
provides data on the values survey respondents place on the provision of habitat for wild plants
and animals. As part of the NSRE, the United States Forest Service (USFS) and the National
Oceanographic and Atmospheric Administration (NOAA) jointly surveyed Americans, age 16
and over, for their report on Uses and Values of Wildlife and Wilderness in the United States.
The NSRE specifically asked respondents to rank the importance of water quality as a benefit of
wilderness. Ninety one percent of respondents ranked water quality protection as either
extremely or very important; less than 1 percent of respondents ranked this service as not
important at all.
522 Pollination
The Os ISA (U.S. EPA, 2013) identifies Os as a possible agent affecting the travel
distance of and the loss of specificity of volatile organic compounds emitted by plants, some of
which act as scent cues for pollinators. While it is not possible to explicitly calculate the loss of
pollination services resulting from this negative effect on scent cues, the loss is reflected in the
current estimated value of $18.3 billion (2010$) for all pollination services, managed and wild,
in North America (U.S., Canada, and Bermuda) (Gallai et al., 2009).
523 Fire Regulation
Fire regime regulation is also negatively affected by Os exposure. Grulke et al. (2009)
reported various lines of evidence indicating that Os exposure may be anticipated to contribute to
southern California forest susceptibility to wildfires by increasing leaf turnover rates and litter.
This, in turn, creates increased fuel loads on the forest floor, Os-increased drought stress, and
increased susceptibility to bark beetle attacks.
5-4
-------
According to the National Interagency Fire Center
(http ://www. nifc. gov/firelnfo/firelnfo stati sties .html), in 2010 in the United States over 3
million acres burned in wildland fires. Over the 5-year period from 2004 to 2008, Southern
California alone experienced, on average, over 4,000 fires per year burning, on average, over
400,000 acres per fire (National Association of State Foresters [NASF], 2009).
The short-term benefits of reducing the anticipated Os-related fire risks include the value
of avoided residential property damages; avoided damages to timber, rangeland, and wildlife
resources; avoided losses from fire-related air quality impairments; avoided deaths and injury
due to fire; improved outdoor recreation opportunities; and savings in costs associated with
fighting the fires and protecting lives and property.
For example, the California Department of Forestry and Fire Protection (CAL FIRE)
estimated that average annual losses to homes due to wildfire from 1984 to 1994 were $226
million (CAL FIRE, 1996) and were over $263 million in 2007 (CAL FIRE, 2008) in inflation
adjusted 2010$. In fiscal year 2008, CAL FIRE's budgeted costs for fire suppression activities
were nearly $304 million 2010 dollars (CAL FIRE, 2008).
CAL FIRE also estimates fire risk in the state on a -1 to 5 scale, with 2 being moderate
risk. Using GIS, we developed maps that overlay the area of California with mixed conifer
forest, an ecosystem that contains Os-sensitive species, and the fire risk area calculated by CAL
FIRE. We then generated maps overlaying the current ambient Os conditions and the modeled
alternative scenarios with the areas of mixed conifer forest that have a fire risk in the moderate
and higher range. These maps allow us to calculate the area of mixed conifer forests with
moderate to high fire risk and correlate that with high W126 index values under various
scenarios. Figure 5-3 shows W126 index values after just meeting the existing and alternative
5-5
-------
standard levels in areas in California with fire risk greater than 2 on CAL FIRE's scale.
Existing Standard
W12615ppm-hr
W12611 and7ppra-hr
Ozone Concentration x
I 10-7
7-11
11-15
Figure 5-3 Overlap of W126 Index Values for the Existing Standard and Alternative
W126 Standard Levels, Fire Threat > 2, and Mixed Conifer Forest
The highest fire risk and highest W126 index values are correlated with each other, and
with significant portions of mixed conifer forest. Under recent conditions, over 97 percent of
mixed conifer forests (21,800 square kilometers) have W126 index values over 7 ppm-hrs and a
moderate to severe fire risk, and 74 percent (16,500 square kilometers) have W126 index values
over 15 ppm-hrs with moderate to severe fire risk. When we simulate just meeting the existing
standard almost all of the area of mixed conifer forest where there is a moderate to high fire
threat sees a reduction in Cb to below a W126 index value of 7 ppm-hrs. At the adjusted
alternative W126 standard level of 15 ppm-hrs all but 40 km2 are under a W126 index value of 7
ppm-hrs and at 1 lor 7 ppm-hrs all of the moderate to high fire threat area is under 7 ppm-hrs.
Table 5-1 summarizes the reductions in areas of moderate to high-fire threat, mixed conifer
forests at the existing and alternative standard levels.
5-6
-------
Table 5-1 Area of Moderate to High-Fire Threat, Mixed Conifer Forest for Existing
and Alternative Standard Levels (in km2)
Recent Conditions
Existing Standard
(75 ppb)
15 ppm-hrs
llppm-hrs
7 ppm-hrs
<7ppm-hrs
482
22,180
22,257
22,297
22,297
7-llppm-hrs
2,542
117
40
0
0
11-15 ppm-hrs
5,271
0
0
0
0
>15 ppm-hrs
16,544
0
0
0
0
In the long term, decreased frequency of fires could result in an increase in property
values in fire-prone areas. Mueller et al. (2007) conducted a hedonic pricing study to determine
whether increasing numbers of wildfires affect house prices in southern California. They
estimated that house prices would decrease 9.7 percent after one fire and 22.7 percent after a
second wildfire within 1.75 miles of a house in the study area. After the second fire, housing
prices took between 5 and 7 years to recover.
Figure 5-4 shows the locations of fires in the mixed conifer forest range in 2010. There
were 961 fires detected in these areas, including many in the national parks. While we can't
conclude that Cb reductions would have prevented these fires because there are many
contributing factors, we can conclude that air quality adjusted to just meet the existing standard
will, in many areas, decrease the anticipated role of Os as a contributing factor by reducing the
W126 index value to below 7 in most areas. Meeting alternative W126 standards results in small
to no additional reductions in the area of forests above a 7 ppm-hrs W126 standard level.
Additionally, long- term decreases in wildfire would be expected to yield outdoor recreation
benefits consistent with the discussion of scenic beauty in subsequent sections.
5-7
-------
WhEkey*MAjr>S haste-Trjprtjr Hati
Figure 5-4 Location of Fires in 2010 in Mixed Conifer Forest Areas (under Recent
Conditions)
5.3 SUPPORTING SERVICES
Supporting services are the services needed by all of the other ecosystem services. Other
categories of services have relatively direct or short-term impacts on humans, while the impacts
on public welfare from supporting services are generally either indirect or occur over a long
time. The next sections describe potential impacts of Os on some of these supporting services.
5.3.1 Net Primary Productivity
Primary productivity underlies the provision of many subsequent ecosystem services that
are highly valued by the public, including provision of food and timber. The Os ISA determined
that biomass loss due to Os exposure may reduce net primary productivity (NPP). According to
the Os ISA (U.S. EPA, 2013), when compared to 1860's era preindustrial conditions, NPP in
5-S
-------
U.S. Mid-Atlantic temperate forests decreased 7-8 percent per year from 1991-2000 due to Os
exposure, even with growth stimulation provided by elevated carbon dioxide and nitrogen
deposition. Also, compared to a presumed pristine condition in 1860, NPP for the conterminous
U.S from 1950-1995 decreased as much as 13 percent in some areas in the agricultural region of
the Midwest during the mid-summer. While there are models available to help quantify changes
in NPP and in the hydrologic cycle discussed in Section 5.2.1 we were not able to attempt
quantification of NPP or hydrology due to resource limitations. Additionally these services are
more difficult to interpret in ways that are meaningful to people.
532 Community Composition and Habitat Provision
Community composition or structure is also affected by Os exposure. Since species vary
in their response to Os, those species that are more resistant to the negative effects of Os are able
to out-compete more susceptible species. For example, according to studies cited in the Os ISA
(U.S. EPA, 2013), the San Bernardino area community composition in high-Os sites has shifted
toward Os-tolerant species such as white fir, sugar pine, and incense cedar at the expense of
ponderosa and Jeffrey pine. Changes in community composition underlie possible changes in
associated services such as herbivore grazing, production of preferred species of timber, and
preservation of unique or endangered communities or species, among others. Table 5-2
summarizes the responses to survey questions regarding the value of wildlife habitat and
preservation of unique or endangered species.
Table 5-2 Responses to NSRE Wildlife Value Questions
Service
Wildlife Habitat
Preserving Unique Wild Plants and
Animals
Protecting Rare or Endangered
Species
Percent of Respondents Considering the Service
Important
Extremely
Important
51
44
50
Very
Important
36
36
33
Moderately
Important
9
13
11
Total*
96
93
94
*The remaining respondents felt these services were not important.
5-9
-------
There exist meta-analyses on the monetary values Americans place on threatened and
endangered species. One such study (Richardson and Loomis, 2009) estimates the average
annual willingness to pay (WTP) for a number of species. The authors report a wide range of
values dependent on the change in the size of the species population, type of species, and
whether visitors or households are valuing the species. The average annual WTP for surveyed
species ranged from $9/year for striped shiner for Wisconsin households to $26I/year for
Washington state households value for anadromous fish, such as salmon, in constant 2010$.
5.4 PROVISIONING SERVICES
Provisioning services include market goods, such as forest and agricultural products. The
direct impact of Os-induced biomass and yield loss can be predicted for the commercial timber
and agriculture markets, respectively, using the Forest and Agriculture Optimization Model
(FASOM). This model provides a national-scale estimate of the effects of Os on these two
market sectors, including producer and consumer surplus estimates (see Section 6.3 for a
discussion of producer and consumer
surplus). Chapter 6 of this document
provides detailed analyses of the potential
impact of biomass and yield loss on these
services. Non-timber forest products
(NTFP), such as foliage and branches used
for arts and crafts or edible fruits, nuts, and
berries, can be affected by the impact of Os
through biomass loss, foliar injury, insect
attack, fire regime changes, and effects
Figure 5-5 Southern Pine Beetle Damage
Courtesy: Ronald F. Billings, Texas Forest
Service. Bugwood.org
on reproduction. Acknowledging that
services lost in this sector can be the
result of interacting effects of Os with other stressors, we also have included details for the
magnitude of the NTFP services in Chapter 6.
In addition to the direct effects of Os on tree growth, Os is anticipated to cause increased
susceptibility to infestation by some chewing insects (U.S. EPA, 2006). This potentially
5-10
-------
includes tree species that are not considered sensitive to either biomass loss or foliar injury such
as Douglas fir.
Chewing insects include the southern pine beetle and western bark beetle, species that are
of particular interest to commercial timber producers and consumers. These infestations can
cause economically significant damage to tree stands and the associated timber production.
Figure 5-5 and Figure 5-6 illustrate the
damage caused by southern pine beetles in
parts of the south.
According to the USFS Report on
the southern pine beetle (Coulson and
BB Klepzig, 2011), "Economic impacts to
.
timber producers and wood-products firms
x : are essential to consider because the SPB
f J»LM^—«ML^£L
wwivrFtraa causes extensive mortality in forests that
Figure 5-6 Southern Pine Beetle Damage , ... . . . . . .
r* . T, u T^ T»-II- T" T-I • have high commercial value in a region with
Courtesy: Ronald F. Billings, Texas Forest
Service. Bugwood.org the most active timber market in the world."
The economic impacts of beetle outbreaks
are multidimensional. In the short-term, the surge in timber supply caused by owners harvesting
damaged timber depresses prices for timber and benefits consumers. In the long-term, beetle
outbreaks reduce the stock of timber available for harvest, raising timber prices to the benefit of
producers and the detriment of consumers.
The USFS further reports that over the 28 years covered in their analysis (1977-2004),
because of beetle outbreaks, timber producers have incurred losses of about $1.4 billion, or about
$49 million per year, and wood-using firms have gained about $966 million, or about $35
million per year. This results in a $15 million per year net negative economic impact. (All
dollar values are reported in constant 2010$.) These annual figures mask that most of the
economic impacts result from a few catastrophic outbreaks, causing the impacts to pulse through
the system in large chunks rather than being evenly distributed over the years. It is not possible
to attribute a portion of these impacts resulting from the effect of Os on trees' susceptibility to
insect attack; however, such losses are already reflected in the losses cited, and any welfare gains
from decreased Os would positively impact the net economic impact.
5-11
-------
In the western United States, Os-sensitive ponderosa and Jeffrey pines are subject to
attack by bark beetles. Ozone exposure is anticipated to increase susceptibility to these insect
infestations in sensitive species.
Figure 5-7 shows areas considered 'at risk' of losing 25 percent or more basal area in the
contiguous United States to the top seven pine beetle species over the next 15 years (pine beetle
projections were calculated by the Forest Health Technology Enterprise Team). Under recent
conditions, approximately 48,000 km2 have W126 index values above 15 ppm-hrs. After just
meeting the existing standard, all areas are under a W126 index value of 7 ppm-hrs with the
exception of about 4,000 km2 in Arizona and Colorado. After just meeting an alternative
standard level of 15 ppm-hrs, no area is above 7 ppm-hrs. Table 5-3 and Table 5-4 provide
summaries of areas at risk of higher pine beetle loss and millions of square feet of basal tree area
at high risk at various W126 index values.
5-12
-------
Ozone
Concentration
7-11
11-15
Ek v
W126: 15, n,and7ppm-hr
Figure 5-7 W126 Index Values for Just Meeting the Existing and Alternative Standard
Levels in Areas Considered 'At Risk' of High Basal Area Loss (>25% Loss)
Table 5-3 Area (km2)' At Risk' of High Pine Beetle Loss at Various W126 Index Values
Recent Conditions
Existing Standard
(75 ppb)
15 ppm-hrs
llppm-hrs
7 ppm-hrs
<7 ppm-hrs
3,456
80,640
84,528
84,528
84,528
7-llppm-hrs
19,440
3,888
0
0
0
11-15 ppm-hrs
13,536
0
0
0
0
>15 ppm-hrs
48,096
0
0
0
0
5-13
-------
Table 5-4 Tree Basal Area Considered 'At Risk' of High Pine Beetle Loss By W126
Index Values after Just Meeting the Existing and Alternative Standard Levels (in millions
of square feet)
Recent Conditions
Existing Standard
(75 ppb)
15 ppm-hrs
llppm-hrs
7ppm-hrs
<7 ppm-hrs
90
982
1,091
1,091
1,091
7-llppm-hrs
368
110
0
0
0
11-15 ppm-hrs
145
0
0
0
0
>15 ppm-hrs
488
0
0
0
0
In 2006, California was the largest producer of ponderosa and Jeffrey pine timber from
public lands. California accounted for 99 million board feet of saw logs - almost 40 percent of
the total U.S. production (U.S. Forest Service, 2009, available at:
http://srsfia2.fs.fed.us/php/tpo_2009/tpo_rpa_int2.php). California also experiences high W126
index values that may contribute to susceptibility to bark beetle attack. It is not possible to
attribute a quantified impact of Os exposure to economic loss from bark beetle damage because
that impact is already reflected in the loss attributed to bark beetle infestation. Reducing Os
impacts would potentially reduce economic loss to California timber production.
Figures 5-5 and 5-6 also illustrate the impact insect outbreaks can have on aesthetic
values such as scenic beauty, as well as to the impacts on timber production. As shown in the
NOx/SOx Policy Assessment (U.S. EPA, 201 le), the value of the impact of Os and insect attack
susceptibility on aesthetic values may be even greater than the market value of the timber. We
will address timber production effects from reduced growth rates in Chapter 6 and effects of
foliar injury on related ecosystem services in Chapter 7.
5.5 CULTURAL SERVICES
Cultural services include non-use values (i.e., existence and bequest values) that can be
directly or indirectly impacted by Os exposure. According to responses to the NSRE, a large
majority of Americans wishes to preserve natural or pristine areas, even if they do not intend to
visit these areas. Outdoor recreation is another cultural service that may be affected by Os
exposure. Foliar injury caused by Os exposure and insect attack aided by Os exposure may have
negative impacts on people's satisfaction with outdoor activities, especially those activities
associated with the quality of natural environments.
5-14
-------
According to the National Report on Sustainable Forests (USDA, 2011) there are
approximately 751 million acres of forest lands in the U.S., one-third of which is federally
owned (Figure 5-8). All of these lands are assumed to be protected to some degree, but specific
protections apply to wilderness areas, which comprise about 20 percent of public land. Of the
remaining lands, 7 percent is protected as national parks; 13 percent is designated as wildlife
refuges; and 60 percent is protected, managed forests, including national forests, Bureau of Land
Management lands, and other state and local government lands. The protections afford
preservation of cultural, social, and spiritual values.
Other corporate, Coip0rate forest ^Local, 1.5%
11.5%A industry, 6.8% ^
State, 9.2%
Family
individual,
V15.1% Federal,
33.1%
Other noncorporate,
2.9%
Forest land ownership (percent)
Figure 5-8 Percent of Forest Land in the US by Ownership Category, 2007
Source: USFS (Almost all forest lands are open for some form of recreation, although
access may be restricted.)
5.5.1 Non-Use Services
The NSRE surveys also track American's attitudes toward various benefits derived from
the environment, including non-use values. When people value a resource even though they may
never visit the resource or derive any tangible benefit from it, they perceive an existence service.
When the resource is valued as a legacy to future generations, a bequest service exists.
Additionally, there exists an option value to knowing that you may visit a resource at some point
in the future. Data provided by the NSRE indicates that Americans have very strong preferences
for existence, bequest, and option services related to forests. Significantly, according to the
survey, only 5 percent of Americans rate wood products as the most important value of public
5-15
-------
forests and wilderness areas, and for private forests, only 20 percent of respondents rated wood
products as most important. Table 5-5 details the survey responses to these questions.
Table 5-5 NSRE Responses to Non-Use Value Questions For Forests
Service
Existence
Option
Bequest
Percent of Respondents Considering the Service Important
Extremely
Important
36
36
81
Very Important
38
37
12
Moderately
Important
18
17
4
Total*
92
90
97
*Remaining respondents felt these services were not important.
Studies (Haefele et al., 1991, Holmes and Kramer, 1995) indicate that the American
public places a high value on protecting forests and wilderness areas from the damaging effects
of air pollution. These studies assess willingness-to-pay (WTP) for spruce-fir forest protection in
the southeast from air pollution and insect damage and confirm that the non-use values held by
the survey respondents were in fact greater than the use or recreation values. The survey
presented respondents with a sheet of color photographs representing three stages of forest
decline and explained that, without forest protection programs, high-elevation spruce forests
would all decline to worst conditions. Two potential forest protection programs were proposed.
The first program (minimal program) would protect the forests along road and trail corridors
spanning approximately one-third of the ecosystem at risk. This level of protection may be most
appealing to recreational users. The second level of protection (more extensive program) was for
the entire ecosystem and may be most appealing to those who value the continued existence of
the entire ecosystem. Median household WTP was estimated to be roughly $29 (in 2007 dollars)
for the minimal program and $44 for the more extensive program. Respondents were then asked
to decompose their value for the extensive program into use, bequest, and existence values. The
results were 13 percent for use value, 30 percent for bequest, and 57 percent for existence value
(Table 5-6).
While these studies are specific to damage due to excess nitrogen deposition and the
wooly balsam adelgid (a pest in Fraser fir), the results are relevant to Os exposure in forests
because the effects are similar. In the southeast, loblolly pine is a prevalent species and Os foliar
5-16
-------
injury can cause visible damage. Ozone exposure is also anticipated to result in trees more
susceptible to insect attack, which in the southeast would include damage caused by the southern
pine beetle.
Table 5-6 Value Components for WTP for Extensive Protection Program for Southern
Appalachian Spruce-Fir Forests
Type of Value
Use
Bequest
Existence
Total
Proportion of WTP
0.13
0.30
0.57
1.0
Component Value ($2007)
5.72
13.20
25.08
44.00
5.6 QUALITATIVE ASSESSMENT OF UNCERTAINTY
As noted in Chapter 3, we have based the design of the uncertainty analysis for this
assessment on the framework outlined in the WHO guidance (WHO, 2008). For this qualitative
uncertainty analysis, we have described each key source of uncertainty and qualitatively assessed
its potential impact (including both the magnitude and direction of the impact) on risk results, as
specified in the WHO guidance. In general, this assessment includes qualitative discussions of
the potential impact of uncertainty on the results (WHO Tierl) and quantitative sensitivity
analyses where we have sufficient data (WHO Tier 2).
Table 5-7 includes the key sources of uncertainty identified for the Os WREA. For each
source of uncertainty, we have (a) provided a description, (b) estimated the direction of influence
(over, under, both, or unknown) and magnitude (low, medium, high) of the potential impact of
each source of uncertainty on the risk estimates, (c) assessed the degree of uncertainty (low,
medium, or high) associated with the knowledge-base (i.e., assessed how well we understand
each source of uncertainty), and (d) provided comments further clarifying the qualitative
assessment presented. The categories used in describing the potential magnitude of impact for
specific sources of uncertainty on risk estimates (i.e., low, medium, or high) reflect our
consensus on the degree to which a particular source could produce a sufficient impact on risk
5-17
-------
estimates to influence the interpretation of those estimates in the context of the secondary O3
NAAQS review. Where appropriate, we have included references to specific sources of
information considered in arriving at a ranking and classification for a particular source of
uncertainty.
5-18
-------
Table 5-7 Summary of Qualitative Uncertainty Analysis in Semi-Quantitative Ecosystem Services Assessments
Source
Description
Potential influence of
uncertainty on risk
estimates
Direction
Magnitude
Knowledge-
Base
Comments (KB: knowledge base, INF: influence of
uncertainty on risk estimates)
A. National W126
surfaces
The fire risk and bark beetle
analyses in this chapter use the
national W126 surfaces for
recent conditions and adjusted to
just meet the existing standard
and alternative W126 standards.
Both
Low-
Medium
Low-Medium
KB and INF: See Chapter 4 for more details.
B. Incremental impact
of O3 on ecosystem
services
Many ecosystem services
affected by Os exposure are
discussed qualitatively or semi-
quantitatively, including
supporting services (e.g., net
primary productivity and
community composition),
regulating services (e.g.,
hydrologic cycle and
pollination), and cultural
services (e.g., recreation and
non-use).
Under
High
Low
KB: The Os ISA concludes that there is a causal relationship
between Os exposure and productivity in terrestrial ecosystems
and biogeochemical cycles, and a likely to be causal
relationship between O3 exposure and terrestrial water cycling
and terrestrial community composition (U.S. EPA, 2011).
However, we do not have sufficient data, methods, or resources
to adequately quantify the incremental effects of changes in Os
on many ecosystem services.
INF: For many services, we can estimate the current total
magnitude and, for some, we can estimate the current
monetized value. The estimates of current service provision
will reflect the loss of services occurring from historical and
current O3 exposure and provide context for the importance of
any potential impacts of Os on those services, e.g., if the total
value of a service is small, the total value of the likely impact of
Os exposure will also be small. Likewise, if the total value is
large, there is a higher potential for significant damage, even if
the relative contribution of Os as a stressor is small.
5-19
-------
Source
Description
Potential influence of
uncertainty on risk
estimates
Direction
Magnitude
Knowledge-
Base
Comments (KB: knowledge base, INF: influence of
uncertainty on risk estimates)
C. Areas with fire risk
in California
Maps of areas with moderate and
higher fire risk have uncertainty,
and thus the potential overlap
with areas with higher W126
index values and mixed conifer
forests are also uncertain.
Unknown
Medium
High
KB: California's fire risk maps are systematically developed
including consideration of factors such as defensible space,
non-flammable roofs, and ignition resistant construction reduce
fire risk. (See
http://www.fire.ca.gov/firejrevention/firejrevention wildlan
d zones developmentphp).
INF: In 2010, over 3 million acres burned in wildland fires
(MFC, 2010). The economic value of homes lost due to
wildfire and fire suppression activities can be hundreds of
millions of dollars per year in California (CAL Fire, 2006,
2007, 2008).
D. Areas at risk due to
bark beetle
In the western U.S., Os-sensitive
ponderosa and Jeffrey pines are
subject to attack by bark beetles.
Maps that identify areas
considered 'at risk' of losing 25
percent or more basal area to
pine beetle have uncertainty, and
thus the potential area of overlap
with areas with higher W126
index values are also uncertain.
Unknown
Medium
Medium
KB: Ozone causes increased susceptibility to infestation by
some chewing insects (U.S. EPA, 2006, 2013), including the
southern pine beetle and the western bark beetle. It is not
possible to attribute a portion of these impacts resulting from
the effect of Os on trees' susceptibility to insect attack;
however, such losses are already reflected in the losses cited,
and any welfare gains from decreased Os would positively
impact these numbers.
INF: Insect infestations can cause economically significant
damage to tree stands and the associated timber production.
USFS estimates a$15 million per year net negative economic
impact due to bark beetle infestations (Coulson and Klepzig,
2011).
5-20
-------
5.7 KEY OBSERVATIONS
Ozone damage to vegetation and ecosystems from recent conditions causes widespread
impacts on an array of ecosystem services. Biomass loss impacts numerous services, including
supporting and regulating services such as net primary productivity, community composition,
habitat, and climate regulation. The provisioning services of timber production can be affected
by the increased susceptibility to insect attack caused by Os exposure. Non-use values, including
existence and bequest values, are also affected by the damage to scenic beauty caused by insect
attack (an indirect effect of Os) and foliar injury (a direct effect). Below we offer a few
observations on the challenges of explicitly valuing ecosystem services, highlight the importance
of continuing to consider the services in our assessments, and indicate where additional analyses
and discussion on valuing the ecosystem services are located in this document.
• Most of the impacts of Os exposure on ecosystem services cannot be specifically
quantified, but it is very important to provide an understanding of the magnitude and
significance of the services that may be harmed by Os exposure. For many ecosystem
services, we can estimate the current total magnitude and, for some, we can estimate
the current value of the services in question.
• Regulating ecosystem services include hydrologic cycle, pollination, and fire
regulation. Hydrologic, or water cycling in forests is affected by Os exposure and has
impacts on ecosystem services associated with both water quality and quantity.
While the NSRE results show that 91 percent of respondents rank water quality
protection as either extremely important or very important, because of data and
methodology limitations, quantifying the loss of value to the public from incremental
changes in Os exposure on water cycling is not currently feasible. Tor pollination
services, it is not currently feasible to explicitly calculate the loss of pollination
resulting from Os exposure, but the loss is reflected in the current total estimated
value of $18.3 billion (2010$) for pollination services in North America. Lastly, fire
regulation may be negatively affected by Os exposure through forest susceptibility to
wildfires, drought stress, and insect attack. The value of this ecosystem service is
5-21
-------
reflected in avoided damage to residential property, timber, rangeland, and wildfire
fighting resources, as well as improved outdoor recreation opportunities. As an
example, the California Department of Forestry and Fire Protection (CAL FIRE)
estimated that average annual losses to homes due to wildfire from 1984 to 1994 were
$163 million (CAL FIRE, 1996) and were over $250 million in 2007 (CAL FIRE,
2008). In fiscal year 2008, CAL FIRE's costs for fire suppression activities were
nearly $300 million (CAL FIRE, 2008).
The impacts on public welfare from supporting services are generally either indirect
or occur over a long time. The Os ISA determined that biomass loss due to Cb
exposure may have adverse effects on net primary productivity. But because of data
and methodology limitations, quantifying the loss of value to the public from
incremental changes in Os exposure on NPP on a national level is not feasible at this
time. Also, it is currently not feasible to quantify the impacts of Os exposure on
community composition.
Provisioning services include market goods, such as forest and agriculture products.
The direct impact of Os-induced biomass loss can be predicted for the commercial
timber and agriculture markets using the Forest and Agriculture Optimization Model.
Chapter 6 of this document provides detailed analyses of the potential impact of
biomass and yield loss on these services. In addition, non-timber forest products
(NTFP), such as foliage and branches used for arts and crafts or edible fruits, nuts,
and berries, can be affected by the impact of Cb through biomass loss, foliar injury,
insect attack, fire regime changes, and effects on reproduction. We include details for
the magnitude of the NTFP services in Chapter 6.
In addition, to estimate the magnitude of insect attacks related to Os exposure on
provisioning services, such as forest products, we reviewed the USFS Report on the
Southern Pine Beetle (Coulson and Klepzig, 2011). The USFS further reports that
over the 28 years covered in their analysis (1977-2004), because of beetle outbreaks,
timber producers have incurred losses of about $1.4 billion, or about $49 million per
year, and wood-using firms have gained about $966 million, or about $35 million per
5-22
-------
year. This results in a $15 million per year net negative economic impact.1 While it
is not currently feasible to attribute a portion of these impacts resulting from the
effect of Os on trees' susceptibility to insect attack, these losses are reflected in the
values cited.
Outdoor recreation is a cultural service that may be affected by Os exposure. Foliar
injury caused by Os exposure and insect attack aided by Os exposure may have
negative impacts on people's satisfaction with outdoor activities, especially those
activities associated with the quality of natural environments. These impacts are
discussed in Chapter 7 on foliar injury. In addition, some cultural services, such as
existence or bequest services, lend themselves to evaluating total importance and
measuring total value, but assessing the impact of Os effects on these services is not
currently feasible.
1 All values are reported in constant 2010$.
5-23
-------
6 BIOMASS LOSS
6.1 INTRODUCTION
The previous O3 AQCDs (U.S. EPA, 1996, 2006) and current O3 ISA (U.S. EPA, 2013)
concluded that there is strong and consistent evidence that ambient Os decrease photosynthesis
and growth in numerous plant species, but the magnitude of the effects are variable both across
species and across regions of the U.S.
The ecosystem services most directly affected by biomass loss include: (1) habitat
provision for wildlife, particularly habitat for threatened or endangered wildlife, (2) carbon
storage, (3) provision of food and fiber, and (4) pollution removal (see Figure 6-1). Although we
cannot quantify reduction in habitat provision due to Os exposure on either a national or case
study scale, there is evidence that this service is important to the public.
Provlalofiifn Serultct
• Ngn.Timb«r Uws
" CcTimi^nifll Npr-Trnber
Fa raa Product:
iffril Pr
/"~~ ~"\
Etoirslem-Level Effects
• WfieKtwi BlofP«s
UM
•Sptcies DhwrsltK
•QynrnuKiiiv Structure
Functioning
•Pollution ft«ntov*l
'. ii Fciluriji- R
• Tlmbet PI .j-.i .i-.'. . .
• ImputE on Producor: ami
fonnmtrs
\S*(
• CntngK it: Mtkin.il Yield ind Prlui
• Impact! on Fr oouior; and
R
-------
the approximate loss of services and the marginal benefits of alternative levels of a W126
standard.
We included national parks at the case-study scale, as well as Class I areas. Class I areas
are designated as areas in which visibility has been determined to be of important value (C.F.R.
40, 81.400). The determination is primarily based on air quality limitations on visibility, but in
this assessment we are using them in the context of protected areas of interest to address
potential impacts. The national parks are meant to be preserved for the enjoyment of present and
future generations, as well as for the unique or sensitive ecosystems and species in the parks.
The parks are not a source of food or fiber production and are not included in the analysis of
those services. And although the parks do provide carbon sequestration and storage and
pollution removal, neither of the models for these ecosystem services available for this review
was able to include national parks. The model used for the urban case study areas allows
analysis of carbon sequestration and storage and pollution removal services; it does not include
habitat provision or food and fiber production.
The remainder of this Chapter includes Section 6.2 - Relative Biomass Loss; Section 6.3
- Commercial Timber Effects; Section 6.4 - Non-Timber Forest Products; Section 6.5 -
Agriculture; 6.6 - Climate Regulation; Section 6.7 - Urban Case Study Air Pollution Removal;
and Section 6.8 - Ecosystem Level Effects.
6.2 RELATIVE BIOMASS LOSS
The 1996 and 2006 Os AQCDs relied extensively on results from analyses conducted on
commercial crop species for the National Crop Loss Assessment Network (NCLAN) and on
analyses of tree seedling species conducted by the EPA's National Health and Environmental
Effects Laboratory Western Ecology Division (NHEERL/WED). Results from these studies
have appeared in numerous publications, including Lee et al. (1994; 1989, 1988b, 1987), Hogsett
et al. (1997), Lee and Hogsett (1999), Heck et al. (1984), Rawlings and Cure (1985), Lesser et al.
(1990), and Gumpertz and Rawlings (1992). Those analyses concluded that a three-parameter
Weibull model is the most appropriate model for the response of absolute yield and growth to Os
exposure because of the interpretability of its parameters, its flexibility (given the small number
of parameters), and its tractability for estimation. See equation 6-1 for an example of a three-
parameter Weibull model.
6-2
-------
Y = a e'
Equation 6-1
In addition, if the intercept term, a, is removed, the model estimates relative yield or
biomass without any further reparameterization. Formulating the model in terms of relative yield
or biomass loss (RBL) in relation to the 3-month W126 index is essential for comparing
exposure-response across species or genotypes or for experiments for which absolute values of
the response may vary greatly. See equation 6-2 for the reformulated model.
RBL =1- exp[-(W126/tt)ls]
Equation 6-2
In the 1996 and 2006 Os AQCDs, the two-parameter model of RBL was used to derive
common models for multiple species, multiple genotypes within species, and multiple locations.
Relative biomass loss (RBL) functions for the 12 tree species used in this assessment are
presented in Table 6-1 (see the Os ISA (U.S. EPA, 2013) for a more extensive review of the
calculation of the E-R functions), and RBL functions for the 10 crop species used in this
assessment are presented in Table 6-2. Relative biomass loss is annual.
6-3
-------
Table 6-1 Relative Biomass Loss Functions for Tree Species
Species
Red Maple (Acer rubrum)
Sugar Maple (Acer saccharum)
Red Alder (Alnus rubra)
Tulip Poplar (Liriodendron tulipifera)
Ponderosa Pine (Pinus ponderosa)
Eastern White Pine (Pinus strobus)
Loblolly Pine (Pinus taeda)
Virginia Pine (Pinus virginiana)
Eastern Cottonwood (Populus deltoides)
Quaking Aspen (Populus tremuloides)
Black Cherry (Prunus serotina)
Douglas Fir (Pseudotsuga menzeiesii)
RBL Function
exp[ (W126/T,) ]
ri(ppm)
318.12
36.35
179.06
51.38
159.63
63.23
3,966.3
1,714.64
10.10
109.81
38.92
106.83
P
1.3756
5.7785
1.2377
2.0889
1.1900
1.6582
1.0000
1.0000
1.7793
1.2198
0.9921
5.9631
Table 6-2 Relative Biomass Loss Functions for Crop Species
Species
Barley
Field Corn
Cotton
Kidney Bean
Lettuce
Peanut
Potato
Grain Sorghum
Soybean
Winter Wheat
RBL Function
exp[ (W126/T,) ]
TI (ppm)
6,998.5
97.9
96.1
43.1
54.6
96.8
99.5
205.3
110.2
53.4
P
1.388
2.968
1.482
2.219
4.917
1.890
1.242
1.957
1.359
2.367
Figure 6-2 shows a comparison of W126 median RBL response functions for the tree
species used in this assessment, and Figure 6-3 shows a comparison of W126 median RBL
response functions for the crop species used in this assessment. The figures illustrate how the
two parameters affect the shape of the resulting curves. Differences in the shapes of these curves
are important for understanding differences in the analyses presented later in this chapter. The
6-4
-------
two parameters of the RBL equation (Equation 6-2) control the shape of the resulting curve. The
value of r| in the RBL function affects the inflection point of the curve, and P affects the
steepness of the curve. Species with smaller values of P (e.g., Virginia Pine) or species with r|
values that are above the normal range of ambient W126 measurements (e.g., Ponderosa Pine
and Red Alder) have response functions with more gradual and consistent slopes. This results in
a more constant rate of change in RBL over a range of Os exposure consistent with ambient
exposure concentrations.
In contrast, the species with larger P values (e.g., Sugar Maple) have response functions
that behave more like thresholds, with large changes in RBL over a small range of W126 index
values and relatively small changes at other index values. In these cases the "threshold" is
determined by the t] parameter of the model.
00
ci
CD
ci
GO
a:
CN
ci
p
ci
Red Maple
Sugar Maple
Red Alder
Tulip Poplar
Ponderosa Pine
White Pine
Loblolly Pine
Virginia Pine
Cottonwood
Aspen
Black Cherry
Douglas Fir
10 20 30
W126 (ppm-hrs)
40
50
Figure 6-2 Relative Biomass Loss Functions for 12 Tree Species
6-5
-------
00
o
CD
O
CN
O
q
o
10 20 30
W126 (ppm-hrs)
40
50
Figure 6-3 Relative Yield Loss Functions for 10 Crop Species
The shape of curves presented in Figure 6-2 and Figure 6-3 also determine how sensitive
the RBL value is to changes in Os. In addition, Figure 6-4 illustrates the elasticity in RBL
relative to W126. The percent change in RBL relative to a 1 percent change in W126 is plotted
on the y-axis across a range of W126 values. Two species, Loblolly Pine (dark grey line) and
Virginia Pine (yellow line) have E-R functions that are linear within the W126 range represented
on the x-axis, meaning that a 1 percent change in W126 produces an equal change in RBL. Black
Cherry (blue line) has an E-R function that is asymptotic (Figure 6-2), which produces a smaller
change in RBL relative to the change in W126. The E-R function for Cottonwood (turquoise
line) produces large changes in RBL at W126 values below 10, but then rapidly levels off. The
remaining species all have E-R functions that produce consistent percent changes in RBL
relative to changes in W126.
6-6
-------
p
oi
CD
CD
D)
I
CD
CL
p
O
10
I
15
I
20
W126(ppm-hrs)
Figure 6-4 Elasticity in Relative Biomass Loss Compared to Changes in W126
[The line colors correlate with the colors used in Figure 6-2]
6.2.1 Species-Level Analyses
6.2.1 1 Comparison of seedling to adult tree biomass loss
The response functions for tree species used in this analysis are all based on seedlings
grown in open top chambers (OTC). Since the 2006 Os AQCD (U.S. EPA, 2006), several studies
were published based on the Aspen Free-Air Carbon Dioxide Enrichment (FACE)1 experiment
using "free air," Os and CCh exposures in a planted forest in Wisconsin. Overall, the studies at
the Aspen FACE experimental site were consistent with many of the open-top chamber (OTC)
studies that were the foundation of previous Os NAAQS reviews. These results strengthen our
1 The Aspen FACE experiment is a multidisciplinary study to assess the effects of increasing tropospheric O3 and
carbon dioxide levels on the structure and function of northern forest ecosystems.
6-7
-------
understanding of Os effects on forests and demonstrate the relevance of the knowledge gained
from Aspen tree seedlings grown in OTC studies.
In the 2006 AQCD (U.S. EPA, 2006), the TREGRO and ZELIG models were used to
simulate growth of adult trees. For this analysis we did not conduct new TREGRO or ZELIG
simulations. We used several existing publications, which modeled tree species used in this
analysis. For this analysis, we calculated the W126 index values from the hourly concentrations
at the monitors used in the studies. The seedling RBL was calculated from this W126 and
compared to the study results and adjusted to reflect an annualized RBL. The results are
summarized below in Table 6-3.
Table 6-3 Comparison of Adult to Seedling Biomass Loss
Study
Constable
and Taylor,
1997
Weinstein et
al., 2001
W126
0.18
8.98
46.37
89.40
149.22
0.32
15.38
59.17
0.32
15.38
59.17
0.32
15.38
59.17
Adult RBL
TREGRO
0%
0.3%
3.1%
6.4%
12.1%
Tulip Polar
4.7%
10.6%
16.8%
Red Maple
2.5%
4.9%
8.2%
Black Cherry
0.2%
0.3%
0.5%
Adult RBL
ZELIG
N/A
Tulip Polar
+3.2%
5.3%
11.2%
Red Maple
0%
15.6%
15.6%
Black Cherry
11.2%
4.2%
+9.1%
Seedling RBL
0.03%
3.2%
20.5%
39.5%
60.3%
Tulip Poplar
0%
7.7%
73.89%
Red Maple
0.01%
1.5%
9.4%
Black Cherry
0.9%
32.8%
78.0%
Comments
This study used
TREGRO and included
the western and eastern
subspecies ofPondemsa
Pine. Ozone data were
not available for the
western subspecies,
which was found to be
more sensitive than the
eastern subspecies. The
seedling E-R function
used does not
differentiate between
subspecies.
This study used
TREGRO and ZELIG to
model Tulip Poplar, Red
Maple, and Black
Cherry.
6-8
-------
These studies indicate that overall, the seedling biomass loss values are much more
consistent with the adult loss, as estimated by TREGRO and ZELIG, at lower W126 index
values. The Constable and Taylor (1997) study implies that for the eastern subspecies of
Ponderosa Pine, the seedling RBL rate overestimates the adult RBL rate. Ozone data for the
western subspecies were not available, but Constable and Taylor (1997) found the western
subspecies to be more sensitive. The Weinstein et al. (2001) study indicates that the seedling
RBL estimates are comparable to the adult estimates, except at higher W126 index values of Os
for Tulip Poplar. The Black Cherry results are an exception, which tells us that this species is
possibly less sensitive as an adult than as a seedling. As such, the seedling RBL rate would
overestimate RBL loss in adult trees. Another study (Samuelson and Edwards, 1993) on Red
Oak, another hardwood species, found the exact opposite pattern - adult trees are much more
sensitive to Os-related biomass loss than seedlings.
Mclaughlin et al. (2007) assessed the interactive effects of Os and climate on tree growth
and water use. We used the monitored Os concentrations in this study to calculate the W126
index value and then used these values to compare the predicted seedling RBL to observations in
the study. The study did not use absolute biomass loss, instead relying on measurements of
circumference to address growth, so can only be used as a general comparison to estimates of
RBL. In addition, the results were presented as comparisons in growth in 2002 and 2003 relative
to 2001. Table 6-4 presents a summary of the results.
Table 6-4 Comparison of Seedling Biomass Loss to Adult Circumference
Species
Tulip Poplar
Tulip Poplar
Tulip Poplar
Black Cherry
Red Maple
Sugar Maple
W126
2001
23.31
19.78
14.71
14.71
14.71
14.71
2002
39.82
32.14
17.50
17.50
17.50
17.50
2003
20.15
11.25
9.22
9.22
9.22
9.22
Study Results
(% change in
circumference)
2002
-26%
-49.6%
-62%
-75%
-59.6%
-43.5%
2003
-38%
7.5%
N/A
N/A
N/A
N/A
RBL (seedling)
2001
-17.5%
-12.7%
-7.1%
-31.7%
-1.5%
-0.5%
2002
-44.4%
-31.3%
-10.0%
-36.4%
-1.8%
-1.5%
2003
-13.9%
-4.1%
-2.7%
-21.3%
-0.8%
-0.04%
Comparison
2002
-60.7%
-59.4%
-72.8%
-41.5%
-58.4%
-97.5%
2003
32.4%
210%
N/A
N/A
N/A
N/A
6-9
-------
Relative to the observed changes in circumference, the seedling RBL estimates are mixed
for Tulip Polar. A loss was overestimated estimated in 2002 (as compared to 2001) but was
underestimated in 2003. The results for Sugar Maple were similar to Tulip Poplar, with loss
overestimated in 2002. In contrast to the TREGRO results presented above, the results in this
study found much greater loss in Black Cherry, and the seedling RBL underestimated the change
for adult trees in 2002. The results for Red Maple were very similar for 2002. Table 6-5
summarizes the uncertainty for all species used in this study.
Table 6-5 Summary of Uncertainty in Seedling to Adult Tree Biomass Loss
Comparisons
Species
Red Maple (Acer rubrum)
Sugar Maple (Acer saccharum)
Red Alder (Alnus rubra)
Tulip Poplar (Liriodendron
tulipifera)
Ponderosa Pine (Pinus
ponderosa)
Eastern White Pine (Pinus
strobus)
Loblolly Pine (Pinus taeda)
Virginia Pine (Pinus virginiana)
Eastern Cottonwood (Populus
deltoides)
Summary of Seedling-Adult Uncertainty
Seedling E-R functions underestimated RBL relative to estimates of adult
biomass loss from TREGRO and ZELIG. The seedling RBL was
comparable to field results of changes in circumference.
No TREGRO data were available. Seedling RBL overestimated loss
compared to field results of changes in circumference.
No data were available.
Seedling E-R functions underestimated RBL relative to results from
TREGRO and ZELIG at lower W126 index values of O3, and overestimated
RBL at the very high index values. Seedling RBL overestimated loss
compared to field results of changes in circumference in 2002, but
underestimated loss in 2003.
Seedling E-R functions overestimated RBL relative to TREGRO results for
the eastern subspecies. Data were not available for the western subspecies,
but the western subspecies is known to be more sensitive.
No data were available.
No comparable data were available; however this species is very non-
sensitive as measured by the seedling E-R function, so the risk of
overestimating loss is low.
No comparable data were available; however this species is very non-
sensitive as measured by the seedling E-R function, so the risk of
overestimating loss is low.
No data were available for this species. Two studies on the closely related
Black Poplar (Populus nigra) found that species to be highly sensitive to Os
exposure as measured by tree ring width (Novak et al., 2010) and growth
rate (Bortier et al., 2000). This supports the high sensitivity in Eastern
Cottonwood as measured by the seedling E-R function, but there is risk of
overestimating loss in this species.
6-10
-------
Quaking Aspen (Populus
tremuloides)
Ozone gradient studies and FACE experiments have found effects on adult
trees consistent with earlier OTC studies on seedlings (Karnosky et al,
1999).
Black Cherry (Prunus serotina)
Seedling E-R functions overestimated RBL relative to results from
TREGRO and ZELIG, except the ZELIG results at the lowest W126 index
values. Seedling RBL underestimated loss relative to field results of changes
in circumference.
Douglas Fir (Pseudotsuga
menzeiesii)
No comparable data were available; however this species is very non-
sensitive as measured by the seedling E-R function, so the risk of
overestimating loss is low.
6.2.1.2 W126 for Different levels of Biomass Loss
The E-R functions can be plotted as a function of the percent biomass loss against
varying W126 index values. This allows us to compare the W126 index values associated with a
range of biomass loss values. Figure 6-5 and Figure 6-6 reflect two separate graphical
representations of these results for trees and crops respectively.
In each graph, the red line represents the median W126 index value associated with the
percent biomass value on the x-axis when all 54 crop studies or 52 tree seedling studies are
included. The green line is the value when only the composite E-R function is used for each of
the species included (10 crop species and 12 tree species). The grey lines are included as
sensitivity analyses to assess the effect of within-species variability. For each grey line, a E-R
function for each species was randomly selected from the available studies, with the resulting
line representing the median value of the 12 tree species and 10 crops. For some species only one
study was available (e.g., Red Maple), and for other species there were as many as 11 studies
available (Ponderosa Pine). The process was repeated 1,000 times, and the median value is
plotted as the red points for biomass loss values of 1 percent to 7 percent, and 10 percent. The
error bar associated with the points represents the 25th and 75th percentiles. For tree and crop
species, the median W126 index values are similar, when using all of the studies or just the
composite E-R function for each species; however, the median value is higher when within-
species variability is included.
6-11
-------
1 I I I I
0 1 2 3 4
I I I I I I I I I I I
5 6 7 8 9 10 11 12 13 14 15
Percent Biomass Loss
Figure 6-5 W126 Index Values for Alternative Percent Biomass Loss for Tree Species
! I , '
01234567
| . ,
9 10 11 12 13 14 15
Percent Biomass Loss
Figure 6-6 W126 Index Values for Alternative Percent Biomass Loss for Crop Species
6-12
-------
6.2.1.3 Individual Species Analyses
Using GIS (ESRI®, ArcMAP™ 10), we used the E-R functions listed in Table 6-1 to
generate RBL surfaces for the 12 trees species. We created the surfaces using recent ambient Os
conditions based on monitored data from 2006 through 2008 and the four Os rollback surfaces
simulating just meeting the existing 8-hr secondary standard of 75 ppb (4th highest daily
maximum) and three alternative W126 scenarios of 7, 11 and 15 ppm-hrs (see Chapter 4 for a
more detailed description of the Os surfaces). We present the maps for one species, Ponderosa
Pine, to illustrate the results (see Figure 6-7, Figure 6-8, Figure 6-9, Figure 6-10, and Figure
6-11). RBL surfaces for 10 species are presented in Appendix 6 A (Maps of Individual Tree
Species). It is important to note that these maps represent the RBL value for one tree species
within each CMAQ grid cell represented, so these maps should be interpreted as indicating
potential risk to individual trees of that species growing in that area.
We based the ranges for the species on data from the Forest Health Technology
Enterprise Team (FHTET) of the USFS (http://www.fs.fed.us/foresthealth/technology/). These
data provide modeled predictions of stand density and basal area. The modeled data were
estimated in 1,000 square meter grids for individual tree species, as well as total basal area. We
summed these values into the larger CMAQ grid cells (12 km x 12 km) used for the Cb surfaces.
For the individual species analyses, these data were used only as a predictor of presence or
absence. In the ecosystem level analysis presented in Section 6.8 these values were used to scale
the biomass loss by the proportion of total basal area for each species.
Overall, the western tree species have more fragmented habitats than the eastern species.
The areas in southern California have the highest W126 index values, which can be seen as the
very high areas of RBL in Figure 6-7. The eastern tree species had less fragmented ranges and
areas of elevated RBL that were more easily attributed to urban areas (e.g., Atlanta, GA and
Charlotte, NC) or to the Tennessee Valley Authority region. In addition to the two western
species not illustrated here, we include maps for the eastern species in Appendix 6A.
6-13
-------
Ponderosa Pine (Pinus ponderosa) (Recent Conditions)
RBL
I 00066&0- 0.020314
| 0020315-0032036
0-032037 - 0 045889
0045S90 - 0.062250
0062261- 0082266
0 082267 -0.113164
| 0113165-0164512
I 0164613-0.143435
Figure 6-7 Relative Biomass Loss (RBL) of Ponderosa Pine (Pinus ponderosa) Seedlings
under Recent Ambient W126 Index Values (2006 - 2008)
Ponderosa Pine (Pinus ponderosa) (Current Standard)
RBL
| 0 000767 - 0.003953
I 000395J-0005914
0005816-0008482
0 008483 -0.011989
0011990- 0.015976
0.015977-0.021390
| 0021391-0028521
I 0.028522-0040549
Figure 6-8 Relative Biomass Loss (RBL) of Ponderosa Pine with
Adjusted to Meet the Existing (8-hr) Primary Standard (75 ppb)
Exposure After
6-14
-------
Ponderosa Pine (Pings ponderosa) (15 ppm-hr$)
RBL
| 0.000729-0003815
| 0 003616 • 0 005624
0.005625-0007219
0 007220 • 0 009466
0 009489- 0,012511
| 0012512 -0015953
| 0015954-0.020754
I 0020755-0.032525
Figure 6-9 Relative Biomass Loss (RBL) of Ponderosa Pine with Os Exposure After
Adjusted to Meet an Alternative Secondary Standard of 15 ppm-hrs (after Meeting
Existing Os Standard)
Ponderosa Pine (Pinus ponderosa) (11 ppm-hrs)
RBL
| 0 000687 - 0 002635
| 0 002636 - 0.004073
0 004074.0 005543
0 005544 • 0 007192
0007193-0009559
0009560-0012992
| 0012993-0019336
I 0 019337 - 0 032525
Figure 6-10 Relative Biomass Loss (RBL) of Ponderosa Pine with Os Exposure After
Adjusted to Meet an Alternative Secondary Standard of 11 ppm-hrs (after Meeting
Existing Os Standard)
6-15
-------
Ponderosa Pine (Pinus ponderosa) (7 ppm-hrs)
RBL
I 0.000643-0.002341
| 0.002342 • 0.003536
0.003537-0.004630
0,004691 - 0.005921
0 005922 - 0.007591
0.007592-0.010159
| 0.010160-0.014979
I 0014980-0024900
Figure 6-11 Relative Biomass Loss (RBL) of Ponderosa Pine with Os Exposure After
Adjusted to Meet an Alternative Secondary Standard of 7 ppm-hrs (after Meeting Existing
O3 Standard)
6-16
-------
Table 6-6 Individual Species Relative Biomass Loss Values - Median, 75th Percentile, Maximum Percentages
Species
Red Maple
Sugar Maple
Red Alder
Tulip Poplar
Ponderosa Pine
White Pine
Loblolly Pine
Virginia Pine
Cottonwood
Aspen
Black Cherry
Douglas Fir
Relative Biomass Loss
(Median/75th Percentile/Maximum Percentages)
Recent Conditions
0.95/1.25/3.49
0.06/0.22/3.96
0.83/1.15/10.10
5.20/6.88/24.68
3.71/5.93/24.34
3.33/5.58/14.70
0.30/0.36/0.71
0.77/0.88/1.63
58.32/74.03/99.79
3.71/6.54/27.51
23.97/28.54/51.51
<0.01/<0.01/0.46
75ppb
0.08/0.17/0.77
<0.01/<0.01/0.07
0.32/0.40/0.78
0.17/0.35/2.79
0.67/1.18/4.05
0.10/0.40/2.66
0.05/0.07/0.17
0.15/0.20/0.54
5.93/11.97/65.90
0.47/1.14/5.85
4.89/7.94/23.90
<0.01/<0.01/<0.01
15 ppm-hrs
0.08/0.17/0.77
<0.01/<0.01/0.07
0.32/0.40/0.78
0.17/0.35/2.79
0.65/0.94/3.25
0.10/0.40/2.66
0.05/0.07/0.17
0.15/0.20/0.54
5.87/11.68/65.90
0.46/1.03/4.22
4.89/7.94/23.90
<0.01/<0.01/<0.01
11 ppm-hrs
0.08/0.13/0.70
<0.01/<0.01/0.01
0.32/0.40/0.78
0.12/0.21/2.40
0.56/0.69/3.25
0.10/0.30/2.05
0.05/0.06/0.15
0.12/0.16/0.50
5.26/8.06/53.33
0.45/0.82/3.89
4.51/6.31/19.42
<0.01/<0.01/<0.01
7 ppm-hrs
0.05/0.08/0.39
<0.01/<0.01/<0.01
0.31/0.39/0.78
0.05/0.09/0.93
0.50/0.58/2.49
0.09/0.17/1.60
0.04/0.05/0.09
0.08/0.10/0.32
3.74/5.06/35.29
0.43/0.72/3.03
3.41/4.41/13.68
<0.01/<0.01/<0.01
6-17
-------
Table 6-6 above includes individual species relative biomass loss values at the median,
the 75th percentile, and the maximum for the 12 tree species for which we have E-R functions.
The values in the table are median, 75th percentile, and maximum percentages. We include the
relative biomass loss values for each species at recent conditions, when adjusted to just meet the
existing standard of 75 ppb, and when adjusted to meet potential alternative standard levels of
15, 11, and 7 ppm-hrs.2 For Ponderosa Pine, at recent conditions, the median value is 3.71
percent RBL, the 75th percentile value is 5.93 percent RBL, and the maximum value is 24.24
percent RBL. When adjusted to just meet the existing standard, the median value is 0.67 percent
RBL, the 75th percentile value is 1.18 percent RBL, and the maximum value is 4.05 percent
RBL; when adjusted to meet a potential alternative standard level of 15 ppm-hrs, the median
value is 0.65 percent RBL, the 75th percentile value is 0.94 percent RBL, and the maximum
value is 3.25 percent RBL; and when adjusted to meet a potential alternative standard level of 7
ppm-hrs, the median value is 0.50 percent RL, the 75th percentile value is 0.58 percent RBL, and
the maximum value is 2.49 percent RBL. In addition, RBL values for each scenario can be
viewed across the entire distribution within each species (Figure 6-12) or as a proportion of the
current standard (Figure 6-13). Figure 6-12 and Figure 6-13 use Ponderosa Pine as an example -
plots for the other 11 species are included in Appendix 6A. In Figure 6-12, the number of
exceedances above 1 percent RBL declines across W126 index values.
Table 6-7 below, summarizes the number of species exceeding 2 percent RBL under
recent Os conditions and under the four air quality scenarios. The maximum number of species
that exceed 2 percent RBL in any one county is five, which only occurs under recent Os
conditions. These data are presented as the number of counties with five, four, three, two, one,
and no species, counties in which the median species exceeds 2 percent, and the total number of
counties (out of 3,109) with at least one species exceeding 2 percent RBL. Because Cottonwood
and Black Cherry are highly sensitive species and to provide a reference for the effect of these
2 W126 calculations are slightly modified in the case of the model adjustment scenarios described in Chapter 4,
Section 4.3.4. When calculating W126 for the model adjustment cases, we first found the three-year average of each
three-month period, and then selected the three-month period with the highest three-year average using the same
three-month period for each of the three years. In this way, the five scenarios are for recent air quality, air quality
adjusted to just meet the existing standard, and air quality further adjusted to just meet three different W126 index
values: 15 ppm-hrs, 11 ppm-hrs, and 7 ppm-hrs.
6-18
-------
species, the data are also presented excluding Cottonwood and excluding Cottonwood and Black
Cherry.
Table 6-7 Number of Counties w/Tree Species Exceeding 2 Percent Relative Biomass Loss
Number of
Species
Exceeding 2
Percent RBL
5
4
3
2
1
0
Median Species
Total Exceeding
Number of Counties (3,109 Total)
Recent
Conditions
134
387
765
882
593
348
2,237
2,761
75 ppb
-
3
24
994
1,292
796
685
2,313
15 ppm-hrs
-
3
22
981
1,273
830
670
2,279
11 ppm-hrs
-
-
14
972
1,238
885
651
2,224
7 ppm-hrs
-
-
5
924
1,277
903
627
2,206
Excluding Cottonwood
5
4
3
2
1
0
Median Species
Total Exceeding
15
180
680
933
610
691
1,604
2,418
-
-
3
46
1,880
1,180
239
1,929
-
-
O
32
1,857
1,217
221
1,892
-
-
-
14
1,818
1,277
204
1,832
-
-
-
5
1,812
1,292
172
1,817
Excluding Cottonwood and Black Cherry
5
4
3
2
1
0
Median Species
Total Exceeding
-
15
187
856
920
1,131
666
1,978
-
-
-
29
95
2,985
36
124
-
-
-
15
72
3,022
18
87
-
-
-
2
19
3,088
6
21
-
-
-
1
8
3,100
2
9
6-19
-------
Recent Conditions
0.00
0.05
0.10
0.15
0.20
0.25
75 |)|)l) Scenario
0.00
0.01 0.02 0.03 0.04
0.05
Is ^
0)
15 |)|)ni-ln Scenario
0.00 0.01
0.02 0.03 0.04 0.05
11 |)|)in -hi Scenario
j-m 11 m-p-^
0.00 0.01
0.05
7 |>|>in-ln Scenario
0.00 0.01 0.02 0.03 0.04 0.05
Relative Biomass Loss, Ponderosa pine
Figure 6-12 Relative Biomass Loss of Ponderosa Pine at the Existing Primary and
Alternative Secondary Standards [RBL in this figure is plotted as a proportion relative to
no Os exposure.]
6-20
-------
o
o
Recent Conditions
I
k
1 1 1 1 1 1 1
0 10 20 30 40 50 60
-------
existing standard of 75 ppb. Within each region we calculated both the W126 value at each
monitor in the region for each year and the three-year average W126 value using the method
described in Chapter 4. The results, depicted in Figure 6-14 below, show that the use of the
three-year average W126 index value may underestimate RBL values slightly, but the approach
does not account for moisture levels or other environmental factors that could affect biomass
loss.
6-22
-------
CD
CN
CO
CD
"CD
3 Year Compounded Relative Biomass Loss
Black Cherry
Central
0.0
0.2
0.4
0.6
0.8
1.0
o
in
8
o
Black Cherry
Northeast
0.00
0.05
° -I
O
in
o
o
o
s .
o
8 .
Tulip Poplar
Southeast
0.00
0.05
0.10
0.15
0.20
0.10
0.15
0.20
O _
O
8 -I
o
White Rne
East North Central
0.00
0.02
0.04
0.06
0.08
8 H
o
q -
o
Fted Alder
Northwest
\ i i i i i r
0.000 0.005 0.010 0.015 0.020 0.025 0.030
° -I
8 -
Ponderosa Rne
Southwest
0.00 0.05 0.10 0.15 0.20 0.25 0.30
ro _
o
8 .
RDnderosa Rne
West
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35
p _
o
8 .
S .
O -
O
8 .
Aspen
West North Central
0.00
0.02
0.04
0.06
0.08
Averaged W126
Figure 6-14 Three-Year Compounded Relative Biomass Loss, by Region
6.3 COMMERCIAL TIMBER EFFECTS
We used the Forest and Agricultural Sectors Optimization Model with Greenhouse Gases
(FASOMGHG) (Adams et al., 2005) to calculate the resulting market-based welfare effects of Os
6-23
-------
exposure in the forestry and agriculture sectors of the United States under the scenarios outlined
below. This section provides a summary of the results of those analyses. The current crop/forest
budgets, which include all inputs to production and the resulting products, included in
FASOMGHG are considered the budgets under recent ambient conditions. To model the effects
of changing W126 index values on the forestry sector, two primary and three alternative
scenarios were constructed and run through the model:
• a base scenario, consistent with recent ambient conditions;
• a scenario with crop and forest yields for Os exposures after simulating just meeting
the existing standard of 75 ppb (4th highest daily maximum) and
• three scenarios that represent Os exposure after just meeting alternative W126-based
standard levels - 15, 11, and 7 ppm-hrs.
We used the Cb E-R functions for tree seedlings to calculate relative yield loss (RYL),
which is equivalent to relative biomass loss, for FASOMGHG trees over their entire life span. To
derive the FASOMGHG region-level RYLs for trees under each scenario, we used FASOMGHG
region Os values along with the mapping in Table 6-8. For additional details on FASOMGHG,
including a map of the FASOMGHG regions, see Appendix 6B (FASOMGHG Full Results).
We calculate the FASOMGHG region-level RYLs for each tree species listed in the first
column of Table 6-8 by extracting county-level W126 concentrations from the CMAQ air quality
surfaces, using only the portion of each county that is identified as forested in the GIS data
utilized and used the simple average across county Os values (forested portions of each county)
for all counties falling in a given FASOM region to represent the region-level Os impacts on
forests. Then the region-level W126 Os values are applied to tree species present in that region to
calculate RYLs. Then, we calculate a simple average of RYLs for each tree species mapped to a
FASOMGHG forest type in a given region. The mapping of tree species to FASOMGHG forest
types is based on "Atlas oj'UnitedStates Trees" (Little, 1971, 1976, 1977, 1978).
6-24
-------
Table 6-8 Mapping O3 Impacts to FASOMGHG Forest Types
Tree Species Used for
Estimating Os Impacts
Black Cherry, Tulip Poplar
Douglas Fir
Eastern White Pine
Ponderosa Pine
Quaking Aspen
Quaking Aspen, Black Cherry, Red Maple,
Sugar Maple, Tulip Poplar
Red Alder
Red Maple
Virginia Pine
Virginia Pine, Eastern White Pine
Virginia Pine, Eastern White Pine
FASOMGHG Forest Type
Upland Hardwood
Douglas Fir
Softwood
Softwood
Hardwood
Hardwood
Hardwood
Bottomland Hardwood
Natural Pine, Oak-Pine,
Planted Pine
Natural Pine, Oak-Pine,
Planted Pine
Softwood
FASOMGHG Region(s)
SC, SE
PNWW
CB,LS
PNWE, PNWW, PSW, RM
RM
CB, LS, NE
PNWE, PNWW, PSW
SC, SE
SC
SE
NE
Table 6-9 presents the region-specific RYLs for the forest types by region. At the
existing standard the highest yield loss occurs in upland hardwood forests in the South Central
and Southeast regions at over three percent per year. The next highest yield losses at the existing
standard occur in Corn Belt hardwoods with just over two percent loss per year and in hard- and
softwoods of the Rocky Mountain region at an average loss across all sensitive forests of slightly
over 1 percent loss per year. With the exception of the Rocky Mountain region, which has yield
losses reduced to under 1 percent per year, yield losses do not appreciably change at the 15 ppm-
hrs alternative. This is primarily because most areas have W126 index values lower than 15
ppm-hrs after just meeting the existing standard. The Corn Belt forests remain at about 1.5
percent loss at 11 ppm-hrs and the South Central and Southeastern forests continue to experience
yield losses between 1 and 2 percent even after just meeting an alternative standard level of 7
ppm-hrs.
6-25
-------
Table 6-9
Scenarios
Percent Relative Yield Loss for Forest Types by Region for Modeled
Forest Type
Douglas Fir
Natural Pine
Oak/Pine
Other Softwoods
Planted Pine
Softwoods
Bottomland Hardwoods
Hardwoods
Upland Hardwoods
Region
PNWW
SC
SE
SC
SE
PNWW
SC
SE
CB
LS
NE
RM
PSW
PNWE
SC
SE
PNWW
CB
LS
NE
RM
PSW
PNWE
SC
SE
Existing Standard
(75 ppb)
0.00
0.15
0.28
0.15
0.28
0.49
0.15
0.28
0.78
0.13
0.05
1.13
0.40
0.52
0.13
0.12
0.34
2.10
0.69
0.41
1.59
0.27
0.36
3.25
3.07
W126
15 ppm-hrs
0.00
0.15
0.28
0.15
0.28
0.48
0.15
0.28
0.78
0.13
0.05
0.91
0.36
0.50
0.13
0.12
0.34
2.10
0.69
0.41
1.27
0.25
0.35
3.25
3.07
11 ppm-hrs
0.00
0.12
0.24
0.12
0.24
0.48
0.12
0.24
0.46
0.13
0.04
0.64
0.31
0.48
0.10
0.10
0.34
1.51
0.69
0.33
0.88
0.22
0.34
2.71
2.79
7 ppm-hrs
0.00
0.09
0.13
0.09
0.13
0.48
0.09
0.13
0.23
0.13
0.02
0.53
0.28
0.47
0.06
0.06
0.33
0.98
0.67
0.25
0.73
0.19
0.33
2.00
1.85
6-26
-------
Table 6-10 Percent Relative Yield Gain for Forest Types by Region with Respect to the
Existing Standard
Forest Type
Douglas Fir
Natural Pine
Oak/Pine
Other Softwoods
Planted Pine
Softwoods
Bottom Hardwoods
Hardwoods
Upland Hardwoods
Region
PNWW
SC
SE
SC
SE
PNWW
SC
SE
CB
LS
NE
RM
PSW
PNWE
SC
SE
PNWW
CB
LS
NE
RM
PSW
PNWE
SC
SE
W126
15 ppm-hrs - ES
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.23
0.04
0.02
0.00
0.00
0.00
0.00
0.00
0.00
0.35
0.03
0.01
0.01
0.01
11 ppm-hrs -ES
0.00
0.02
0.04
0.02
0.04
0.01
0.02
0.04
0.35
0.00
0.01
0.52
0.09
0.04
0.03
0.01
0.01
0.65
0.00
0.09
0.77
0.06
0.03
0.65
0.34
7 ppm-hrs - ES
0.00
0.06
0.16
0.06
0.16
0.01
0.06
0.16
0.59
0.00
0.02
0.63
0.13
0.05
0.06
0.06
0.01
1.22
0.02
0.17
0.93
0.09
0.04
1.48
1.48
6-27
-------
Yield gains associated with meeting alternative W126 standards compared to meeting the
existing standard are relatively small on a percentage change basis, especially in the 15 ppm-hrs
scenario where the highest change is 0.35 percent per year. At 11 ppm-hrs the yield gains are
larger with gains between 0.35 and 0.77 percent for the most affected regions. The 7 ppm-hrs
scenario generates yield gains between 0.59 and 1.48 percent for the Corn Belt, Rocky Mountain,
South Central, and Southeast regions. These results are presented in Table 6-10 and graphically
in Figure 6-15 and Figure 6-16. While the yield gains for the alternative scenarios are small
relative to the baseline of the existing standard, when applied nationally to forest production they
result in increased forest production at every alternative in all years until the last period modeled
in 2040 as shown in Table 6-11. The change in relative yield between the existing standard and
the alternative scenarios results in changes in timber harvests and prices, as shown in Table 6-11.
In general, harvests increase and prices decrease with resulting changes in consumer and
producer welfare.
6-28
-------
FASOM Region
Legend
Figure 6-15 RYG for Softwoods by Region
6-29
-------
Legend
FASOM Region
°->0 *%
Figure 6-16 RYG for Hardwoods by Region
6-30
-------
"able 6-11 Percenta
Product
Hardwood saw logs
15 ppm-hrs
11 ppm-hrs
7ppm-hrs
Hardwood pulp logs
15 ppm-hrs
11 ppm-hrs
7 ppm-hrs
Softwood saw logs
15 ppm-hrs
11 ppm-hrs
7 ppm-hrs
Softwood pulp logs
15 ppm-hrs
11 ppm-hrs
7 ppm-hrs
§e Changes in National Timber Prices
Policy
75 ppb
2010
0.69
2020
0.65
2030
0.39
2040
0.19
Change with Respect to Existing Standard
75 ppb
-0.28
-0.79
-1.59
0.24
0.13
0.13
-2.60
0.44
-0.16
-2.52
-8.72
0.22
0.94
-1.51
-7.12
0.12
Change with Respect to Existing Standard
75 ppb
0.00
-0.87
-2.10
2.31
-0.15
-1.95
-3.52
1.91
-0.08
-2.06
-4.92
1.60
-0.08
-2.64
-6.23
1.31
Change with Respect to Existing Standard
75 ppb
-0.09
-0.26
-0.46
1.42
-0.33
-1.24
-1.54
1.12
-0.44
-1.32
-1.91
1.34
-0.69
-1.40
-2.28
0.94
Change with Respect to Existing Standard
-0.14
-0.43
-1.03
0.12
0.13
-0.42
0.15
-0.19
-0.82
0.18
-0.51
-2.17
Table 6-12 shows the estimated welfare changes brought about by the simulation
scenarios. Consumer and producer welfare in the forest sector are more affected by the
alternative scenario environments than the agricultural sector (see Section 6.5). In general,
consumer welfare increases in both the forest and agricultural sectors as higher productivity
tends to increase total production and reduce market prices. Because demand for most forestry
and agricultural commodities is inelastic, producer welfare tends to decline with higher
productivity as the effect of falling prices on profits more than outweighs the effects of higher
production levels. In other words consumers do not increase their demand for the product enough
in response to the falling prices created by increases production to offset the producer's loss of
6-31
-------
revenue. The increase in consumer welfare is much larger than the loss of producer welfare
resulting in net welfare gains in the forestry sector nationally.
Welfare economics focuses on the optimal allocation of resources and goods and how
those allocations affect total social welfare. Total welfare is also referred to as economic surplus,
which is the overall benefit a society, composed of consumers and producers, receives when a
good or service is bought or sold, given a quantity provided and a market price. Economic
surplus is divided into two parts: consumer and producer surplus.
Consumers like to feel like they are getting a good deal on the goods and services they
buy, and consumer surplus is an economic measure of this satisfaction. For example, assume a
consumer goes out shopping for a CD player and he or she is willing to spend $250. When the
shopper finds that the CD player is on sale for $150, economists would say that this shopper has a
consumer surplus of $100, e.g., the difference between the $150 sale price and the $250 the
consumer was willing to spend.
Producer surplus refers to the benefit a producer receives from providing a good or
service at a market price when they would have been willing to sell that good or service at a lower
price. For example, if the amount the producer is willing to sell the CD player for is $75, and the
producer sells the CD player for $150, the producer surplus is $75, e.g., the $150 sale price less
the $75 price at which the producer was willing to sell.
Price
Supply
Demand
Quantity
6-32
-------
Table 6-12 Consumer and Producer Sur
Product
Consumer
surplus
15 ppm-hrs
11 ppm-hrs
7ppm-hrs
Producer
surplus
15 ppm-hrs
11 ppm-hrs
7 ppm-hrs
Policy
75ppb
75ppb
2010
721,339
2015
793,234
plus in Forestry, Million $2010
2020
809,271
2025
826,375
2030
875,620
2035
894,705
2040
934,882
Change with Respect to Existing Standard
7
44
86
93,322
31
48
187
121,476
118
360
694
153,997
105
202
224
146,275
2
688
734
145,913
6
56
91
146,115
597
712
779
133,132
Change with Respect to Existing Standard
-11
-41
-136
-7
20
-48
-141
-503
-892
-161
-178
-37
15
-880
-786
-46
55
156
-839
-858
-766
Key uncertainties in this approach are discussed in Section 6.6.1. It should be noted that
since public lands are not affected within the model, the estimates presented would likely be
higher if public lands were included.3 See Appendix 6B for a full discussion of the model and
methodology.
6.4 NON-TIMBER FOREST PRODUCTS
Non-timber forest products (NTFP) such as foliage and branches used for arts and crafts,
or edible fruits, nuts, and berries can be affected by the impact of Os through biomass loss, foliar
injury, insect attack, fire regime changes, and effects on reproduction. Commercial gathering
activities in national forests are allowed by permit holders. The USDA has assessed the harvest
and market value of these products in commercial markets (Emery, 2003). A significant portion
of NTFP is also valuable to subsistence gatherers. Subsistence practices are much more difficult
to assess because these forest users are not required to obtain a permit for use of federal public
lands; as such the impacts are more difficult to enumerate. Because permits or contracts are not
3 The FASOMGHG model includes 348.6 million acres of private, managed forests. The USFS estimates that there
are approximately 751 million forest acres in the United States (USDA, 2011).
6-33
-------
required for gathering activities for personal use the analyses done by USDA are not able to
account for the subsistence use of NTFP.
In Table 6-13 we list some of the uses of the tree species known to be sensitive to the
effects of Os on biomass. These species have a wide variety of uses ranging from the value of
the timber produced to medicinal uses.
Table 6-13
Sensitive Trees and Their Uses
Tree Species
Os Effect
Uses
Black Cherry
Prunus serotina
Biomass loss,
Visible foliar
injury
Cabinets, furniture, paneling, veneers, crafts, toys
Cough remedy, tonic, sedative
Flavor for rum and brandy
Wine making and jellies
Food for song birds, game birds, and mammals
Douglas Fir
Pseudotsuga
menziesii
Biomass loss
Commercial timber
Medicinal uses, spiritual and cultural uses for several Native American
tribes
Spotted owl habitat
Food for mammals including antelope and mountain sheep
Eastern
Cottonwood
Populus deltoides
Biomass loss
Containers, pulp, and plywood
Erosion control and windbreaks
Quick shade for recreation areas
Beaver dams and food
Eastern White
Pine
Pinus strobus
Biomass loss
Commercial timber, furniture, woodworking, and Christmas trees
Medicinal uses as expectorant and antiseptic
Food for song birds and mammals
Used to stabilize strip mine soils
Hemlock
Tsuga canadensis
Biomass loss
Commercial logging for pulp
Habitat for deer, ruffled grouse, and turkeys
Important ornamental species
Hickory
Biomass loss
Used in furniture and cabinets, fuelwood, and charcoal
Edible nuts
Food for ducks, quail, wild turkeys and many mammals
Ponderosa Pine
Pinus ponderosa
Biomass loss,
Visible foliar
injury
Lumber for cabinets and construction
Ornamental and erosion control use
Recreation areas
Food for many bird species, including the red-winged blackbird,
chickadee, finches, and nuthatches
6-34
-------
Tree Species
Os Effect
Uses
Quaking Aspen
Populus
tremuloides
Biomass loss,
Visible foliar
injury
Commercial logging for pulp, flake-board, pallets, boxes, and plywood
Products including matchsticks, tongue depressors, and ice cream sticks
Valued for its white bark and brilliant fall color
Important as a fire break
Habitat for variety of wildlife
Traditional native American use as a food source
Red Alder
Alnus rubra
Biomass loss,
Visible foliar
injury
Commercial use in products such as furniture, cabinets, and millwork
Preferred for smoked salmon
Dyes for baskets, hides, moccasins
Medicinal use for rheumatic pain, diarrhea, stomach cramps - the bark
contains salicin, a chemical similar to aspirin
Roots used for baskets
Food for mammals and birds - dam and lodge construction for beavers
Conservation and erosion control
Red Maple
Acer rubrum
Biomass loss
Revegetation and landscaping especially riparian buffer
Red Oak
Quercus rubra
Biomass loss
Important for hardwood lumber for furniture, flooring, cabinets
Food, cover, and nesting sites for birds and mammals
Bark used by Native Americans for medicine for heart problems, bronchial
infections or as an astringent, disinfectant, and cleanser
Short Leaf Pine
Pinus echinata
Biomass loss
Second only to loblolly pine in standing timber volume
Used for lumber, plywood, pulpwood, boxes, crates, and ornamental
vegetation
Habitat and food for bobwhite quail, mourning dove, other song birds and
mammals
Older trees with red heart rot provide red-cockaded woodpecker cavity
trees
Sugar Maple
Acer saccharum
Biomass loss
Commercial syrup production
Native Americans used sap as a candy, beverage - fresh or fermented into
beer, soured into vinegar and used to cook meat
Valued for its fall foliage and as an ornamental
Commercial logging for furniture, flooring, paneling, and veneer
Woodenware, musical instruments
Food and habitat for many birds and mammals
Virginia Pine
Pinus virginiana
Biomass loss,
Visible foliar
injury
Pulpwood, strip mine spoil banks and severely eroded soils
Nesting for woodpeckers, food for songbirds and small mammals
Yellow (Tulip)
Poplar
Liriodendron
tulipifera
Biomass loss,
Visible foliar
injury
Furniture stock, veneer, and pulpwood
Street, shade, or ornamental tree - unusual flowers
Food for wildlife
Rapid growth for reforestation projects
Sources: USDA-NRCS, 2013; Burns, 1990; Hall and Braham, 1998.
6-35
-------
6.4.1 Commercial Non-Timber Forest Products
In addition to timber, forests provide many other products that are harvested for
commercial or subsistence activities. These products include:
• edible fruits, nuts, berries, and sap,
• foliage, needles, boughs, and bark,
• transplants,
• grass, hay, alfalfa, and forage,
• herbs and medicinals,
• fuelwood, posts and poles, and
• Christmas trees.
For the 2010 National Report on Sustainable Forests (USDA, 2011) these products were
divided into several categories including nursery and landscaping uses; arts, crafts, and floral
uses; regeneration and silviculture uses. Table 6-14 details selected categories of NTFP
harvested by permit in 2007. These harvests are reported in measures relevant to the specific
articles, i.e., bushels of cones, tons of foliage or boughs, or individual transplants. The harvests
quantified in the table are only for permitted activities in national forests and do not include
those activities that occur on private or state- or locally-owned property.
6-36
-------
Table 6-14 Quantity of NTFP Harvested on U.S.
Management Land
Forest Service and Bureau of Land
Product Category
Arts, crafts, and florals
Christmas trees
Edible Fruits, nuts, berries, and sap
Fuelwood
Grass, hay, and alfalfa
Forage
Herbs and medicinals
Nursery and landscape
Regeneration and silviculture
Posts and poles
Unit
Bushels
Pounds
Tons
Each
Lineal foot
Bushels
Pounds
Syrup Taps
ccf
Cords
Pounds
Tons
Pounds
Each
Pounds
Tons
Bushels
ccf
Each
Pounds
Tons
ccf
Each
Lineal foot
Harvest All U.S.
70,222
3,442,125
620,773
151,274
94.758
250
1,614,565
10,686
35,800
417,692
4,265,952
480
101,365
766,645
25,689
316
7,627
8
21,265
247,543
110,873
5,281
1,684,618
326,312
Note: ccf = 100 cubic feet Source: USD A 2011
According to the Os ISA, Os exposure causes biomass loss in sensitive woody and
herbaceous species, which in turn could affect forest products used for arts, crafts, and florals.
For example, Douglas Fir and Red Alder, among others, are used on the Pacific Coast for arts
and crafts, particularly holiday crafts and decorations. The effects of Os on plant reproduction
6-37
-------
(see Cb ISA, Table 9-1, 2013) could affect the supply of seeds, berries, and cones. Foliar injury
impacts on Os-sensitive plants would potentially affect the harvest of leaves, needles, and
flowers from these plants for decorative uses. The visible injury and early senescence caused by
Os in some evergreens may also reduce the value of a whole tree such as Christmas trees.
Likewise the same Cb effects would reduce the harvest of edible fruits, nuts, berries, and sap.4
The use of native grasses as forage is a significant aspect of forest-land management in the
western U.S. (Alexander et al. 2002). Ozone effects on community composition, particularly
changes in the ratio of grasses to forbs (broad-leaved herbs other than a grass), and nutritive
quality of grasses can have effects on rangeland quality for some herbivores (Krupa et al., 2004,
Sanz et al., 2005) and therefore effects on grazing efficiency. The negative impacts of Cb on
plants would similarly affect the harvest in the rest of the categories.
According to the U.S. Census Bureau's County Business Patterns data from 2006, this
activity is captured in the industry code 1132 — forest nurseries and gathering of forest products -
- and employed 2,098 people, accounting for an annual payroll of $71,657,000 ($2006) with an
average annual income of $34,155 (U.S. Census Bureau, 2006).
The USD A estimates the proportion of the national supply of NTFP represented by USFS
and BLM lands is approximately 10 percent. Retail values for NTFPs harvested on USFS and
BLM lands are approximately $1.4 billion (2010$). These estimates are very rough and are based
only on permit or contract sales. These estimates could be low due to harvests taken without
permit or contract and sold through complex commodity chains that can combine wild-harvested
and agriculturally grown commodities. It is important to note that while we cannot estimate the
loss of production and value to this sector due to Cb exposures, these losses are already reflected
in the harvest and values reported.
6.4.2 Informal Economy or Subsistence Use of Non-Timber Forest Products
Most people gathering NTFPs are doing so for personal use (Baumflek et al., 2010;
USD A, 2011). By one estimate (Baumflek et al., 2010) up to 80 percent of the people collecting
NTFPs in Oregon and Washington are collecting for personal reasons. Such personal use may be
characterized either as part of the informal economy or as subsistence activity. Participants in
the informal economy may earn a wage or salary and participate in gathering NTFPs for reasons
1 To name a few, this category includes blueberries, pine nuts, and sap for maple syrup.
6-38
-------
other than recreation (Brown et al., 1998). The term subsistence has usually been applied to
special groups such as Native Americans or the Hmong people and has generally been
understood to imply extreme poverty such that these activities are essential to the necessities of
life (Freeman, 1993). However, Freeman points out researchers stress that economic goals are
only a part of the impetus for these activities.
Brown et al. (1998) proposed a composite definition for the terms that captures both the
informal economy, as practiced by those who are not necessarily a part of a special population,
and subsistence, as generally referenced to those special populations.
"Subsistence refers to activities in addition to, not in place of, wage labor engaged in on a
more or less regular basis by group members known to each other in order to maintain a
desired and/or normative level of social and economic existence."
This definition allows consideration of the cultural and social aspects of subsistence lifestyles.
These non-economic benefits range from maintenance of social ties and relationships through
shared activity to family cohesiveness to retreatism and a sense of self-reliance for the individual
practitioner (Brown et al., 1998).
While there is general acknowledgement of subsistence activities by Native Americans
and specific treaty rights for tribes guaranteeing access to lands for hunting, fishing, and
gathering, there has been a lack of research focused on other populations (Emery and Pierce,
2005). However, there are some studies that clarify that subsistence activities provide valued
resources for a variety of people in the coterminous United States. Baumflek et al. (2010) and
Alexander et al. (2011) have documented the collection and use of culturally and economically
important NTFPs in Maine and the eastern United States, respectively. Brown et al. (1998)
reports on subsistence activities among residents of the Mississippi Delta. Emery (2003) and
Hufford (2000) examine activities in the Appalachians, and Pena (1999) reports activities by
Latinos in the Southwest.
As with the commercial harvest of NTFPs, subsistence gathering of these forest products
can potentially be affected by the adverse effects of Os on growth, reproduction, and foliar injury
to the sensitive plants in use for nutrition, medicine, cultural, and decorative purposes. It is
important to note that some plants may have more than one use or significance. For example, the
6-39
-------
Mi'kmaq and Maliseet Indian tribes in Maine do not differentiate between blueberries'
nutritional, medicinal, and spiritual uses. Blueberries are a food and a medicine that is often
incorporated into ceremonies (Baumflek et al., 2010). And while we cannot quantify the size of
the harvest of subsistence-gathered items or monetize the loss of benefit due to Os effects, a
comparison to the commercial harvest detailed in section 6.4.1 may provide perspective on the
significance of these activities to the people who engage in them.
6.5 AGRICULTURE
6.5.1 Commercial Agriculture
Because the forestry and agriculture sectors are related, and trade-offs occur between the
sectors based on individual decisions given agriculture and forestry market conditions, we used
the same FASOMGHG model runs outlined in the forestry/timber section (Section 6.3) to
calculate the resulting market-based welfare effects of Os exposure in the agricultural sector of
the United States. This section provides a summary of the results of the agricultural sector
analyses. We have included results at the national scale for both sectors and at the regional and
subregional scale for agriculture. Table 6-15 defines the production and market regions available
in FASOMGHG. The regional-scale analysis provides an estimate of the changes due to
alternative levels of the standard for 63 subregions and indicates the disparate results between
regions. The full model results, including a county-level analysis and a full explanation of
interactions between the forestry and agriculture sectors, are reported in Appendix 6B. Of note
in the county-level analysis is that the relative yield loss estimates mirror the associated
subregion. Under recent conditions, there are significant numbers of counties with greater than 5
percent yield loss — for soybeans 1,718 out of 1,729, or 99 percent, of soybean-producing
counties. When adjusting air quality scenarios to just meet the existing standard of 75 ppb, no
counties have relative yield losses above 5 percent for any crop.
6-40
-------
Table 6-15 Definition of FASOMGHG Production Regions and Market Regions
Key
NE
LS
CB
GP
SE
SC
SW
RM
PSW
PNWE
PNWW
Market Region
Northeast
Lake States
Corn Belt
Great Plains (agriculture
only)
Southeast
South Central
Southwest (agriculture
only)
Rocky Mountains
Pacific Southwest
Pacific Northwest — East
side
Pacific Northwest — West
side (forestry only)
Production Region (States/Subregions)
Connecticut, Delaware, Maine, Maryland, Massachusetts, New
Hampshire, New Jersey, New York, Pennsylvania, Rhode Island,
Vermont, West Virginia
Michigan, Minnesota, Wisconsin
All regions in Illinois, Indiana, Iowa, Missouri, Ohio (IllinoisN, IllinoisS,
IndianaN, IndianaS, lowaW, lowaCent, lowaNE, lowaS, OhioNW,
OhioS, OhioNE)
Kansas, Nebraska, North Dakota, South Dakota
Virginia, North Carolina, South Carolina, Georgia, Florida
Alabama, Arkansas, Kentucky, Louisiana, Mississippi, Tennessee,
Eastern Texas
Oklahoma, All of Texas but the Eastern Part (Texas High Plains, Texas
Rolling Plains, Texas Central Blacklands, Texas Edwards Plateau, Texas
Coastal Bend, Texas South, Texas Trans Pecos)
Arizona, Colorado, Idaho, Montana, Nevada, New Mexico, Utah,
Wyoming
All regions in California (CaliforniaN, CaliforniaS)
Oregon and Washington, east of the Cascade mountain range
Oregon and Washington, west of the Cascade mountain range
Using the modeled W126 index values in each subregion under the scenarios, for crops,
we first calculated the RYL in the 63 subregions that have E-R functions. For those crops that do
not have E-R functions, we assign them RYLs for each scenario based on the crop proxy
mapping shown in Table 6-16. In addition, for oranges, rice, and tomatoes, which have Os E-R
functions that are not W126-based (they are defined based on alternative measures of Os
concentrations), we directly used the median RYG values under the "13 ppm-hrs" Os
concentration reported in Table G-7 of Lehrer et al. (2007). In addition, we updated RYLs for
crops with county-level production data and specific E-R functions by using production-
weighting. Production weighting applies a county's share of the region's total production to the
average so that counties with less production have a smaller impact on the average.
The RYLs for proxy crops were calculated for each FASOMGHG subregion so they
could be used in calculating the yield losses for other crops that occur in those regions. The
6-41
-------
weighted RYLs that were used for corn, cotton, winter wheat (hard winter wheat and soft red
winter wheat), sorghum, and soybeans in the model scenarios were calculated only for their
production regions. The values calculated in all 63 regions were weighted by production for
these crops, which eliminated regions with no production.
Table 6-16 Mapping of O3 Impacts on Crops to FASOMGHG Crops
CROPS
FASOMGHG Crops
W126 Crops
Corn
Cotton
Potatoes
Winter Wheat
Sorghum
Soybeans
Aspen (tree)
Corn
Cotton
Potatoes
Soft White Wheat, Hard Red Winter Wheat, Soft Red Winter Wheat, Durum Wheat, Hard Red
Spring Wheat, Oats, Barley, Rye, Wheat Grazing, and Improved Pasture
Sorghum, Silage, Hay, Sugarcane, Sugar Beet, Switchgrass, Energy Sorghum, and Sweet
Sorghum
Soybeans, Canola
Hybrid Poplar, Willow (FASOMGHG places short-rotation woody biomass production in the crop
sector rather than in the forest sector)
Non-W126 Crops
Oranges
Rice
Tomatoes
Orange Fresh/Processing, Grapefruit Fresh/Processing
Rice
Tomato Fresh/Processing
The following figures (Figure 6-17 and Figure 6-18) present the yield loss relative to the
existing standard and yield gains for corn and soybeans under the various adjusted air quality
scenarios. We are using corn and soybeans to illustrate some of the interactions that occur
between crop responses to Os reductions, production, prices, producer cropping decisions, and
welfare effects for both producers and consumers. For full model results for all crops included in
the analysis see Appendix 6B. In general, the RYL and RYG are unchanged between the
existing 75 ppb standard and the 15 ppm-hrs W126 scenarios. Also, note that in many cases,
subregions that show no change in yield for a given crop have no production of that crop in that
subregion in FASOMGHG. For example, soybeans are relatively sensitive to Os and there are
large reductions in Os in California, but there are no impacts on soybean yields in that region
because no soybeans are produced in California in FASOMGHG.
6-42
-------
Corn is relatively insensitive to Os-induced yield losses at the existing standard or 15
ppm-hrs. The highest loss occurs in California at 0.88 percent, while in the Corn Belt, Lake
States, and Great Plains the highest loss occurs in southern Ohio with 0.34 percent. Because the
yield losses are small due to corn's insensitivity to Os under the alternative W126 standard
scenarios, the yield losses are virtually eliminated at all three alternative W126 standards. Yield
gains associated with the alternative scenarios are almost nonexistent; the highest gain occurs in
Arizona at 0.02 percent at the 7ppm-hrs level.
Soybeans, on the other hand, are relatively sensitive to Os-induced yield losses. The
highest losses at the existing standard or 15 ppm-hrs occur in Colorado, southern Indiana,
Kentucky, and northwest Ohio at over 1 percent. Yield losses remain under all scenarios for
W126, although for the 7 ppm-hrs scenario all losses are less than 0.6 percent. Yield gains
across the alternative W126 standard levels generally range between 0.54 percent and 0.84
percent with northeast Ohio, Tennesse, Kentucky, Illinois, and Indiana on the high end.
Colorado has the highest gain at 1.01 percent at the 7 ppm-hrs level and most soybean producing
states have at least small gains at every W126 scenario.
6-43
-------
Legend
Region
Sub region
Figure 6-17 Percentage Changes in Corn RYG with Respect to 75 ppb
6-44
-------
Legend
| | FASOM Region
| | FASOM Subregion
Figure 6-18 Percentage Changes in Soybean RYG with Respect to 75 ppb
6-45
-------
In general, increased yield leads to increased supply and lower prices. Because corn does
not lose or gain very much under any scenario one could expect that prices would remain
relatively stable. Soybeans, however, would experience yield gains in any scenario, and prices
would likely fall. In the modeled scenarios soybean prices fall, and since consumer demand does
not increase enough to offset the loss of revenue due to price decreases there is a net decrease in
producer welfare, but consumers always benefit from falling prices. In response to falling
soybean prices, the model predicts that producers would switch to less Os-sensitive crops with
stable prices, such as corn, thereby increasing corn production. See Appendix 6C for an
explanation of the supply curve shift.
Overall, across the full agriculture sector, these changes in production are small, seldom
above 0.5 percent and usually 0.01 percent or less. The production increases lead to generally
lower prices, with price decreases greater than the change in production. The drop in market
prices, while a loss for producers, represents a gain for consumers. In terms of producer and
consumer welfare across the agriculture sector, in nearly all cases producer welfare is negatively
affected. Table 6-17 presents the overall welfare gains and losses. For producers, the W126
alternatives occasion welfare gains in the middle years, 2020-2030, and welfare losses in all
other years. For consumers, however, the changes in production and prices occasion welfare
gains in all scenarios in all years.
Since the forestry and agriculture sectors are interlinked and factors affecting one sector
can lead to changes in the other, it is important to consider the overall effect of Os changes in the
context of producer and consumer welfare across both sectors. The impacts on consumer surplus
are positive for both sectors, with benefits increasing with lower W126 alternatives. For producer
surplus, however, impacts are negative for the 15 ppm-hrs and 11 ppm-hrs scenarios and positive
for the 7 ppm-hrs case. Table 6-18 presents the annualized surplus for both sectors.
6-46
-------
Table 6-17 Consumer and Producer Surplus in Agriculture (Million 2010$)
Product
Consumer
Surplus
Producer
Surplus
Policy
75ppb
15 ppm-
hrs
1 1 ppm-
hrs
7ppm-
hrs
75ppb
15 ppm-
hrs
1 1 ppm-
hrs
7ppm-
hrs
2010
1,918,082
2015
1,940,673
2020
1,968,142
2025
1,995,346
2030
2,023,022
2035
2,050,791
2040
2,076,018
Change with Respect to Existing Standard
15
19
-31
725,364
-2
24
46
831,565
1
13
36
815,072
6
51
104
863,165
-7
42
90
878,986
10
20
26
836,692
3
13
46
863,308
Change with Respect to Existing Standard
612
1,474
269
-1,255
-2,197
-1,873
980
1,013
1,780
-961
230
423
90
232
264
41
-3,413
-1,052
697
2,189
2,991
6-47
-------
Table 6-18 Annualized Changes in Consumer and Producer Surplus in Agriculture and
Forestry, 2010-2040, Million 2010$ (4% Discount Rate)
Product
Consumer surplus
Policy
75 ppb
Agriculture
NA
Forestry
NA
Total
NA
Change with Respect to Existing Standard
Producer surplus
15 ppm-hrs
1 1 ppm-hrs
7ppm-hrs
75 ppb
4.5
25.4
36.7
NA
88.1
236.9
344.0
NA
92.5
262.3
380.7
NA
Change with Respect to Existing Standard
Total surplus
15 ppm-hrs
1 1 ppm-hrs
7 ppm-hrs
75 ppb
-4.7
-4.6
194.4
NA
-112.2
-264.4
-318.4
NA
-116.9
-269.0
-124.0
NA
Change with Respect to Existing Standard
15 ppm-hrs
1 1 ppm-hrs
7 ppm-hrs
-0.2
20.8
231.1
-24.2
-27.5
25.6
-24.4
-6.7
256.7
6.6 CLIMATE REGULATION
Biomass loss due to Os exposure affects climate regulation by ecosystems by reducing
carbon sequestration and storage. More carbon stays in the atmosphere because carbon uptake
by forests is reduced. The studies cited in the Cb ISA demonstrate a consistent pattern of
reduced carbon uptake because of Os damage, with some of the largest reductions projected over
North America. In one simulation (Sitch et al., 2007) the indirect radiative forcing due to Os
effects on carbon uptake by plants are shown as even greater than the direct effect of Os on
climate change.
6.6.1 National Scale Forest Carbon Sequestration
FASOMGHG can calculate the difference in carbon sequestration by forests and
agriculture due to biomass loss caused by Os exposure. By comparing equilibriums under the
different scenarios outlined in Section 6.3, we can calculate changes in carbon sequestration
6-48
-------
potential over time. Details of FASOMGHG and the methodology for the analyses done for this
risk and exposure assessment are available in Appendix 6B.
The impacts of the simulations of meeting the existing and alternative secondary Os
standards on carbon sequestration potential in U.S. forest and agricultural sectors are presented
in Table 6-19, where numbers indicate increased sequestration. As shown in the table, much
greater sequestration changes are projected in the forest sector than in the agricultural sector. The
15 ppm-hrs scenario does not appreciably increase carbon storage relative to just meeting the
existing standard. The vast majority of the enhanced carbon sequestration potential under the
alternative secondary standard scenarios lies in the forest biomass increases over time at the 11
and 7 ppm-hrs standard levels. The forest carbon sequestration potential would increase between
593 and 1,602 million tons of CCh equivalents over 30 years after meeting the 11 or 7 ppm-hrs
standard level, respectively, compared to just meeting the existing Os standard. On an annual
basis when just meeting the 11 ppm-hrs W126 standard level, total forestry and agriculture
carbon storage is increased by about 20 million tons per year relative to just meeting the existing
Os standard; equivalent to taking about 4 million cars off the road as calculated by the EPA
Greenhouse Gas Equivalencies Calculator5 (U.S. EPA, 2013b). When meeting the 7 ppm-hrs
W126 standard level, the increased annual carbon storage is about 53 million tons relative to just
meeting the existing Os standard, or approximately 11 million fewer cars on the road.
The baseline stock of carbon storage decreases over time for agriculture because the
agriculture sector GHG emissions sources are released every year and soil carbon sequestration
stabilizes over the 30-year period. There are only small increases in net carbon storage compared
to the existing standard for each of the alternative scenarios modeled.
5 Available at http://www.epa.gov/cleanenergv/energv-resources/calculator.html.
6-49
-------
Table 6-19 Increase in Carbon Storage, MMtCOie, Cumulative over 30 years
Product
Forestry
Agriculture
Policy
75ppb
2010
74,679
2020
79,171
2030
84,863
2040
89,184
Change with Respect to Existing Standard
15 ppm-hrs
1 1 ppm-hrs
7 ppm-hrs
75 ppb
1
19
50
18,748
0
103
305
15,363
16
312
832
12,002
13
593
1,602
8,469
Change with Respect to Existing Standard
15 ppm-hrs
1 1 ppm-hrs
7 ppm-hrs
0
2
3
1
5
4
1
6
6
4
10
9
6.6.2 Urban Case Study Carbon Storage
Urban forests are subject to the adverse effects of Os exposure in the same ways as
forests in rural areas. These urban forests provide a range of ecosystem services such as carbon
sequestration, pollution removal, building energy savings, and reduced stormwater runoff. The
analyses in this section focus on carbon sequestration. Pollution removal services are discussed
in section 6.7. The i-Tree model6 used in this analysis is a peer-reviewed suite of software tools
provided by USFS. We used data from five urban areas to estimate the effects of Os (based on
CMAQ modeled W126 index surfaces) on carbon storage. We used the i-Tree Forecast model to
estimate tree growth and ecosystem services provided by trees over a 25-year period, using for
the base case the measured inventory of trees in the area and standard growth rates over the 25-
year period. The growth rates in the model were standardized from measurements of forest
stands, park trees, and open space trees in their ambient Os conditions at the time of
measurement. We adjusted the tree growth downward from the standard growth rates using the
reduced growth factors for the species present in each area for which we have E-R functions
(only species with W126 E-R functions were reduced). Unlike the FASOMGHG model, E-R
functions were not assigned to species in the study areas that do not have specific E-R functions
available from the literature because the model does not account for dynamic interactions in the
' Available at http://www.itreetools.org/.
6-50
-------
community composition based on increased or decreased competitiveness of the species present.
We contrasted the differences between the scenarios for the 25-year period. We ran six scenarios
simulating a scenario without Os-induced changes in biomass, recent ambient conditions, a
simulation of "just meeting" the existing standard, and just meeting three alternative W126
standards of 15, 11, and 7 ppm-hrs. The model assumed an annual influx of between one and six
trees/hectare/year and a three to four percent annual mortality rate. See Appendix 6D for details
of the model and the methodology employed for these case studies.
We chose the five urban areas based on data availability and presence of species with a
W126 E-R function. No urban area with available vegetation data had more than three qualified
species present. The selected study areas include Baltimore, Syracuse, the Chicago region,
Atlanta, and the urban areas of Tennessee. Table 6-20 shows details of the tree species present,
the percent of sensitive trees in the top ten species present, and the percent of sensitive trees in
the total species in each study area.
6-51
-------
Table 6-20 Tree Species with Available E-R Functions in Selected Urban Study Areas
Study Area
Top Ten
Occurring
Species
1
2
3
4
5
6
7
8
9
10
Species w/E-R
Function ~
% of Top 10
Species w/E-R
Function ~
% of Total Trees
Baltimore
American beech
Black locust
American elm
Tree of heaven
White ash
Black cherry
White mulberry
Northern red
oak
Red maple
White oak
8.5
11.2
Syracuse
European
buckthorn
Sugar maple
Black cherry
Boxelder
Norway maple
Northern white
cedar
Norway spruce
Staghorn sumac
Eastern
cottonwood
Eastern
hophornbeam
18.5
20.2
Chicago
Region
European
buckthorn
Green ash
Boxelder
Black cherry
Hardwood
American elm
Sugar maple
White ash
Amur
honeysuckle
Silver maple
7.7
10.5
Atlanta
Sweetgum
Loblolly pine
Flowering
dogwood
Tulip tree
Water oak
Boxelder
Black cherry
White oak
Red maple
Southern red
oak
6.6
8.9
Tennessee
Chinese privet
Virginia pine
Eastern red
cedar
Hackberry
Flowering
dogwood
Amur
honeysuckle
Winged elm
Red maple
Black tupelo
American beech
9.3
17.4
Bold - species with E-R function, Italics - species known to be sensitive, no E-R function
6-52
-------
The largest differences in the modeled air quality are between the recent ambient
conditions and meeting the existing standard. The distribution of Os air quality is not changed in
most areas in the eastern U.S. when simulating meeting an alternative W126 standard of 15 ppm-
hrs relative to the scenario of just meeting the existing Cb standard. There are small incremental
differences between just meeting the existing Os standard and just meeting alternative W126
standards of 11 and 7 ppm-hrs.
The model results for changes in carbon storage show substantial reductions in the
capacity of these urban forests to sequester carbon for the simulation of "just meeting" the
existing standard. Estimates for the five modeled areas at the existing standard or an alternative
standard of 15 ppm-hrs are about 3.5 million tons of carbon storage lost over 25 years (about
140,000 tons /year). At an alternative standard of 11 ppm-hrs, loss of carbon sequestration is
128,000 metric tons per year, and at an alternative standard of 7 ppm-hrs, the estimated loss is
112,000 metric tons per year of carbon storage services.
Three of the urban areas show gains in carbon storage at alternative W126 standards
below 15 ppm-hrs. Syracuse and Baltimore do not realize gains because they are currently very
close to meeting the alternative standards. Of the five areas modeled, the combined urban areas
of Tennessee have the largest estimated gains in carbon storage at almost 20,000 tons per year
when meeting the alternative standard of 7 ppm-hrs. The Chicago region gains about 6,400 tons
per year of carbon sequestration when meeting the alternative standard of 7 ppm-hrs. See Table
6-21 for details.
Compared to other activities, the yearly carbon storage gains at 11 ppm-hrs for Atlanta
are only equivalent to taking 50 cars per year off the road or recycling about 90 more tons of
waste every year. At the 7 ppm-hrs standard level, Atlanta would need to remove 250 cars per
year to be equivalent to the gains from reduced Cb. The Chicago region would need to take 417
cars per year off the road. At 7 ppm-hrs, Chicago would need to remove more than 1,300 cars.
The urban areas of Tennessee would need about 1,800 fewer cars per year at the 11 ppm-hrs
standard level. To reach the carbon sequestration provided by the urban forests in Tennessee at
the 7 ppm-hrs standard level, Tennessee would need 4,000 fewer cars every year.
Baltimore and Syracuse would realize no gains at the alternative standard levels chosen
for this analysis. Chicago and Atlanta are in the middle of the range of results. In Tennessee, at
recent ambient conditions, the urban areas are all above a W126 standard of 15 ppm-hrs and
6-53
-------
comprise a much larger area than the other four case study areas with a far larger tree population.
Thus the relative gains in carbon storage in Tennessee are far larger than the other case study
areas. Keeping in mind that of the 11 tree species for which we have E-R functions, only two to
three species were present in a given area comprising at most 18.5 percent of the total trees
present. It seems reasonable to conclude that the actual effect on carbon storage because of Os
exposure would be higher than the estimates modeled here.
These results should not be combined with the results from the FASOMGHG model
discussed in Section 6.7.1. The methodology employed for the FASOMGHG runs assigned
values for Os exposure E-R functions for species that do not have a function calculated in the Os
ISA. We did this to ensure the dynamic trade-offs in the model functioned properly. The i-Tree
model does not provide trade-offs between species, so the species that do not have a E-R
function were not assigned values. This could lead to an underestimation of the carbon storage
losses in i-Tree if the other species in the study area are sensitive to Os exposure effects.
Alternatively assigning E-R functions to species as we did for the FASOMGHG runs would
likely produce an overestimation since many species, even within the same genus, may not be
sensitive to Os effects.
Table 6-21 Os Effects on Carbon Storage for Five Urban Areas over 25 Years (in
millions of metric tons)
Region
Atlanta
Baltimore
Chicago
Region
Syracuse
Tennessee
Total
NoOs
Adjustment
(NOA)
1.426
0.578
19.560
0.169
20.568
42.302
Existing
Standard/15
ppm-hrs
(ES/15)
1.315
0.571
17.053
0.169
19.668
38.607
ES/15 v
NOA
-0.112
-0.007
-2.508
-0.0015
-0.900
-3.528
11 ppm-
hrs v
NOA
-0.106
-0.007
-2.457
-0.0015
-0.676
-3.247
7 ppm-
hrs v BC
-0.081
-0.007
-2.346
-0.0015
-0.410
-2.845
ESv
llppm-
hrs
0.006
0.00
0.05
0.00
0.22
0.276
ESv 7
ppm-hrs
0.03
0.00
0.16
0.00
0.49
0.68
ES = Existing Standard
NOA = No Adjustment to Growth Rates using Os-related E-R Functions
6-54
-------
In addition to its direct impacts on vegetation, Os is a well-known GHG that contributes
to climate warming (U.S. EPA, 2013). A change in the abundance of tropospheric Os perturbs
the radiative balance of the atmosphere, an effect quantified by the radiative forcing metric. The
IPCC (2007) reported a radiative forcing of 0.35 W/m2 for the change in tropospheric Os since
the preindustrial era, ranking it third in importance after the greenhouse gases CCh (1.66 W/m2)
and methane (CH4) (0.48 W/m2). The earth-atmosphere-ocean system responds to the radiative
forcing with a climate response, typically expressed as a change in surface temperature. Finally,
the climate response causes downstream climate-related ecosystem effects, such as redistribution
of ecosystem characteristics because of temperature changes. While the global radiative forcing
impact of Os is generally well understood, the downstream effects of the Os-induced climate
response on ecosystems remain highly uncertain.
Since Os is not emitted directly but is photochemically formed in the atmosphere, it is
necessary to consider the climate effects of different Os precursor emissions. Controlling
methane, CO, and non-methane VOCs may be a promising means of simultaneously mitigating
climate change and reducing global Os concentrations (West et al. 2007). Reducing these
precursors reduces global concentrations of the hydroxyl radical (OH), their main sink in the
atmosphere, feeding back on their lifetime and further reducing Os production. NOx reductions
decrease OH, leading to increased methane lifetime and increased Os production globally in the
long-term. The resulting positive radiative forcing from increased methane may cancel or even
slightly exceed the negative forcing from decreased Os globally (West et al. 2007). Of the Os
precursors, methane abatement reduces climate forcing most per unit of emissions reduction, as
methane produces Os on decadal and global scales and is itself a strong climate forcer. Since
they may have different effects on concentrations of different species in the atmosphere, all Os
precursors must be considered in evaluating the net climate impact of emission sources or
mitigation strategies.
6.7 URBAN CASE STUDY AIR POLLUTION REMOVAL
In addition to sequestering and storing carbon, urban forests also remove pollutants from
the local atmosphere. The reduction in growth rates resulting from Os exposure would reduce the
current and future amount of pollutants removed by these forests. We used the i-Tree model
6-55
-------
described in Section 6.5.2 to estimate the removal of air pollutants by the forests in the urban
areas discussed.
The preliminary results for changes in air pollution removal estimates for carbon
monoxide, nitrogen dioxide, Os, and sulfur dioxide show reduced capacity for these urban forest
canopies to remove pollution (1) at recent ambient Os conditions and (2) after adjusting air
quality to just meeting the existing standards and alternative standards. These analyses show that
even at the lowest scenario urban forest capacity to remove pollution is still reduced compared to
a no ozone scenario. Because of the limitations in the availability of E-R functions for all of the
common tree species in urban areas, and because of the limited number of urban areas for which
the i-Tree model has been applied, these reductions only reflect a portion of the impacts on
pollution removal by urban forests in the U.S. Though the model does include estimates for
particulate matter (PM), we do not include those estimates because the model does not yet
distinguish between PMio and PIVfo.s, and this distinction is important for evaluating the potential
health and welfare effects associated with PM. Estimates suggest that after meeting the existing
standard about 1,535 tons of air pollution removal capacity is lost annually (or about 38,384 tons
over 25 years) in the five areas modeled. As in the simulations for carbon storage, Syracuse and
Baltimore see the least change in capacity with the urban areas of Tennessee reporting the largest
changes. Syracuse and Baltimore have no change in pollution removal when meeting the
existing and the modeled alternatives. Atlanta and Chicago gain about 470 and 6,500 metric tons
of additional pollution removal after meeting the alternative W126 standard of 7 ppm-hrs
compared to meeting the existing standard, while Tennessee gains almost 12,000 tons of
potential pollution removal annually for the same comparison. For the 7 ppm-hrs scenario, about
51 percent of the pollution removal capacity lost under the existing standard is regained. See
Table 6-22 for details.
We performed a simple analysis of the Os removal potential to show how this process
might affect ambient air quality values. The analysis makes some general assumptions to
estimate order of magnitude effects of Os removal by trees on Os concentrations in the five urban
areas. To make this calculation, the metric tons of Cb removed listed in Table 6-20 are spread
evenly over every hour in the 25-year tree lifetime to achieve an hourly Os removal. Using the
ideal gas-law, this mass can be converted to an equivalent volume of gas assuming standard
temperature and pressure. Each urban area is treated as a well-mixed volume with the height
6-56
-------
determined as the average maximum daytime boundary layer height7 extracted from an April-
October 2007 Weather Research Forecasting (WRF) model simulation for each area of interest.
The ratio of the Os volume to the urban area air volume multiplied by 109 gives an equivalent
concentration in ppbv. The effects on Os concentration are generally small; deposition to tree
surfaces results in ambient Os concentration reductions ranging from 0.08 ppbv in Tennessee to
0.52 ppbv in Chicago. Differences between the scenarios are minute. The base case numbers
are consistent with previously published values from Song et al. (2008) who used a
photochemical model to show that changes in land use from development in Austin, TX, might
lead to a 0-0.3 ppbv change in Os concentration due solely to deposition differences. Some
additional benefit may be achieved from cumulative effects, which are not accounted for here
(i.e., Os removed at 9am will not only decrease concentrations instantaneously, but will also
decrease the starting concentration to some degree at 10am, 1 lam, etc. throughout the day). In
addition, changing the boundary layer height based on variability in this value could increase or
decrease the ppbv estimates by a factor of two. But in any case, the values would still be small.
7 The maximum daytime boundary layer height is the depth in the atmosphere over which air is well-mixed in the
afternoon. The WRF modeling simulation showed that this depth was approximately 1700m in Atlanta, 1500m in
Baltimore, 1150m in Chicago, 1350m in Syracuse, and 1750m in Tennessee.
6-57
-------
Table 6-22 Comparison of Pollutant Removal Between an Unadjusted Scenario and
Alternatives and Gains Between the Existing Standard and Alternatives (metric tons)
No Os Adjustment
(NOA)
Existing
Standard (75
ppb)/15 ppm-hrs
(ES/15)
NOA
V
ES/15
NOA v
11 ppm-
hrs
NOAv
7 ppm-
hrs
ES/15 v
11 ppm-
hrs
ES/15 v
7 ppm-
hrs
CO
Atlanta
Baltimore
Chicago
Syracuse
Tennessee
1,482
186
8,620
55
12,854
1,429
186
8,001
55
12,626
-54
0
-619
0
-227
-50
0
-569
0
-97
-34
0
-476
0
62
3
0
142
0
131
9
0
235
0
290
NO2
Atlanta
Baltimore
Chicago
Syracuse
Tennessee
6,852
1,968
104,247
50
54,381
6,605
1,963
96,766
50
53,419
-248
-5
-7,481
0
-962
-231
-5
-6,883
0
-408
-159
-5
-5,758
0
263
16
5
598
0
554
88
5
1,723
0
1,226
Os
Atlanta
Baltimore
Chicago
Syracuse
Tennessee
25,495
6,262
243,701
1,544
393,205
24,574
6,247
226,214
1,541
386,247
-921
-15
-17,488
-4
-6,957
-861
-15
-16,090
-4
-2,953
-591
-15
-13,460
-4
1902
60
0
1,398
0
4,004
331
0
4,028
0
8,860
SO2
Atlanta
Baltimore
Chicago
Syracuse
Tennessee
3,380
852
29,675
71
59,371
3,257
850
27,546
71
58,320
-122
-2
-2,129
0
-1,050
-114
-2
-1,959
0
-446
-78
-2
-1,639
0
287
8
0
170
0
605
44
0
490
0
1,338
Total
Atlanta
Baltimore
Chicago
Syracuse
Tennessee
37,209
9,268
386,242
1,721
519,810
35,865
9,246
358,527
1,717
510,613
-1,344
-22
-27,817
-4
-9,197
-1,825
-22
-25,501
-4
-3,904
-862
-22
-21,333
-4
2,514
87
5
2,308
0
5,294
472
0
6,476
0
11,714
ES = Existing Standard
NOA = No Adjustment to Growth Rates Using Os-related E-R Functions
6-58
-------
6.8 ECOSYSTEM-LEVEL EFFECTS
To assess the risk to ecosystems from biomass loss, as opposed to the potential risk to
individual tree species, we attempted to combine the RBL values into one metric. One factor in
assessing the risk to ecosystems is a measure of the overall abundance of each species. As a
measure of overall abundance, we used the basal area estimates described in Section 6.2.1 to
calculate the proportion of basal area for each of the 12 species assessed. Table 6-23 reflects, by
region, the total basal area covered by the 12 tree species assessed. We separated the total basal
area covered into different categories of percent cover of the species assessed. For example, in
the Southwest region, 13 percent of the total basal area assessed had less than 10 percent cover of
the 12 tree species; 7.1 percent of the total basal area assessed had between 10 and 25 percent
cover of the 12 tree species; 8.8 percent of the total basal area assessed had between 25 and 50
percent cover of the 12 tree species; and 64.9 percent of total basal area assessed had no data on
percent cover of the 12 tree species. The Southwest and West regions had the largest
percentages of total basal area assessed with no data on percent cover of tree species, and the
Central and Northeast regions had the smallest percent of total basal area assessed with no data
on percent cover of tree species.
Table 6-23 Percent of Total Basal Area Covered by 12 Assessed Tree Species
Region
Central
East North Central
Northeast
Northwest
South
Southeast
Southwest
West
West North
Central
All Regions
Percent of Total Basal Area Covered by 12 Assessed Tree Species
^10%
38.4%
33.4%
7.0%
4.5%
28.6%
16.0%
13.0%
10.0%
20.2%
20.3%
10% to
25%
32.0%
25.7%
22.1%
7.7%
4.0%
14.2%
7.1%
3.7%
8.0%
12.0%
25% to
50%
26.6%
27.5%
47.9%
20.0%
7.7%
48.1%
8.8%
7.0%
9.7%
19.1%
50% to
75%
2.2%
8.9%
22.2%
24.5%
7.7%
17.7%
5.1%
5.5%
8.2%
10.0%
> 75%
<0.1%
0.1%
0.5%
15.0%
0.9%
0.3%
1.2%
0.2%
6.5%
2.7%
No Data
0.7%
4.3%
0.3%
28.3%
51.2%
3.8%
64.9%
73.5%
47.4%
35.9%
6-59
-------
Data on basal area were available in over 64 percent of the cover area assessed, as
measured by the number of grid cells. To understand the potential W126 index values in the
percent of cover area not assessed, Table 6-24 includes information on the (i) number of grid
cells with no data on basal area above a certain amount and (ii) total number of grid cells with no
data on basal area. For those grid cells with no data on basal area, the table also shows, under
recent conditions, the number of grid cells with W126 index values that would exceed potential
alternative standards of 15, 11, and 7 ppm-hrs. In the Southwest, under recent conditions, 52
percent of the grid cells with no data have W126 index values above 15 ppm-hrs, 95 percent
have W126 index values above 11 ppm-hrs, and 100 percent have W126 index values above 7
ppm-hrs. In contrast, in the East North Central, under recent conditions, no grid cells with no
data have W126 index values above 15 ppm-hrs, 1 percent have W126 index values above 11
ppm-hrs, and 3.5 percent have W126 index values above 7 ppm-hrs.
Table 6-24 Grid Cells With No Data That Exceed W126 Index Values under Recent
Conditions
Region
Central
East North
Central
Northeast
Northwest
South
Southeast
Southwest
West
West North
Central
All Regions
Number of Grid
Cells with No Data
35
198
11
1,256
5,239
200
4,904
3,550
4,013
19,406
Number of Grid Cells w/No Data that Exceed
W126 Index Values Under Recent Conditions
> 7 ppm-hrs
34
7
11
779
4,638
59
4,904
3,452
1,870
15,754
> 11 ppm-hrs
11
2
11
451
1,945
15
4,662
3,274
1,158
11,529
> 15 ppm-hrs
3
0
6
189
27
3
2,572
2,680
283
5,763
We used the proportion of total basal area for each species to weight the RBL value for
that species in each grid cell. The weighted values for all species present in each grid cell were
summed to generate a weighted RBL value for each grid cell. Table 6-25 provides a summary of
6-60
-------
the percent of total basal area that exceeds a 2 percent weighted biomass loss under recent
conditions, at just meeting the existing standard (75 ppb) and at potential alternative W126
standard levels of 15, 11, and 7 ppm-hrs. The data are also presented excluding Cottonwood,
which is a very sensitive species. Note that for biomass loss, CASAC recommended that EPA
should consider options for W126 standard levels based on factors including a predicted one to
two percent biomass loss for trees and a predicted five percent loss of crop yield (Frey and
Samet, 2012b). Small losses for trees on a yearly basis compound over time and can result in
substantial biomass losses over the decades-long lifespan of a tree. We chose to use the 2
percent biomass loss recommendation in this analysis; however, the weighted RBL value is not
the same as the individual species analysis (Section 6.2.1.3). These data are interpreted in a
more relative manner where higher values represent a larger potential impact on the overall
ecosystem.
The data in Table 6-25 shows that the total area exceeding two percent biomass loss
decreases, as expected, across air quality scenarios. For example, for the Central region under
recent conditions, a total of 23.4 percent of total basal area assessed would exceed a 2 percent
biomass loss and when adjusted to just meet the existing standard, a total of 2.7 percent of total
basal area assessed would exceed a 2 percent biomass loss. When adjusted to meet potential
alternative standard levels of 15, 11, and 7 ppm-hrs, 2.7 percent, 1 percent and 0.1 percent,
respectively, of total basal area assessed would exceed a 2 percent biomass loss.
While it is not possible to predict overall effects, the results from these analyses show the
weighted RBL to be a potential predictor of risk in areas with species present for which E-R
functions were available.
6-61
-------
Table 6-25 Percent of Area Exceeding 2 Percent Weighted Biomass
Region
Central
East North Central
Northeast
Northwest
South
Southeast
Southwest
West
West North Central
All Regions
Region
Central
East North Central
Northeast
Northwest
South
Southeast
Southwest
West
West North Central
All Regions
Percent of Area Exceeding 2 Percent Weighted Biomass Loss
(12 Assessed Tree Species)
Recent
Conditions
23.4%
13.6%
18.0%
2.7%
2.2%
9.2%
11.1%
4.8%
15.4%
10.8%
75 ppb
2.7%
0.6%
0.2%
0.0%
0.2%
0.0%
0.5%
0.0%
2.2%
0.8%
15 ppm-hrs
2.7%
0.6%
0.2%
0.0%
0.2%
0.0%
0.2%
0.0%
2.0%
0.7%
11 ppm-hrs
1.0%
0.4%
0.0%
0.0%
0.1%
0.0%
0.1%
0.0%
1.8%
0.5%
7 ppm-hrs
0.1%
0.3%
0.0%
0.0%
0.1%
0.0%
<0.1%
0.0%
1.0%
0.2%
(11 Tree Species, excluding Cottonwood)
Recent
Conditions
15.4%
8.0%
17.2%
2.7%
0.2%
9.1%
10.7%
4.8%
6.9%
7.6%
75 ppb
0.9%
0.6%
0.2%
0.0%
0.0%
0.0%
0.3%
0.0%
0.2%
0.2%
15 ppm-hrs
0.9%
0.6%
0.2%
0.0%
0.0%
0.0%
0.0%
0.0%
0.2%
0.2%
11 ppm-hrs
0.2%
0.4%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.2%
0.1%
7 ppm-hrs
<0.1%
0.3%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
<0.1%
<0.1%
6-62
-------
Two important things to note with respect to the weighted RBL analysis. First, the
proportional basal area values do not account for total cover, only for the relative cover of the
tree species present. This is most noticeable with Cottonwood and Ponderosa pine, which are
near 100 percent cover in some areas; however, the absolute cover is very different. Ponderosa
pine occurs in relatively high density in some grids, exceeding 100 square feet per acre, while
Cottonwood is often less than 10 square feet per acre. This affects the direct interpretation of the
values presented because the overall ecosystem effect may be very different, although equally
important. It is important to remember with this data set that these numbers are only useful as a
very general estimate of potential effects. Second, this analysis only accounts for the 12 tree
species with E-R functions; other species are known to be sensitive to Os exposure, but E-R
functions were not available. It is also possible other species that are not sensitive may be
indirectly affected through changes in community composition and competitive interactions.
6.8.1 Potential Biomass Loss in Federally Designated Areas
6.8.1.1 Class I Areas
We analyzed federally designated Class I areas in relation to the W126 air quality surface
and the weighted RBL values. We completed the analyses of Class I areas in the same manner as
the analyses across the entire range of data; however, we present the results as a count of the
Class I areas and not as a percentage of area. We treated each Class I area as an individual
geographic endpoint and calculated an average weighted RBL for all Class I areas with at least
one grid cell that had a non-zero weighted RBL. Data were available in 145 of the 156 Class I
areas. A complete list of Class I areas and the weighted RBL values at the existing standard and
alternative W126 standard levels is included in Appendix 6E.
Table 6-26 summarizes the number of Class I areas exceeding 1 percent and 2 percent
weighted RBL across varying percent cover of species and under recent conditions and when
adjusted to just meet the existing standard and potential alternative standard levels of 15, 11, and
7 ppm-hrs. The number of areas exceeding 1 percent and 2 percent decreases across air quality
scenarios.
6-63
-------
Table 6-26 Weighted RBL and Percent Cover in Class I Areas
Percent
of Total
Basal
Area
No Data
<10
10 to 25
25 to 50
50 to 75
>75
Total
Areas
Class I
Areas
Covered
11
54
35
48
6
2
156
Number of Class I Areas Exceeding 1%
Weighted RBL
Recent
Conditions
-
6
8
20
1
1
36
75
ppb
-
2
0
1
0
1
4
15
ppm-
hrs
-
2
0
1
0
1
4
11
ppm-
hrs
-
2
0
0
0
1
3
7
ppm-
hrs
-
1
0
0
0
1
2
Number of Class I Areas Exceeding 2%
Weighted RBL
Recent
Conditions
-
2
2
7
1
1
13
75
ppb
-
1
0
0
0
1
2
15
ppm-
hrs
-
1
0
0
0
1
2
11
ppm-
hrs
-
1
0
0
0
1
2
7
ppm-
hrs
-
0
0
0
0
1
1
6.9 QUALITATIVE ASSESSMENT OF UNCERTAINTY
As noted in Chapter 3, we have based the design of the uncertainty analysis for this
assessment on the framework outlined in the WHO guidance (WHO, 2008). For this qualitative
uncertainty analysis, we have described each key source of uncertainty and qualitatively assessed
its potential impact (including both the magnitude and direction of the impact) on risk results, as
specified in the WHO guidance. In general, this assessment includes qualitative discussions of
the potential impact of uncertainty on the results (WHO Tierl) and quantitative sensitivity
analyses where we have sufficient data (WHO Tier 2).
Table 6-27 includes the key sources of uncertainty identified for the Os WREA. For each
source of uncertainty, we have (a) provided a description, (b) estimated the direction of influence
(over, under, both, or unknown) and magnitude (low, medium, high) of the potential impact of
each source of uncertainty on the risk estimates, (c) assessed the degree of uncertainty (low,
medium, or high) associated with the knowledge-base (i.e., assessed how well we understand
each source of uncertainty), and (d) provided comments further clarifying the qualitative
assessment presented. The categories used in describing the potential magnitude of impact for
specific sources of uncertainty on risk estimates (i.e., low, medium, or high) reflect our
consensus on the degree to which a particular source could produce a sufficient impact on risk
estimates to influence the interpretation of those estimates in the context of the secondary Os
6-64
-------
NAAQS review. Where appropriate, we have included references to specific sources of
information considered in arriving at a ranking and classification for a particular source of
uncertainty.
6-65
-------
Table 6-27 Summary of Qualitative Uncertainty Analysis in Relative Biomass Loss Assessments
Source
Description
Potential influence of
uncertainty on risk
estimates
Direction
Magnitude
Knowledge-
Base
Comments (KB: knowledge base, INF: influence of
uncertainty on risk estimates)
A. National W126
surfaces
The biomass loss analyses in
this chapter use the national
W126 surfaces for recent
conditions and adjusted to just
meet the existing standard and
alternative W126 standards.
Both
Low-
Medium
Low-medium
KB and INF: See Chapter 4 for more details.
B. Shape of the E-R
function for biomass
loss for different
species
Biomass loss and yield loss
estimates are highly sensitive to
the parameters in the E-R
function.
Unknown
High
Medium
KB: We conducted sensitivity analyses for 10 crops (in 54
studies) and 12 tree species (in 52 studies), which showed
high intraspecific and interspecific variability. Some species
only had one study, while other species had many studies.
INF: The resulting E-R functions for the included species
were mostly of intermediate sensitivity, with only a few
species considered very sensitive and several that showed
little or no sensitivity to Os. This range of sensitivities was
consistent with the additional studies included in the Os ISA,
but further studies are needed to determine how accurately
this reflects the larger suite of tree species in the U.S.
C. Absence of E-R
functions for many
Os-sensitive species
E-R functions are available for
only 12 tree species, thus the
majority of trees in the modeled
urban areas and Class I areas
were not incorporated.
Under
Medium-
High
Medium-Low
KB: We are certain that there are additional sensitive species
based on studies cited in the Os ISA that reported effects.
However, the studies of additional sensitive species did not
provide sufficient information to generate E-R functions.
Therefore, we are certain that we are underestimating tree
biomass loss in urban areas and Class I areas.
INF: Eighty to 90 percent of the total trees in the urban case
study areas are excluded from the analysis. There are 2 tree
species in the case study areas that we know are sensitive but
for which no E-R function is available. The magnitude of the
influence is dependent on the community composition in each
area.
6-66
-------
Source
Description
Potential influence of
uncertainty on risk
estimates
Direction
Magnitude
Knowledge-
Base
Comments (KB: knowledge base, INF: influence of
uncertainty on risk estimates)
D. Using E-R
functions for tree
seedlings rather than
adult trees
E-R functions for trees are
based on analyses of tree
seedlings, but most biomass
impacts are from effects on
adult trees.
Both
Low-
Medium
Medium
KB and INF: In general, estimates of relative biomass loss
(RBL) in tree seedlings are comparable to the estimates for
adult trees, with a few exceptions such as black cherry. Some
species overestimate RBL in adult trees and some species
underestimate RBL.
E. Urban tree
inventory in iTree
The base inventory of urban
trees, including species and
distribution, in iTree has
uncertainty.
Unknown
Low
High
KB: The urban tree inventories included in the iTree analyses
are based on field counts and measurements of trees in the
specific urban areas analyzed (personal communication,
Nowak, 6/2011). Tree census data (e.g., Baltimore, Syracuse,
Chicago, and Atlanta) are generally considered less uncertain
than modeled tree inventories (e.g., urban areas of
Tennessee).
INF: The iTree model estimates carbon sequestration and
pollution removal services provided by urban forests. These
services are based on tree growth and pollution removal
functions that are specific to the forest structure in each urban
area, including the species composition, number of trees, and
diameter distribution of trees. Uncertainties in the tree
inventory are propagated into the estimates of carbon
sequestration and pollution removal based on those
inventories.
F. Pollution removal
functions in iTree
The functions applied in iTree
to estimate pollution removal
are uncertain and vary by
species.
Unknown
Medium
Medium
KB: Pollution removal is calculated based on field, pollution
concentration, and meteorological data. The pollution
removal functions in iTree are from Nowak et al. (2006).
INF: iTree estimates that 1,535 tons/year of pollution are
removed from the urban case study areas at the existing
standard. Nowak et al. (2006) provides an indication of the
ranges of pollution removal in the literature.
6-67
-------
Source
Description
Potential influence of
uncertainty on risk
estimates
Direction
Magnitude
Knowledge-
Base
Comments (KB: knowledge base, INF: influence of
uncertainty on risk estimates)
G. VOC emissions
from trees
Many tree species are biogenic
sources of volatile organic
compounds (VOC) that
contribute to formation of
ozone. Additional VOC
emissions associated with
biomass gains are not
addressed.
Over
(generally)
Medium
High
KB: According to the O3 ISA (U.S. EPA, 2013, section
3.2.1), vegetation emits substantial quantities of VOCs, and
the 2005 NEI approximately 29 MT/year of VOC emissions
were from biogenic sources.
INF: Vegetation may account for as much as two-thirds of the
VOC production (Guenther et al., 2006). Carlton et al. (2010)
found, however, that if man-made pollutants were not present,
O3 attributable to biogenic emissions would drop by as a
much as 50 percent.
H. Carbon
sequestration
functions in iTree and
FASOM
The functions applied in the
models to estimate carbon
sequestration are uncertain and
vary by species.
Unknown
Medium
Medium
KB: The studies in the Os ISA show a consistent pattern of
reduced carbon uptake due to Os damage, with large
reductions projected over North America. The forest carbon
accounting component of FASOMGHG is largely derived
from the U.S. Forest Service's Forestry Carbon (FORCARB)
modeling system, which is an empirical model of forest
carbon budgets simulated across regions, forest types, land
classes, forest age classes, ownership groups, and carbon
pools. Multiple equations for individual species were
combined to produce one predictive equation for a wide range
of diameters for individual species. Formulas were combined
to prevent disjointed sequestration estimates that can occur
when calculations switch between individual biomass
equations. If no allometric equation could be found for an
individual species, the average of results from equations of
the same genus is used. If no genus equations are found, the
average of results from all broadleaf or conifer equations is
used.
INF: We estimate that carbon storage would increase by 13
million metric tons and 1.6 billion metric tons over 40 years
after just meeting the existing and the alternative standard
level of 7 ppm-hrs, respectively. The process of combining
the individual formulas produced results that were typically
within 2% of the original estimates.
6-68
-------
Source
Description
Potential influence of
uncertainty on risk
estimates
Direction
Magnitude
Knowledge-
Base
Comments (KB: knowledge base, INF: influence of
uncertainty on risk estimates)
I. Use of median E-R
functions for crops in
FASOM
FASOMGHG incorporates
median parameters from Lehrer
etal. (2007) in the E-R
functions for oranges, rice, and
tomatoes. Using alternative E-R
functions would result in lower
or higher Os impacts on crop
and tree species biomass
productivity, which would
potentially lead to different
economic equilibrium
outcomes.
Both
Low
Low
KB: These 3 crops have C-R functions based on Os metrics
other than W126, as reported in Lehrer (2007).
INF: Use of the median function could affect the estimates for
those crops specifically. No other crop estimates are based on
these functions.
J. Crop proxy and
forest type
assumptions
The crops/tree species modeled
are only a subset of species
present in U.S. agriculture and
forestry systems. Actual
impacts may differ from those
of the crop proxy or the forest
type. Further, FASOMGHG
modeling used a simple average
of tree RYLs for all forest types
within a region.
Both
Medium-
High
Low
KB: Aggregation of crop and tree species was conducted
based on recommendations from CASAC (Frey and Samet,
2012a). As stated by CASAC, it is not feasible to obtain E-R
functions for all species, and there is no reliable mechanism
to infer E-R relationships in a novel species even from
knowledge of a closely-related species.
INF: Total economic surplus is estimated to decrease by $24
million or increase by as much as $257 million between 2010
and 2040. It is unclear how using actual species information
rather than proxy species would affect these estimates.
However, consistent with CASAC recommendation, we did
not assign the most sensitive E-R relationships to the proxy
species.
6-69
-------
Source
Description
Potential influence of
uncertainty on risk
estimates
Direction
Magnitude
Knowledge-
Base
Comments (KB: knowledge base, INF: influence of
uncertainty on risk estimates)
K. FASOMGHG
does not model
agriculture/ forestry
on public lands
Because public lands are not
affected within the model, the
estimates of changes in
consumer and producer surplus
would likely be higher if public
lands were included.
Under
Medium
Medium-Low
KB: The model assumes that O3 biomass effects would have
little influence on harvest decisions because timber harvests
on public lands are set by the relevant government regulating
body (Forest Service, Bureau of Land Management, etc).
INF: The FASOMGHG model includes 349 million acres of
private, managed forests. The USFS estimates that there are
approximately 751 million forest acres in the U.S., but only a
small portion of this public land is logged for timber.
L. Forest adaptation
toO3
FASOMGHG modeling does
not reflect changes in tree
species mixes within a forest
type made by natural adaptation
and adaptive management by
landowners due to Os. Less
sensitive tree species may gain
relative advantage over more
sensitive species.
Unknown
Low
Low
KB: The Os ISA finds that the evidence is sufficient to
conclude that Os causes changes in community composition
favoring Os tolerant species over sensitive species. The KBs
for natural adaptation and adaptive management are different,
and the relative dominance of one over the other would differ
depending on the degree of active management.
INF: Over time, the Os impacts on forests may be reduced as
forests adapt to Os environments through forest management
or natural processes.
M. International trade
projections in
FASOMGHG
FASOMGHG reflects future
international trade projections
by USDA based on recent Os
conditions. Soybeans and wheat
are major crop exports and have
relatively large responses to Os,
which are not reflected in the
trade projections.
Both
Medium
Medium
KB: Although FASOMGHG includes international trade for
major commodities, the international trade projections do not
reflect the potential for increased exports associated with
increased yield from reduced Os exposure. The world trade
quantities data in the model have been updated to reflect more
recent trade data for specific commodities in the literature
since the original data from the USDA SWOPSDVI model
(Roningen, 1986).
INF: Increased exports could increase producer surplus but
the impacts on consumer surplus are unclear.
6-70
-------
Source
Description
Potential influence of
uncertainty on risk
estimates
Direction
Magnitude
Knowledge-
Base
Comments (KB: knowledge base, INF: influence of
uncertainty on risk estimates)
N. Estimates of tree
basal area used to
assess larger scale
ecosystem effects
Estimates of basal area were
modeled by the FHTET at a
scale of 240 m2. These values
were aggregated to the 144
(12x12) km2 CMAQ grid.
Unknown
Low-
Medium
Low
KB: USDA's FHTET has been actively working to refine
their models to estimate basal area for individual tree species
and total basal area nationwide.
INF: The effect on risk estimates would vary between
ecosystems, depending on community composition, total
basal area and the ecosystem services being affected. Due to
the overall large number of CMAQ cells included for each
species, the overall estimates presented here would likely be
small.
6-71
-------
6.10 KEY OBSERVATIONS
Relative Biomass Loss:
• We compared seedling RBL to results from several studies with mature trees with
mixed results. The studies indicate that overall the seedling biomass loss values are
much more consistent with the adult loss at lower W126 index values.
• The Constable and Taylor (1997) study implies that for the eastern subspecies of
Ponderosa Pine, the seedling RBL rate could possibly overestimate the adult RBL
rate.
• The Weinstein et al. (2001) study indicates that the seedling RBL estimates are
comparable to the adult estimates, except at higher W126 index values for Tulip
Poplar. The Black Cherry results are an exception, which tells us that this species is
much less sensitive as an adult than as a seedling.
• Another study (Samuelson and Edwards, 1993) on Red Oak found the exact opposite
pattern — adult trees are much more sensitive to Os-related biomass loss than
seedlings.
• Overall, the western tree species have more fragmented habitats than the eastern
species. The areas in southern California have the highest W126 index values. The
eastern tree species had less fragmented ranges and areas of elevated RBL that were
more easily attributed to urban areas (e.g. Atlanta, GA and Charlotte, NC) or to the
Tennessee Valley Authority Region.
Commercial Timber Effects:
• At the existing standard of 75 ppb the highest yield loss occurs in upland hardwood
forests in the South Central and Southeast regions at over 3 percent per year. The next
highest yield losses occur in Corn Belt hardwoods with just over 2 percent loss per
year and in hard- and softwoods of the Rocky Mountain region at an average loss
across all sensitive forests of slightly over 1 percent loss per year. With the exception
of the Rocky Mountain region, yield losses do not appreciably change when meeting
6-72
-------
the 15 ppm-hrs alternative incremental to meeting the existing standard. Yield gains
associated with meeting alternative W126 standards are relatively small on a
percentage change basis, especially in the 15 ppm-hrs scenario where the highest
change is 0.35 percent per year.
• Consumer and producer welfare in the forest sector are more affected by meeting
alternative W126 standards incremental to meeting the existing standard than the
agricultural sector. In general, consumer welfare increases in both the forest and
agricultural sectors as higher productivity tends to increase total production and
reduce market prices. Because demand for most forestry and agricultural
commodities is inelastic, producer welfare tends to decline with higher productivity
as the effect of falling prices on profits more than outweighs the effects of higher
production levels.
Climate Regulation:
• For national-scale carbon sequestration, much greater changes in carbon sequestration
are projected in the forest sector than in the agricultural sector. The 15 ppm-hrs
scenario does not appreciably increase carbon storage relative to just meeting the
existing standard. The vast majority of the enhanced carbon sequestration potential
under the scenarios is from increased forest biomass due to the yield increases
accruing to forests over time at the 11 and 7 ppm-hrs alternative W126 standards.
The forest carbon sequestration potential would increase between 593 and 1,602
million tons of CCh equivalents over 30 years after meeting the 11 or 7 ppm-hrs
W126 standard level, respectively.
• For the urban case study areas, estimates suggest that in the five modeled areas
relative to recent conditions, at the existing standard or at an alternative W126
standard level of 15 ppm-hrs about 3.5 million tons of carbon storage will be lost over
25 years (about 140,000 tons/year). At an alternative W126 standard level of 11
ppm-hrs, loss of carbon sequestration is approximately 128,000 metric tons per year,
and meeting an alternative W126 standard of 7 ppm-hrs results in the loss of 112,000
metric tons per year of carbon storage services.
6-73
-------
• Of the five areas modeled, the combined urban areas of Tennessee have the largest
estimated gains in carbon storage at almost 20,000 tons per year when meeting an
alternative W126 standard of 7 ppm-hrs relative to the existing standard.
Urban Case Study Air Pollution Removal:
• Estimates from i-Tree indicate that at the existing standard about 1,535 tons of air
pollution removal capacity is lost annually in the five areas modeled. Syracuse and
Baltimore have no change in pollution removal when meeting the existing standard
and the modeled alternatives. Atlanta and Chicago gain about 470 and 6,500 metric
tons of additional pollution removal when meeting the 7 ppm-hrs W126 alternative
standard compared to the existing standard, while Tennessee gains almost 12,000 tons
of potential pollution removal annually for this scenario. Under the 7 ppm-hrs
scenario, about 51 percent of the pollution removal capacity lost under the existing
standard is regained.
Agriculture:
• Among the major crops, winter wheat and soybeans are more sensitive to ambient Os
levels than corn and sorghum. California, the Northeast, and the Rocky Mountain
regions generally have the highest yield losses.
• For winter wheat, the highest loss occurs in California at 15 percent. In the
Northeast, the losses range from 7.65 percent in Maryland to 3.69 percent in
Pennsylvania, with 6.43 percent in Delaware and 6.55 percent in New Jersey. In the
Rocky Mountain region, the losses in Utah are 7.26 percent. When the W126
scenarios are modeled, the yield losses are almost eliminated at all values of W126.
• For soybeans, the highest loss occurs in Maryland at 8.3 percent. In the Northeast,
the losses range from 8.3 percent in Maryland to 5.38 percent in Pennsylvania, with
7.65 percent in Delaware and 7.76 percent in New Jersey. In the Corn Belt the
highest loss occurs in southern Indiana at 5.1 percent. In the Rocky Mountain region,
the losses in Colorado are 6.73 percent. Yield losses remain under all scenarios for
W126, although for the 7 ppm-hrs scenario all losses are less than 0.6 percent.
6-74
-------
For corn, the highest loss occurs in California at 0.88 percent. In the Northeast, the
losses range from 0.68 percent in Maryland to 0.26 percent in Pennsylvania, with
0.56 percent in Delaware and 0.48 percent in New Jersey. In the Corn Belt, Lake
States, and Great Plains the highest loss occurs in southern Ohio at 0.34 percent. And
in the Rocky Mountain region, the losses range from 0.67 percent in Utah to 0.42
percent in Nevada, with 0.45 percent in Colorado. When the W126 scenarios are
modeled, the yield losses are virtually eliminated at all values of W126 and
subsequent yield gains are almost nonexistent.
In general, increased yield leads to increased supply and lower prices. Because corn
does not lose or gain very much under any scenario prices are likely to remain
relatively stable. Soybeans, however, would experience yield gains in any scenario
and prices would likely fall. In response to falling soybean prices, the model predicts
that producers would switch to less Os-sensitive crops with stable prices, such as
corn, thereby increasing corn production.
For producers, the W126 alternatives results in welfare gains in the middle years,
2020-2030, and welfare losses in all other years. For consumers, however, the
changes in production and prices results in welfare gains in all scenarios in all years.
6-75
-------
7 VISIBLE FOLIAR INJURY
7.1 INTRODUCTION
Visible foliar injury resulting from exposure to ozone (Os) has been well characterized
and documented over several decades on many tree, shrub, herbaceous, and crop species (U.S.
EPA, 2013, 2006, 1996, 1984, 1978). Visible foliar injury symptoms are considered diagnostic
as they have been verified experimentally in exposure-response studies using exposure
methodologies such as continuous stirred-tank reactors (CSTRs), open-top chambers (OTCs),
and free-air fumigation (see Section 9.2 of the Os ISA for more detail on exposure
methodologies). Although the majority of Os-induced visible foliar injury occurrence has been
observed on seedlings and small plants, many studies have reported visible injury of mature
coniferous trees, primarily in the western U.S. (Arbaugh et al., 1998), and of mature deciduous
trees in eastern North America (Schaub et al., 2005; Vollenweider et al., 2003; Chappelka et al.,
1999a; Chappelka et al., 1999b; Somers et al., 1998; Hildebrand et al., 1996).
The ecosystem services most likely to be affected by Os-induced foliar injury are
aesthetic value and outdoor recreation. Aesthetic value and recreation services depend on the
perceived scenic beauty of the environment. Studies of Americans' perception of scenic beauty
are quite consistent (Ribe, 1994) in their findings - people tend to have a reliable set of
preferences for forest and vegetation with fewer damaged or dead trees and plants. Aesthetic
value not related to recreation includes the scenic value of vistas observed as people go about
their daily lives and the scenic value of the views of open space near and around homes. Many
outdoor recreation activities directly depend on the scenic value of the area, in particular scenic
viewing, wildlife watching, hiking, and camping. These activities are enjoyed by millions of
Americans every year and generate millions of dollars in economic value (OIF, 2012; NFS,
2002a, 2002b, 2002c). Figure 7-1 illustrates the relationship between foliar injury and ecosystem
services as discussed in this chapter.
7-1
-------
Ambient
Ozone
Exposure
Ecological Effect
Visible Foliar Injury
Ecosystem Level Effects
' National-Seal* Analysis of Foliar Injury
Cultural Services
•Recreational Use
• National Values of Trip and
Equipment-Related Expenditures
for Wildlife-Watching, Trail, and
Camping Ac ti vi ti es
* Screening-Level Assessment
of Foliar Injury in National
Parks
Cultural Services
•Recreational Use
• For 3 National Parks Case Studies,
Data on Activities, Travel and Local
Expenditures, and Local Economic
Impact
Figure 7-1 Relationship between Visible Foliar Injury and Ecosystem Services
The significance of Os injury at the leaf and whole-plant levels depends on how much of
the total leaf area of the plant has been affected, as well as the plant's age, size, developmental
stage, and degree of functional redundancy among the existing leaf area. Previous Os Air
Quality Criteria Documents (AQCDs) and the Os Integrated Science Assessment (Os ISA) for
have noted the difficulty in relating visible foliar injury symptoms to other vegetation effects
such as individual plant growth, stand growth, or ecosystem characteristics (U.S. EPA, 2013,
2006, 1996). As a result, it is not currently possible to determine, with consistency across
species and environments, what degree of injury at the leaf level has significance to the vigor of
the whole plant. However, in some cases, visible foliar symptoms have been correlated with
decreased vegetative growth (Somers et al., 1998; Karnosky et al., 1996; Peterson et al., 1987;
Benoit et al., 1982) and with impaired reproductive function (Chappelka, 2002; Black et al.,
2000). Conversely, the lack of visible injury does not always indicate a lack of phytotoxic effects
from Os or a lack of non-visible Os effects (Gregg et al., 2006).
The National Park Service (NFS) published a list of trees and plants considered sensitive
because they exhibit foliar injury at or near ambient concentrations in fumigation chambers or
they have been observed to exhibit symptoms in the field by more than one observer. This list
includes many species not included in Table 6-10, such as various milkweed species, asters,
7-2
-------
coneflowers, huckleberry, evening primrose, Tree-of-heaven, redbud, blackberry, willow, and
many others. Many of these species are important for non-timber forest products, recreation, and
aesthetic value among other ecosystem services.
Based on the NFS sensitive species list (NFS, 2003), data from the Forest Health
Technology Enterprise Team of the U.S. Forest Service (described in Chapter 6, Section 6.2.1.3)
were available for 15 tree species. Figure 7-2 illustrates the percent of total basal area that is
accounted for by these 15 species, which include Ponderosa Pine, Loblolly Pine, Virginia Pine,
Red Alder, Tulip Poplar, Aspen, Black Cherry, Jack Pine, Table Mountain Pine, Pitch Pine,
White Ash, Green Ash, Sweetgum, California Black Oak, and Sassafrass.
15 Tree Species Sensitive to Foliar Injury
Percent of Basal Area
D%
| 0.1% - 10%
10.1%-25%
25.1%-50%
^B 50.1%-75%
^H 75.1%- 100%
Figure 7-2 Tree Species Sensitive to Foliar Injury
Table 7-1 summarizes the overall cover of the 15 tree species and the percent of area in
each cover category that exceeds varying W126 index values. It is important to note that there
7-3
-------
are additional tree species that are known to be sensitive for which cover data were not available,
and there are many non-tree species listed in the NFS report that are not addressed in this
analysis.
Table 7-1 Percent of Cover Category Exceeding W126 Index Values
Cover Category
(percent of total basal area accounted
for by the 15 species included)
None Present
Less than 10%
10% to 25%
25% to 50%
50% to 75%
Greater than 75%
National
Distribution
34.5%
26.0%
17.0%
12.8%
7.9%
1.8%
Percent of Cover Category Area
Exceeding W126 Index Values
> 7 ppm-hrs
85.0%
65.4%
73.4%
79.3%
83.6%
82.4%
>10 ppm-hrs
71.1%
39.7%
52.7%
60.9%
57.2%
41.6%
> 15 ppm-hrs
31.4%
9.4%
13.9%
20.7%
15.7%
9.5%
In addition to direct impacts on foliar injury, Os exposure contributes to trees'
susceptibility to insect infestation (Grulke et al., 2009, U.S. EPA, 2006). These infestations can
affect scenic beauty and the services associated with the perceived beauty of the environment.
Foliar injury and insect attack can occur separately or in conjunction with one another and are
briefly discussed together in the next section of this chapter, Section 7.1.1, on ecosystem services
impacts. The remainder of this chapter provides details on the analyses we conducted and
includes Section 7.2 -National-Scale Analysis of Foliar Injury; Section 7.3 -Screening-level
Assessment of Visible Foliar Injury in National Parks; and Section 7.4 - National Park Case
Study Areas, including Great Smoky Mountains National Park, Rocky Mountain National Park,
and Sequoia and Kings Canyon National Parks. The national park case studies include
discussions of the potential value of the ecosystem services affected by foliar injury resulting
from Os exposure.
7.1.1 Ecosystem Services
7.1.1.1 Aesthetic Value
Aesthetic value services not related to recreation include the view of the landscape from
houses, as individuals commute, and as individuals go about their daily routine in a nearby
7-4
-------
community. Studies find that scenic landscapes are capitalized into the price of housing and also
document the existence of housing price premiums associated with proximity to forest and open
space (Acharya and Bennett, 2001; Geoghegan, Wainger, and Bockstael, 1997; Irwin, 2002;
Mansfield et al., 2005; Smith et al., 2002; Tyrvainen and Miettinen, 2000). In addition,
according to Butler (2008), approximately 65 percent of private forest owners rate providing
scenic beauty as either a very important or important reason for their ownership of forest land.
These aesthetic value services are at risk of impairment because of Os-induced damage:
directly due to foliar injury, and indirectly due to increased susceptibility to insect attack. Data
are not available to explicitly quantify these negative effects; however, the damage is included in
the price premium mentioned. In other words, without such damage, the associated price
premium for scenic beauty that is incorporated into housing prices is likely higher.
7.1.1.2 Recreation
With few exceptions, publicly owned forests are open for some form of recreation.
Based on the analysis done for the USDA National Report on Sustainable Forests (USDA, 2011),
almost all of the 751 million acres of forest lands in the U.S. are at least partially managed for
recreation. Of these 751 million acres, 44 percent are publicly owned (federal, state, or local).
Scenic quality has been found to be strongly correlated to recreation potential and the likelihood
of visiting recreation settings, and the correlations apply to both active and passive recreational
pursuits (Ribe, 1994). According to Ribe (1994), differences in scenic beauty account for 90
percent of the variation in participant satisfaction across all recreation types.
Americans enjoy a wide variety of outdoor pursuits many of which are subject to
negative impacts resulting from Os exposure, especially the effects on foliage, insect
susceptibility, habitat, and community composition. The effects related to scenic beauty (foliar
injury and insect damage) affect not only the scenery viewing, but also satisfaction with other
scenery-dependent activities. Ninety-seven percent of National Survey on Recreation and the
Environment (NSRE) survey respondents rated scenic beauty as an important or extremely
important aspect of their wilderness experience.
Perceptions of scenic beauty depend on a number of forest attributes, including the
appearance of forest health, the effects of air pollution and insect damage, visual variety, species
variety, and lush ground cover (Ribe, 1989). The Os ISA concludes that there is a causal
7-5
-------
relationship between Os exposure and visible foliar injury. Figure 7-3 shows the effects of foliar
injury on ponderosa pine, milkweed, and tulip poplar.
The presence of downed wood, whether caused by Os mortality, insect attack, or other
causes, has a negative impact on scenic beauty assessments (Ribe, 1989; Buyhoff, et al., 1982).
Figure 7-4 shows the effects of southern bark beetle damage. Species composition of forests
may also influence preferences. According to Ribe (1994) these preferences may be affected by
cultural, regional, or contextual expectations, which would include the expectation of the
presence of certain species in specific areas (e.g., the presence of ponderosa pine in California).
In addition, there is a positive effect on preferences for ground cover rather than bare or
disturbed soil (Brown and Daniel, 1984, 1986). Thus, the reduced value of scenic beauty from
Os-induced effects on sensitive plants, by way of foliar injury, extends beyond large trees to the
grasses, forbs, ferns, and shrubs that comprise the understory of a forest setting.
In Peterson et al. (1987), where Os-exposure had resulted in foliar injury to ponderosa
pines in the San Bernardino Forest, survey participants were asked to: (1) rank preferences for
scenic views, (2) rate their recreation experiences, (3) state how decreases in tree quality would
affect their visitation, and (4) specify whether they would be willing to pay for programs to
mitigate the damage. This survey showed that visible foliar injury had a negative impact on
perceptions of scenic beauty and a nonzero value for willingness to pay for programs to improve
forest aesthetics damaged by Os.
Figure 7-3 Examples of Foliar Injury from Os Exposure
Courtesy: NFS, Air Resources Division
7-6
-------
Figure 7-4 Examples of Southern Bark Beetle Damage
Courtesy: Ronald F. Billings, Texas Forest Service. Bugwood.org
The NSRE provides estimates of participation for many recreation activities. According
to the survey some of the most popular outdoor activities are walking, including day hiking and
backpacking; camping; bird watching; wildlife watching; and nature viewing. Participant
satisfaction with these activities depends wholly or partially on the quality of the natural scenery.
Table 7-2 summarizes the survey results, for these and other popular activities, including the
percent participation and the number of participants nationally, the number of days participants
engage in recreation activities annually, and their willingness-to-pay (WTP) for their
participation.
7-7
-------
Table 7-2 National Outdoor Activity Participation
Activity
Day Hiking
Backpacking
Picnicking
Camping (Developed
and Primitive Sites)
Visit a Wilderness Area
Birdwatching /
Photography
Wildlife Watching /
Photography
Natural Vegetation
Viewing / Photography
Natural Scenery Viewing
/ Photography
Sightseeing
Gathering (Mushrooms,
Berries, Firewood)
Percent
Participation
32.4
10.4
54.9
42.3
32.0
31.8
44.2
43.9
59.6
50.8
28.6
Number of
Participants
(in millions)
69.1
22.2
116.9
90.1
68.2
67.7
94.2
93.6
126.9
108.2
60.9
Number of
Activity Days
(in millions)
2,508
224.0
935.2
757.5
975.4
5,828.1
3,616.5
5,720.8
7,119.7
2,055.0
852.7
Mean
WTP/Day
(in 2010$)
$60.63
$13.33
$20.70
$19.98
N/A
$49.74
$48.72
N/A
N/A
$45.94
N/A
Mean Total
Participation
Value
(in millions of
2010$)
$152,060
$2,986
$19,359
$15,135
N/A
$289,773
$176,196
N/A
N/A
$94,407
N/A
Source: NSRE 2010 and 2003 National Report on Sustainable Forest Management. 2003 National Report:
Documentation for Indicators 35, 36, 37, 42, and 43 available at:
http://warnell.forestrv.uga.edu/nrrt/NSRE/MontrealIndDoc.PDF and Recreation Values Database available at:
http://recvaluation.forestry.oregonstate.edu/
N/A = not available
The relationship between scenic beauty and recreation satisfaction for camping was
quantified by Daniel et al. (1989) in a contingent valuation study. The authors surveyed campers
regarding their perceptions of scenic beauty, as indicated by a photo array of scenes along a
spectrum of scenic beauty, and their WTP to camp in certain areas. All else being equal, scenic
beauty and WTP demonstrated a nearly perfect linear relationship (correlation coefficient of
0.96). This suggests that campers would likely have a greater WTP for recreation experiences in
areas where scenic beauty is less damaged by Os. Since as mentioned previously, Ribe (1994)
found that scenic beauty plays a strong role in recreation satisfaction and explains 90 percent of
the difference in recreation satisfaction among all types of outdoor recreation, there is reason to
7-8
-------
believe that this linear relationship between scenic beauty and WTP would hold across all
recreation types. We believe that it would follow that decreases in Ch damage would generate
benefits to all recreators. We cannot estimate the incremental impact of reducing Os damage to
scenic beauty and subsequent recreation demand; however, given the large number of outdoor
recreation participants and their substantial WTP for recreation, even very small increments of
change in WTP or activity days should generate significant benefit to these recreators.
Another resource for estimating the economic value of consumers' recreation experiences
is the data available on actual expenditures for recreation and the total economic impact of
recreation activities. Economic impacts across the national economy can be estimated using the
IMPLAN® model (MIG Inc, 1999).1 For this document we refer to analyses done for the 2011
National Survey of Fishing, Hunting, and Wildlife-Associated Recreation (FHWAR) (U.S.
Department of the Interior, U.S. Fish and Wildlife Service, and U.S. Department of Commerce,
2011) and an analysis performed by Southwick and Associates for the Outdoor Industry
Foundation (OIF), The Economic Contribution of Active Outdoor Recreation - Technical Report
on Methods and Findings (OIF, 2012).
The FHWAR and the OIF report provide estimates of trip and equipment-related annual
expenditures for wildlife watching activities in the U.S. The OIF report also provides estimates
of recreators' annual expenditures on trail-related activities, camping, bicycling, snow-related
sports, and paddle sports. For this review, we include the data on trail-related activities and
camping as the most relevant for analysis of Os-related damages.
As shown in Table 7-3, the total expenditures across wildlife watching activities, trail-
based activities, and camp-based activities are approximately $230 billion dollars annually.
While we cannot estimate the magnitude of the impacts of Os damage to the scenic beauty, the
losses are reflected in the values reported.
1 IMPLAN® is a commercially available input-output model that has been used by the Department of Interior, the
National Park Service, and other government agencies in their analyses of economic impacts.
7-9
-------
Table 7-3 National Expenditures for Wildlife Watching, Trail, and Camp-Related
Recreation (in billions of 2010$)
Expenditure Type
Trip-Related
Equipment & Services
Other Expenditures
Total for All Expenditures
Wildlife-Watching"
$16.7
$26.3
$10.2
Trailb
$53.7
$6.3
N/R
Campb
$109.3
$8.3
N/R
Totalb
$179.7
$40.9
$10.2
$230.8
a Data from 2011 FHWAR
b Data from 2012 OIF report
N/R = not reported
The impact of these expenditures has a multiplier effect through the economy, which was
estimated by OIF using the IMPLAN® model.2 The model estimates the flow of goods and
money through the economy at scales from local to national. According to the OIF report
(2012), trail activities generated over $190 billion in total economic activity, including $97
billion in salaries, and wages. The same report estimates the total economic activity generated
by camping-related recreation at $346 billion, including $175 billion in salaries, and wages. The
total economic activity estimates also include state and federal tax revenues.
7.2 NATIONAL-SCALE ANALYSIS OF FOLIAR INJURY
To assess foliar injury at a national scale, we compared data from the Forest Health
Monitoring Network (USFS, 2011) with Os exposure estimates for individual years, described in
Section 4.3.1.2, and soil moisture data, which was estimated using NOAA's Palmer Z drought
index (NCDC, 2012b).
7.2.1 Forest Health Monitoring Network
The only national-scale data set pertaining to foliar injury is from the USDA Forest
Service's (USFS) Ozone Biomonitoring Program (OBP). This effort was completed as part of
the Forest Inventory and Analysis (FIA) and Forest Health Monitoring (FHM) programs (see
Figure 7-5 for the locations of the Os biomonitoring sampling sites, or "biosites"). A biosite is
a plot of land on which data was collected regarding the incidence and severity of visible foliar
2 Assumptions and Caveats to the IMPLAN®Results: Statistics on the precision of the final economic impacts were
not produced by OIF because of feasibility issues. Harris Interactive survey results combine several parameters
from the data, and outside data from the U.S. Bureau of the Census' population estimates and IMPLAN multipliers
were used.
7-10
-------
injury on a variety of Os-sensitive plant species. The OBP used a number of bioindicator
species (Os-sensitive plants) to monitor the potential impacts of Os on our nation's forests. The
field methods, sampling procedures, and analytical techniques were consistent across sites and
between years (USFS, 2011).
We obtained data on foliar injury from the USFS for the five years from 2006 to 2010.
Because of privacy laws that require the exact location information of biosites to not be made
public, the data were assigned to the CMAQ grid used for the Os exposure surface by the USFS
(USFS, 2013). Data were not available for California, Oregon, and Washington, so we used the
publically available data. In those states we assigned the data to the CMAQ grid based on the
publically available geographic coordinates, which are masked for privacy concerns as
mentioned above; the data in those states have additional uncertainty relating the Os and Palmer
Z drought index data to the foliar injury data. Also, because sampling was discontinued in
some states prior to this analysis, we did not include data for most of the western states
(Montana, Idaho, Wyoming, Nevada, Utah, Colorado, Arizona, New Mexico, Oklahoma, and
portions of Texas).
The "biosite index" is a measure of the severity of Os-induced visible foliar injury
observed at each biosite. The biosite index is calculated from a combination of the proportion
of leaves affected on individual bioindicator plants. In order to calculate the biosite index, at
least 30 individual plants of two bioindicator species must be present at each biosite. The mean
severity of symptoms ranges from a score of zero to a score of 100 (USFS, 2011), with a score
greater zero indicating the presence of foliar injury.
7-11
-------
Figure 7-5 Os Biomonitoring Sampling Sites ("Biosites")
Note: Sites are shown as the CMAQ grid cell in which they occur. Some biosites were sampled in more than one
year, but are indicated on this figure only as the most recent year sampled.
7.2.2 NOAA Palmer Z Drought Index
The Palmer Z drought index represents the difference between monthly soil moisture and
long-term average soil moisture (Palmer, 1965). These data typically range from -4 to +4, with
positive values representing more wetness than normal and negative values representing more
dryness than normal. Values between -1.25 and +1.0 could be interpreted as normal soil
moisture, whereas values beyond the range from -2.75 to +3.5 could be interpreted as extreme
drought and extremely moist, respectively (NCDC, 2012c).
The soil moisture index is calculated for each of the 344 climate regions divisions within
the contiguous U.S. defined by the National Climatic Data Center (NCDC) (NOAA, 2012a).
Because we did not have soil moisture data outside of the continental contiguous U.S., we did
not evaluate parks in Alaska, Hawaii, Puerto Rico, or Guam. We identify the NCDC climate
divisions with Palmer Z data in Figure 7-6.
7-12
-------
Figure 7-6 344 Climate Divisions with Palmer Z Soil Moisture Data
Source: NCDC, 2012a
7.2.3 Results of National-Scale Analysis
Data were available for a total of 5,284 biosites across the five years from 2006 - 2010
(Table 7-4, Figure 7-5). Table 7-4 summarizes the biosite index values for each year. The
categories used in Table 7-4 follow the USFS risk categories with the exception of including a
separate category for a biosite index of zero. We defined and use the zero category as a measure
of the presence or absence of foliar injury, without relying on potentially subjective
categorization of the biosite index values. We included the data to highlight that across all of the
sites, over 81 percent of the observations recorded no foliar injury. This percentage was similar
across all of the years, with a low value of 78 percent and a high value of 85 percent. The data
showed no clear relationship between Os and biosite index (Figure 7-7), as well as no clear
relationship between Os and the Palmer Z drought index (measured as an average value of the
months from April to August (Figure 7-8)).
7-13
-------
Table 7-4 Summary of Biosite Index Values for 2006 to 2010 Os Biomonitoring Sites
Categories modified from USFS (Smith et al., 2008)
Biosite
Index
0
<5
5 to 15
15 to 25
>25
Damage
None
Very Light
Light
Moderate
Heavy
Total
2006
744
139
41
15
12
951
2007
769
131
29
6
4
939
2008
796
98
29
8
4
935
2009
902
135
61
6
8
1,112
2010
1,075
183
65
12
12
1,347
Total
4,286
686
225
47
40
5,284
x
-------
o
LO -
o
o
X
-------
Table 7-5 Censored Regression Results
Coefficient
Intercept
W126
Palmer Z (Apr-Aug)
W126: Palmer Z
W126
Palmer Z (Apr-Aug)
W126: Palmer Z
Intercept Estimate
-22.5967
0.7307
1.8357
0.1357
Marginal Effect
0.1178
0.2960
0.0219
Std. Error
0.8934
0.0613
0.4850
0.0437
0.0099
0.0777
0.0070
t-value
-25.293
11.919
3.785
3.104
11.918
3.812
3.093
P
< 0.0001
0.0001
0.0002
0.0019
O.0001
0.0001
0.0020
To further assess the relationship between Os and foliar injury, we conducted a
cumulative analysis (Figure 7-9 through Figure 7-11). In these analyses, we ordered the data by
W126 index value, then for each W126 index value we calculated the proportion of sites
exceeding the selected biosite index value for all observations at or below that W126 index
value. In this analysis, we split the data into individual years, as well as into moisture categories;
the moisture categories followed NOAA's Palmer Z drought index, with values less than -1.25
considered dry, values greater than or equal to 1 considered wet, and values between those
considered normal.
When looking only at presence/absence of foliar injury ("any injury") (Figure 7-9), with
the exception of 2008, the proportion of sites across all W126 index values exceeds 15 percent;
in 2006, it exceeds 20 percent, while in 2008 the proportion of sites with foliar injury across all
W126 index values was just below 15 percent.
There are two important observations that can be made in these analyses: (1) the
proportion of sites exhibiting foliar injury rises rapidly at increasing W126 index values below
10 ppm-hrs, and (2) there is relatively little change in the proportions above W126 index values
of 20 ppm-hrs.
7-16
-------
Biosites with Foliar Injury
in
o
o
c\|
ci
in T-
.2 d
m
Q. *-
O O
Q-
in
q
ci
o
q
ci
All
2006
2007
2008
2009
2010
I
10
I
20
W126(ppm-hrs)
30
I
40
Figure 7-9 Cumulative Proportion of Sites with Foliar Injury Present, by Year
When categorized by moisture categories, as defined by the average Palmer Z drought
index, the data show a more distinct pattern. Similar to the analysis by individual years, the most
rapid increase in the proportion occurs at W126 index values below 10 ppm-hrs, but the moisture
category has a much greater effect on the overall proportion (Figure 7-10). In addition, there is
relatively little change in the proportion beyond a W126 of 20 ppm-hrs in normal and dry years.
The data for normal moisture sites are very similar to the dataset as a whole, with an
overall proportion of close to 18 percent for presence/absence. Sites classified as wet (average
Palmer Z > 1) have much higher overall proportions at any injury and a much more rapid
increase in proportion of sites with foliar injury present, exceeding 20 percent at W126 index
values under 5 ppm-hrs. At sites considered dry (average Palmer Z < -1.25), the overall
proportions are much lower, around 10 percent for presence/absence of foliar injury. This
indicates that drought does provide protection from foliar injury as discussed in the Os ISA (U.S.
EPA, 2013), but not entirely.
7-17
-------
Biosites with Foliar Injury
in
o
o
ci
in
.2
m
c
o
'
Q. *-
O O
Q-
in
q -
ci
o
q -
ci
I
10
I
20
W126(ppm-hrs)
I
30
I
40
Figure 7-10 Cumulative Proportion of Sites with Foliar Injury Present, by Moisture
Category
In Figure 7-11, we provide the data separated by NOAA climate regions (Karl and Koss,
1984). Although we had data for most regions of the contiguous U.S., we did not have data for
the Southwest and limited data for the West and West North Central regions. For example, from
2006 to 2010, there were over 1,000 biosite index values each for the Northeast and Central
regions and no biosite index values for the Southwest. When viewed by region, the pattern
observed nationally is not as clear. This is possibly due to the relationship between Os and
moisture, which can vary between regions.
7-18
-------
Biosites with Foliar Injury
in
d
in
a)
'S m-
o o
in
c
o
'•e
o
Q.
O
q
ci
All
Central
East North Central
Northeast
Northwest
South
Southeast
West
West North Central
o o o o
I
10
I
20
W126(ppm-hrs)
30
I
40
Figure 7-11 Cumulative Proportion of Sites with Foliar Injury Present, by Climate
Region
7.3 SCREENING-LEVEL ASSESSMENT OF VISIBLE FOLIAR INJURY IN 214
NATIONAL PARKS
In order to assess the potential for foliar injury risk in national parks, we considered the
approach in Kohut (2007). This study assessed the risk of Cb-induced visible foliar injury on Os
bioindicators (i.e., Os-sensitive vegetation) in 244 parks managed by the NFS. Specifically,
Kohut (2007) estimated Cb exposure using hourly Os monitoring data collected at 35 parks from
1995 to 1999, estimated Os exposure at 209 additional parks using kriging (a spatial interpolation
technique), and qualitatively assessed risk. Kohut applied a subjective evaluation based on three
criteria: (1) the frequency of exceedance of foliar injury "thresholds"4 using several Os exposure
metrics (i.e., SUM06, W126 and N100), (2) the extent that low soil moisture constrains Os
4 Kohut (2007) uses the term "foliar injury thresholds". In this assessment, we use the term "benchmarks" in order to
avoid implying that foliar injury could not occur below these levels.
7-19
-------
uptake during periods of high exposure, and (3) the presence of Os sensitive species within each
park. Based on these criteria, Kohut (2007) concluded that the risk of visible foliar injury was
high in 65 parks (27 percent), moderate in 46 parks (19 percent), and low in 131 parks (54
percent).5
In this assessment, we applied a modified screening-level approach using more recent Os
exposure and soil moisture data for 214 parks in the contiguous U.S.6 Consistent with advice
from CAS AC (Frey and Samet, 2012a), we modified the approach used by Kohut (2007) to
apply the W126 metric alone, and in doing so we chose foliar injury benchmarks derived from
the FHM analysis described in section 7.2 that assesses soil moisture quantitatively.7
7.3.1 Screening-Level Assessment Methods
7.3.1.1 O3 Exposure
As described in Section 4.3.3, we used recent Os monitoring data (2006-2010) to create
spatial surfaces of Os exposure using the Voronoi Neighbor Averaging (VNA) interpolation
method, which covers the contiguous U.S. with a spatial resolution of 12 km by 12 km for each
of the five years. This method allowed us to assess parks in the contiguous U.S., including parks
without Os monitors located within their park boundaries. We provide the W126 index values
estimated for each park by year in Appendix 7A.
7.3.1.2 Soil Moisture
As described in section 9.4.2 of the Os ISA (U.S. EPA, 2013), soil moisture is a major
modifying factor for Os-induced visible foliar injury. Low soil moisture generally decreases
stomatal conductance of plants and, therefore, limits the amount of Os entering the leaf that can
5 Kohut (2007) assigned a risk rating of "high" to parks likely to experience foliar injury in most years (e.g., in at
least three of the five years evaluated), a rating of "moderate" to parks likely to experience injury at some point
(e.g., in one or two of the five years evaluated), and a rating of "low" to parks not likely to experience injury (e.g.,
no years of the five years evaluated).
6 We did not include all of the 244 parks managed by NFS that were assessed in Kohut. Most of the excluded parks
are outside of the contiguous U.S., and a few others were not identified in the shapefile of park boundaries. The
parks assessed here include lands managed by the NFS in the continental U.S., which includes National Parks,
Monuments, Seashores, Scenic Rivers, Historic Parks, Battlefields, Reservations, Recreation Areas, Memorials,
Parkways, Military Parks, Preserves, and Scenic Trails.
7 We applied different foliar injury benchmarks in this assessment after further investigation into the benchmarks
applied in Kohut (2007), which were derived from biomass loss rather than visible foliar injury. Kohut cited a
threshold of 5.9 ppm-hrs for highly sensitive species from Lefohn (1997), which was based on the lowest W126
estimate corresponding to a 10% growth loss for black cherry. For soil moisture, Kohut (2007) qualitatively
assessed whether there appeared to be an inverse relationship between soil moisture and high Os exposure.
7-20
-------
cause injury. Dry periods tend to decrease the incidence and severity of foliar injury. However,
injury could still occur because plants must open their stomata even during these dry periods. We
are unaware of a clear threshold for soil moisture below which visible foliar injury would not
occur. To incorporate short-term soil moisture into the screening-level assessment, we applied
Palmer Z data for 2006 to 2010 (NCDC, 2012b). Consistent with the FHM analysis in Section
7.2, we categorized soil moisture as wet, normal, and dry (NOAA, 2012c). These data are for the
contiguous U.S. only.
Short-term estimates of soil moisture are highly variable over time, even from month to
month within a single year. For this reason, we used an average estimate of soil moisture to
reflect the cumulative nature of foliar injury in each park in each year. To determine the
appropriate timeframe for the soil moisture average, we identified the months corresponding to
the highest W126 index value estimated for each park with an Os monitor. The highest 3-month
W126 index value for 98 percent of monitored parks occurred between March and September
across all years, which roughly corresponds to the growing season (see Figure 7A-8 in the
appendix). Only 70 percent of monitored parks had the highest W126 between April and August.
Based on this information, we applied the 7-month soil moisture average from March to
September for each year in the core screening-level assessment for all parks. For parks with Os
monitors, we also conducted sensitivity analyses applying the 5-month soil moisture average
from April to August and the 3-month soil moisture average corresponding to the specific 3-
months with the highest W126 estimate at that monitor (see results in section 7.3.3.2 and
underlying data in Table 7A-1 in the appendix). We also evaluated the variability in soil moisture
averages across the 7-month, 5-month, and 3-month average timeframes by year for each
monitored park (see Figures 7A-9 through 7A-11 in the appendix).
7.3.1.3 GIS Analysis
Using GIS (ESRI® ArcMAP™ 10), we spatially overlaid the interpolated Cb exposure
surfaces and soil moisture data (NCDC, 2012b) with the NFS boundaries (USGS, 2003) to link
these data to each park. First, we dissolved all of the internal boundaries for each park such that
each park only had one park boundary. Next, we spatially joined the soil moisture data and the
gridded W126 data with the park boundaries, creating an average soil moisture estimate and
W126 index value estimated for each park. To identify the parks with Os monitors, we spatially
7-21
-------
overlaid the Os monitor data with the NFS park boundaries and included only those monitors
located within the park boundaries.8 We excluded all parks outside of the contiguous U.S.
because of the absence of soil moisture data, resulting in 42 parks with Os monitors and 214
parks with Os exposure estimated from the interpolated W126 surface.9 In Figure 7-12 we
provide the distribution of Os exposure and average soil moisture estimates for the 214 parks for
each year in this assessment, noting the range of "near normal" soil moisture conditions as
defined by NCDC (NOAA, 2012c).
8 There are 57 Os monitors located within NFS parks, and an additional 7 monitors are located within 1km of the
park boundaries. Some monitors (e.g., at Rocky Mountain National Park) have addresses that imply locations
within park boundaries but are actually located just outside the NFS boundary. We did not include the monitors
located just outside of the parks in the monitored park assessment. In addition, nine parks contained more than one
Os monitor. We provide the Os exposure and soil moisture data for the 57 monitors located within NFS parks in
Appendix 7A.
9 Along coastlines, the shapefile for soil moisture is more generalized than the shapefile for O3 exposure. Therefore,
we manually linked the soil moisture data to (a) 8 seashore parks in order to include them in the 214 park
assessment and (b) 4 park monitors for the 42 park assessment.
7-22
-------
4.00 i
Wet
3.00 -
f 2-00
| 1.00
oj
a
5 o.oo
^J -1.00
01
E
"3 -2.00
2006
-4.00
Dry
10 20 30
W126
4.00 -\
Wet
3.00 -
~ 2.00
1.00 -
0.00 -
-1.00 -
-2.00 -
-3.00 -
2007
-4.00
Dry 0
10 20 30
W126
4.00 -\
Wet
3.00 -
i i.oo
<
&
E 0.00
1
17-1.00
I -2.00
a.
-3.00 -
2008
2009
-4.00
Dry 0
20 30
W126
2010
2006-2010
Wet
3.00 -
2.00 -
g 1.00
&
5 0.00
N -1.00
oj
I -2.00
-4.00
Dry 0
4.00
Wet
3.00
| 2.00
i i.oo
I
I 0.00
N -1.00
OJ
E
"3 -2.00
-4.00
Dry
•-. "
0 10 20 30
W126
• 2006
•2007
2008
• 2009
•2010
(Shaded area represents "near normal" soil moisture (-1.25 > Palmer Z > 1)
Figure 7-12 Distribution of Os and Soil Moisture in 214 Parks by Year
7-23
-------
7.3.1.4 Sensitive Vegetation Species
NFS (2003) defines a sensitive species as "species that typically exhibit foliar injury at or
near ambient ozone concentrations in fumigation chambers and/or are species for which ozone
foliar injury symptoms in the field have been documented by more than one expert observer."
According to NFS (2003), the lists of sensitive species is limited in number of species because
few species from natural ecosystems have been fumigated in chambers or examined in the field
for Os symptoms.
We identified the parks containing Os-sensitive vegetation species (NFS, 2003, 2006b)
and considered the results for parks without species as potential until species are identified in
future field surveys at these parks. In addition, we conducted a sensitivity analysis where parks
without sensitive species are assumed to not exceed the benchmark criteria (see results in section
7.3.3.2). Based on the NFS lists, 95 percent of the parks in this assessment contain at least one
sensitive species. (See Figure 7A-7 and Table 7A-3 in the appendix for the parks with and
without currently identified sensitive species.)
7.3.1.5 W126 Benchmarks for Visible Foliar Injury
For each park, we evaluated whether Os exposure exceeded certain foliar injury
benchmark criteria in each year between 2006 and 2010. Specifically, we derived W126
benchmarks for five scenarios from the national-scale foliar injury analysis using FHM data
described in section 7.2. These benchmarks do not indicate thresholds below which no foliar
injury would be expected to occur. Rather, these benchmarks provide an indication of the risk of
foliar injury based on analysis of the FFDVI data.
All scenarios assessed in the screening-level assessment reflect the special status of parks
as areas designated for protection, and thus apply benchmarks corresponding to the presence of
any visible foliar injury. The "base scenario" represents the W126 index value where the slope of
exposure-response relationship changes for FHM biosites. As shown in Figure 7-10, the
percentage of biosites showing injury levels off at approximately 17.7 percent when considering
all biosites in all soil moisture categories, and we used this point to derive the W126 benchmark
(10.46 ppm-hrs) for the base scenario. At W126 index values above this benchmark, the
percentage of FFDVI biosites showing foliar injury remains relatively constant. The other four
scenarios explicitly consider soil moisture categorization, and these benchmarks represent the
7-24
-------
W126 index values corresponding to different percentages of FHM biosites with injury present
(i.e., 5 percent, 10 percent, 15 percent, and 20 percent) when the data is segregated by soil
moisture category. In total, we evaluated ten different W126 benchmarks associated with the
five scenarios.10
Table 7-6 provides the W126 benchmarks and the soil moisture categories for each of the
five scenarios. In the appendix, we provide the figures showing the derivation of the W126
benchmarks for each scenario (see Figures 7A-1 through 7A-5 in the appendix).
Table 7-6 W126 Benchmarks by Relative Soil Moisture Category in Five Scenarios
Scenario
Base
5% of
biosites
10% of
biosites
15% of
biosites
20% of
biosites
Description
17.7% of all FHM biosites showed any
injury (higher W126 index values have a
relatively constant percentage of FHM
biosites showing injury)
5% of FHM biosites showed any injury,
reflects soil moisture categorization
10% of FHM biosites showed any injury,
reflects soil moisture categorization
15% of FHM biosites showed any injury,
reflects soil moisture categorization
20% of FHM biosites showed any injury,
reflects soil moisture categorization
W126 Benchmark (in ppm-hrs)
Wet
(Palmer Z>1)
Normal Moisture
(Palmer Z between
-1.25 and 1)
Dry
(Palmer Z < -
1.25)
10.46
(soil moisture not considered)
3.76
4.42
4.69
5.65
3.05
5.94
8.18
N/A
6.16
24.61
N/A
N/A
N/A = Not available. We were unable to derive W126 benchmarks because of the limited number of biosites
showing injury in these categories.
7.3.2 Screening-Level Assessment Results and Discussion
To assess the potential for foliar injury risk in each parks we evaluated the frequency that
Os exposure exceeded certain W126 benchmarks and the average soil moisture conditions in
10 For some scenarios, we were unable to derive W126 benchmarks for all soil moisture categories because of the
limited number of biosites showing injury in those categories. For example, fewer than 15 percent of FHM biosites
categorized as "dry" showed any injury (see Figure 7-10). Therefore, we do not have a W126 benchmark for "dry"
for the 15 percent scenario.
7-25
-------
each year from 2006 to 2010. As shown in Table 7-7, in this assessment of 214 parks based on
the interpolated W126 surface, 11 percent of the parks exceeded the W126 benchmark in the
base scenario (10.46 ppm-hrs) for all five years evaluated, 39 percent for at least four years, 58
percent for at least three years, 70 percent for at least two years, and 83 percent for at least one
year. Table 7-7 also shows the results for each of the four scenarios that reflect soil moisture. In
general, scenarios for higher percentages of FHM biosites showing foliar injury have fewer parks
that exceed the benchmarks for those scenarios across multiple years. For example, nearly all
parks exceeded the W126 benchmarks for at least three years in the 5 percent scenario, but only a
few parks exceed the benchmarks for the 20 percent scenario.
As shown in Table 7-8, the number of parks exceeding the benchmarks in any given year
varies by scenario because Os exposure and average soil moisture vary by year. For example,
fewer parks exceeded the benchmarks in 2009 (a comparably dry, low Os year) than other years.
Figure 7-13 shows the national map of the results to highlight the geographic differences
for the base scenario. We also provide the detailed results for each park, including additional
figures to highlight the geographical differences in the other scenarios (see Table 7A-3 and
Figures 7A-12 through 7A-23 in the appendix).
Table 7-7
(Cumulative)
Parks Exceeding W126 Benchmarks in Five Scenarios from 2006 to 2010
Scenario
Base
5% of biosites
10% of biosites
15% of biosites
20% of biosites
Cumulative Number of Parks that Exceed Benchmarks (% of 214 parks)
All 5
years
23(11%)
195 (91%)
58 (27%)
23(11%)
0 (0%)
At least 4
years
84 (39%)
27 (13%)
127 (59%)
98 (46%)
0 (0%)
At least 3
years
124 (58%)
209 (98%)
172 (80%)
145 (68%)
4 (2%)
At least 2
years
149 (70%)
209 (98%)
193 (90%)
175 (82%)
20 (9%)
At least 1
year
177 (83%)
210 (98%)
204 (95%)
192 (90%)
72 (34%)
No years
37 (17%)
4 (2%)
10 (5%)
22 (10%)
142 (66%)
7-26
-------
Table 7-8 Parks Exceeding W126 Benchmarks in Five Scenarios in Individual Years
from 2006 to 2010
Scenario
Base
5% of biosites
10%ofbiosites
15% of biosites
20% of biosites
Number of Parks that Exceed Benchmarks in Each Year (% of 214 parks)
2006
171 (80%)
139 (65%)
207 (97%)
173 (81%)
164 (77%)
2007
147 (69%)
66(31%)
205 (96%)
119(56%)
103 (48%)
2008
125 (58%)
93 (43%)
203 (95%)
177 (83%)
155 (72%)
2009
26 (12%)
15 (7%)
206 (96%)
114(53%)
71 (33%)
2010
88 (41%)
63 (29%)
206 (96%)
171 (80%)
140 (65%)
Key: All 5 years 4 years 5 years 2 years 1 year No years
Figure 7-13 Foliar Injury Results Maps for the Base Scenario in 214 Parks
(Parks identified by park code. Not all park labels shown due to overlap. National Parks are prioritized in mapping.
Maps for additional scenarios and park code explanations available in Appendix 7A.)
7-27
-------
In the assessment of 42 parks with Ch monitors based on the interpolated surface, 24
percent of parks exceeded the W126 benchmark for the base scenario for all five years, 36
percent for at least four years, 57 percent for at least three years, 69 percent for at least two years,
and 81 percent for at least one year. These results are generally similar to the results for the 214
park assessment for the base scenario, except that the monitored park analysis showed a higher
fraction of parks that exceeded the benchmark criteria for all five years rather than at least four
years. This result may be because parks with consistently higher Os concentrations may be more
likely to have an Os monitor. Table 7-9 provides the results of the monitored park assessment.
We also evaluated three different methods for assigning Os exposure to parks with monitors:
interpolated surface, highest monitor, and average monitor. The results of this sensitivity analysis
are discussed in more detail in section 7.3.3.1.
7-28
-------
Table 7-9 Screening-level
Methods for Assigning Os Ex
Park Name
Acadia National Park
Agate Fossil Beds National
Monument
Badlands National Park
Big Bend National Park
Blue Ridge Parkway
Canyonlands National Park
Cape Cod National Seashore
Carlsbad Caverns National Park
City of Rocks National Reserve
Colorado National Monument
Congaree National Park
Cowpens National Battlefield
Craters of the Moon National
Monument
Cumberland Gap National Historical
Park
Death Valley National Park
Devil's Tower National Monument
Dinosaur National Monument
Glacier National Park
Grand Canyon National Park
Great Basin National Park
Great Smoky Mountains National
Park
Indiana Dunes National Lakeshore
Joshua Tree National Park
Lassen Volcanic National Park
Foliar Injury Results in 42 Parks with an Os Monitor using 3
posure to Each Park in Base Scenario
State
ME
NE
SD
TX
NC
UT
MA
NC
ID
CO
sc
sc
ID
KY
CA
WY
CO
MT
AZ
NV
TN
IN
CA
CA
Years with
Monitoring
Data
(# years)
5
3
5
5
5
5
5
4
1
4
5
5
4
4
5
o
6
4
5
5
5
5
5
5
5
Years Exceeding W126 Benchmarks for Base
Scenario (# years)
Interpolated
Surface
0
2
1
1
3
5
o
5
i
4
o
J
3
2
o
J
3
5
2
4
0
5
5
o
J
1
5
4
Highest
Monitor
1
N/A
1
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
0
N/A
N/A
4
N/A
5
N/A
Average
Monitor
0
N/A
0
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
0
N/A
N/A
3
N/A
5
N/A
Onlyl
Monitor
in Park
N/A
1
N/A
o
3
i
5
3
2
0
2
2
2
1
1
5
0
2
N/A
4
4
N/A
1
N/A
o
3
7-29
-------
Park Name
Mesa Verde National Park
Mojave National Preserve
Mount Rainier Wilderness
Olympic National Park
Padre Island National Seashore
Petrified Forest National Park
Pinnacles National Monument
Saguaro National Park
Saratoga National Historical Park
Scotts Bluff National Monument
Sequoia-Kings Canyon National Park
Shenandoah National Park
Theodore Roosevelt National Park
Tonto National Monument
Voyageurs National Park
Wind Cave National Park
Yellowstone National Park
Yosemite National Park
Summary Results by
Os Exposure
Method*
State
CO
CA
WA
WA
TX
AZ
CA
AZ
NY
NE
CA
VA
ND
AZ
MN
SD
WY
CA
All 5 years
At least 4 years
At least 3 years
At least 2 years
At least 1 year
No years
Years with
Monitoring
Data
(# years)
5
4
5
1
2
5
5
5
5
1
5
5
5
5
5
5
5
5
71%
86%
93%
95%
100%
0%
Years Exceeding W126 Benchmarks for Base
Scenario (# years)
Interpolated
Surface
5
5
0
0
0
4
3
4
0
3
5
2
0
5
0
2
1
5
24%
36%
57%
69%
81%
19%
Highest
Monitor
N/A
N/A
N/A
0
N/A
N/A
N/A
N/A
N/A
N/A
5
N/A
0
N/A
N/A
N/A
N/A
5
19%
36%
43%
60%
76%
24%
Average
Monitor
N/A
N/A
N/A
0
N/A
N/A
N/A
N/A
N/A
N/A
5
N/A
0
N/A
N/A
N/A
N/A
5
19%
33%
43%
60%
71%
29%
Onlyl
Monitor
in Park
5
4
0
N/A
0
4
4
5
0
0
N/A
4
N/A
5
0
2
2
N/A
N/A
* Summary results assume that parks with only one monitor exceeded the W126 benchmarks the same number of
years using either the highest or average monitor method.
N/A = Not applicable.
7-30
-------
7.3.3 Sensitivity Analyses for Screening-Level Assessment
7.3.3.1 O3 Exposure
Monitoring provides the most accurate assessment of Os exposure in specific locations,
but a single monitor may not reflect the differences in exposure throughout a park. For this
reason, we compared the results of the assessment for parks with Os monitors located within the
park boundaries using the interpolated surface with the results based on Os monitor data. As
shown in Table 7-9, the results using the highest monitor and average monitor were generally
similar to each other and to the results using the interpolated surface. For the 30 parks with all
five years of monitoring data, 17 parks had the same results using all three methods, five parks
had more years exceeding the benchmark for the base scenario using the interpolation, five parks
had more years exceeding that benchmark using either monitor method, and three parks had
more years exceeding using the highest monitor.
It can be informative to apply alternative screening criteria based on Os exposure alone.
For this sensitivity analysis, we identified the parks that exceeded W126 index values that were
consistent with the range of alternative standard levels considered in the Policy Assessment.
Table 7-10 shows that 23 percent of parks exceeded 15 ppm-hrs for at least three years from
2006 to 2010, while 80 percent of parks exceeded 7 ppm-hrs for at least three years.
Because W126 index values can be highly variable from year to year, evaluation of
different years could lead to different results. In Table 7-10, we provide the sensitivity of the
results for the base scenario by splitting the data into two timeframes. In general, more parks had
higher Os exposure during the first three years of the assessed timeframes (i.e., 2006-2008) than
the last three years (i.e., 2008-2010).
7-31
-------
Table 7-10 Foliar Injury Sensitivity Analyses for 214 Parks
Alternative Screening Criteria
03
Exposure
only
Timeframe
Sensitive
Species
W126>15
W126>13
W126>11
W126>9
W126>7
Base scenario using
2006-2008 only
Base scenario using
2008-20 10 only
Base scenario
assuming park does
not exceed if no
sensitive species in
park
Number of Parks Exceeding W126 Benchmark in 2006-2010 (% of 214
parks)
All 5
years
6 (3%)
7 (3%)
17 (8%)
46 (21%)
88 (41%)
N/A
N/A
22 (10%)
At least 4
years
20 (9%)
44 (21%)
71 (33%)
117
(55%)
156
(73%)
N/A
N/A
80 (37%)
At least 3
years
49 (23%)
76 (36%)
111
(52%)
150
(70%)
171
(80%)
117
(55%)
23(11%)
116
(54%)
At least 2
years
83 (39%)
110
(51%)
142
(66%)
165
(77%)
183
(86%)
149
(70%)
86 (40%)
141
(66%)
At least 1
year
125
(58%)
155
(72%)
174
(81%)
186
(87%)
196
(92%)
177
(83%)
130
(61%)
168
(79%)
No years
89 (42%)
59 (28%)
40 (19%)
28 (13%)
18 (8%)
37 (17%)
84 (39%)
46 (21%)
N/A = Not applicable
7.3.3.2
Soil Moisture
Evaluating soil moisture is more subjective than evaluating Os exposure because of its
high spatial and temporal variability within the Os season. Although we are unable to quantify
the within-region variability in soil moisture for the relatively large NCDC climate regions, we
can evaluate the sensitivity of the results to different averaging times for soil moisture data.
Specifically, we compared the results using the 7-month average from the main analysis with
alternative 5-month and 3-month soil moisture averages at Os monitors in parks. As shown in
Table 7-11, the results for the 57 Os monitors in parks are not very sensitive to the different
timeframes for soil-moisture data for the five scenarios. On balance, we believe that the impact
of the variability in the spatial resolution of the data likely exceeds the impact of the temporal
resolution of the data, and thus this assessment is likely to underestimate the potential of foliar
injury that could occur in some localized areas such as stream banks.
7-32
-------
Table 7-11 Soil Moisture Sensitivity Analyses in 57 Oa Monitors in Parks*
Scenario and Soil Moisture
Timeframe
7-month
Palmer Z
(Mar-Sept)
5%ofbiosites
10% of biosites
15% of biosites
20% of biosites
Parks Exceeding Benchmark Criteria in 2006-2010 (% of 57 Parks)
All 5 years
31(54%)
8 (14%)
2 (4%)
0 (0%)
At least 4
years
42 (74%)
24 (42%)
15 (26%)
1 (2%)
At least 3
years
45 (79%)
34 (60%)
26 (46%)
1 (2%)
At least 2
years
49 (86%)
43 (75%)
39 (68%)
7 (12%)
At least 1
year
55 (96%)
49 (86%)
46 (81%)
18 (32%)
No years
2 (4%)
8 (14%)
11(19%)
39 (68%)
Change in Parks Exceeding Benchmark Criteria for Alternative Soil Moisture Timeframes (% of 57 Parks)
5 -month
Palmer Z
(Apr-Aug)
Specific 3-
Month
Palmer Z
(based on
monitor)
5% of biosites
10% of biosites
15% of biosites
20% of biosites
5% of biosites
10% of biosites
15% of biosites
20% of biosites
-1 (-2%)
-1 (-2%)
NC
NC
+2 (+4%)
-2 (-4%)
-1 (-2%)
NC
-1 (-2%)
-5 (-9%)
-5 (-9%)
-1 (-2%)
NC
-5 (-9%)
-5 (-9%)
NC
NC
-1 (-2%)
-1 (-2%)
NC
NC
-2 (-4%)
-4 (-7%)
-1 (-2%)
+1 (+2%)
NC
-1 (-2%)
-3 (-5%)
NC
-1 (-2%)
-1 (-2%)
NC
NC
-1 (-2%)
-1 (-2%)
+5 (+9%)
NC
+1 (+2%)
+1 (+2%)
-5 (-9%)
NC
+1 (+2%)
+1 (+2%)
-5 (-9%)
NC
-1 (-2%)
-1 (-2%)
+5 (+9%)
* Includes multiple monitors in 9 parks. The base scenario is not included in this table because this scenario does not
include screening criteria for soil moisture.
NC=No change.
7.3.3.3 Evaluation of Existing Standard and Alternative W126
Standards
This screening-level assessment does not evaluate the model-adjusted W126 spatial
surfaces for the scenarios of just meeting the existing 75 ppb (4th highest daily maximum)
standard or alternative W126 standards. Because this screening-level assessment relies on year-
by-year estimates of Os exposure and soil moisture, it would not be possible to evaluate these
year-by-year impacts using the W126 surfaces derived from three years of model-adjusted W126
data. Nevertheless, we can make a few observations regarding the potential implications of just
meeting the existing and alternative standards. For example, as shown in Table 7-10, 42 percent
of parks did not exceed 15 ppm-hrs during 2006-2010 using annual W126 data. In addition, none
of the 214 parks would exceed the annual benchmark criteria for the base scenario (W126>10.46
ppm-hrs) after adjustments to just meet the existing standard (adjustments based on 3-year
average W126 data). Similarly, only eight parks would exceed 7 ppm-hrs using the 3-year
7-33
-------
average model-adjusted surfaces that just meet the existing standard. We provide the W126
index values for each of the 214 parks after just meeting the existing standard and alternative
W126 standards in Appendix 7A.
7.4 NATIONAL PARK CASE STUDY AREAS
The national parks represent a set of resources the public has agreed are special areas in
need of protection for this and future generations to experience and enjoy.11 Because of this
status risks to park resources are of special concern, particularly for bequest and option services
because these services are specifically referenced in the creation of the parks. The NFS is
responsible for the protection of all resources within the national park system. These resources
include those that are related to and/or dependent upon good air quality, such as whole
ecosystems and ecosystem components.
Several laws and policies protect the natural resources in national parks. The NFS, in its
Organic Act (16 U.S.C. 1), is directed to conserve the scenery, natural and historic objects and
wildlife and to provide for the enjoyment of these resources unimpaired for current and future
generations. The Wilderness Act of 1964 (Public Law 88-577, 16 U.S. C. 1131-1136) asserts
wilderness areas will be administered in such a manner as to leave them unimpaired and preserve
them for the enjoyment of future generations. NFS Management Policies (2006) guide all NFS
actions including natural resources management. In general, the NFS Management Policies
reiterate the NFS Organic Act's mandate to manage the resources "unimpaired." Although we
have not quantified the monetary value of the bequest or option services given the data and
methodology limitations inherent in such an effort, the status afforded these special areas through
these laws and policies is indicative of their value to the public.
The ecosystem service we can quantify, with some qualifications, is the recent monetary
value of the total recreation opportunity provided by the parks. We cannot quantify the loss in
monetary value for these services associated with Os; however, the magnitude of the overall
value is informative in understanding the potential significance of any Os damage (see Chapter 5
for more discussion). The NFS has collected data on visitation, recreational activities, and
11 C.F.R. 40, 81.400 provides for visibility protection for federal Class I areas.
7-34
-------
expenditures for trips to parks and modeled the economic impacts to local communities around
parks. The NSRE provides WTP estimates for the value of recreation activities specific to the
regions where parks are located. Together these data allow us to estimate the magnitude of the
recreation services provided by parks. The loss of service provision or visitor satisfaction due to
Os injury to sensitive species in the case study parks is reflected in these estimates.
The three parks we are highlighting for case study analysis, Great Smoky Mountains NP,
Rocky Mountain NP, and Sequoia/Kings Canyon NP, represent different regions of the country,
different ecosystems, and Os conditions. Each park contains species sensitive to Os injury. The
text boxes accompanying each section highlight some of the reasons these parks were chosen for
special protection.
For the case study areas, we used the Os-sensitive species list from the preceding section
and cover data from VegBank plots (see Section 7.2). The resulting maps give cover estimates
for Os-sensitive species at the finer scale of the NFS vegetation map. It is important to note that
the cover estimates are separated into vegetation stratum (e.g., herb, shrub, tree) and it is possible
to have more than one vegetation strata present in a location. As such, it is possible to have
sensitive species cover at a higher cumulative proportion than is shown here. We also used the
benchmarks presented in section 7.2 to assess the effect of just meeting the existing and
alternative standards on W126 index values in the case study parks. We used a benchmark of 10
percent of biosites exhibiting foliar injury in a normal year as the basis for the analysis, which is
depicted in Figure 7-14.
7-35
-------
Biosite Index > 0
O
c
g
t
o
Q.
o
c\i
o
o
r\|
d
LO
ci
o
ci
LO
P -
ci
o
o _
10
\
20
I
30
40
W126(ppm-hrs)
Figure 7-14 Identification of W126 Index Value where 10 Percent of Biosites Show Any
Foliar Injury
7-36
-------
7.4.1 Great Smoky Mountains
National Park
In 2010, the Great Smoky Mountains National
Park (GRSM) welcomed approximately 9.5 million
visitors (NFS, 2010) making it the most visited national
park in America.
The "whole park" services affected by potential
Os impacts include the existence, option, and bequest
values and habitat provision discussed in Chapter 5.
Recreation value specific to the park is discussed later in
this section.
The extent of sensitive species coverage in
GRSM is substantial. Showing the percent cover of
species sensitive to foliar injury and focusing the analysis
on areas where recreation services are provided can
provide some perspective on the potential level of harm
to scenic beauty and recreation satisfaction within the
Park.
The NFS 2002 Comprehensive Survey of the
American Public, Southeast Region Technical Report
includes responses from recent visitors to southeast parks
about the activities they pursued during their visits (NFS,
2002a). Using the 2010 annual visitation rate from the
NFS survey (NFS, 2010) and the regional results from
the Kaval and Loomis (2003) report on recreational use
values compiled for the NFS, we estimated visitors'
WTP for various activities; we present the estimates in
Table 7-12. In addition to the activities listed in the
table, 19 percent, or 1.8 million park visitors, benefited
from educational services offered at the park by
Mount Le Conte, Summer
Great Smoky Mountains National Park
Courtesy: NPS
http://www.nps.gov/grsm/photosmultimedia/index.htm
Great Smoky Mountains National Park is
the most visited national park in America
and a UNESCO World Heritage Site. The
Park is valued for the diversity of its
vegetation and wildlife; the scenic beauty
of its mountains, including the famous fogs
that give the Smoky Mountains their name;
and the preservation of the remnants of
Southern Appalachian culture. It is also
subject to high ambient Os levels. The
park has recent W126 index values of 10 -
18 ppm-hrs with a mean of 14.7 ppm-hrs.
7-37
-------
participating in a ranger-led nature tour, which suggests that visitors wish to understand the
ecosystems preserved in the park.
Table 7-12 Value of Most Frequent Visitor Activities at Great Smoky Mountains
National Park
Activity
Sightseeing
Day Hiking
Camping
Picnicking
Total
Percent
Participation
82
40
19
50
Number of
Participants
(thousands)
7,790
3,800
1,805
4,750
Mean WTP
(in 2010$)
53.34
69.93
29.87
42.42
Total Value of
Participation
(millions of 2010$)
416
266
54
201
937
The report Economic Benefits to Local Communities from National Park Visitation and
Payroll (NFS, 2011) provides estimates of visitor spending and economic impacts for each park
in the system. Visitor spending and its economic impact to the surrounding area are provided in
Table 7-13 for the GRSM. In addition, Table 7-14 includes data on the median value that
visitors spend on food, gas, lodging, and other items.
Table 7-13 Visitor Spending and Local Area Economic Impact of Great Smoky
Mountains National Park
Public Use Data
2010 Recreation
Visits
9,463,538
2010 Overnight
Stays
393,812
Visitor Spending 2010a
All
Visitors
$818,195
Non-Local
Visitors
$792,547
Impacts on Non-Local Visitor
Spending
Jobs
11,367
Labor
Income"
$303,510
Economic
Impact"
$504,948
a ($OOOs)
Source: NFS (2011)
7-38
-------
Table 7-14 Median Travel Cost for Great Smoky Mountains National Park Visitors
Expense/Visit
Gas and Transportation
Lodging
Food and Drinks
Clothes, Gifts, and Souvenirs
Total Per Visitor Party
Median Expenditures (2010$)
$73
$182
$73
$61
$389
Source: NFS (2002a)
Each of the activities discussed above is among those shown in the national-scale
analysis to be strongly affected by visitor perceptions of scenic beauty. As discussed in Section
7.1.1.2 for visible Os damage (Peterson, 1987) and for visible nitrogen and adelgid damage (a
pest in Fraser fir) (Haefele et al., 1991 and Holmes and Kramer, 1996) visitors have a non-zero
WTP for reductions in the described scenic impairments. As in the national analysis, it is not
possible to assess the extent of loss of services from impairment of scenic beauty by Os;
however, for the park these losses are captured in the estimated values for spending, economic
impact, and WTP.
GRSM is prized, in part, for its rich species diversity. The large mix of species includes
37 Os-sensitive species across vegetative strata, and many areas contain several sensitive species.
For instance, there may be a sensitive tall shrub occurring under the canopy of a sensitive tree
and various sensitive short shrubs or herbaceous plants occurring in the area of the tall shrub. In
areas where sensitive species overlap, it is possible to have sensitive species coverage
substantially higher than coverage for any one category of vegetation. Figure 7-15 shows the
park coverage of various sensitive species. Nearly 40 percent of the Park's 2,185 km2 total area
has sensitive tree cover (canopy and subcanopy) greater than 20 percent. Of that, 232 km2 has
sensitive tree species cover between 20 percent and 40 percent. Shrubs account for 491 km2 of
sensitive vegetation, with over 100 km2 having over 80 percent of the species present as
sensitive. While sensitive herbaceous species occur throughout the park, the percent cover rarely
exceeds 20 percent.
We can quantify the extent of the hiking trails in areas where sensitive species are at risk
for foliar injury. Of the approximately 1,287 km of trails in GRSM, including approximately
114 km of the Appalachian Trail, over 1,040 km, or about 81 percent of trail area, are in areas
7-39
-------
where species sensitive to foliar injury occur. Figure 7-16 shows a summary of the overlap of
the hiking trails in the GRSM, including a portion of the Appalachian Trail, with the species
cover index. The accompanying pie charts in Figure 7-17 show the number of trail kilometers in
each cover category. The categories likely most visible to hikers are subcanopy trees, shrubs,
and herbaceous vegetation. There are 311 km, or about 24 percent, of trail area where sensitive
subcanopy tree cover accounts for over 20 percent of the tree species present. Sensitive shrubs
cover over 20 percent of 549 km of trail area, or about 43 percent of total area.
Although we cannot quantify the incremental loss of hiker satisfaction with their
recreation experience, this analysis illustrates that very substantial numbers of trail kilometers
are potentially at risk. With 3.8 million hikers using the trails every year and those hikers willing
to pay over $266 million for that activity, even a small benefit of reducing Os damage in the park
could result in a significant value.
7-40
-------
Canopy
Tall Shrub
- - --
Herbaceous
Sub Canop
Short Shrubs
.
"
No data
0%
<20%
2D%-40%
40%-60%
60%-80%
>80%
Figure 7-15 Cover of Sensitive Species in Great Smoky Mountains National Park
7-41
-------
Canopy
Sub Canopy
Tall Shrubs
Herbaceous
No data
0%
<20%
20%-40%
40%-60%
60%-80%
>80%
Appalachian Trail
Figure 7-16 Percent of Sensitive Species Near Trails in Great Smoky Mountains National Park
7-42
-------
Tree Canopy
,-,,-, • No Data
177.4 135.2
H 0%
^^^^^k
• <20%
315-6 ' 20% to 40%
738 5 • 40% to 60%
• 60 % to 80%
• >80%
Tall Shrub
640135.2 • No Data
155.0
• B0%
• <20%
20% to 40%
1032'5 • 40% to 60%
• 60 % to 80%
• >80%
Tree Subcanopy
176.2 135.2 "No Data
• 0%
^^k
• <20%
iy '4 20% to 40%
948.9 • 40% to 60%
• 60 % to 80%
• >80%
Short Shrub
640
55.6 • 1
\ Il35.2 • No Data
155.0 M .112.1 mo%
<20%
20% to 40%
1,133.7 B40%to60%
• 60 % to 80%
• >80%
Herbaceous
55.6135.2 • No Data
• 0%
^253.9 1 <20%
20% to 40%
1,211.0 • 40% to 60%
• 60 % to 80%
• >80%
Figure 7-17 Trail Kilometers of Sensitive Species by Cover Category in Great Smoky
Mountains National Park
One of the amenities provided by GRSM is the scenic views from the roads and trails —
the views from the scenic overlooks are one of the major park attractions. On a day with natural
viewing conditions visitors can see about 150 km across the mountain ridges of North Carolina
and Tennessee, far outside the borders of the park itself. On average viewing days visitors can
still see about 40 km, again outside the park itself. Figure 7-18 shows the sensitive tree canopy
cover within a 3 km buffer of the overlooks. Within these small buffers 78 km2 have sensitive
species cover over 20 percent. While there are no data on the number of visitors stopping at the
overlooks, almost 8 million visitors identify sightseeing as one of their activities in the Park.
With their collective WTP for this activity over $400 million, it seems reasonable to conclude
that park visitors substantially value the scenic quality of the overlooks. Ozone concentrations in
GRSM have been among the highest in the eastern U.S., sometimes twice as high as neighboring
cities such as Atlanta and Knoxville. Under recent conditions 44 percent, or 959 km2, of the park
has W126 index values above 15 ppm-hrs. After just meeting the existing standard at 75 ppb,
W126 index values are reduced such that no area is over 7 ppm-hrs. Just meeting the alternative
7-43
-------
of 15 ppm-hrs produces the same result as meeting the existing standard. The lower alternative
standards of 11 and 7 ppm-hrs result in the park having W126 index values under 3 ppm-hrs for
the entire park, with most of the park under 2 ppm-hrs after just meeting the 7 ppm-hrs standard
level. See Table 7-15 for additional details.
Figure 7-18 Sensitive Vegetation Cover in Great Smoky Mountains National Park
Scenic Overlooks (3km)
Table 7-15 Geographic Area of Great Smoky Mountains National Park after Just
Meeting Existing and Alternative Standard Levels (km2)
Recent conditions
(2006-2008)
Just meeting 75 ppb
15 ppm-hrs
1 1 ppm-hrs
7 ppm-hrs
Under 5.94
ppm-hrs
0
2,185
2,185
2,185
2,185
Between 5.95
and 7 ppm-hrs
0
0
0
0
0
Between 7-
llppm-hrs
48
0
0
0
0
Between 11-15
ppm-hrs
1,178
0
0
0
0
Over 15
ppm-hrs
959
0
0
0
0
7-44
-------
7.4.2 Rocky Mountain National Park
In 2010 Rocky Mountain National Park (ROMO) welcomed
3 million visitors (NFS, 2010) to its 1,075 km2 of mountain
ecosystems. ROMO allows visitors to enjoy vegetation and
wildlife unique to these ecosystems along over 483 km of hiking
trails.
The NFS 2002 Comprehensive Survey of the American
Public, Intermountain Region Technical Report includes responses
from recent visitors to intermountain parks about the activities they
pursued during their visit (NFS, 2002b). As in the GRSM case
study, using the 2010 visitation rate from the NFS survey (NFS,
2010) and the regional results from the Kaval and Loomis (2003)
report on recreational use values compiled for the NFS, we present
estimates for visitors' WTP for various activities in Table 7-16.
Table 7-16 Value of Most Frequent Visitor Activities at Rocky
Mountain National Park
Activity
Sightseeing
Day Hiking
Camping
Picnicking
Total
Percent
Participation
85
51
27
38
Number of
Participants
(thousands)
2,550
1,520
810
1,140
Mean
WTP
(in
2010$)
$28.17
$46.03
$41.47
$33.77
Total Value of
Participation
(millions of
2010$)
$72
$70
$34
$38
$214
In addition to the activities listed in Table 7-16, 11 percent
of, or 330,000, park visitors took advantage of educational services
offered at the park by participating in a ranger-led nature tour.
Each of the activities discussed above are among those
shown in the national-scale analysis to be strongly affected by
visitor perceptions of scenic beauty. As in the national analysis it is
7-45
Sheep Lakes
Courtesy: NFS
http://www.nps.gov/romo/photosmulti
media/index.htm
Rocky Mountain National Park
features riparian ecosystems
with 150 lakes and 450 stream
miles that support lush
vegetation. The montane
ecosystem includes pine
forests and grasslands, while
subalpine elevations present
spruce and fir trees weathered
by the elements. The alpine
ecosystems are too harsh for
trees, but support low growing
plants. The park has recent
W126 index values of 2-54
ppm-hrs with a mean of 14.2
ppm-hrs.
|
-------
not possible to assess the extent of loss of services due to impairment of scenic beauty due to Os
damage; however those losses are captured in the estimated values for spending, economic
impact, and WTP for the park. If Os impacts were lower these estimated values would likely be
higher.
The report Economic Benefits to Local Communities from National Park Visitation and
Payroll (NFS, 2011) provides estimates of visitor spending and economic impacts for each park
in the system. Visitor spending and its economic impact to the surrounding area are given in
Table 7-17 for the ROMO. Table 7-18 includes data on the median value that visitors spend on
food, gas, lodging, and other items.
Table 7-17 Visitor Spending and Local Area Economic Impact of Rocky Mountain
National Park
Public Use Data
2010
Recreation
Visits
2,955,821
2010
Overnight
Stays
174,202
Visitor Spending 2010
All Visitors
229,032
Non-Local
Visitors
221,896
Impacts on Non-Local Visitor Spending
Jobs
3,316
Labor
Income"
$89,975
Economic
Impact"
$ 155,157
a($OOOs)
Source: NFS (2011)
Table 7-18 Median Travel Cost for Rocky Mountain National Park Visitors
Expense/Visit
Gas and Transportation
Lodging
Food and Drinks
Clothes, Gifts, and Souvenirs
Total per Visitor Party
Median Expenditures (in 2010$)
$63
$100
$63
$45
$271
Source: NFS (2002b)
Unlike GRSM, only 7 sensitive species provide cover in ROMO as depicted in Figure
7-19. The most notable of these is Quaking Aspen, or Populus tremuloides. This is significant
in that many of the visitors to ROMO visit specifically to see this tree in its fall foliage. In some
areas of the park, cover of this species can reach 80 percent. The species is found, along with the
other sensitive tree species silver wormwood and Scouler's willow, in all vegetative layers in the
park. Sensitive species cover in just the tree canopy, subcanopy, and tall shrub layers is over 40
percent in 328 km2, or 30 percent, of the park.
7-46
-------
We were able to quantify the extent of the hiking trails present in areas where sensitive
species are at risk for foliar injury. Of the approximately 562 km of trails in ROMO, including
approximately 87 km of the Continental Divide National Scenic Trail, over 242 km, or about 43
percent of trail area, are in areas where species sensitive to foliar injury in the canopy, subcanopy
or tall shrub category occur in greater than 20 percent coverage. Figure 7-20 maps the hiking
trails in ROMO, including the relevant portion of the Continental Divide National Scenic Trail
overlaid with the species cover index. The accompanying pie charts in Figure 7-21 show the
number of trail km in each cover category.
Again, although we are not able to quantify the impact of this scenic damage on hiker
satisfaction, given 1.5 million hikers in ROMO and their $70 million WTP for the hiking
experience, even a small improvement in the scenic value could be significant. While we did not
map the scenic overlooks in ROMO, given the 2.5 million visitors who come to the park to
sightsee and the $72 million they are willing to pay for this activity, it is reasonable to conclude
that any improvement in the scenic quality of the vistas at the overlooks would be of significant
value.
Under recent conditions, all 1,067 km2 of the park have W126 index values over 15 ppm-
hrs. Meeting the existing standard would bring about 59 percent of the Park into the 7-15 ppm-
hrs range, with the remaining 440 km2 under 7 ppm-hrs. Assessing an alternative standard of 15
ppm-hrs would bring the entire park under 7 ppm-hrs. See Table 7-19 for a summary of full
results.
7-47
-------
Table 7-19 Geographic Area of Rocky Mountain National Park after Just Meeting
Existing and Alternative Standard Levels (km2)
Recent conditions
(2006-2008)
Just meeting 75 ppb
15 ppm-hrs
1 1 ppm-hrs
7 ppm-hrs
Under 5.94
ppm-hrs
0
37
986
1,067
1,067
Between
5.95-7
ppm-hrs
0
403
81
0
0
Between 7-11
ppm-hrs
0
627
0
0
0
Between 11-15
ppm-hrs
0
0
0
0
0
Over 15
ppm-hrs
1,067
0
0
0
0
7-48
-------
Emergent Tree,-
\
Dwarf Shrub
Herbaceous
Sub Canopy
Short Shrub
Canopy
Tall Shrub
: '- ^
No Data
0%
20%-40%
40%-60%
60%-80%
>80%
Figure 7-19 Sensitive Species Cover in Rocky Mountain National Park
7-49
-------
Emergent Tree
Dwarf Shrub
Herbaceous
Sub Canopy
Canopy
Figure 7-20 Percent Cover of Sensitive Species Near Trails in Rocky Mountain National Park
7-50
-------
Trees Canopy Tree Subcanopy
11.7 19.1-.
•s: Y°
24-3->^ • No Data 4-°
• No Data
'°% 39'9| .0%
„, ,- B<20%
"1-b 106.4 B<20%
2°%t04°% ^ 20% to 40%
•4°%t06°% .40%to 60%
"6°%t08°% ^ .60%to80%
• >80%
• >80%
394n Tall Shrubs
Short Shrubs
<20%
60 % to 80%
No Data
10%
I <20%
20% to 40%
40% to 60%
60 % to 80%
I >80%
Herbaceous
<20%
20% to 40%
60 % to 80%
Emergent Trees
• <20%
20% to 40%
40% to 60%
60 % to 80%
• >80%
Dwarf Shrubs
• No Data
• 0%
• <20%
20% to 40%
40% to 60%
60 % to 80%
• >80%
Figure 7-21 Trail Kilometers of Sensitive Species by Cover Category in Rocky
Mountain National Park
7-51
-------
7.4.3 Sequoia and Kings Canyon
National Parks
Sequoia and Kings Canyon National Parks (SEKI)
are located in the southern Sierra Nevada Mountains east of
the San Joaquin Valley in California. The two parks
welcomed 1.6 million visitors in 2010 (NFS, 2010) to
experience the beauty and diversity of some of California's
iconic ecosystems.
The NFS 2002 Comprehensive Survey of the
American Public, Pacific West Region Technical Report
includes responses from recent visitors to western parks
about the activities they pursued during their visit (NFS,
2002c). By using the 2010 annual visitation rate from the
NFS survey and the regional results from the Kaval and
Loomis (2003) report on recreational use values compiled
for the NFS, we estimated visitors' WTP for various
activities; the results are presented in Table 7-20.
7-52
Kings Canyon
Courtesy: NFS,
http ://www. nps. gov/seki/photo smu
Itimedia/index. htm
The Sequoia and Kings
Canyon National Parks share a
boundary and natural
resources. The natural
resource features include the
giant sequoia trees (and other
species, including ponderosa
and Jeffrey pine). The varied
ecosystems from the top of
Mount Whitney to the marble
caverns provide habitat for a
rich diversity of species. The
park has recent W126 index
values of 34 - 53 ppm-hrs with
a mean of 43ppm-hrs.
-------
Table 7-20 Value of Most Frequent Visitor Activities at Sequoia and Kings Canyon
National Parks
Activity
Sightseeing
Day Hiking
Camping
Picnicking
Total
Percent Participation
81
58
33
45
Number of Participants (thousands)
1,300
928
528
720
Mean WTP
(in 2010$)
$24.21
$27.77
$124.65
$76.72
Total Value of
Participation
(millions of
2010$)
$31
$26
$66
$55
$178
In addition to the activities listed in Table 7-20, 14 percent of, or 224,000 park visitors
availed themselves of educational services offered at the park by participating in a ranger-led
nature tour, which suggests that visitors wish to understand the ecosystems preserved in the park.
Each of the activities discussed above is among the activities shown in the national-scale
analysis to be strongly affected by visitor perceptions of scenic beauty. As in the national
analysis, it is not possible to assess the extent of loss of services resulting from impairment of
scenic beauty due to Os damage; however, these losses are captured in the estimated values for
spending, economic impact, and WTP for the parks. If Os impacts were lower these estimated
values would likely be higher.
The report Economic Benefits to Local Communities from National Park Visitation and
Payroll (NFS, 2011) provides estimates of visitor spending and economic impacts for each park
in the system. Visitor spending and its economic impact to the surrounding area are provided in
Table 7-21 for SEKI. In addition, Table 7-22 includes data on the median value that visitors
spend on good, gas, lodging, and other items.
Table 7-21 Visitor Spending and Local Area Economic Impact of Sequoia and Kings
Canyon National Parks
Public Use Data
2010
Recreation
Visits
1,320,156
2010
Overnight
Stays
438,677
Visitor Spending 2010a
All Visitors
$97,012
Non-Local
Visitors
$89,408
Impacts on Non-Local Visitor Spending
Jobs
1,283
Labor
Income"
$37,299
Economic
Impact"
$60,504
a($OOOs)
Source: NFS (2011)
7-53
-------
Table 7-22 Median Travel Cost for Sequoia and Kings Canyon National Parks Visitors
Expense/Visit
Gas and Transportation
Lodging
Food and Drinks
Clothes, Gifts, and Souvenirs
Total per Visitor Party
Median Expenditures (in 2010$)
$75
$150
$98
$63
$386
Source: NFS (2002c)
There are 12 identified sensitive species in SEKI. The percent coverage of these species
is depicted in Figure 7-22. Areas of the parks with sensitive species cover of over 20 percent in
the canopy comprise 646 km2, or about 20 percent of the total area of SEKI. This area
encompasses about 285 km of the 1,287 km (22 percent) of hiking trails available to
approximately 928,000 hikers in the parks. Figure 7-23 depicts the sensitive species cover across
the trail system, including the portion of the John Muir Trail that crosses the Parks' 19 km,
which has sensitive species coverage over 20 percent. Figure 7-24 shows the sensitive species
by type.
Again, although we are not able to quantify the impact of this scenic damage on hiker
satisfaction for hikers in SEKI and their $26 million WTP for the experience, even a small
improvement in the scenic value could be significant.
As in the previous case studies, moving from recent conditions to meeting the existing Os
standard results in a large change in the area of the parks with exposures above 15 ppm-hrs. For
SEKI, this means the parks move from all areas experiencing exposures above 15 ppm-hrs to all
areas in the SEKI having exposures below 7 ppm-hrs. At lower alternative standards, SEKI
moves to exposures below 3 ppm-hrs. See Table 7-23 for additional details.
7-54
-------
Table 7-23 Geographic Area of Sequoia and Kings Canyon National Parks after Just
Meeting Existing and Alternative Standard Levels (km2)
Recent conditions
(2006-2008)
Just meeting 75 ppb
15 ppm-hrs
1 1 ppm-hrs
7 ppm-hrs
Under 5.94
ppm-hrs
0
3,466
3,466
3,466
3,466
Between 5.95-7
ppm-hrs
0
0
0
0
0
Between 7-11
ppm-hrs
0
0
0
0
0
Between 11-15
ppm-hrs
0
0
0
0
0
Over 15
ppm-hrs
3,466
0
0
0
0
7-55
-------
Canopy Tall Shrub Herbaceous
No Data
0%
<20%
20%-40%
40%-SO%
69%-80%
>80%
Figure 7-22 Sensitive Species Cover in Sequoia and Kings Canyon National Parks
7-56
-------
Canopy
Tall Shrub
Herbaceous
Sensitive Species Cover
Tall Shrub
^^— No Data
- 0%
<20%
20°/80%
John Muir Trail
Figure 7-23 Percent Cover of Sensitive Species Near Trails in Sequoia and Kings Canyon National Parks
7-57
-------
Tree Canopy
Tall Shrub
Herbaceous
25.1
107.6
20% to 40%
40% to 60%
60 % to 80%
1.9
ii, fL9
547.2
A 507.4
5.0
<20%
20% to 40%
40% to 60%
60% to 80%
681.8
• <20%
20% to 40%
40% to 60%
• 60 % to 80%
Figure 7-24 Trail Kilometers of Sensitive Species by Cover Category in Sequoia and
Kings Canyon National Parks
7.5 QUALITATIVE ASSESSMENT OF UNCERTAINTY
As noted in Chapter 3, we have based the design of the uncertainty analysis for this
assessment on the framework outlined in the WHO guidance (WHO, 2008). For this qualitative
uncertainty analysis, we have described each key source of uncertainty and qualitatively assessed
its potential impact (including both the magnitude and direction of the impact) on risk results, as
specified in the WHO guidance. In general, this assessment includes qualitative discussions of
the potential impact of uncertainty on the results (WHO Tierl) and quantitative sensitivity
analyses where we have sufficient data (WHO Tier 2).
Table 7-24 includes the key sources of uncertainty identified for the Os WREA. For each
source of uncertainty, we have (a) provided a description, (b) estimated the direction of influence
(over, under, both, or unknown) and magnitude (low, medium, high) of the potential impact of
each source of uncertainty on the risk estimates, (c) assessed the degree of uncertainty (low,
medium, or high) associated with the knowledge-base (i.e., assessed how well we understand
each source of uncertainty), and (d) provided comments further clarifying the qualitative
assessment presented. The categories used in describing the potential magnitude of impact for
specific sources of uncertainty on risk estimates (i.e., low, medium, or high) reflect our
consensus on the degree to which a particular source could produce a sufficient impact on risk
estimates to influence the interpretation of those estimates in the context of the secondary Os
NAAQS review. Where appropriate, we have included references to specific sources of
information considered in arriving at a ranking and classification for a particular source of
uncertainty.
7-58
-------
Table 7-24 Summary of Qualitative Uncertainty Analysis in Visible Foliar Injury Assessments.
Source
Description
Potential influence of
uncertainty on risk
estimates
Direction
Magnitude
Knowledge-
Base
Comments (KB: knowledge base, INF: influence of
uncertainty on risk estimates)
A. National W126
surfaces
The foliar injury analyses in
this chapter use the interpolated
W126 surfaces for individual
years (2006-2010), as well as
the surfaces for recent
conditions and adjusted to just
meet the existing standard and
alternative W126 standards.
Both
Low-
Medium
Low-medium
KB and INF: See Chapter 4 for more details.
B. Surveys of
recreational activities
Survey estimates of
participation rates, visitor
spending/economic impacts,
and willingness-to-pay are
inherently uncertain. These
surveys potential double-count
impacts based on the allocation
of expenditures across activities
but also potentially exclude
other activities with economic
value.
Both
Medium
Medium
KB: Each survey (NSRE, FHWAR, OIF, NFS, etc) uses
different survey methods, so it is not appropriate to generalize
across the surveys. In general, the national level surveys apply
standard approaches, which minimize potential bias.
INF: Since the surveys are in agreement that there are millions
of outdoor recreationists and billions of recreation days across
various recreation types even small changes induced by
changes in recreation satisfaction due to Os injury to
recreation sites could potentially result in large changes in the
value of outdoor recreation.
C. Os sensitive
species
Only species identified as Os-
sensitive by NFS are included
in the analyses.
Under
Medium
Medium
KB: Relatively few vegetation species have been evaluated for
Os-sensitive foliar injury in the field and continuing fieldwork
will likely identify additional sensitive species (NFS, 2003).
INF: The identification of additional sensitive species would
likely increase the extent of foliar injury in additional
locations and the percentage of injured vegetation at a
location. Due to the small number of parks without sensitive
species (i.e., only 11 parks, or 5 percent) and on-going
fieldwork, the magnitude of this uncertainty is likely to be
small for the screening-level assessment.
7-59
-------
Source
Description
Potential influence of
uncertainty on risk
estimates
Direction
Magnitude
Knowledge-
Base
Comments (KB: knowledge base, INF: influence of
uncertainty on risk estimates)
D. Spatial assignment
of foliar injury
biosite data to 12x12
km grids
Because of privacy laws that
require the exact location
information of sampling sites to
not be made public, the data
were assigned to the CMAQ
grid by the USFS. Data in
California, Oregon, and
Washington were assigned to
the CMAQ grid based on
publically available geographic
coordinates; thus, these data
have a higher level of
uncertainty.
Both
Low
Medium-Low
KB: The FHM biosites are small relative to the 12x12 km
CMAQ grids. The publically available data have the latitude
and longitude fuzzed by up to 7km in any direction, so in
California, Washington and Oregon so it is possible these sites
were assigned to the wrong CMAQ grid. In the remaining
states, the CMAQ grid was assigned from the actual locality
data.
INF: Having precise geographic locations would reduce
uncertainty, but the direction is unclear. The sites would most
likely be assigned to an adjacent CMAQ grid cell. Due to the
interpolation of the surfaces, differences between adjacent
cells are relatively small, so the magnitude of this effect is
likely small.
E. Availability of
biosite sampling data
Because sampling was
discontinued in some states
prior to this analysis, we did not
include data for many western
states (Montana, Idaho,
Wyoming, Nevada, Utah,
Colorado, Arizona, New
Mexico, Oklahoma, and
portions of Texas).
Unknown
Medium
Low
KB: Due to the discontinued sampling, data are not available
in these areas. It appears unlikely that sampling will resume in
those regions at this time.
INF: It is unclear how the addition of biosites from these
states would affect the risk estimates. The absence of biosite
sampling data in the southwest region and limited data in the
west and west north central region results in national
benchmarks that may not be applicable to these region. The
southwest in particular has generally higher W126 index
values than other regions, so data from that region would be
important. In addition, the southwest has many national parks.
F. Soil moisture
threshold for foliar
injury
Low soil moisture reduces the
potential for foliar injury, but
injury could still occur because
plants must open their stomata
even during periods of drought.
Over
High
Medium
KB: We are unaware of a clear threshold for drought below
which visible foliar injury would not occur. The national-scale
foliar injury analysis did not provide any evidence of a soil
moisture threshold for injury.
INF: If there is a threshold for drought, we may overestimate
foliar injury at lower levels of soil moisture.
7-60
-------
Source
Description
Potential influence of
uncertainty on risk
estimates
Direction
Magnitude
Knowledge-
Base
Comments (KB: knowledge base, INF: influence of
uncertainty on risk estimates)
G. Spatial resolution
of soil moisture data
Some vegetation such as along
riverbanks may experience
sufficient soil moisture during
periods of drought to exhibit
foliar injury. In addition, we did
not have soil moisture data for
Alaska, Hawaii, Puerto Rico, or
Guam.
Under
Medium
Medium
KB: Soil moisture has substantial spatial variation. The data
source for soil moisture are NOAA's 344 climate divisions,
which can be hundreds of miles wide. The inability to capture
within-division variability in soil moisture adds some
uncertainty to this assessment, particularly along riverbanks.
However, we are currently unable to quantify the magnitude
of this uncertainty.
INF: Soil moisture can vary, even within small geographic
areas. It is most likely that soil moisture is underestimated in
areas considered to be in drought conditions, so if plants in
these areas exhibited foliar injury, the soil moisture would be
underestimated, which underestimate the importance of soil
moisture's effect on foliar injury.
H Time period for
soil moisture data
Short-term estimates of soil
moisture are highly variable
over time, even from month to
month within a single year.
Using averages contributes to a
potential temporal mismatch
between soil moisture and
injury.
Unknown
Low-
Medium
Low
KB: The average of monthly values is sensitive to skew by a
single very wet or very dry month within that timeframe or
even a single precipitation episode within a month. As shown
in a sensitivity analysis, parks are not very sensitive to the
different timeframes for soil-moisture data.
INF: Without much more precise sampling, it is difficult to
assess the effect of the soil moisture sampling period, but the
overall effect of averaging appears to normalize both very
high and very low moisture conditions, which would affect
these results in opposite directions.
I. Drought categories
The soil moisture categories
used to derive the foliar injury
benchmarks (i.e., wet, normal,
and dry) are uncertain.
Unknown
Unknown
Low
KB: NOAA's categorization for Palmer Z soil moisture data
has been described as "rather arbitrary" (Karl, 1986).
INF: Using a different categorization would lead to different
benchmark criteria for Os exposure associated with foliar
injury, but it is not clear whether this uncertainty could
underestimate or overestimate the potential foliar injury.
7-61
-------
Source
Description
Potential influence of
uncertainty on risk
estimates
Direction
Magnitude
Knowledge-
Base
Comments (KB: knowledge base, INF: influence of
uncertainty on risk estimates)
J. Spatial resolution
for combining soil
moisture, biosite, and
O3 exposure data
For the national-scale foliar
injury assessment, we combined
data from different spatial
resolutions.
Unknown
Medium
Low
KB: In general, the biosite data is at a finer spatial resolution
(usually ~ .02 km2 than the O3 data (144 km2) and the soil
moisture data (hundreds of miles across).
INF: We used data at the finest spatial resolution available to
minimize this uncertainty.
K. Maps of
vegetation and
recreational areas
within parks
Maps of vegetation and
recreational areas that overlap
with areas with higher W126
index values are uncertain.
Unknown
Low
High
KB and INF: VegBank is the vegetation plot database of the
Ecological Society of America's Panel on Vegetation
Classification, and it consists of (1) actual plot records, (2)
vegetation types, and (3) all plant taxa. (See
http://vegbank.org/vegbank/general/info.html) Even though
the data quality of the vegetation maps are high, extrapolating
across the park using plant communities is uncertain due to
unqualified variation in the defined community. The spatial
resolution of the vegetation maps is higher than the gridded O3
exposure maps (12km2).
7-62
-------
7.6 KEY OBSERVATIONS
National-Scale Analysis of Foliar Injury:
• Using the FHM data on biosites and the Palmer Z drought index, across all of the
biosites (5,284 over five years from 2006-2010) over 81 percent of observations
showed no foliar injury. Using the full dataset including all observations with or
without injury, the analysis showed no clear relationship between Os and the biosite
index and no clear relationship between Os and the Palmer Z drought index. This
largely reflects the fact that Os is not a good predictor of the presence or absence of
foliar injury, but not necessarily that there is no relationship between the degree of
injury and Os in plants that do show injury.
• To better understand the relationship between Os and those biosites that did show
foliar injury, we conducted a cumulative analysis. When analyzed by individual year
and looking at the presence/absence of foliar injury, the proportion of sites exhibiting
foliar injury rises rapidly (over 20 percent in 2010) at increasing W126 index values
up to 10 ppm-hrs.
• When categorized by moisture category, the results show a more distinct pattern.
Looking at the presence/absence of foliar injury, there is a rapid increase in the
proportion of sites exhibiting foliar injury at Os below a W126 index value of 10
ppm-hrs. Sites classified as wet have much higher overall proportions at any injury
and a much more rapid increase in proportion of sites with foliar injury present. At
sites considered dry, the overall proportions are much lower for presence/absence,
potentially indicating that drought may provide protection from foliar injury as
discussed in the Os ISA.
• This analysis suggests that reductions in W126 index values at or above the W126
benchmark of 10.46 ppm-hrs are unlikely to substantially reduce the prevalence of
foliar injury. Similarly, this analysis suggests that reductions in W126 index values
below the base scenario benchmark are likely to relatively sharply reduce the
prevalence of foliar injury.
7-63
-------
Screening-level Assessment of Visible Foliar Injury in National Parks:
• Based on NFS lists, 95 percent of the parks contain at least one Os-sensitive species.
• During 2006 to 2010, 58 percent of parks exceeded the W126 benchmark
corresponding to the base scenario (W126>10.46 ppm-hrs, 17.7 percent of all biosites
in all soil moisture categories) for at least three years.
• During 2006 to 2010, 98 percent, 80 percent, 68 percent and 2 percent of parks would
exceed the W126 benchmarks corresponding to the 5 percent, 10 percent, 15 percent,
and 20 percent scenarios for at least 3 years.
• During 2006-2010, 42 percent of parks did not exceed 15 ppm-hrs.
• None of the 214 parks would exceed the W126 benchmark for the base scenario
(W126>10.46 ppm-hrs) after adjustments to just meet the existing standard at 75 ppb.
Only 8 parks exceed 7 ppm-hrs after adjustments to meet the existing standard at 75
ppb.
National Park Case Study Areas:
• GRSM is prized, in part, for its rich species diversity. The large mix of species
includes 37 Os-sensitive species and many areas contain several sensitive species.
With 3.8 million hikers using the trails every year and those hikers willing to pay over
$266 million for that activity, even a small benefit of reducing Os damage in the park
could result in a significant value.
• W126 index values in GRSM have been among the highest in the eastern U.S. - at
times twice as high as neighboring cities such as Atlanta. Under recent conditions, 44
percent of the Park has W126 index values over 15 ppm-hrs. After just meeting the
existing standard, W126 index values are reduced such that no area is over 7
ppm-hrs.
• Unlike GRSM, sensitive species cover in ROMO is driven by a few Os-sensitive
species (7 species) and most notably by Quaking Aspen. This is significant in that
many of the visitors to ROMO visit specifically to see this tree in its fall foliage.
7-64
-------
Given 1.5 million hikers in ROMO and their $70 million WTP for the hiking
experience, even a small improvement in the scenic value could be significant.
• Under recent Os conditions, all 1,067 km2 of ROMO have W126 index values over 15
ppm-hrs. Meeting the existing standard would bring about 59 percent of the Park into
the 7-15 ppm-hrs range, with the remaining 41 percent under 7 ppm-hrs. Assessing
an alternative standard of 15 ppm-hrs would bring the entire park under 7 ppm-
hrs.
SEKI is home to 12 identified sensitive species. Again, although we are not able to
quantify the impact of this scenic damage on hiker satisfaction for hikers in SEKI and
their $26 million WTP for the experience, even a small improvement in the scenic
value could be significant.
• As in the previous national park case studies, moving from recent conditions to
meeting the existing Os standard of 75 ppb results in a large change in the area of
SEKI with exposures above 15 ppm-hrs. For SEKI this means the parks move
from all areas experiencing exposures above 15 ppm-hrs to the SEKI having
exposures below 7 ppm-hrs.
7-65
-------
8 SUMMARY OF ANALYSES AND SYNTHESIS OF RESULTS
8.1 Introduction
The goals for this welfare risk and exposure assessment include characterizing ambient
ozone (Os) exposure and its relationship to ecological effects and estimating the resulting
impacts to several ecosystem services. In particular, we characterize ambient Os exposures,
using the W126 metric1, on two important ecological effects - biomass loss and foliar injury -
and estimate impacts to the following ecosystem services: supporting, regulating, provisioning,
and cultural services. In the assessment, we conduct national- and regional-scale analyses to (1)
characterize ambient Os exposure (Chapter 4); (2) quantify the effects of insect damage related to
foliar injury (Chapter 5); (3) consider the overall risk to a subset of ecosystem services by
combining the relative biomass loss (RBL) rates for multiple tree species into one metric and
evaluating weighted RBL rates (Chapter 6); (4) estimate the market effects of biomass loss on
timber production and agricultural harvesting and quantify the associated economic effects
(Chapter 6); (5) estimate the effect of biomass loss on carbon sequestration (Chapter 6); (6)
estimate the effect of foliar injury and its impact on national recreation (Chapter 7); (7) derive
W126 benchmarks representing the prevalence of visible foliar injury and soil moisture
considerations; and (8) apply these benchmarks to a screening-level assessment of foliar injury in
214 national parks (Chapter 7). In addition, we conduct case study-scale analyses to (1)
characterize the effect of foliar injury on forest susceptibility and fire regulation in California
(Chapter 5); (2) quantify the effects of biomass loss on carbon sequestration and pollution
removal in five urban areas (Chapter 6); (3) characterize the effects of relative biomass loss in
Class I areas (Chapter 7); and (4) assess the impacts of foliar injury on recreation in three
national parks (Chapter 7). In addition, in Chapters 5, 6, and 7 we qualitatively assess additional
ecosystem services, including regulating services such as hydrologic cycle and pollination;
provisioning services such as commercial non-timber forest products; and cultural services with
aesthetic and non-use values.
1 The W126 metric is a seasonal sum of hourly Os concentrations, designed to measure the cumulative effects of Os
exposure on vulnerable plant and tree species. The W126 metric uses a sigmoidal weighting function to place less
emphasis on exposure to low concentrations and more emphasis on exposure to high concentrations.
8-1
-------
To evaluate risk associated with just meeting the existing daily maximum 8-hour average
standard2 and alternative W126 standards in this welfare risk and exposure assessment, we (1)
quantified ecological effects based on relationships between ecological effect and the W126
metric, (2) quantified the impact of these ecological effects on ecosystem services, and (3)
qualitatively assessed potential impacts to several additional ecosystem services. The results
from these assessments will help inform consideration of the adequacy of the existing Os
standards and potential risk reductions associated with adjustments to meeting several alternative
levels of the standard, using the W126 form. In addition, the assessment (1) includes
information (e.g., foliar injury analyses) that could be relevant to a three-year average of a W126
standard, (2) addresses how air quality just meeting alternative W126 standard levels would
affect exposures and welfare risks and associated ecosystem services, and (3) addresses
uncertainties and limitations in the available data.
To facilitate interpretation of these results, this chapter provides a summary of the
analyses and a synthesis of the various results, focusing on comparing and contrasting results to
identify common patterns or important differences. These comparisons focus on patterns across
different geographic areas of the U.S., across years of analysis, and across alternative W126
standard levels. We evaluate the degree to which the integrated results are representative of
overall patterns of exposure and risk across different types of ecosystems. We also summarize
overall confidence in the results, as well as relative confidence between the different analyses.
The chapter concludes with an overall characterization of risk in the context of key policy
relevant questions. The remainder of this chapter summarizes the results (Section 8.2) and
includes discussions on patterns of risk (Section 8.3), representativeness (Section 8.4),
confidence in the results (Section 8.5), and integrated risk characterization (Section 8.6).
8.2 Summary of Analyses and Key Results
We conducted a variety of analyses to assess Os welfare risk and exposure and to
estimate the relative change in risk and exposure resulting from air quality adjustments to just
meet existing and alternative standards. These analyses included national- and case study-scale
analyses addressing air quality, biomass loss, foliar injury, insect damage, fire risk, and
2 The existing secondary standard for O3 is identical to the existing primary health-based standard, which is set at 75
ppb for the fourth-highest daily maximum 8-hour average, averaged over three years.
8-2
-------
recreation. The remainder of this section briefly summarizes the national- and case study-scale
analyses and key results.
8.2.1 National-Scale Analyses
8.2.1.1 Air Quality Analyses
The analyses used ambient air quality data from 2006 through 2008, as well as air quality
data adjusted to meet the current and potential alternative secondary standard levels.3 A Higher
Order Decoupled Direct Method, or HDDM, adjustment, similar to the one used in the Health
Risk and Exposure Assessment (see Chapter 4, Section 4.3.4.1 for a discussion of the
methodology), independently adjusted air quality for nine climate regions as defined by the
National Oceanic and Atmospheric Administration (NOAA) and shown in Figure 8-1 below
(Figure 4-6 from Chapter 4).4 We considered these regions an appropriate delineation for our
analyses because geographic patterns of both Os and plant species are often largely driven by
climatic features such as temperature and precipitation patterns. The NOAA climate regions
were used for all of the adjustments between observed air quality concentrations and air quality
adjusted to just meet the existing and potential alternative W126 standard levels.
In the air quality analyses in Chapter 4, we consider the changes across the distribution of
W126 index values after adjusting air quality to just meet the existing standard and just meet
alternative W126 standard levels, all three-year averages. As indicated above, each climate
region was adjusted independently such that the entire region was adjusted based on the
magnitude of across-the-board reductions in U.S. anthropogenic NOx emissions required to
bring the highest monitor down to the targeted level.5 For the biomass loss analyses, we
generated a national-scale air quality surface that just meets the existing standard using the
Voronoi Neighbor Averaging (VNA) interpolation technique to fill in values between monitor
3 W126 calculations are slightly modified in the case of the model adjustment scenarios described in Chapter 4,
Section 4.3.4. When calculating W126 for the model adjustment scenarios, we first found the three-year average of
each three-month period, and then selected the three-month period with the highest three-year average using the
same three-month period for each of the three years. In this way, the five scenarios are for recent air quality, air
quality adjusted to just meet the current standard, and air quality further adjusted to just meet three different W126
index values: 15 ppm-hrs, 11 ppm-hrs, and 7 ppm-hrs.
4 Many of the models and analytical tools used in the analyses include different definitions of geographic areas. To
the extent possible, we will refer to geographic areas by the nine climate regions based on National Oceanic and
Atmospheric Administration (NOAA) National Climate Data Center (NCDC) regions in this chapter and note where
definitions differ.
5 All of the climate regions required adjustments to just meet the existing standard of 75 ppb.
8-3
-------
locations. VNA national surfaces were also created for monitors adjusted to meet the existing
standard and for monitors adjusted to meet alternative W126 standard levels of 15, 11, and 7
ppm-hrs. During the last Os National Ambient Air Quality Standards review, the Clean Air
Scientific Advisory Committee (CASAC) recommended and supported a range of alternative
W126 standard levels from 7 to 15 ppm-hrs. The adjusted surfaces, based on monitored, three-
year average W126 index values from 2006 through 2008, are used as inputs to several
assessments (described below), including the geographic analysis to assess the effects of insect
damage related to foliar injury, the national- and case study-scale biomass loss assessments, and
the national park case studies for foliar injury. For the national-scale and screening-level foliar
injury analyses, to better match air quality data with short-term soil moisture data we generated
five national-scale air quality surfaces from the monitored annual W126 index values
(unadjusted) for the individual years from 2006 through 2010, also using VNA. See Chapter 4,
Section 4.3 for more detailed discussions of the air quality analyses.
The largest reduction in W126 index values occurs when moving from recent ambient
conditions to meeting the existing secondary standard of 75 ppb (daily maximum 8-hour
average). After adjusting to just meet the existing standard, only two of the nine U.S. regions
remain above 15 ppm-hrs (West — 18.9 ppm-hrs and Southwest - 17.7 ppm-hrs). The Central
region would meet an alternative W126 standard level of 15 ppm-hrs, but further air quality
adjustment would be needed for the Central region to meet alternative standards of 11 and 7
ppm-hrs. In addition, when adjusting to just meet the existing standard, four regions (East North
Central, Northeast, Northwest, and South) would meet 7 ppm-hrs, and two regions (Southeast
and West North Central) have index values between 9 and 12 ppm-hrs (Southeast - 11.9 ppm-hrs
and West North Central - 9.3 ppm-hrs).
8-4
-------
Figure 8-1 Map of the 9 NOAA Climate Regions (Karl and Koss, 1984) used in the
Welfare Risk and Exposure Assessment
8.2.1.2 Forest Susceptibility to Insect Infestation
In Chapter 5, we review information on Os exposure and the increased susceptibility of
forests to insect infestations. Os exposure is anticipated to result in increased susceptibility to
infestation by some chewing insects, including the southern pine beetle and western bark beetle.
These infestations can cause economically significant damage to tree stands and the associated
timber production. In the short term, the immediate increase in timber supply that results from
the additional harvesting of damaged timber depresses prices for timber and benefits consumers.
In the longer term, the decrease in timber available for harvest raises timber prices, harming
consumers and potentially benefitting some producers. The United States Forest Service (USFS)
reports timber producers have incurred losses of about $1.4 billion (2010$), and wood-using
§-5
-------
firms have gained about $966 million, due to beetle outbreaks between 1977 to 2004 (Coulson
and Klepzig, 2011). It is not possible to attribute a portion of these impacts resulting from the
effect of Os on trees' susceptibility to insect attack; however, the losses are embedded in the
estimates cited and any welfare gains from decreased Os would positively impact the net
economic impact.
In addition, in Chapter 5 we provide summaries of area at risk of high pine beetle loss
(i.e., high loss due to pine beetle damage), as well as millions of square feet of tree basal area at
risk of high pine beetle loss after just meeting the existing and alternative standards. For area at
risk of high pine beetle loss, under recent ambient conditions approximately 57 percent of the at-
risk area is at or above 15 ppm-hrs; approximately 16 percent of the at-risk area is between 15
and 11 ppm-hrs; approximately 23 percent of the at-risk area is between 11 and 7 ppm-hrs;
and approximately four percent of the at-risk area is below 7 ppm-hrs. After adjusting to just
meet the existing standard, approximately five percent of the at-risk area is between 11 and 7
ppm-hrs, and no at-risk area is above 11 ppm-hrs. When adjusting to a potential alternative
standard level of 15 ppm-hrs, no at-risk area is above 7 ppm-hrs. In terms of millions of square
feet of tree basal area at risk of high pine beetle loss, under recent ambient conditions,
approximately 45 percent of the "at-risk square feet" is at or above 15 ppm-hrs; approximately 13
percent of "at-risk square feet" is between 15 and 11 ppm-hrs; approximately 34 percent is
between 11 and 7 ppm-hrs; and approximately eight percent is below 7 ppm-hrs. After adjusting
to just meet the existing standard, approximately ten percent of the "at-risk square feet" is
between 11 and 7 ppm-hrs, and no square feet are above 11 ppm-hrs.
8.2.1.3 Biomass Loss
We reviewed several studies that modeled vegetation growth for several tree and crop
species. For trees, we calculated seedling RBL associated with W126 index values and
compared the seedling RBL values to the study results for adult trees. Overall, seedling biomass
loss values are much more consistent with adult biomass loss at lower W126 index values. For
example, for Tulip Poplar, at 15 ppm-hrs, the adult biomass loss rate is estimated to be 10.5
percent, and the seedling biomass loss rate is estimated to be 7.7 percent. See Chapter 6, Section
6.2.1.1 for additional information.
-------
For biomass loss, CAS AC recommended that EPA should consider options for W126
standard levels based on factors including a predicted one to two percent biomass loss for trees
and a predicted five percent loss of crop yield. Small losses for trees on a yearly basis compound
over time and can result in substantial biomass losses over the decades-long lifespan of a tree
(Frey and Samet, 2012b). To assess overall ecosystem-level effects from biomass loss, we
weighted the RBL values for multiple tree species using basal area6 and combined them into a
weighted RBL value and considered the weighted value in relation to the proportion of basal area
accounted for by the tree species. A weighted RBL value is a relatively straight-forward metric
to attempt to understand the potential ecological effect on some ecosystem services. We
summarized the percent of total basal area that exceeds a two percent weighted biomass loss
under recent conditions, at just meeting the existing standard (75 ppb) and at potential alternative
W126 standard levels of 15, 11, and 7 ppm-hrs.7 The data indicate that the total area exceeding
two percent biomass loss decreases across air quality scenarios. For example, for the Central
region under recent conditions, a total of 23.4 percent of total basal area assessed would exceed a
two percent biomass loss, and when adjusted to just meet the existing standard, a total of 2.7
percent of total basal area assessed would exceed a two percent biomass loss. It is important to
note that the proportional basal area values do not account for total cover, but rather the relative
cover of the tree species present. See Chapter 6, Section 6.8 for additional information. We also
analyzed federally designated Class I areas by calculating an average weighted RBL value for
145 of the 156 Class I areas and present the results as a count of the Class I areas and not as a
percentage of area. The number of areas exceeding one percent and two percent biomass loss
decreases across air quality scenarios. See Chapter 6, Section 6.8.1 for additional information.
Using the exposure-response (E-R) functions for tree seedlings and crops, we determined
the range of biomass loss associated with just meeting the existing daily maximum 8-hour
average standard and alternative W126 standard levels. We plotted the E-R functions as a
function of the percent biomass loss against varying W126 index values. For a one percent
biomass loss for tree seedlings, the estimated W126 index values were between 4 and 10 ppm-
6 Basal area is the term used in forest management that defines the area of a given section of land that is occupied by
a cross-section of tree trunks and stems at their base. This typically includes a measurement taken at the diameter at
breast height of a tree above the ground and includes the complete diameter of every tree, including the bark.
7 We also present the data excluding Cottonwood, which is a very sensitive species.
8-7
-------
hrs; for a two percent biomass loss for tree seedlings the estimated W126 index values were
between 7 and 14 ppm-hrs; and for a five percent biomass loss for crops the estimated W126
index values were between 12 and 17 ppm-hrs. See Chapter 6, Section 6.2.1.2 for additional
information.
Using the Forest and Agricultural Optimization Model with Greenhouse Gases
(FASOMGHG), we conducted national-scale analyses to quantify the effects of biomass loss on
timber production and agricultural harvesting, as well as on carbon sequestration.8 We used the
Os E-R functions for tree seedlings and crops to calculate relative yield loss (RYL), which is
equivalent to relative biomass loss. Because the forestry and agriculture sectors are related, and
trade-offs occur between the sectors, we simultaneously calculated the resulting market-based
welfare effects of Os exposure in the forestry and agriculture sectors.
In the analyses for commercial timber
production, because most areas are lower than
15 ppm-hrs when simulating meeting the
existing standard (based on reducing
nationwide emissions of NOx), RYLs are
below one percent, with the exception of the
Southwest, Southeast, Central, and South
regions (see text box for clarification on region
names). Relative yield losses remain above one
percent for the parts of the Southeast, Central,
and South regions at alternative W126 standard
The states included in the NOAA NCDC regions
and the states included in the FASOMGHG model
regions differ slightly. Below we align the
different region names. To be consistent across
summary discussions, we use the NCDC region
names.
NCDC FASOMGHG
West primarily Pacific Southwest
Southwest primarily Rocky Mountain
Central primarily Cornbelt
South primarily South West and South
Central
Southeast primarily South Central and
Southeast
Northeast primarily Northeast
levels of 15 and 11 ppm-hrs, and for the Southeast and South regions at an alternative W126
standard level of 7 ppm-hrs.
In the analyses for agricultural harvest, the largest yield changes occur when comparing
recent ambient conditions to just meeting the existing standard. Under recent ambient
conditions, the West, Southwest, and Northeast regions generally have the highest yield losses.
8 FASOMGHG is a national-scale model that provides a complete representation of the U.S. forest and agricultural
sectors' impacts of meeting alternative standards. FASOMGHG simulates the allocation of land over time to
competing activities, e.g., production of different crops or livestock, in both the forest and agricultural sectors.
-------
At alternative W126 standard levels of 15, 11, and 7 ppm-hrs, for winter wheat9 relative yield
losses are less than the five percent loss recommended by CASAC, as well as less than one
percent. For soybeans, when the W126 scenarios are modeled, yield losses above both five and
one percent remain at 15 ppm-hrs for the Southwest and Central regions. Yield losses are
reduced to below one percent at alternative W126 standard levels of 11 and 7 ppm-hrs.
In addition to estimating changes in forestry and agricultural yields, FASOMGHG
estimates the changes in consumer and producer/farmer surplus associated with the change in
yields.10 Changes in yield affect individual tree species and crops, but the overall effect on forest
ecosystem productivity depends on the composition of forest stands and the relative sensitivity of
trees within those stands. Overall effect on agricultural yields and producer and consumer
surplus depends on the (1) ability of producers/farmers to substitute other crops that are less Os
sensitive and (2) responsiveness, or elasticity, of demand and supply. Relative to just meeting
the existing standard, W126 index values decrease in the Southwest, West, Central, Southeast,
South, East North Central, and West North Central regions at alternative standard levels of 15,
11, and 7 ppm-hrs. These decreases in W126 index values are estimated to result in changes in
patterns for agricultural production and resulting consumer and producer surplus. For example,
with reductions in W126 index values, wheat crops would likely increase in one of its major
production regions, the Southwest region. This expansion of wheat production may result in a
decrease in wheat production in the East North Central region. The East North Central region
would likely see production changes for other crops because the contraction in wheat production
makes room for alternatives. Soybean production in the East North Central region would likely
expand, and this expansion would induce regional shifts of soybean production at the national
level, including decreases in soybean production in the West North Central and Central regions.
Generally the crop producers' surplus in the Central and Southwest regions would increase and
in the South region would decrease. Crop producers' surplus in the West North Central and East
North Central regions would fluctuate over time.
Economic welfare impacts resulting from just meeting the existing and alternative
standards were largely similar between the forestry and agricultural sectors — consumer surplus,
9 Among the major crops, because winter wheat and soybeans are more sensitive to ambient O3 levels than other
crops we include these crops for this discussion.
10 See Chapter 6, Section 6.3 for a brief discussion of economic welfare and consumer and producer surplus.
8-9
-------
or consumer gains, generally increased in both sectors because higher productivity under lower
W126 index values increased total yields and reduced market prices. Because demand for most
forestry and agricultural commodities is not highly responsive to changes in price, there were
more cases where producer surplus, or producer gains, decline. In some cases, lower prices
reduce producer gains more than can be offset by higher yields. For example, in 2040, the year
with maximum changes in consumer and producer surplus, in the forestry sector at just meeting
the existing standard, total producer surplus is estimated to be $133 billion and total consumer
surplus is estimated to be $935 billion, or 7 times greater than producer surplus. For the forestry
sector, when adjusting to meeting alternative W126 standard levels of 15, 11, and 7 ppm-hrs,
consumer surplus increases $597 million, $712 million, and $779 million (i.e., 0.06, 0.08, and
0.08 percent), respectively, while producer surplus decreases $839 million, $858 million, and
$766 million, (i.e., about 0.6 percent), respectively. All estimates are in 2010$ for the U.S.
only.11
In the analysis for changes in carbon sequestration related to biomass loss, relative to just
meeting the existing standard, the 15 ppm-hrs W126 alternative standard level does not
appreciably increase carbon sequestration (meeting the existing 8-hour standard of 75 ppb
increases carbon sequestration by 2,972 million metric tons per year). The majority of the
enhanced carbon sequestration potential is in the forest biomass increases over time under
alternative secondary W126 standard levels at 11 and 7 ppm-hrs. In the forestry sector, relative
to just meeting the existing standard (with sequestration of 89 billion metric tons of CCh
equivalents), at alternative W126 standard levels of 11 and 7 ppm-hrs carbon sequestration
potential is projected to increase 593 million and 1.6 billion metric tons of CCh equivalents over
30 years (i.e., 0.66 and 1.79 percent) respectively. For the agricultural sector, relative to just
meeting the existing standard (with sequestration of 8 billion metric tons of CCh equivalents), at
alternative W126 standard levels of 11 and 7 ppm-hrs carbon sequestration potential is projected
to increase 9 and 10 million metric tons of CCh equivalents respectively over 30 years, or about
0.1 percent.
11 FASOMGHG is an international model and the increase in productivity caused by a reduction in O3 results in a i
increase in the present value of total global economic surplus (consumer + producer surplus). For any given year,
there may be a decline in global consumer and producer surplus due to the effects on the dynamics of planting and
harvesting decisions in the forestry sector.
8-10
-------
8214 Visible Foliar Injury
To assess the effects of visible foliar injury on recreation, we reviewed the National
Survey on Recreation and the Environment (NSRE), as well as the 2006 National Survey of
Fishing, Hunting, and Wildlife-Associated Recreation (FHWAR) and a 2006 analysis done for
the Outdoor Industry Foundation (OIF). According to the NSRE, some of the most popular
outdoor activities are walking, including day hiking and backpacking; camping; bird watching;
wildlife watching; and nature viewing. Participant satisfaction with these activities can depend
on the quality of the natural scenery, which can be adversely affected by Os-related visible foliar
injury. According to the FHWAR and the OIF reports, the total expenditures across wildlife
watching activities, trail-based activities, and camp-based activities are approximately $200
billion dollars annually. While we cannot quantify the magnitude of the impacts of Os damage
to the scenic beauty and outdoor recreation, the existing losses associated with current Os-related
foliar injury are reflected in reduced outdoor recreation expenditures.
To assess foliar injury at a national-scale, we conducted several analyses using a national
data set on foliar injury from the USFS's Forest Health Monitoring (FHM) Network. We
conducted the analyses representing the prevalence (i.e., presence/absence) of foliar injury across
years and different soil moisture categories in NOAA climate divisions.12 Across years, when
assessing the presence or absence of foliar injury, at an alternative W126 standard level of 15
ppm-hrs between 12 and over 20 percent of biosites indicated the presence of foliar injury; at an
alternative W126 standard level of 11 ppm-hrs between 12 and over 20 percent of biosites
indicated the presence of foliar injury; and at an alternative W126 standard level of 7 ppm-hrs
between 4 and over 20 percent of biosites indicated the presence of foliar injury.13 Generally, the
results of all of these foliar injury analyses demonstrate a similar pattern - the proportion
of biosites14 showing foliar injury increases steeply with W126 index values up to approximately
10 ppm-hrs and is relatively constant above 10 ppm-hrs. This analysis suggests that reductions
in W126 index values at or above this benchmark (W126 > 10.46 ppm-hrs) are unlikely to
12 See Chapter 7, Section 7.2 for a more detailed discussion of the data on biosites and foliar injury from the USFS
and the Palmer Z drought index data from NOAA.
13 See Chapter 7, Section 7.2.3 for additional discussion and Figure 7-8 for additional information. The proportion
of sites with foliar injury present varies by year, creating these ranges for percent of sites with foliar injury present.
14 A biosite is a plot of land on which data was collected regarding the incidence and severity of visible foliar injury
on a variety of Os-sensitive plant species.
8-11
-------
substantially reduce the prevalence of foliar injury. Similarly, this analysis suggests that
reductions below 10 ppm-hrs are likely to relatively sharply reduce the prevalence of foliar
injury. Figure 8-2, which originally appears as Figure 7-10 in Chapter 7, shows the pattern seen
in the foliar injury analyses stratified by soil moisture category. We see a similar pattern when
the foliar injury is stratified by year. See Section 7.2.3 for a more detailed discussion of the
national-scale analyses. In addition, in Appendix 7A (Table 7A-27) we include the percentage of
all biosites across all years (2006 - 2010) showing foliar injury at alternative secondary standard
levels. At an alternative secondary standard of 15 ppm-hrs, 18.1 percent of all biosites show
foliar injury; at 11 ppm-hrs, 17.8 percent of all biosites show foliar injury; and at 7 ppm-hrs, 15.8
percent of all biosites show foliar injury.
Biosites with Foliar Injury
10
20
W126 (ppm-hrs)
30
40
Figure 8-2 Cumulative Proportion of Biosites with Visible Foliar Injury Present, by
Moisture Category
Enjoyment of recreation in national parks can be adversely affected by visible foliar
injury, and national parks are areas designated for protection. We used the results of the national
analysis to derive benchmarks for visible foliar injury that we apply in a screening-level
assessment and case studies of national parks. We define five scenarios for evaluating potential
8-12
-------
W126 benchmarks, representing the full range of the percentages of biosites showing visible
foliar injury, including four scenarios considering soil moisture. We defined the W126
benchmark for the "base scenario" as the W126 index value where the slope of exposure-
response relationship changes for all FHM biosites in all soil moisture categories. We also
looked at additional scenarios based on three different categories of soil moisture (i.e., wet,
normal, and dry) and the W126 index values associated with four different prevalences (e.g., 5
percent, 10 percent, 15 percent and 20 percent of biosites) of any foliar injury. In total, the
welfare risk and exposure assessment evaluated ten different W126 benchmarks associated with
the five foliar injury risk scenarios. The W126 benchmarks across the five scenarios range from
3.05 ppm-hrs (five percent of biosites, normal moisture, any injury) up to 24.61 ppm-hrs (15
percent of biosites, dry, any injury). See Table 7-6 for the specific benchmark criteria
corresponding to each of the five scenarios.
The general approach in the screening-level assessment of national parks is derived from
Kohut (2007), but we apply more recent Os exposure and soil moisture data for 214 national
parks in the contiguous U.S. combined with the benchmarks derived from the national-scale
analysis. Generally, scenarios for higher percentages of FHM biosites showing foliar injury have
fewer parks that exceed the benchmarks for those scenarios across multiple years. During 2006
to 2010, 58 percent of parks exceeded the W126 benchmark corresponding to the base scenario
(W126>10.46 ppm-hrs, all biosites in all soil moisture categories) for at least three years. In
addition, 98 percent, 80 percent, 68 percent and 2 percent of parks would exceed the benchmark
criteria corresponding to the prevalence scenarios (i.e., 5 percent, 10 percent, 15 percent, and 20
percent) for at least three years within the 2006-2010 period. Because the screening-level
assessment relies on annual estimates of W126 index values and soil moisture, we cannot fully
evaluate just meeting the existing and alternative standards because they are based on the three-
year average air quality surfaces. However, we can observe that after adjusting the W126
surfaces to just meet the existing standard, all of the 214 parks are below 10.46 ppm-hrs, which
corresponds to the W126 benchmark for the base scenario.
8.2.2 Case Study-Scale Analyses
8221 Fire Regulation
As indicated in Chapter 5, fire regime regulation is also negatively affected by Os
exposure. For example, Grulke et al. (2009) reported various lines of evidence indicating that Os
8-13
-------
exposure may contribute to southern California forest susceptibility to wildfires by increasing
leaf turnover rates and litter, increasing fuel loads on the forest floor. According to the National
Interagency Fire Center, in the U.S. in 2010 over 3 million acres burned in wildland fires. From
2004 to 2008, Southern California alone experienced, on average, over 4,000 fires per year
burning, on average, over 400,000 acres per year. The California Department of Forestry and
Fire Protection (CAL FIRE) estimated that losses to homes due to wildfire were over $250
million in 2007 (CAL FIRE, 2008). In 2008, CAL FIRE's costs for fire suppression activities
were nearly $300 million (CAL FIRE, 2008).
We developed maps that overlay the mixed conifer forest area of California with areas
of moderate or high fire risk defined by CAL FIRE and with surfaces of recent conditions and
surfaces adjusted to just meet existing and alternative standards. The highest fire risk and
highest W126 index values overlap with each other, as well as with significant portions of mixed
conifer forest. Under recent conditions, over 97 percent of mixed conifer forest area was over 7
ppm-hrs with a moderate to severe fire risk, and 74 percent was over 15 ppm-hrs with a moderate
to severe fire risk. When adjusted to just meet the existing standard, almost all of the mixed
conifer forest area with a moderate to high fire risk shows a reduction in Os to below 7 ppm-hrs.
At the alternative W126 standard level of 15 ppm-hrs, all but 0.18 percent of the area is below 7
ppm-hrs, and at alternative standard levels of 11 and 7 ppm-hrs all of the moderate to high fire
threat area is below 7 ppm-hrs.
8222 Biomass Loss
Using the iTree model to estimate tree growth and ecosystem services provided by trees
over a 25-year period, we conducted case-study scale analyses to quantify the effects of biomass
loss on carbon sequestration and pollution removal in five urban areas.15 See Appendix 6D for
details on the iTree model and the methodology used for the case study analyses.
We estimated the effects of Os-related biomass loss on carbon sequestration and ran five
scenarios, including current conditions, just meeting the existing standard, and just meeting
alternative W126 standards of 15, 11, and 7 ppm-hrs. While both urban and non-urban forests
have the potential to remove pollutants from the atmosphere, using iTree we also estimated the
15 The iTree model is a peer-reviewed suite of software tools provided by USFS.
8-14
-------
effects of Os-related biomass loss on the potential to remove carbon monoxide, nitrogen dioxide,
Os, and sulfur dioxide pollution in the five urban areas (1) at recent ambient Cb conditions and
(2) after adjusting air quality to just meet the existing standard and alternative W126 standard
levels of 15, 11, and 7 ppm-hrs. As a supplement to the iTree analysis, we also performed a
simple analysis of the Os pollution removal potential to show how this process might affect
ambient air quality values. This analysis made some general assumptions to estimate order of
magnitude effects of Os removal by trees in the five urban areas. The results indicate that the
effects on Os concentrations are small; when meeting the current standard, deposition to tree
surfaces results in ambient Os concentration reductions ranging from 0.08 parts per billion by
volume (ppbv) in Tennessee to 0.52 ppbv in Chicago compared to Os concentrations that would
occur without any deposition to trees in these cities.16 Relative changes in ambient Os
concentrations due to changes in deposition to tree surfaces were much smaller.
Relative to just meeting the existing standard, three of the urban areas (Atlanta, Chicago,
and the urban areas of Tennessee) show gains in carbon sequestration at alternative W126
standard levels of 11 and 7 ppm-hrs. For example, relative to just meeting the existing standard,
Chicago gains about 6,400 tons of carbon sequestration per year at 7 ppm-hrs, and the urban
areas of Tennessee gain about 8,800 tons of carbon sequestration per year at 11 ppm-hrs and
20,000 tons of carbon sequestration per year at 7 ppm-hrs. Syracuse and Baltimore do not
realize gains in carbon sequestration because recent air quality almost meets the alternative
standards levels in those areas. Similar to changes in carbon sequestration, Syracuse and
Baltimore have no change in pollution removal when just meeting the existing standard and the
W126 alternative standards. Atlanta, Chicago, and the urban areas of Tennessee show gains in
potential pollution removal at alternative W126 standard levels of 11 and 7 ppm-hrs compared to
meeting the existing standard. For example, relative to just meeting the existing standard,
Chicago gains about 2,300 metric tons of pollution removal annually at 11 ppm-hrs and 6,500
metric tons of pollution removal annually at 7 ppm-hrs, and the urban areas of Tennessee gain
about 5,300 metric tons of pollution removal annually at 11 ppm-hrs and 11,700 metric tons of
pollution removal annually at 7 ppm-hrs.
16 The ratio of Os volume to urban area air volume multiplied by 10A9 gives the concentration in ppbv.
8-15
-------
8223 Foliar Injury - Three National Parks
In addition to the national-scale analysis, we also assess foliar injury at a case-study scale
because national parks are designated as special areas in need of protection. Specifically, we
assess Os-exposure risk at three national parks - Great Smoky Mountains National Park
(GRSM), Rocky Mountain National Park (ROMO), and Sequoia/Kings Canyon National Parks
(SEKI). For each park, we assess the potential impact of Os-related foliar injury on recreation
(cultural services) by considering information on visitation patterns, recreational activities and
visitor expenditures. We include percent cover of species sensitive to foliar injury and focus on
the overlap between recreation areas within the park and elevated W126 index values.
In GRSM, there are 37 sensitive species across vegetative strata, and 2011 visitor
spending exceeded $800 million. W126 index values in GRSM have been among the highest in
the eastern U.S. — under recent ambient conditions, 44 percent of GRSM is over 15 ppm-hrs.
After adjustments to just meet the existing standard of 75 ppb, no area in GRSM exceeds 7 ppm-
hrs. ROMO has seven sensitive species, including Quaking Aspen. In 2011 visitor spending at
ROMO was over $170 million. Under recent ambient conditions, all of ROMO is over 15 ppm-
hrs. When adjusted to just meet the existing standard, 41 percent of the park would be below 7
ppm-hrs and 59 percent of the park would be between 7 and 11 ppm-hrs. In SEKI there are 12
sensitive species across vegetative strata, and 2011 visitor spending was over $97 million. When
adjusted to just meet the existing standard, no area in SEKI is above 7 ppm-hrs.
8.3 Patterns of Risk
Considering the national- and case study-scale analyses and appropriate benchmarks for
biomass loss and foliar injury, we reviewed whether there were patterns or trends in the risk and
risk reductions - between geographic areas and across years and alternative standards. For
biomass loss, CASAC recommended that EPA should consider options for W126 standard levels
based on factors including a predicted one to two percent biomass loss for trees and a predicted
five percent loss of crop yield. Small losses for trees on a yearly basis compound over time and
can result in substantial biomass losses over the decades-long lifespan of a tree (Frey and Samet,
2012b). For trees, annual W126 index values for a one percent biomass loss range from
approximately 4 to 10 ppm-hrs and for a two percent biomass loss range from approximately 7 to
14 ppm-hrs. For crops, annual W126 index values for a five percent biomass loss range from
8-16
-------
General references to the eastern and western
U.S. and the states included in the NOAA
NCDC regions differ. For ease of discussion,
below we align the general U.S. region and
NCDC region references.
approximately 12 to 17 ppm-hrs. Based on this assessment, the pattern is that crops exceed
CASAC's benchmarks at higher W126 index values than trees, and suggests that meeting
alternative standards that are protective of trees will also protect crops. Unlike biomass, CAS AC
did not recommend a benchmark for foliar injury. As a result, we developed a set of W126
benchmark criteria ("scenarios") associated with the prevalence (i.e., presence/absence) of foliar
injury across years and different soil moisture categories.
8.3.1 Risk Patterns Across or Between Geographic Areas
The geographic or spatial patterns of changes in W126 index values and changes in
ecosystem services and related economic welfare are slightly different. Figure 8-3 and Figure
8-4, which originally appear as Figures 4-9 and 4-11 in Chapter 4, show the W126 index values
after being adjusted to just meet alternative standards of 15 and 11 ppm-hrs. After adjusting to
just meet an alternative standard of 15 ppm-hrs,
the West, Southwest, and Central regions show
the highest W126 index values between 11 and
15 ppm-hrs; after adjusting to just meet an
alternative standard level of 11 ppm-hrs, all
areas show W126 index values below 11 ppm-
hrs. The analyses of biomass loss and affected
timber and agricultural yields show that most of
the remaining risk after adjusting to just meet an
alternative standard level of 15 ppm-hrs is in the
Southwest, South, Southeast, and Central
regions; after adjusting to just meet an
alternative standard level of 11 ppm-hrs, most of the remaining risk is in the South, Southeast,
and Central regions.
There is substantial heterogeneity in plant responses to Os, both within species, between
species, and across regions of the U.S. The Os-sensitivetree species are different in the eastern
and western U.S. — the eastern U.S. has far more total species (see text box for clarification on
region names). Os exposure and risk are somewhat easier to assess in the eastern U.S. because of
the availability of more data and the greater number of species to analyze. In addition, there are
more Os monitors in the eastern U.S. but fewer national parks. In the national-scale analyses for
8-17
General U.S.
Western U.S.
Eastern U.S.
NCDC
Northwest
West
Southwest
West North Central
East North Central
Central
South
Southeast
Northeast
-------
commercial timber production, because most areas are below 15 ppm-hrs after simulating just
meeting the existing standard, RYL are below one percent, with the exception of the Southwest,
Southeast, Central, and South regions. In part because the South and Southeast regions have
more forest land, RYL remain above one percent for parts of those regions even after just
meeting an alternative W126 standard level of 7 ppm-hrs.
15 ppm-hr Scenario
Figure 8-3 National Surface of 2006-2008 Average W126 Index Values Adjusted to
Just Meet the Alternative Standard Level of 15 ppm-hrs
8-18
-------
11 ppm-hr Scenario
Figure 8-4 National Surface of 2006-2008 Average W126 Index Values Adjusted to Just
Meet the Alternative Standard Level of 11 ppm-hrs
The largest improvements in agricultural harvesting resulting from reduced Os exposure
are likely to occur in the West, Southwest, South, Southeast, and Central regions because those
regions (1) have the most sensitive crop species present, (2) have significant agricultural
production, and (3) will experience the most significant air quality improvement between recent
conditions and just meeting the existing secondary standard. For soybeans, when the W126
scenarios are modeled, yield losses above both five and one percent remain at 15 ppm-hrs for the
Southwest and Central regions. For all regions, yield losses are reduced to below five and one
percent at alternative W126 standard levels of 11 and 7 ppm-hrs.
8-19
-------
8.3.2 Risk Patterns Across Years
Using the FASOMGHG model to calculate forestry and agricultural yield changes, we
estimated changes in consumer and producer surplus from 2010 through 2040 for alternative
standard levels of 15, 11, and 7 ppm-hrs. Over the period in the forestry sector, changes in
consumer surplus are always positive and range from <0.01 percent in 2010 for alternative
standard levels of 15 and 11 ppm-hrs up to 0.08 percent in 2040 for alternative standard levels of
11 and 7 ppm-hrs (relative to consumer surplus at just meeting the existing standard of $721
billion in 2010 and $934 billion in 2040 (2010$)). Consumer surplus does not consistently
increase between 5-year periods from 2010 to 2040.17 For example, while always a positive
value, consumer surplus decreases between 2025 and 2030, increases slightly between 2030 and
2035, and increases significantly between 2035 and 2040. Changes in producer surplus are
generally negative and range from <-0.1 percent in 2010 for an alternative standard level of 7
ppm-hrs to -0.6 percent in 2040 for alternative standard levels of 15 and 11 ppm-hrs (relative to
producer surplus at just meeting the existing standard of between $93 billion in 2010 and $133
billion in 2040).
In the agricultural sector over the period, changes in consumer surplus are generally
positive and <0.01 percent (relative to consumer surplus at just meeting the existing standard of
between $1.9 trillion in 2010 and $2.1 trillion in 2040 (2010$)). Changes in producer surplus
vary and range from -0.2 percent in 2015 for alternative standard levels of 11 and 7 ppm-hrs to
0.25 and 0.35 percent in 2040 for alternative standard levels of 11 and 7 ppm-hrs (relative to
producer surplus at just meeting the existing standard of between $725 billion in 2010 and $863
billion in 2040). At just meeting the existing standard, total consumer and producer surplus
values are much higher in the agricultural sector than in the forestry sector. As a result, absolute
changes in consumer and producer surplus values at alternative standard levels are much larger
in the agricultural sector. In the agricultural sector, over time and by alternative standard,
changes in consumer surplus are largely positive, with approximately 15 percent of the estimates
being minor negative changes. Over time and by alternative standard, changes in producer
17 FASOMGHG results include multi-period, multi-commodity results over 60 to 100 years in 5-year time intervals
when running the combined forest-agriculture version of the model.
8-20
-------
surplus are mixed, with approximately 30 percent of the estimates being significant negative
changes. See Section 6.5 and Appendix 6B for additional discussion of these analyses.
In the national-scale assessment to identify foliar injury benchmarks, we conducted
analyses using a national data set on foliar injury. Across years in the data set, we analyzed
presence/absence of foliar injury. Generally, 2010 showed a more dramatic rise in the proportion
of sites showing the presence of foliar injury below 10 ppm-hrs, and 2006 through 2009 showed
a more subtle pattern. Figure 8-5 below, which originally appears as Figure 7-9 in Chapter 7,
shows the pattern for presence/absence of foliar injury across years.
Biosites with Foliar Injury
R -
E
D_
10
20
W126 (ppm-hrs)
30
40
Figure 8-5 Cumulative Proportion of Sites with Foliar Injury Present, by Year
In addition to the above foliar injury analyses, the screening-level assessment for 214
national parks assessed foliar injury in individual years. This assessment, which was based on
W126 index values and soil moisture that varied temporally, concluded that Os-related foliar
injury risk in parks was generally lower in the 2008-2010 time period than in the 2006-2008 time
period. For the base scenario, 2009 represented the year with the lowest percentage of parks
exceeding the benchmark criteria (i.e., only 12 percent of parks) and 2006 represented the year
8-21
-------
with the highest percentage of parks exceeding the benchmark criteria (i.e., 80 percent of parks).
Further, this assessment determined that the three-month timeframe corresponding to the highest
W126 estimates in monitored parks occurred between March and September, which roughly
corresponds to the vegetation growing season.
8.3.3 Risk Patterns Across Alternative W126 Standard Levels
For the ecological effect of biomass loss, Os-related exposure and risk decrease at lower
alternative W126 standard levels. For the ecological effect of foliar injury, changes in Cb-related
exposure and risk at lower alternative W126 standard levels are more challenging to directly
assess because we do not have E-R functions to assess changes in foliar injury across different
W126 index values. However, we observe that after just meeting the existing standard, all of the
214 parks are below 10.46 ppm-hrs, which corresponds to the W126 benchmark for the base
scenario. See Table 8-1 and Table 8-2 for a summary of risk across alternative W126 standard
levels for these two ecological effects.
8-22
-------
Table 8-1 Summary of Os-Exposure Risk Across Alternative W126 Standards Relative to Just Meeting Existing Standard -
National-Scale Analyses
15 ppm-hrs 11 ppm-hrs 7 ppm-hrs
Ecological Effect
Biomass Loss
Ecosystem Services
Provisioning
Regulating
Average Weighted RBL Loss
for Tree Seedlings
(Section 6. 8)
Timber Production
(Section 6. 3)
Consumer and Producer
Surplus (2010$) - Forestry
(Section 6. 3)
Agricultural Harvest
(Section 6. 5)
Consumer and Producer
Surplus (2010$) - Agriculture
(Section 6. 5)
Carbon Sequestration
(Section 6. 6.1)
Percent of Covered Area exceeding 1
and 2 percent weighted RBL
declines by about 0.3 percent
All regions RYL below 1 percent.
Consumer surplus - in 2010 is $7
million, or 0.01% and in 2040 is
$597 million, or 0.06%
Producer surplus - in 2010 is -$1 1
million, or -0.01% and in 2040 is
-$839 million, or -0.6%
For some sensitive crops (soybeans),
RYL remain > 1 percent in the
Southwest and Central regions. All
other regions RYL below 1 percent.
Consumer surplus - in 2010 is $15
million, or <0.01% and in 2040 is $3
million, or <0.01%
Producer surplus - in 2010 is $612
million, or 0.08%; in 2015 is -$1,255
million, or -0.15%; and in 2040 is
$697 million, or 0.08%
Little change compared to just
meeting existing standard
Percent of Covered Area exceeding 1
and 2 percent weighted RBL declines
by between 0.5 and 1.3 percent
All regions RYL below 1 percent.
Consumer surplus - in 2010 is $44
million, or 0.01% and in 2040 is $712
million, or 0.08%
Producer surplus - in 2010 is -$41
million, or -0.04% and in 2040 is -$858
million, or -0.6%
For most sensitive crops, RYL < 1
percent.
Consumer surplus - in 2010 is $19
million, or <0.01% and in 2040 is $13
million, or <0.01%
Producer surplus - in 2010 is $1,474
million, or 0.2%; in 2015 is -$2,197
million, or -0.26%; and in 2040 is
$2,189 million, or 0.25%
In forestry sector, storage potential is
projected to increase 593 million metric
tons of CO2 equivalents (CO2e), or 0.66
percent, over 30 years.
In agricultural sector, storage potential is
projected to increase 9 million metric
tons of CO2e, or about 0. 1 percent, over
30 years.
Percent of Covered Area exceeding
1 and 2 percent weighted RBL
declines by between 0.6 and 2
percent
All regions RYL below 1 percent.
Consumer surplus - in 2010 is $86
million, or 0.01% and in 2040 is
$779 million, or 0.08%
Producer surplus - in 2010 is
-$136 million, or -0. 15% and in
2040 is -$766 million, or -0.6%
For most sensitive crops, RYL < 1
percent.
Consumer surplus - in 2010 is
-$3 1 million, or <0.01% and in
2040 is $46 million, or <0.01%
Producer surplus - in 2010 is $269
million, or 0.04%; in 2015 is
-$1,873 million, or -0.23%; and in
2040 is $2,991 million, or 0.3%
In forestry sector, storage potential
is projected to increase 1.6 billion
metric tons of CO2e, or 1.79
percent, over 30 years.
In agricultural sector, storage
potential is projected to increase 10
million metric tons of CO2e, or 0. 1
percent, over 30 years.
8-23
-------
Table 8-1 Summary of Os-Exposure Risk Across Alternative W126 Standards Relative to Just Meeting Existing Standard -
National-Scale Analyses, continued
15 ppm-hrs
11 ppm-hrs
7 ppm-hrs
Ecological Effect
Foliar Injury
National-Scale Foliar Injury
Analysis18
(Section 7.2 and Appendix
7A)
Depending on year, between 12 and
>20 percent of biosites showed
presence/absence of foliar injury
during 2006 to 2010
Across all years, 18.1 percent of
biosites showed presence/absence of
foliar injury during 2006 to 2010
Depending on moisture category,
between 7 and >20 percent of
biosites showed presence/absence of
foliar injury during 2006 to 2010
Depending on year, between 12 and
>20 percent of biosites showed
presence/absence of foliar injury during
2006 to 2010
Across all years, 17.8 percent of biosites
showed presence/absence of foliar injury
during 2006 to 2010
Depending on moisture category,
between 7 and >20 percent of biosites
showed presence/absence of foliar injury
during 2006 to 2010
Depending on year, between 4 and
> 20 percent of biosites showed
presence/absence of foliar injury
during 2006 to 2010
Across all years, 15.8 percent of
biosites showed presence/absence of
foliar injury during 2006 to 2010
Depending on moisture category,
between 7 and >20 percent of
biosites showed presence/absence of
foliar injury during 2006 to 2010
18 This analysis is not relative to just meeting the existing standard, but is a national-scale analysis that summarizes foliar injury at different levels.
8-24
-------
Table 8-2 Summary of Os-Exposure Risk Across Alternative Standards Relative to Just Meeting Existing Standard -
Case Study-Scale Analyses
15 ppm-hrs
11 ppm-hrs
7 ppm-hrs
Ecosystem Services
Regulating (Biomass
Loss)
Carbon Sequestration
(Section 6.6.2 - Five
Urban Areas)
W126 index values and carbon
storage potential do not change
relative to just meeting existing
standard
Atlanta, Chicago, and the urban areas of
Tennessee show gains in carbon
sequestration. For example, urban areas of
Tennessee gain about 8,800 tons of
sequestration annually.
Syracuse and Baltimore do not realize
gains because recent W126 index values
almost meet the alternative standards
levels.
Atlanta, Chicago, and the urban areas
of Tennessee show gains in carbon
sequestration. For example, urban areas
of Tennessee gain about 20,000 tons of
sequestration annually.
Syracuse and Baltimore do not realize
gains because recent W126 index
values almost meet the alternative
standards levels.
Pollution Removal
(Section 6.7 - Five
Urban Areas)
W126 index values and pollution
potential do not change relative to
just meeting existing standard
Atlanta, Chicago, and the urban areas of
Tennessee show gains in pollution
removal. For example, urban areas of
Tennessee gain about 5,300 tons of
pollution removal annually.
Syracuse and Baltimore do not realize
gains because recent W126 index values
almost meet the alternative standards
levels.
Atlanta, Chicago, and the urban areas
of Tennessee show gains in pollution
removal. For example, urban areas of
Tennessee gain about 11,700 tons of
pollution removal annually.
Syracuse and Baltimore do not realize
gains because recent W126 index
values almost meet the alternative
standards levels.
Ecosystem Services
Cultural (Foliar Injury)
Recreation in National
Parks - Case Studies
(Section 7.4)
Recreation in National
Parks - Screening-Level
Assessment
(Section 7.3)
Rocky Mountain National Park - No
area of park exceeds 15 ppm-hrs
when adjusted to just meet the
existing standard
Great Smoky Mountains National
Park and Sequoia/Kings National
Park — No area of parks exceeds 15
ppm-hrs when adjusted to just meet
the existing standard
In screening-level assessment, of 214
parks, 3 parks remain above 7 ppm-
hrs after adjusted to 15 ppm-hrs
Rocky Mountain National Park - 59
percent of the park would be between 11
and 7 ppm-hrs when adjusted to just meet
the existing standard
Great Smoky Mountains National Park and
Sequoia/Kings National Park — No area of
parks exceeds 11 ppm-hrs when adjusted
to just meet the existing standard
In screening-level assessment, of 214
parks, 2 parks remain above 7 ppm-hrs
after adjusted to 11 ppm-hrs
Rocky Mountain National Park - 59
percent of the park would be between
11 and 7 ppm-hrs when adjusted to just
meet the existing standard
Great Smoky Mountains National Park
and Sequoia/Kings National Park — No
area of parks exceeds 7 ppm-hrs when
adjusted to just meet the existing
standard
In screening-level assessment, of 214
parks, no parks remain above 7 ppm-
hrs after adjusted to 7 ppm-hrs
8-25
-------
8.4 Representativeness
In conducting the national and case-study scale analyses of ecological effects and
resulting impacts on ecosystem services, we worked to reflect appropriate representation of
vegetation species, geographic regions, and timeframes. The following briefly discusses the
representativeness across species, geography, and time in our analyses.
8.4.1 Species Representativeness
To estimate the effect of Os exposure on biomass loss, we used data on 12 tree species
and 10 crop species. The 12 species represent a range of sensitivities normally distributed
around intermediately sensitive species. Several species are not very sensitive, two species are
relatively sensitive, and the remainder of the species represent moderately sensitive species. The
data on the 12 species facilitate representation of other species for which we do not have data.
For tree species, we used data for areas with at least one of the tree species present, resulting in
approximately 46.6 percent of the contiguous U.S. constituting the area being assessed. For 74
percent of the area being assessed, the species we know about made up 50 percent or less of total
basal area cover. For another 12 percent of the area being assessed, the species we know about
made up between 50 and 75 percent of total basal area cover. For the remaining 14 percent of
the area being assessed, the species we know about made up over 75 percent of total basal area
cover. Although we know that there are additional Os-sensitive species, we do not have E-R
functions for those species. We also used these E-R functions for the tree and crop species in
FASOMGHG, and to better employ the dynamic tradeoffs within the model, FASOMGHG
assigns proxy functions for Os exposure E-R functions for additional species. For the iTree case-
study scale analysis on carbon sequestration and pollution removal, we chose the five urban
areas based on data availability and presence of species with a W126 E-R function. No urban
areas with available vegetation data had more than three sensitive species present. Unlike
FASOMGHG, the iTree model does not provide tradeoffs between species, so the species that do
not have an E-R function were not assigned values, and thus were not part of the carbon
sequestration and pollution removal estimates. Therefore, the majority of trees in those urban
areas were not accounted for in the Os damages. For example, there are three tree species
present in these areas that we know are sensitive but for which no E-R function is available,
8-26
-------
excluding 80 to 90 percent of the total trees present in these two study areas. The species include
northern red oak in Baltimore and southern red oak and tulip tree in Atlanta.
We also qualitatively discuss many additional ecological effects and ecosystem services
for which we do not have data to assess quantitatively; those ecological effects and related
ecosystem services include supporting services such as net primary productivity; regulating
services such as hydrologic cycle and pollination; provisioning services such as commercial non-
timber forest products; and cultural services such as recreation, aesthetic and non-use values. In
addition, other ecological effects that are causally or likely causally associated with Cb exposure
are not directly addressed in this risk and exposure assessment. These ecological effects include
terrestrial productivity, water cycle, biogeochemical cycle, and community composition.19
8.4.2 Geographic Representativeness
Nine of the 12 tree species used in the biomass analyses were in the eastern U.S. and
three were in the western U.S., with a few species such as Aspen and Cottonwood in both the
eastern and western U.S. For the biomass loss analyses, by region we include the total basal area
covered by the 12 tree species assessed. In parts of the eastern U.S. - the Central, East North
Central, and Northeast regions — from less than 1 percent to 4 percent of basal area assessed had
no data on percent cover of the 12 tree species. In contrast, in parts of the western U.S. -
Southwest, West, West North Central regions — from 47 percent to 74 percent of basal area
assessed had no data on percent cover of the 12 tree species.
We applied E-R functions for 12 tree species and 10 crop species in FASOMGHG to
estimate nationwide effects on timber production, agricultural harvest, and carbon sequestration.
While we used available E-R functions for tree and crop species, as well as the available models,
we had differential and inconsistent species coverage across the U.S., e.g., data were available
for more species in the eastern U.S. than in the western U.S. In addition, to assess overall
ecosystem-level effects from biomass loss, we combined the RBL values for multiple tree
species into a weighted RBL value and considered the weighted value in relation to proportion of
basal area covered, both nationally and in Class I areas.
19 For additional details on these other ecological effects, see Table 2-4 of the Integrated Science Assessment for
Ozone and Related Photochemical Oxidants (U.S. EPA, 2013).
8-27
-------
Also, in estimating the effect of Ch exposure on biomass loss and resulting changes in
carbon sequestration and pollution removal capacity, for case-study scale analysis we used the
iTree model and data from five urban areas. The urban areas represent diverse geography in the
Northeast, Southeast, and Central regions, but we did not assess an urban area in the western part
of the U.S. Based on the monitored data from 2006 to 2008, Atlanta, Baltimore, and the urban
areas in Tennessee are over 20 ppm-hrs, with Atlanta having the highest W126 index value.
After adjusting to just meet the existing standard, all of the urban areas are between 5 and 7 ppm-
hrs. Because there are more monitors in urban areas in the eastern U.S., we focused on urban
areas in the eastern U.S. for the case-study analyses.
For the national-scale foliar injury analysis, the biosite data covered most of the
contiguous U.S., with less coverage in the Southwest, West and West North Central regions. In
assessing foliar injury at parks, we conducted a screening-level assessment, as well as a case-
study scale analysis of national parks. In assessing foliar injury at the case-study scale, the three
national parks represent diverse geographic areas — in the Southeast/Central (GRSM), the
Southwest (ROMO), and the West (SEKI). In the screening-level assessment of foliar injury, we
included 214 parks, which reflects nearly all of the parks managed by the National Park Service
(NFS) in the contiguous U.S.
8.4.3 Temporal Representativeness
For the national-scale analyses of foliar injury, the national-scale surfaces used
represented the individual years of 2006 through 2010. Monitored Os index values in those years
vary considerably, and those years represent a reasonable range of meteorological conditions that
affect Os formation. The period also includes years with varying categories of soil moisture,
which impacts the sensitivity of plants to foliar injury.
The biomass loss analysis relied upon the national-scale air quality surfaces adjusted to
just meet the existing and alternative standards for 2006 to 2008 (three-year average). Because
the forestry and agriculture sectors are interlinked and factors affecting one sector can lead to
changes in the other, we considered overall effects on producers and consumers associated with
just meeting alternative W126 standard levels over time and across sectors. In estimating the
effect of Os exposure on biomass loss and ecosystem services, we used the E-R functions for 12
tree seedlings to estimate relative yield changes over the entire lifespan of the trees, including
8-28
-------
percentage changes in national timber product market prices through 2040. At the national scale,
we estimated changes in carbon sequestration by forests and agriculture through 2040. At the
case-study scale, we estimated changes in carbon sequestration and pollution removal capacity in
the five urban areas over a 25-year period.
8.5 Overall Confidence in Welfare Exposure and Risk Results
There are several important factors to consider when evaluating the overall confidence
we can express about the estimates of exposures and risks associated with just meeting the
existing and potential alternative W126 secondary standards. Foremost, we must consider the
strength of the underlying body of scientific evidence. In addition, as with any complex analysis
using estimated parameters and inputs from numerous data sources and models, there are many
sources of uncertainty that may affect estimated results. Despite these uncertainties, the overall
body of scientific evidence underlying the ecological effects and associated ecosystem services
evaluated in this assessment is strong, and the methods used to quantify associated risks are
scientifically sound.
The overall effect of the combined set of uncertainties on confidence in the interpretation
of the results of the analyses is difficult to quantify. Due to differences in available information,
the degree to which each analysis was able to incorporate quantitative assessments of uncertainty
differed. In general, we followed the WHO tiered approach to uncertainty characterization,
which includes both quantitative and qualitative assessments. Chapters 4, 5, 6, and 7 include
tables identifying and characterizing the potential impact of key uncertainties on risk estimates,
including the degree to which we were able to quantitatively address those uncertainties. Below
we summarize several key limitations and uncertainties, but these uncertainties do not change
our conclusions regarding overall confidence and confidence in the individual analyses.
8.5.1 Confidence and Key Uncertainties in Air Quality Analyses
Because the W126 estimates generated in the air quality analyses are inputs to the
vegetation risk analyses for biomass loss and foliar injury, our confidence and any uncertainties
in these analyses are propagated into those subsequent analyses. The national W126 surface was
created using spatial interpolation techniques that perform better in areas where the Os
monitoring network is denser. Therefore, we have high confidence in the W126 estimated in
much of the contiguous U.S., and somewhat lower confidence in the rural areas in the West,
8-29
-------
Northwest, Southwest, and West North Central with few or no monitors. A potential source of
bias comes from the adjustment methodology, which used across-the-board NOx emissions cuts
and could mean that exposure in some areas could be slightly underestimated. However, this
approach is reasonable given implementation of EPA regulations such as the Clean Air Interstate
Rule (CAIR) and mobile source rules, both of which will lead to reductions in NOx emissions
from these sources across broad regions of the country in the near future.
8.5.2 Confidence and Key Uncertainties in Biomass Loss Analyses
The scientific evidence suggests that there are additional species adversely affected by
Os-related biomass loss beyond the 12 tree species and 10 crop species with available E-R
functions. This absence of information for additional species likely underestimates the Os-
related biomass loss impacts in trees and crops. The overall confidence in the E-R functions is
high, but varies by species based on the number of studies available for that species. Some
species have low within-species variability (e.g., many agricultural crops) and high
seedling/adult comparability (e.g., Aspen), while other species do not (e.g., Black Cherry). In
the national-scale analyses of agriculture and timber production, we may underestimate impacts
because FASOMGHG does not include agriculture and forestry on public lands, changes in
exports due to Os into international trade projections, or forest adaptation. In the case study
analyses of five urban areas, iTree does not account for the potential additional VOC emissions
from tree growth, which could contribute to Os formation that might somewhat offset the
estimated impacts.
8.5.3 Confidence and Key Uncertainties in Visible Foliar Injury Analyses
Based on the available evidence, we cannot identify a clear threshold for drought below
which visible foliar injury would not occur. On balance, we believe that the spatial and temporal
resolution for the soil moisture data used in the analyses is likely to underestimate the potential
of foliar injury that could occur in some areas. In general, we have high confidence in biosite
injury data in most of the country, but we acknowledge limited biosite data in a few regions,
which affects the benchmarks applied to these regions in the park screening-level analysis. In
general, we have very high confidence in the park mapping supplied by NFS, but there are
potential uncertainties related to the mapping of potential foliar injury, such as park boundaries,
vegetation species cover, and park amenities, such as scenic overlooks and trails.
8-30
-------
8.6 Conclusions
This welfare risk and exposure assessment provides information to further inform the
following policy-relevant questions20: (1) in considering alternative standards, to what extent do
alternative standard levels, averaging times, and forms reduce estimated exposures and welfare
risks attributable to Cb; (2) what range of alternative standard levels should be considered based
on the scientific information evaluated in the ISA, air quality analyses, and the welfare risk and
exposure assessment; and (3) what are the important uncertainties and limitations in the evidence
and assessments and how might those uncertainties and limitations be taken into consideration in
identifying alternative secondary standards for consideration. To develop information to help
inform these questions, we quantified ecological effects of biomass loss and visible foliar injury
based on the relationship with the W126 metric and assessed the associated impacts on
ecosystem services. For some ecosystem services, such as commercial non-timber forest
products, recreation, and aesthetic and non-use values, we qualitatively assessed potential
impacts to services. We assessed impacts on ecosystem services at the national and case-study
scales, as well as across species, U.S. geographic regions and future years.
In conclusion, we estimated that exposures and risks remain after just meeting the
existing standard and that in many cases, just meeting alternative standard levels results in
reductions in those remaining exposures and risks. Overall, the largest reduction in Os exposure-
related welfare risk occurs when moving from recent ambient conditions to meeting the existing
secondary standard of 75 ppb (equal to the existing primary standard). This finding should be
considered in the context of potential uncertainties in the actual responsiveness of W126 values
in all areas to the emissions reductions used in the adjustments to just meet the existing standard.
When using monitored W126 index values (three-year) and adjusting for meeting the existing Os
standard of 75 ppb, only two of the nine U.S. regions remain above 15 ppm-hrs (West - 18.9
ppm-hrs and Southwest - 17.7 ppm-hrs). Four regions (East North Central, Northeast,
Northwest, and South) would meet 7 ppm-hrs, and two regions (Southeast and West North
Central) are between 9 and 12 ppm-hrs (Southeast - 11.9 ppm-hrs and West North Central - 9.3
ppm-hrs). When adjusting to just meet the existing standard, the Central region would meet an
alternative W126 of 15 ppm-hrs, but further air quality adjustment would be needed for the
20 The policy-relevant questions were identified in the Integrated Review Plan for the Ozone National Ambient Air
Quality Standards (IRP, US EPA, 201 la).
8-31
-------
Central region to meet alternative standards of 11 and 7 ppm-hrs - alternate standard levels that
would protect against the recommended one to two percent biomass loss for trees and five
percent for crops. Keeping in mind the potential uncertainties associated with the actual
responsiveness of W126 values to the emissions reductions used in the adjustments to just meet
the existing standard, at an alternative W126 standard level of 15 ppm-hrs, ambient conditions
and related risk are not appreciably different than they are after just meeting the existing standard
of 75 ppb. Meeting alternative standard levels of 11 ppm-hrs and 7 ppm-hrs results in smaller
risk reductions compared to the decreases in risk from meeting the existing standard relative to
recent conditions.
8-32
-------
9 REFERENCES
Acharya, G., and L.L. Bennett. (2001). Valuing Open Space and Land-Use Patterns in Urban
Watersheds. Journal of Real Estate Finance and Economics 22(2/3):221-23 7.
Adams, D.; Alig, R.; McCarl, B.A.; Murray, B.C. (2005). FASOMGHG Conceptual Structure,
and Specification: Documentation. Available at
http://agecon2.tamu.edu/people/faculty/mccarl-bruce/FASOM.html
Alexander, S.J.; Oswalt, S.N.; Emery, M.R. (2011). Nontimber forest products in the United
States: Montreal Process indicators as measures of current conditions and sustainability.
Gen. Tech. Rep. PNW-GTR-852. Portland, OR: U.S. Department of Agriculture, Forest
Service, pacific Northwest Research Station. 36 p. Available at
http://www.fs.fed.us/pnw/pubs/pnw gtr851 .pdf
Alexander, S.J.; Weigand, J.; Blatner, K.A. (2002). Nontimber forest product commerce in
Nontimber forest products of the United States, Jones, E.T.; McLain, R.J.; Weigand, J.,
eds. University of Kansas Press.
Arbaugh, M; Bytnerowicz, A; Grulke, N; Fenn, M; Poth, M; Temple, P; Miller, P. (2003).
Photochemical smog effects in mixed conifer forests along a natural gradient of ozone
and nitrogen deposition in the San Bernardino Mountains. Environ Int 29: 401-406.
Arbaugh, MJ; Miller, PR; Carroll, JJ; Takemoto, BL; Proctor, T. (1998). Relationships of ozone
exposure to pine injury in the Sierra Nevada and San Bernardino Mountains of
California, USA. Environ Pollut 101: 291-301.
Baumflek, M.J.; Emery, M.R.; Ginger, C. (2010). Culturally and economically important
nontimber forest products of northern Maine. Gen.Tech. Rep. NRS-68. Newton Square,
PA: Y.s. Department of Agriculture, Forest Service, Northern Research Station. 74p.
Available at
http://nrs.fs.fed.us/sustaining_forests/conserve_enhance/special_products/maine_ntfp/
Benoit, LF; Skelly, JM; Moore, LD; Dochinger, LS. (1982). Radial growth reductions of Pinus
strobus L correlated with foliar ozone sensitivity as an indicator of ozone-induced losses
in eastern forests. Can J For Res 12: 673-678.
Berrocal, V.J., Gelfand, A.E., Holland, D.M. (2012). Space-Time Data Fusion Under Error in
Computer Model Output: An Application to Modeling Air Quality. Biometrics, 68(3),
837-848.
Black, VJ; Black, CR; Roberts, JA; Stewart, CA. (2000). Impact of ozone on the reproductive
development of plants. New Phytol 147: 421-447.
9-1
-------
Bortier, K.;De Temmerman, L.; Ceulemans, R. (2000). Effects of ozone exposure in open-top
chambers on poplar (Populus nigra) and beech (Fagus sylvaticd): a comparison. Environ
Pollutl09: 509-516.
Brown, R.B.; Xu, X.; Toth, J.F. Jr. (1998). Lifestyle options and economic strategies:
susbsistence activities in the Mississippi Delta. Rural Sociology 63-4: p599-623.
Brown, T. C., and T. C. Daniel. (1984). Modeling forest scenic beauty: Concepts and application
to ponderosa pine. USDA Forest Service Research Paper RM-256. Rocky Mountain
Forest and Range Experiment Station, Fort Collins, Colorado. Brown, T.C., Daniel, T.C.,
1986. Predicting scenic beauty of forest timber stands. Forest Sci. 32, 471-487.
Brown, T.C., and T.C. Daniel. (1986). Predicting scenic beauty of timber stands. Forest Science
32:471-487.
Burns, Russell M., and Barbara H. Honkala, tech. coords. (1990). Silvics of North America: 1.
Conifers; 2. Hardwoods. Agriculture Handbook 654. U.S. Department of Agriculture,
Forest Service, Washington, DC. vol.2, 877 p.
Butler, BJ. (2008). Family forest owners of the United States. (2006). Gen. Tech. Rep. NRS-
GTR-27. Newton Square, PA: U.S. Department of Agriculture, Forest Service, Northern
Research Station. 72 p.
Buyhoff, G.J.; Wellman, J.D.; Daniel, T.C. (1982). Predicting scenic quality for mountain pine
beetle and western spruce budworm damaged forest vistas. Forest Science 28:827-838.
Byun, D.W. and J.K.S. Ching. (1999). Science algorithms of the EPA Models-3 Community
Multiscale Air Qualty (CMAQ) Modeling System, (EPA/600/R-99/030 0). U.S. EPA.
Byun, D.W. and K.L. Schere. (2006). Review of the governing equations, computational
algorithms, and other components of the Models-3 Community Multiscale Air Quality
(CMAQ) modeling system. Applied Mechanics Reviews, 59. 51-77'.
CAL FIRE (California Department of Forestry and Fire Protection). (1996). California Fire Plan.
Available at http://cdfdata.fire.ca.gov/fire er/fpp_planning cafireplan
CAL FIRE (California Department of Forestry and Fire Protection). (2008). CAL FIRE 2007
Wildland Fire Summary. Available at
www.fire.ca.gov/communications/downloads/fact sheets/2007 Summary.pdf.
Carlton, A.G., R.W. Finder, P.V. Bhave and G.A. Pouliot. (2010). To What Extent Can Biogenic
SOA be Controlled, Environmental Science & Technology 44:3376-3380.
9-2
-------
Chappelka, A; Skelly, J; Somers, G; Renfro, J; Hildebrand, E. (1999a). Mature black cherry used
as a bioindicator of ozone injury. Water Air Soil Pollut 116: 261-266.
Chappelka, A; Somers, G; Renfro, J. (1999b). Visible ozone injury on forest trees in Great
Smoky Mountains National Park, USA. Water Air Soil Pollut 116: 255-260.
Chappelka, AH. (2002). Reproductive development of blackberry (Rubus cuneifolius) as
influenced by ozone. NewPhytol 155: 249-255.
Chen, J., Zhao, R., Li, Z. (2004). Voronoi-based k-order neighbor relations for spatial analysis.
ISPRS J Photogrammetry Remote Sensing, 59(1-2), 60-72.
Constable, J. V. H. and G. E. Taylor. (1997). Modeling the effects of elevated tropospheric Os on
two varieties of Pinus ponderosa. Canadian Journal of Forest Research 27:527-53
Coulson, R.N.; Klepzig, K.D. (2011). Southern Pine Beetle II. Gen. Tech. Rep. SRS-140.
Asheville, NC: U.S. Department of Agriculture Forest Service, Southern Research
Station, 512p. Available at http://www.srs.fs.usda.gov/pubs/gtr/gtr_srs 140/gtr_srs 140.pdf
Daniel, T.C.; Brown, T.C.; King, D.A.; Richards, M.T.; Stewart, W.P. (1989). Perceived scenic
beauty and contingent valuation of forest campgrounds. Forest Science 35:76-90.
Darbah, JNT; Kubiske, ME; Neilson, N; Oksanen, E; Vaapavuori, E; Karnosky, DF. (2007).
Impacts of elevated atmospheric CCh and Os on paper birch (Betula papyrifera):
Reproductive fitness. ScientificWorldJournal 7: 240-246.
Darbah, JNT; Kubiske, ME; Nelson, N; Oksanen, E; Vapaavuori, E; Karnosky, DF. (2008).
Effects of decadal exposure to interacting elevated CCh and/or Os on paper birch (Betula
papyrifera) reproduction. Environ Pollut 155: 446-452.
Dickson, RE; Lewin, KF; Isebrands, JG; Coleman, MD; Heilman, WE; Riemenschneider, DE;
Sober, J; Host, GE; Zak, DR; Hendrey, GR; Pregitzer, KS; Karnosky, DF. (2000). Forest
Atmosphere Carbon Transfer and Storage (FACTS-II) the Aspen Free-Air CO2 and O3
Enrichment (FACE) project: An overview. (General Technical Report NC-214). St. Paul,
MN: U.S. Dept. of Agriculture, Forest Service.
Emery, M.R.; Ginger, C.; Newman, S.; Giammusso, M.R.B. (2003). Special forest products in
context: gatherers and gathering in the eastern United States., USDA Forest Service,
Northeastern Research Station, Newton Square, PA. Available at
http://www.fs.fed.us/ne/newtown_square/publications/technical_reports/pdfs/2003/gtrne3
06.pdf
9-3
-------
Emery, M.R.; Pierce, A.R. (2005). Interrupting the telos: locating subsistence in contemporary
U.S. forests. Environment and Planning, 37: p. 981-993.
ESRI. (2011). ArcGIS Desktop: Release 10. Redlands, CA: Environmental Systems Research
Institute.
Felzer, B; Kicklighter, D; Melillo, J; Wang, C; Xhuang, Q; Prinn, R. (2004). Effects of ozone on
net primary production and carbon sequestration in the conterminous United States using
a biogeochemistry model. Tellus B Chem Phys Meteorol 56: 230-248.
Freeman, Milton R. (1993). "The International Whaling Commission, Small-type Whaling, and
Coming to Terms with Subsistence." Human Organization 52:243-51.
Frey, H.C. and Samet, J.M. (2012a). CASAC Review of the EPA's Health Risk and Exposure
Assessment for Ozone (First External Review Draft - Updated August 2012) and Welfare
Risk and Exposure Assessment for Ozone (First External Review Draft - Updated August
2012). EPA-CASAC-13-002. Available at
http://vosemite.epa.gov/sab/sabproduct.nsf/264cbl227d55e02c85257402007446a4/38CO
lBlEDD14C59E85257ABC004AOE18/$File/EPA-CASAC-13-002+unsigned.pdf
Frey, H.C. and Samet, J.M. (2012b). CASAC Review of the EPA's Policy Assessment for the
Review of the Ozone National Ambient Air Quality Standards (First External Review
Draft- August 2012). ((PDF, 73 pp., 355,203 bytes). EPA-CASAC-13-003. Available at
http://vosemite.epa.gov/sab/sabproduct.nsf/264cbl227d55e02c85257402007446a4/E670
94C7FBBECD8685257AC200727082/$File/EPA-CASAC-13-003+unsigned.pdf
Frey, H.C. (2014). CASAC Review of the EPA's Welfare Risk and Exposure Assessment for
Ozone (Second External Review Draft). ((PDF, 46 pp., 252,241 bytes). EPA-CASAC-14-
003. Available at
http://yosemite.epa.gOv/sab/sabproduct.nsf/0/cd07f81a4105139d85257ab60056d531IOpe
nDocument&TableRow=2.3#2.
Gallai, N; Salles, JM; Settele, J.;Vaissiere, B.E. (2009). Economic Valuation of the vulnerability
of world agriculture confronted with pollinator decline. Environmental Economics
68:810-821.
Geoghegan, J., L.A. Wainger, andN.E. Bockstael. (1997). "Spatial Landscape Indices in a
Hedonic Framework: An Ecological Economics Analysis Using GIS." Ecological
Economics 23:251-264.
Gold, C. (1997). Voronoi methods in GIS. In: Algorithmic Foundation of Geographic
Information Systems (va Kereveld M., Nievergelt, J., Roos, T., Widmayer, P., eds).
Lecture Notes in Computer Sicnece, Vol 1340. Berlin: Springer-Verlag, 21-35.
9-4
-------
Gregg, JW; Jones, CG; Dawson, TE. (2006). Physiological and developmental effects of O3 on
cottonwood growth in urban and rural sites. Ecol Appl 16: 2368-2381.
Grulke, N. E.; Minnich, R.A.; Paine, T.D.; Seybold, S.J.; Chavez, D.J.; Fenn, M.E.; Riggan, P.J.;
Dunn, A. (2009). Air pollution increases forest susceptibility to wildfires: a case study in
the San Bernardino Mountains in southern California in Developments in Environmental
Science, Volume 8. A. Bytnerowicz, M. Arbaugh, A. Riebau, and C. Anderson (eds.)
Guenther, A.; Karl, T.; Harley, P.; Wiedinmyer, C.; Palmer, P. I; Geron, C. (2006). Estimates of
global terrestrial isoprene emissions using MEGAN (model of emissions of gases and
aerosols from nature). Atmos. Chem. Phys. 6, 3181-3210.
Gumpertz, ML; Rawlings, JO. (1992). Nonlinear regression with variance components:
Modeling effects of ozone on crop yield. Crop Sci 32: 219-224.
Haefele, M., R.A. Kramer, and T.P. Holmes. (1991). Estimating the Total Value of a Forest
Quality in High-Elevation Spruce-Fir Forests. The Economic Value of Wilderness:
Proceedings of the Conference. Gen. Tech. Rep. SE-78 (pp. 91-96). Southeastern For
Exper. Station. Asheville, NC: USD A Forest Service.
Hall, Karen R. and Richard R. Braham. (1998). Native Pines of Eastern North America. North
Carolina State University Department of Forestry. Available at
http://www.ncsu.edu/proj ect/dendrology/
Heck, WW; Cure, WW; Rawlings, JO; Zaragoza, LJ; Heagle, AS; Heggestad, HE; Kohut, RJ;
Kress, LW; Temple, PJ. (1984). Assessing impacts of ozone on agricultural crops: II.
Crop yield functions and alternative exposure statistics. J Air Pollut Control Assoc 34:
810-817.
Henderson, R. (2006a). Letter from CASAC Chairman Rogene Henderson to EPA Administrator
Stephen Johnson, February 16, 2006, EPA-CASAC-06-003. Available at
http://vosemite.epa.gov/sab/sabproduct.nsf/69FBBlE21FBlE4428525712D004BA05D/$
File/casac con 06 003 .pdf
Henderson, R. (2006b). Letter from CASAC Chairman Rogene Henderson to EPA Administrator
Stephen Johnson, June 5, 2006, EPA-CASAC-06-007. Available at
http://vosemite.epa.gov/sab/sabproduct.nsf/0202D7053AC6E2AC852571870075ClD2/$
File/casac-06-007.pdf
Henderson, R. (2007). Clean Air Scientific Advisory Committee's (CASAC) Review of the
Agency's Final Ozone Staff Paper. EPA-CASAC-07-002. March 26. Available at
http://vosemite.epa.gov/sab/sabproduct.nsf/FE915E916333D776852572AC007397B5/$F
ile/casac-07-002.pdf.
9-5
-------
Hildebrand, E; Skelly, JM; Fredericksen, TS. (1996). Foliar response of ozone-sensitive
hardwood tree species from 1991 to 1993 in the Shenandoah National Park, Virginia. Can
J For Res 26: 658-669.
Hogsett, WE; Weber, JE; Tingey, D; Herstrom, A; Lee, EH; Laurence, JA. (1997).
Environmental auditing: An approach for characterizing tropospheric ozone risk to
forests. J Environ Manage 21: 105-120.
Holmes, T., and R. Kramer. (1995). "An Independent Sample Test of Yea-Saying and
StartingPoint Bias in Dichotomous-Choice Contingent Valuation." Journal of
Environmental Economics and Management 28:121-132.
Hufford, M. (2000). Holding up the mountains: talk as a historical discourse. Smithsonian
Folklife Center, Library of Congress, Washington D.C.
IPCC (Intergovernmental Panel on Climate Change). (2007). Summary for policymakers. In: The
Physical Science Basis. Contribution of Working Group I to the Fourth Assessment
Report of the Intergovernmental Panel on Climate Change. Cambridge, United Kingdom
and New York, NY, USA: Cambridge University Press. Available at
http://www.ipcc.ch/publications_and_data/ar4/wg2/en/spm.html
Irwin, E.G. (2002). "The Effects of Open Space on Residential Property Values." Land
Economics 78(4):465-480.
Jaegle, L., DJ. Jacob, W.H. Brune, and P.O. Wenberg. (2001). Chemistry of HOx radicals in the
upper troposphere. Atmos. Environ., 35, 469-489.
Karl, T.R. (1986). The sensitivity of the Palmer Drought Severity Index and Palmer's Z-Index to
their calibration coefficients including potential evapotranspiration. J. Climate Appl.
Meteor., 25, 77-86.
Karl, T.R. andKoss, W.J. (1984). Regional and National Monthly, Seasonal, and Annual
Temperature Weighted by Area, 1895-1983. Historical Climatology Series 4-3, National
Climatic Data Center, Asheville, NC, 38 pp.
Karnosky, DF; Gagnon, ZE; Dickson, RE; Coleman, MD; Lee, EH; Isebrands, JG. (1996).
Changes in growth, leaf abscission, biomass associated with seasonal tropospheric ozone
exposures of Populus tremuloides clones and seedlings. Can J For Res 26: 23-37.
Karnosky, DF; Mankovska, B; Percy, K; Dickson, RE; Podila, GK; Sober, J; Noormets, A;
Hendrey, G; Colman, MD; Kubiske, M; Pregitzer, KS; Isebrands, JG. (1999). Effects of
tropospheric Os on trembling aspen and interaction with CO2: Results from and Os-
gradient and FACE experiment. Water, Air, and Soil Pollution 116:311-322.
9-6
-------
Kaval, P. and J. Loomis. (2003). Updated outdoor recreation use values with emphasis on
National Park recreation. Final Report, Cooperative Agreement 1200-99-009, Project
number IMDE-02-0070. Fort Collins, CO: Colorado State University, Department of
Agricultural and Resource Economics.
Kohut, R. (2007). Assessing the risk of foliar injury from ozone on vegetation in parks in the US
National Park Service's Vital Signs Network. Environ Pollut 149: 348-357.
Krupa, S.; Muntifering, R.B.; Chappelka, A.H. (2004). Effects of ozone on plant nutritive
quality characteristics for ruminant animals. Botanica 54, 129-140.
Kubiske, ME; Quinn, VS; Heilman, WE; McDonald, EP; Marquardt, PE; Teclaw, RM; Friend,
AL; Karnoskey, DF. (2006). Interannual climatic variation mediates elevated CCh and Os
effects on forest growth. Global Change Biol 12: 1054-1068.
Kubiske, ME; Quinn, VS; Marquardt, PE; Karnosky, DF. (2007). Effects of elevated
atmospheric CCh and/or Cb on intra- and interspecific competitive ability of aspen. Plant
Biol (Stuttg) 9: 342-355.
Lee, EH; Hogsett, WE. (1999). Role of concentrations and time of day in developing ozone
exposure indices for a secondary standard. J Air Waste Manag Assoc 49: 669-681.
Lee, EH; Hogsett, WE; Tingey, DT. (1994). Attainment and effects issues regarding alternative
secondary ozone air quality standards. J Environ Qual 23: 1129-1140.
Lee, EH; Tingey, DT; Hogsett, WE. (1987). Selection of the best exposure-response model using
various 7-hour ozone exposure statistics. Research Triangle Park, NC: U.S.
Environmental Protection Agency.
Lee, EH; Tingey, DT; Hogsett, WE. (1988b). Evaluation of ozone exposure indices in exposure-
response modeling. Environ Pollut 53: 43-62.
Lee, EH; Tingey, DT; Hogsett, WE. (1989). Interrelation of experimental exposure and ambient
air quality data for comparison of ozone exposure indices and estimating agricultural
losses. (EPA/600/3-89/047). Corvallis, OR: U.S. Environmental Protection Agency.
Lefohn, A. S., Laurence, J. A., & Kohut, R. J. (1988). A comparison of indices that describe the
relationship between exposure to ozone and reduction in the yield of agricultural crops.
Atmospheric Environment (1967), 22(6), 1229-1240.
Lehrer, J.A., M. Bacou, B. Blankespoor, D. McCubbin, J. Sacks, C.R. Taylor, and D.A.
Weinstein. (2007). Technical Report on Ozone Exposure, Risk, and Impact Assessments
for Vegetation. EPA 452/R-07-002.
9-7
-------
Lesser, VM; Rawlings, JO; Spruill, SE; Somerville, MC. (1990). Ozone effects on agricultural
crops: Statistical methodologies and estimated dose-response relationships. Crop Sci 30:
148-155.
Little, E.L., Jr. (1971). Atlas of United States trees, volume 1, conifers and important hardwoods:
Misc. Pub. 1146. Washington, D.C.: U.S. Department of Agriculture. 9 p., 200 maps.
Little, E.L., Jr. (1976). Atlas of United States Trees, Volume 3, Minor Western Hardwoods. U.S.
Department of Agriculture Miscellaneous Publication 1314, 13 p., 290 maps.
Little, E.L., Jr. (1977). Atlas of United States Trees, Volume 4, Minor Eastern Hardwoods. U.S.
Department of Agriculture Miscellaneous Publication 1342, 17 p., 230 maps.
Little, E.L., Jr. (1978). Atlas of United States Trees, Volume 5, Florida. U.S. Department of
Agriculture Miscellaneous Publication 1361, 262 maps.
Mansfield, C.A., S.K. Pattanayak, W. McDow, R. MacDonald, and P. Halpin. (2005). Shades of
Green: Measuring the Value of Urban Forests in the Housing Market. Journal of Forest
Economics 11(3): 177-199.
McLaughlin, SB; Nosal, M; Wullschleger, SD; Sun, G. (2007a). Interactive effects of ozone and
climate on tree growth and water use in a southern Appalachian forest in the USA. New
Phytol 174: 109-124.
MEA (Millennium Ecosystem Assessment Board). (2005). Ecosystems and Human Well-being:
Current State and Trends, Volume 1. Edited by R. Hassan, R. Scholes, and N. Ash.
Washington: Island Press. Available at http://www.millenniumassessment.Org/8
documents/document. 766. aspx. pdf
Minnesota EVIPLAN Group (MIG, Inc). (1999). EVIPLAN System (data and software), 502 2nd
Street, Suite 301, Hudson, WI 54016. Available at www.implan.com
Morgan, PB; Bernacchi, CJ; Ort, DR; Long, SP. (2004). An in vivo analysis of the effect of
season-long open-air elevation of ozone to anticipated 2050 levels on photosynthesis in
soybean. J Plant Physiol 135: 2348-2357.
Morgan, PB; Mies, TA; Bollero, GA; Nelson, RL; Long, SP. (2006). Season-long elevation of
ozone concentration to projected 2050 levels under fully open-air conditions substantially
decreases the growth and production of soybean. New Phytol 170: 333-343.
9-8
-------
Mueller, J., J. Loomis, and A. Gonzalez-Caban. (2007). Do Repeated Wildfires Change
Homebuyers' Demand for Homes in High-Risk Areas? A Hedonic Analysis of the Short
and Long-Term Effects of Repeated Wildfires on House Prices in Southern California.
Journal of Real Estate Finance and Economics, 1-18.
National Association of State Foresters (NASF). (2009). Quadrennial Fire Review 2009.
Washington, DC: NASF. Quadrennial Fire and Fuel Review Final Report 2009.National
Wildfire Coordinating Group Executive Board January 2009.
National Climatic Data Center (NCDC). (2012a).U.S. Climatological Divisions. National
Oceanic and Atmospheric Administration. Available at http://www.ncdc.noaa.gov/temp-
and-precip/us-climate-divisions.ph p
National Climatic Data Center (NCDC). (2012b). Historical Palmer Drought Indices. National
Oceanic and Atmospheric Administration. Available at
http://wwwl.ncdc.noaa.gov/pub/data/cirs/
National Climatic Data Center (NCDC). (2012c). Drought. National Oceanic and Atmospheric
Administration. Available at http://www.ncdc.noaa.gov/sotc/drought/
National Interagency Fire Center (NIFC). (2012). Statistics. Available at
http://www.nifc.gov/firelnfo/firelnfo statistics.html
National Park Service (NPS). (2002a). Comprehensive Survey of the American Public, Southeast
Region Technical Report. Available at
http://www.nature.nps.gov/socialscience/docs/archive/SER Tech Rep.pdf
National Park Service (NPS). (2002b). Comprehensive Survey of the American Public,
Intermountain Region Technical Report. Available at
http://www.nature.nps.gov/socialscience/docs/archive/IMR_Tech_Rep.pdf
National Park Service (NPS). (2002c). Comprehensive Survey of the American Public Pacific
West Region Technical Report. Available at
http://www.nature.nps.gov/socialscience/docs/archive/PWR Tech Rep.pdf
National Park Service (NPS). (2003). Ozone Sensitive Plant Species on National Park Service
and U.S. Fish and Wildlife Service Lands: Results of a June 24-25, 2003 Workshop.
Baltimore, MD. Available at http://www.nature.nps.gov/air/pubs/pdf/baltfmalreportl.pdf
National Park Service (NPS). (2006b). Ozone Sensitive Plant Species, by Park, November 2006.
Available at
http://www.nature.nps.gov/air/Permits/ARIS/docs/Ozone Sensitive BvPark 3600.pdf
9-9
-------
National Park Service (NFS). (2010). Visitor Use Statistics: Annual Recreation Visitation by
Park Type or Region for: 2010. Available at https://irma.nps.gov
National Park Service (NPS). (2011). Economic benefits to Local Communities from National
Park Visitation and Payroll, 2010. Natural Resources Report NPS/NRSS/EGD/NRR-
2011/481. Available at
http://www.nature.nps.gov/socialscience/docs/NPSSystemEstimates2009.pdf
National Park Service. (NPS). (2006). Management Policies 2006. Available at
http://www.nps.gov/policy/MP2006.pdf
Novak, K.; Cherubini, P.; Saurer, M.; Fuhrer, J; Skelly, J.M.; Krauchi, N.; Schaub, M. (2007).
Ozone air pollution effects on tree-ring growth, 513 C, visible foliar injury and leaf gas
exchange in three ozone-sensitive woody plant species. Tree Physiol 27: 941-949.
Oregon State University, College of Forestry. Recreation Use Values Database. Available at
http: //recvaluati on. fore stry. oregonstate. edu/
Outdoor Industry Foundation (OIF). (2012). The Economic Contribution of Active Outdoor
Recreation - Technical Report on Methods and Findings. Available at
http://www.outdoorindustry.org/images/ore_reports/outdoorrecreationeconomy-
technicalreport2012-oia.pdf
Palmer, W.C. (1965). Meteorological drought. Research Paper No. 45. U.S. Weather Bureau.
NOAA Library and Information Services Division. Washington, D.C. Available at http://
ncdc.noaa.gov/temp-and-precip/drought/docs/palmer.pdf.
Payton, R. (2007). Effects of Wildfire Smoke on UV Ozone Instruments, Regional Applied
Research Effort, U.S. EPA Region 8, October 2007.
Pena, D.G. (1999). Cultural landscapes and biodiversity: the ethnoecology of an Upper Rio
Grandewatershed commons in Ethnoecology: Situated Knowledge/Located Lives Ed.
V.D. Nazarea, University of Arizona Press, Tucson, AZ, pp 107-132.
Peterson, DL; Arbaugh, MJ; Wakefield, VA; Miller, PR. (1987). Evidence of growth reduction
in ozone-injured Jeffrey pine (Pinus jeffreyi Grev and Balf) in Sequoia and Kings
Canyon National Parks. J Air Waste Manag Assoc 37: 906-912.
Pregitzer, KS; Burton, AJ; King, JS; Zak, DR. (2008). Soil respiration, root biomass, and root
turnover following long-term exposure of northern forests to elevated atmospheric Co-2
andtroposphericO-3.NewPhytol 180: 153-161.
9-10
-------
Rawlings, JO; Cure, WW. (1985). The Weibull function as a dose-response model to describe
ozone effects on crop yields. Crop Sci 25: 807-814.
Rea et al, (2012). Using Ecosystem Services To Inform Decisions on U.S. Air Quality Standards.
Environ. Sci. Technol.46 (12), 6481-6488.
Ribe, R. G. (1989). The Aesthetics of Forestry: What has empirical preference research taught
us? Environmental Management 13: 55-74.
Ribe, R.G. (1994). Scenic Beauty along the ROS. Journal of Environmental Research 42:199-
221.
Richardson, L.; Loomis, J. (2009). The total economic value of threatened, endangered and rare
specis: and updated meta-analysis. Ecological Economics, 68:1535-1548.
Rizzo, M. (2005). A Comparison of Different Rollback Methodologies Applied to Ozone
Concentrations. November 7, 2005. Available at
http://www.epa.gov/ttn/naaqs/standards/ozone/s_o3_cr_td.html
Rizzo, M. (2006). A Distributional Comparison between Different Rollback Methodologies
Applied to Ambient Ozone Concentrations. May 31, 2006. Available at
http://www.epa.gov/ttn/naaqs/standards/ozone/s_o3_cr_td.html
SAB CVPESS (Science Advisory Board Committee on Valuing the Protection of Ecological
Systems and Services). (2009). Valuing the Protection of Ecological Systems and
Services. EPA-SAB-09-012. May. Available at
http://vosemite.epa.gov/sab/sabproduct.nsf/WebBOARD/SAB-09-
012/$File/SAB%20Advisorv%20Report%20full%20web.pdf
Samet. JM. (2009). Consultation on Ambient Air Monitoring Issues Related to the Ozone
NAAQS. EPA-CASAC-09-005. March. Available at
http://vosemite.epa.gov/sab/sabproduct.nsf/64B88B99C37A68CF852575710072D8CO/$
File/EP A-CASAC-09-005-unsigned.pdf
Samet. JM. (2011). Consultation on EPA's Ozone National Ambient Air Quality Standards:
Scope and Methods Plan for Health Risk and Exposure Assessment (April 2011) and
Ozone National Ambient Air Quality Standards: Scope and Methods Plan for Welfare
Risk and Exposure Assessment (April 2011). EPA-CASAC-11-008. June. Available at
yosemite.epa.gov/sab/sabproduct.nsf/0594FCC1374FCC5D852578B60069CDB2/$File/E
P A-C AS AC-1 l-008-unsigned.pdf
Samuelson, L. J. & Edwards, G. S. (1993). A comparison of sensitivity to ozone in seedlings and
trees ofQuercus rubra L. New Phytol, 125, 373-9.
9-11
-------
Sanz, J.; Muntifering, R.B.; Bermejo, V.; Gimeno, B.S.; Elvira, S. (2005). Ozone and increased
nitrogen supply effects on the yield and nutritive quality of Trifolium subterraneum.
Atmospheric Environment 39: 5899-5907.
Schaub, M; Skelly, JM; Zhang, JW; Ferdinand, JA; Savage, JE; Stevenson, RE; Davis, DD;
Steiner, KC. (2005). Physiological and foliar symptom response in the crowns of Prunus
serotina, Fraxinus americana and Acer rubrum canopy trees to ambient ozone under
forest conditions. Environ Pollut 133: 553-567.
Sitch, S; Cox, PM; Collins, WJ; Huntingford, C. (2007). Indirect radiative forcing of climate
change through ozone effects on the land-carbon sink. Nature 448: 791-794.
Smith, G., J. Coulston, and B. O'Connell. (2008). Ozone bioindicators and forest health: A guide
to the evaluation, analysis, and interpretation of the ozone injury data in the Forest
Inventory and Analysis program. Gen. Tech. Rep. NRS-34. Newtown Square, PA: U.S.
Department of Agriculture, Forest Service, Northern Research Station. Available at
http://www.treesearch.fs.fed.us/pubs/19036.
Smith, V.K., C. Poulos, and H. Kim. (2002). Treating Open Space as an Urban Amenity.
Resource and Energy Economics 24:107-129.
Somers, GL; Chappelka, AH; Rosseau, P; Renfiro, JR. (1998). Empirical evidence of growth
decline related to visible ozone injury. For Ecol Manage 104: 129-137.
Southwick Associates. (2012). The Economic Contribution of Active Outdoor Recreation -
Technical Report on Methods and Findings For: Outdoor Industry Foundation. Available
at
http://www.outdoorfoundation.org/pdf/ResearchRecreationEconomyTechnicalReport.pdf
Taylor, C.R., K.H. Reichelderfer and S.R. Johnson. (1993). Agricultural Sector Models for the
United States: Descriptions and Selected Policy Applications. Iowa State University
Press, Ames, IA, 1993.
Timin B, Wesson K, Thurman J. Application of Model and Ambient Data Fusion Techniques to
Predict Current and Future Year PM2.5 Concentrations in Unmonitored Areas. (2010). Pp.
175-179 in Steyn DG, Rao St (eds). Air Pollution Modeling and Its Application XX.
Netherlands: Springer.
Tyrvainen, L., and A. Miettinen. (2000). "Property Prices and Urban Forest Amenities." Journal
of Economics and Environmental Management 39:205-223.
U.S. Census Bureau. 2006. County Business Patterns. Available at
http://www.census.gov/econ/cbp/
9-12
-------
U.S. Department of Agriculture (USD A). (2004). National Report on Sustainable Forests-2003
Documentation for the 2003 National Report: Indicators 35, 36, 37, 42, & 43 Available at
http://warnell.forestry.uga. edu/nrrt/NSRE/MontreallndDoc.PDF
U.S. Department of Agriculture (USD A). (2010). National Survey on Recreation and the
Environment (NSRE): 2000-2002. The Interagency National Survey Consortium,
Coordinated by the USDA Forest Service, Recreation, Wilderness, and Demographics
Trends Research Group, Athens, GA and the Human Dimensions Research Laboratory,
University of Tennessee, Knoxville, TN.
U.S. Department of Agriculture (USDA). (2011). National Report on Sustainable Forests-2010
Available at http://www.fs.fed.us/research/sustain/
U.S. Department of Agriculture National Resources Conservation Service (USDA-NRCS).
(2013). Fact Sheets & Plant Guides. Available at http://plants.usda.gov/java/factSheet
U.S. Department of the Interior, U.S. Fish and Wildlife Service, and U.S. Department of
Commerce, U.S. Census Bureau. (2012). 2011 National Survey of Fishing, Hunting, and
Wildlife-Associated Recreation. Available at
http ://www. census. gov/prod/2012pub s/fhw 11 -nat.pdf
U.S. EPA (U.S. Environmental Protection Agency). (1978). Air quality criteria for ozone and
other photochemical oxidants (pp. 373). (EPA/600/8-78/004). Washington, DC.
U.S. EPA (U.S. Environmental Protection Agency). (1984). Air quality criteria for ozone and
other photochemical oxidants, Vol. 3 (pp. 405). (EPA/600/8-84/020A). Research Triangle
Park, NC.
U.S. EPA (U.S. Environmental Protection Agency). (1996). Review of national ambient air
quality standards for ozone: Assessment of scientific and technical information: OAQPS
staff paper. (EPA/452/R-96/007). Research Triangle Park, NC.
U.S. EPA (U.S. Environmental Protection Agency). (2006). Air quality criteria for ozone and
related photochemical oxidants (pp. 2118). (EPA/600/R-05/004AF). Research Triangle
Park, NC.
U.S. EPA (U.S. Environmental Protection Agency). (2007). Review of the national ambient air
quality standards for ozone: Policy assessment of scientific and technical information:
OAQPS staff paper. (EPA/452/R-07/007). July. Research Triangle Park, NC. Available at
http://www.epa.gov/ttn/naaqs/standards/ozone/data/2007 01 ozone staff paper.pdf
9-13
-------
U.S. EPA (U.S. Environmental Protection Agency). (2009). The Risk and Exposure Assessment
for Review of the Secondary National Ambient Air Quality Standards for Oxides of
Nitrogen and Oxides of Sulfur. (EPA-452/R-09-008a). September. Available at
http://www.epa.gov/ttn/naaqs/standards/no2so2sec/data/NOxSOxREASep2009MainCont
ent.pdf
U.S. EPA (U.S. Environmental Protection Agency). (2009a). Integrated Review Plan for the
Ozone National Ambient Air Quality Standards - External Review Draft. (EPA 452/D-
09-001). September. Available at
http://www.epa.gov/ttn/naaqs/standards/ozone/data/externalreviewdraftO3IRP093 009. pdf
U.S. EPA (U.S. Environmental Protection Agency). (201 la). Integrated Review Plan for the
Ozone National Ambient Air Quality Standards - Final. (EPA 452/R-l 1-006). April.
Available at http://www.epa.gov/ttn/naaqs/standards/ozone/data/201 l_04_OzoneIRP.pdf.
U.S. EPA (U.S. Environmental Protection Agency). (201 Ib). Ozone National Ambient Air
Quality Standards: Scope and Methods Plan for Health Risk and Exposure. (EPA-452/P-
11-001). Research Triangle Park, NC. April. Available at
http://www.epa.gov/ttn/naaqs/standards/ozone/data/2011 _04_HealthRE A. pdf
U.S. EPA (U.S. Environmental Protection Agency). (201 Ic). Ozone National Ambient Air
Quality Standards: Scope and Methods Plan for Welfare Risk and Exposure. (EPA 452/P-
11-002). April. Research Triangle Park, NC. Available at
http://www.epa.gov/ttn/naaqs/standards/ozone/data/2011 04 WelfareREA.pdf
U.S. EPA (U.S. Environmental Protection Agency). (201 le). Policy Assessment for the Review
of the Secondary National Ambient Ai Quality Standards for Oxides of Nitrogen and
Oxides of Sulfur. (EPA/452/R-1 l-005a). February. Research Triangle Park, NC.
Available at
http://vosemite.epa.gOv/sab/sabproduct.nsf/0/A56303ClB3FC092B852577EC004DA8B
F/$File/NOx+SOx+PA+FINAL-main.pdf
U.S. EPA. (U.S. Environmental Protection Agency). (2013). Integrated Science Assessment of
Ozone and Related Photochemical Oxidants (Final Report). U.S. Environmental
Protection Agency, Washington, DC, EPA/600/R-10/076F. Available at
http://cfpub.epa.gov/ncea/isa/recordisplay.cfm?deid=247492#Download
U.S. EPA. (U.S. Environmental Protection Agency). (2013b). Greenhouse Gas Equivalencies
Calculator. Updated September 2013. Available at
http://www.epa.gov/cleanenergv/energy-resources/calculator.html
9-14
-------
U.S. Geological Survey (USGS). (2003). FEDLANP020 - Federal Lands and Indian
Reservations of the United States. Reston, VA. Available at
http://coastalmap.marine.usgs.gov/GISdata/basemaps/boundaries/fedlands/fedlanp020.zip
USFS (U.S. Forest Service). (2011). Forest Health Monitoring Network.
USFS (U.S. Forest Service). (2013). Personal email communication with John Coulston, USFS,
June 2013.
USFS (U.S. Forest Service). Forest Health Technology Enterprise Team. Available at
http://www.fs.fed.us/foresthealth/technology/
Vollenweider, P; Woodcock, H; Kelty, MJ; Hofer, R, -M. (2003). Reduction of stem growth and
site dependency of leaf injury in Massachusetts black cherries exhibiting ozone
symptoms. Environ Pollut 125: 467-480.
Wegman, L. (2012). Updates to information presented in the Scope and Methods Plans for the
O3 NAAQS Health and Welfare Risk and Exposure Assessments. Memorandum from
Lydia Wegman, Division Director, Health and Environmental Impacts Division, Office
of Air Quality Planning and Standards, Office of Air and Radiation, US EPA to Holly
Stallworth, Designated Federal Officer, Clean Air Scientific Advisory Committee, US
EPA Science Advisory Board Staff Office. May 2, 2012. Available at
www.epa.gov/ttn/naaqs/standards/ozone/data/20120502rea.pdf
Weinstein, D.A., R.M. Beloin and R.D. Yanai. (1991). Modeling changes in red spruce carbon
balance and allocation in response to interacting ozone and nutrient stresses. Tree
Physiol. 9:127-146.
Weinstein, DA; Gollands, B; Retzlaff, WA. (2001). The effects of ozone on a lower slope forest
of the Great Smoky Mountain National Park: Simulations linking an individual tree
model to a stand model. Forest Sci 47: 29-42.
Wells, B., Wesson, K., Jenkins, S. (2012). Analysis of Recent U.S. Ozone Air Quality Data to
Support the O3 NAAQS Review and Quadratic Rollback Simulations to Support the First
Draft of the Risk and Exposure Assessment. Available at
http://www.epa.gOv/ttn/naaqs/standards/ozone/s o3 2008 rea.html
West, J. J., A. M. Fiore, V. Naik, L. W. Horowitz, M. D. Schwarzkopf, D. L. Mauzerall (2007).
Ozone air quality and radiative forcing consequences of changes in ozone precursor
emissions. Geophys Res Lett, 34, L06806.
9-15
-------
This page intentionally left blank.
-------
United States Office of Air Quality Planning and Standards Publication No. EPA-452/R-14-005a
Environmental Protection Health and Environmental Impacts Division August 2014
Agency Research Triangle Park, NC
------- |