Technical Support Document (TSD)
Preparation of Emissions Inventories for the Version 7.1
2016 Hemispheric Emissions Modeling Platform

November 2019

Contacts:

Jeff Vukovich, Alison Eyth

U.S. Environmental Protection Agency
Office of Air Quality Planning and Standards
Air Quality Assessment Division
Emissions Inventory and Analysis Group
Research Triangle Park, North Carolina


-------
TABLE OF CONTENTS

LIST OF TABLES	Ill

LIST OF FIGURES	Ill

ACRONYMS	IV

1	INTRODUCTION	1

2	HEMISPHERIC EMISSION INVENTORIES AND MODELING APPROACHES	4

2.1	Global HT AP version 2	10

2.1.1	Spatial Allocation	11

2.1.2	Chemical Speciation	12

2.1.3	Temporal Allocation	13

2.1.4	Vertical Allocation	14

2.1.5	Projection to 2014 via CEDS	15

2.2	THU China	17

2.2.1	Temporal Allocation	19

2.2.2	Chemical Speciation	21

2.2.3	Spatial Allocation	21

2.2.4	Vertical AI location	21

2.3	Global FINN fires (g_ptfire, g_ptagfire)	21

2.4	Other global emissions	23

2.4.1	Biogenic emissions	23

2.4.2	Lightning NOx emissions	25

2.4.3	Ocean chlorine emissions	26

2.5	Changes to North America sectors to support hemispheric modeling	26

2.5.1	Emissions inventory differences	26

2.5.2	Spatial Allocation	27

2.5.3	Vertical Allocation	28

2.6	Final CMAQ-Ready Hemispheric Emissions	28

3	SENSITIVITY RUNS	29

3.1 Sensitivity Run Specifications	30

3.1.1	ZANTH	30

3.1.2	ZROW.	32

3.1.3	ZUSA	33

3.1.4	ZSHIP	35

3.1.5	ZCHN	35

3.1.6	Z1ND	36

3.1.7	ZCANMEX	37

3.1.8	ZFIRE	38

3.1.9	EdgarCHN	38

4	EMISSION SUMMARIES	39

5	REFERENCES	42

11


-------
List of Tables

Table 1-1. Description of the hemispheric platform grid	1

Table 2-1. Sectors for the 2016 hemispheric modeling platform	4

Table 2-2. Summary of emissions differences between 2015 spinup and 2016 base case emissions by sector8

Table 2-3. HTAP sector mappings to platform sectors	10

Table 2-4. Emissions totals by HTAP sector, 108km hemispheric domain, tons/year for 2016	11

Table 2-5. Mapping of CEDS sectors to HTAP sectors	15

Table 2-6. List of THU China sectors and equivalent HTAP sectors	18

Table 2-7. Annual emissions totals by THU China sector, tons/year	18

Table 2-8. Emissions totals for global fire sectors, 108km hemispheric domain, year 2016, tons/year	23

Table 2-9. Annual biogenic emissions totals, 108km hemispheric domain, year 2016, tons/year	25

Table 2-10. Hemispheric spatial allocation approach for all North America sectors	27

Table 3-1. Hemispheric sensitivity run sector table summary	29

Table 3-2. Fraction of 2016 emissions from wildfire and prescribed burns in USA	31

Table 4-1. Domain total emissions by sector for 2016 hemispheric platform, tons/year	39

Table 4-2. Domain total emissions for 2016 hemispheric sensitivity cases, tons/year	41

List of Figures

Figure 1-1. Sample plot of Emissions on the 108km hemispheric modeling grid (log-scale tons/day)	2

Figure 2-1. Map of GEOCODE Country Codes	12

Figure 2-2. Day-of-week temporal profiles for HTAP sectors	13

Figure 2-3. Hour-of-day temporal profiles for HTAP sectors	14

Figure 2-4. Layer fractions for HTAP sectors	14

Figure 2-5. Projection factors from year 2010 to 2014 for POWER(left) and TRANSPORT(right) sectors for

NOx emissions derived from CEDS	16

Figure 2-6. Time allocation factors for the China Inventory. Averages of province-specific allocation

factors are shown	20

Figure 2-7. Diurnal temporal profiles for global wild fires and ag fires	22

Figure 2-8. Layer fractions for global wild fires (g_ptfire) and ag fires (g_ptagfire)	23

Figure 2-9. Spatial extent of 12US1 BEIS emissions on the 108km hemispheric domain	24

Figure 2-10. Lightning NOx Annual emissions	26

Figure 2-11. Annual NOx(left) and S02(right) emissions for base case run	28

Figure 3-1. Annual NOx(left) and S02(right) emissions for ZANTH sensitivity run	32

Figure 3-2. Annual NOx(left) and S02(right) emissions for ZROW sensitivity run	33

Figure 3-3. Annual NOx(left) and S02(right) emissions for ZUSA sensitivity run	34

Figure 3-4. Percent difference from basecase NOx(left) and S02(right) weekday-summer emissions for

ZSHIP sensitivity run	35

Figure 3-5. Annual NOx(left) and S02(right) emissions for ZCHN sensitivity run	36

Figure 3-6. Annual NOx(left) and S02(right) emissions for ZIND sensitivity run	36

Figure 3-7. Annual NOx(left) and S02(right) emissions for ZCANMEX sensitivity run	37

Figure 3-8. Annual NOx(left) and S02(right) emissions for ZFIRE sensitivity run	38

Figure 3-9. Percent difference NOx(left) and S02(right) weekday-summer emissions for EdgarCHN

sensitivity run	39

in


-------
Acronyms

AE5	CMAQ Aerosol Module, version 5, introduced in CMAQ v4.7

AE6	CMAQ Aerosol Module, version 6, introduced in CMAQ v5.0

NBAFM	Naphthalene, Benzene, Acetaldehyde, Formaldehyde and Methanol

BEIS	Biogenic Emissions Inventory System

BELD	Biogenic Emissions Land use Database

C1/C2	Category 1 and 2 commercial marine vessels

C3	Category 3 (commercial marine vessels)

CAP	Criteria Air Pollutant

CARB	California Air Resources Board

CB05	Carbon Bond 2005 chemical mechanism

CEMS	Continuous Emissions Monitoring System

CI	Chlorine

CMAQ	Community Multiscale Air Quality

CMV	Commercial Marine Vessel

CO	Carbon monoxide

EBAFM	Ethanol, Benzene, Acetaldehyde, Formaldehyde and Methanol

ECA	Emissions Control Area

EEZ	Exclusive Economic Zone

EGU	Electric Generating Units

EIS	Emissions Inventory System

EPA	Environmental Protection Agency

EMFAC	Emission Factor (California's onroad mobile model)

FF10	Flat File 2010

FIPS	Federal Information Processing Standards

FHWA	Federal Highway Administration

HAP	Hazardous Air Pollutant

HC1	Hydrochloric acid

Hg	Mercury

IMO	International Marine Organization

MCIP	Meteorology-Chemistry Interface Processor

MOVES	Motor Vehicle Emissions Simulator

NEI	National Emission Inventory

NESCAUM Northeast States for Coordinated Air Use Management

NESHAP	National Emission Standards for Hazardous Air Pollutants

NH3	Ammonia

NIF	NEI Input Format

NLCD	National Land Cover Database

nm	nautical mile

NOAA	National Oceanic and Atmospheric Administration

NOx	Nitrogen oxides

ORD	EPA's Office of Research and Development

PM2.5	Particulate matter less than or equal to 2.5 microns

PM10	Particulate matter less than or equal to 10 microns

ppb, ppm	Parts per billion, parts per million

RBT	Refinery to Bulk Terminal

RWC	Residential Wood Combustion

RVP	Reid Vapor Pressure

SCC	Source Classification Code

iv


-------
SMARTFIRE Satellite Mapping Automated Reanalysis Tool for Fire Incident Reconciliation

SMOKE

Sparse Matrix Operator Kernel Emissions

SOi

Sulfur dioxide

SOA

Secondary Organic Aerosol

TOG

Total Organic Gas

TSD

Technical support document

VOC

Volatile organic compounds

WRF

Weather Research and Forecasting Model

v


-------
1 Introduction

The U.S. Environmental Protection Agency (EPA) developed an air quality emissions modeling platform
for hemispheric modeling with the Community Multiscale Air Quality (CMAQ) for the year 2016. This
emission modeling platform represents an update to the 2011 platform described by Eyth et al. (2016).
Similar to that work, this platform uses publicly available emissions that cover the globe and integrates
these emissions with regional inventories to create a single consistent platform.

The hemispheric modeling platform includes the following components:

In the United States, Canada, and Mexico, emissions are based on the 2016 alpha platform for
regional scale air quality modeling. Details regarding the North America portion of the
hemispheric emissions platform are provided in the 2016 v7.1 (also known as alpha) platform
Technical Support Document).

Outside North America, fire emissions are from the Fire Inventory from NCAR (FINN) for the
year 2016. This includes wildfires, prescribed burning, and agricultural fires.

In China, anthropogenic emissions excluding fires and carbon monoxide are from an emissions
dataset provided by Tsinghua University (THU) for year 2015.

Outside of North America and China, anthropogenic emissions excluding fires use the
Hemispheric Transport of Air Pollution (HTAP) emissions inventory, version 2. HTAP emissions
are for the year 2010, and are adjusted to year 2014 using factors derived from the Community
Emissions Data System (CEDS).

- Biogenic emissions in most of North America are equivalent to those from the regional alpha
platform and are generated by the BEIS model. Elsewhere in the modeling domain, biogenic
emissions are from the Model of Emissions and Gases and Aerosols from Nature (MEGAN).

Hemispheric air quality modeling is performed on a polar stereographic grid which is centered on the
North Pole, extends south to the equator, and has an approximate resolution of 108-km by 108-km per
grid cell. Table 1-1 includes the description (i.e., specification) of the hemispheric grid. Figure 1-1 shows
a visualization of NOx emissions on the 108km hemispheric modeling grid.

Table 1-1. Description of the hemispheric platform grid

Common
Name

Grid
Cell Size

Description

Grid name

Parameters listed in SMOKE grid
description (GRIDDESC) file:
projection name, xorig, yorig, xcell,
ycell, ncols, nrows, nthik*

Hemispheric
108km grid

108 km

Polar stereographic
grid of the Northern
Hemisphere

HEMI_108k

'POL HEMI', -10098000, -10098000,
108.D3, 108.D3, 187, 187, 1

* Corresponds to P

IOJ4 definition "+proj=stere +lat 0=90.0 +lat ts=45.0 +lon 0=-98.0

+y_0=l 0098000.0 +x_0=10098000.0 +a=6370000.0 +b=6370000.0 +to_meter= 108000 +no_defs"

1


-------
Figure 1-1. Sample plot of Emissions on the 108kin hemispheric modeling grid (log-scale tons/day)
2016fe hemi annual emissions total: mrqqrid NOX

tons/year
0 to 1.00
1.00 to 10.0
^ 10.0 to 100
¦¦ 100 to 1000
^ 1000 to 10000
^ 10000 +

The Community Multiscale Air Quality (CMAQ) model was used to model ozone (O3) and particulate
matter (PM) for this project. CMAQ includes support for hemispheric modeling, sometimes referred to as
H-CMAQ modeling, and can be used to develop boundary conditions (BCs) for regional scale air quality
modeling. CMAQ requires hourly and gridded emissions of the following inventory pollutants: carbon
monoxide (CO), nitrogen oxides (NOx), volatile organic compounds (VOC), sulfur dioxide (SO2),
ammonia (NH3), particulate matter less than or equal to 10 microns (PM10), and individual component
species for particulate matter less than or equal to 2.5 microns (PM2.5). In addition, the Carbon bond
version 6 (CB6) with chlorine chemistry used here within CMAQ allows for explicit treatment of the
YOC HAPs naphthalene, benzene, acetaldehyde, formaldehyde and methanol (NBAFM) and includes
anthropogenic HAP emissions of HC1 and CI. However, most emissions inventories in this platform
outside of the United States do not include any HAP emissions.

The primary emissions modeling tool used to create the air quality model-ready emissions was the Sparse
Matrix Operator Kernel Emissions (SMOKE) modeling system version 4.5 (SMOKE 4.5) with some
updates. Most regional scale CMAQ modeling is performed with "inline" plume rise, in which plume rise
calculations for point sources are performed by CMAQ using stack parameters and meteorology.

However, because the HTAP and Till.' China inventories do not include information on individual stacks,
it is not possible to run H-CMAQ with the inline plume rise option. Instead, we must create layered

2


-------
emissions within SMOKE by running the Laypoint program for sources with stack information, and by
running the Layalloc program to apply a fixed vertical allocation to sectors which do not have stack
information.

The hemispheric modeling platform includes multiple sets of model-ready emissions, or "cases":
An annual "base case" for year 2016, abbreviated 2016fe_hemi_cb6_16jh.

An annual "spinup case" for year 2015, abbreviated 2016fe_spinup_cb6_15jh. H-CMAQ
modeling is performed with an 8-month spinup period, and the emissions for the spinup period are
based on an entirely separate emissions case specific to the year 2015. For quality assurance and
for other modeling applications, the spinup case emissions were processed for all of 2015, not just
for the 8-month period required for spinup of 2016 modeling.

Several sensitivity runs in which one or more categories of emissions are zeroed out or modified,
as described in Section 3.

The gridded meteorological model used to provide input data for the emissions modeling was developed
using the Weather Research and Forecasting Model (WRF version 3.8, Advanced Research WRF core
(Skamarock, et al., 2008). The WRF Model is a mesoscale numerical weather prediction system
developed for both operational forecasting and atmospheric research applications. The WRF was run for
2015 and 2016 over the 108-km resolution hemispheric modeling domain with 44 vertical layers. The
WRF run for this platform is given the EPA meteorological case label "16jh." The full case name
includes this abbreviation following the emissions portion of the case name to fully specify the name of
the case as "2016fe_hemi_cb6_16jh."

The CMAQ model requires hourly emissions of specific gas and particle species for the horizontal and
vertical grid cells contained within the modeled region (i.e., modeling domain). To provide emissions in
the form and format required by the model, it is necessary to "pre-process" the "raw" emissions (i.e.,
emissions input to SMOKE). In brief, the process of emissions modeling transforms the emissions
inventories from their original temporal resolution, pollutant resolution, and spatial resolution into the
hourly, speciated, gridded resolution required by the air quality model. Emissions modeling includes
temporal allocation, spatial allocation, and pollutant speciation. For hemispheric modeling, emissions
modeling also includes the vertical allocation of point sources.

The temporal resolutions of the emissions inventories input to SMOKE vary across sectors and may be
hourly, daily, monthly, or annual total emissions. The spatial resolution may be individual point sources,
county/province/municipio totals, or pre-gridded emissions and varies by sector. This section provides
some basic information about the tools and data files used for emissions modeling as part of the modeling
platform.

For most sectors, SMOKE version 4.5 was used to process the raw emissions inventories into emissions
inputs for each modeling sector into a format compatible with CMAQ. For hemispheric modeling, plume
rise is not calculated "inline" during CMAQ modeling; instead, the emissions output from SMOKE and
input to CMAQ are 3-D emissions files and already have plume rise applied. For QA of the emissions
modeling steps, emissions totals by specie for the entire model domain are output as reports that are then
compared to reports generated by SMOKE on the input inventories to ensure that mass is not lost or
gained during the emissions modeling process. Preparation of China emissions from the THU dataset
(section 2.2), biogenic emissions from MEGAN (section 2.4.1), and lightning NOx emissions (section
2.4.2) was not performed in SMOKE, but with other utilities as described in the relevant sections below.

3


-------
The chemical mechanism used for the 2014 platform is the CB6 mechanism (Yarwood, 2010). We used a
particular version of CB6 that we refer to here as "CB6-CMAQ" that breaks out naphthalene (NAPH)
from XYL and PAR as an explicit model species, resulting in model species NAPH and XYLMN instead
of XYL, and revising PAR to remove the naphthalene portion (very small amount). CB6-CMAQ also
uses SOAALK (Pye and Pouliot, 2012), a species produced from TOG speciation that is not used in
CAMX. This platform generates the PM2.5 model species associated with the CMAQ Aerosol Module
version 6 (AE6).

Section 2 of this document describes the global inventories input to SMOKE and preparation of the
emissions for air quality modeling. Section 3 outlines several sensitivity cases in which new sets of
emissions were modeled with a portion of the emissions zeroed out. Data summaries are provided in
Section 4.

2 Hemispheric Emission Inventories and Modeling Approaches

This section describes the emissions data and processing that make up the 2016 hemispheric platform,
excluding the United States, Canada, and Mexico inventories that make up the corresponding regional
modeling platform. Documentation for the North America portion of the platform is in the 2016 alpha
platform TSD.

For the purposes of preparing the air quality model-ready emissions, the platform consists of several
sectors for emissions modeling. The significance of an emissions modeling or platform "sector" is that
the data are run through the SMOKE programs independently from the other sectors except for the final
merge (Mrggrid). The final merge program combines the sector-specific gridded, speciated, hourly
emissions together to create CMAQ-ready emission inputs.

Table 2-1 lists all sectors in the hemispheric platform and their data sources. For North America sectors
originating from the alpha platform for regional modeling, only a brief description of the sector is offered
here, with additional detail in the regional modeling TSD. The platform sector abbreviations are provided
in italics. These abbreviations are used in the SMOKE modeling scripts, inventory file names, and
throughout the remainder of this document. Most sectors which include global (e.g. HTAPv2) emissions
include a "g_" prefix.

Table 2-2 provides a brief by-sector outline of the differences between the emissions used in the 2016
base case and the emissions used in the 2015 spinup case.

Table 2-1. Sectors for the 2016 hemispheric modeling platform

Platform Sector:

abbreviation

Data
Source

Description and resolution of the data input to SMOKE

U.S. EGUs:

ptegu

Alpha
Platform

Electric Generating Units (EGUs) from the US NEI. Annual
resolution, with hourly resolution for sources that match Continuous
Emission Monitoring (CEM) data.

U.S. point source oil
and gas:
pt oilgas

Alpha
Platform

Oil and gas production point sources from the US NEI. Includes
emissions from oil rigs in the Gulf of Mexico. Annual resolution.

U.S. remaining non-
EGU point:

ptnonipm

Alpha
Platform

All point sources from the US NEI not matched to the ptegu or
pt_oilgas sectors. Annual resolution.

4


-------
Platform Sector:

abbreviation

Data
Source

Description and resolution of the data input to SMOKE

U.S. agricultural:

ag

Alpha
Platform

Nonpoint livestock and fertilizer application emissions from the US
NEI. Livestock includes ammonia and other pollutants (except
PM2.5). County and annual resolution.

U.S. agricultural fires
with point resolution:

ptagfire

Alpha
Platform

Agricultural fire sources that were developed by EPA as point sources
for 2016. Daily resolution.

U.S. area fugitive
dust:

afdust

Alpha
Platform

PMio and PM2 5 fugitive dust sources from the 2014NEIv2 nonpoint
inventory; including building construction, road construction,
agricultural dust, and road dust. The NEI emissions are reduced
during modeling according to a transport fraction and a meteorology-
based (precipitation and snow/ice cover) zero-out. Does not include
onroad emissions in Alaska, Hawaii, Puerto Rico, or Virgin Islands
(see othafdust sector). County and annual resolution.

U.S. biogenic:

beis

Alpha
Platform

Year 2016, hour-specific, grid cell-specific emissions generated from
the BEIS3.61 model within SMOKE. Includes emissions in the
Continental U.S. and portions of southern Canada and northern
Mexico. Uses BELD v4.1 land use data.

Hemispheric
biogenic: g biog

MEGAN
model

Year 2016, hour-specific, grid cell-specific emissions generated from
the MEGAN model for the portion of the hemispheric modeling
domain not covered by the BEIS sector.

U.S. Category 1, 2
CMV:

cmv clc2

Alpha
Platform

Category 1 (CI) and category 2 (C2) commercial marine vessel (cmv)
emissions sources based on the 2014NEIv2 nonpoint inventory, with
SO2 emissions reduced by 90% for 2016. Includes emissions in U.S.
State and Federal Waters. County and annual resolution.

U.S. Category 3
CMV:

cmv c3

Alpha
Platform

Category 3 (C3) cmv emissions based on the 2014NEIv2 nonpoint
inventory, converted to point sources, and with SO2 emissions reduced
by 90% for 2016. Includes emissions in U.S. State and Federal Waters
only; CMV emissions beyond U.S. Federal Waters are in the g_ships
sector. Annual resolution.

U.S. locomotives:
rail

Alpha
Platform

Rail locomotives emissions from the 2014NEIv2. County and annual
resolution.

U.S. remaining
nonpoint:

nonpt

Alpha
Platform

2014NEIv2 nonpoint sources not included in other platform sectors.
County and annual resolution.

U.S. nonpoint source
oil and gas:
np oilgas

Alpha
Platform

2014NEIv2 nonpoint sources from oil and gas-related processes,
projected to 2016. County and annual resolution.

U.S. Residential
Wood Combustion:

rwc

Alpha
Platform

2014NEIv2 nonpoint sources from residential wood combustion
(RWC) processes. County and annual resolution.

U.S. Nonroad:

nonroad

Alpha
Platform

2016 nonroad equipment emissions developed with the MOVES2014a
model. MOVES was used for all states except California, which
submitted their own emissions. County and monthly resolution.

U.S. Onroad:

onroad

Alpha
Platform

2016 onroad mobile source gasoline and diesel vehicles from moving
and non-moving vehicles that drive on roads, along with vehicle
refueling. Includes the following modes: exhaust, extended idle,
auxiliary power units, evaporative, permeation, refueling, and brake
and tire wear. For all states except California, developed using winter
and summer MOVES emissions tables produced by MOVES2014a.
Does not include onroad emissions in Alaska, Hawaii, Puerto Rico, or
Virgin Islands (see onroad can sector).

5


-------
Platform Sector:

abbreviation

Data
Source

Description and resolution of the data input to SMOKE

U.S. Onroad
California:

onroad ca adj

Alpha
Platform

2016 California-provided CAP and metal HAP onroad mobile source
gasoline and diesel vehicles submitted to the NEI, gridded and
temporalized using MOVES2014a. Volatile organic compound
(VOC) HAP emissions derived from California-provided VOC
emissions and MOVES-based speciation.

U.S. point source
fires: ptflre

Alpha
Platform

Point source day-specific wildfires and prescribed fires for 2016
computed using SMARTFIRE2 and Blue Sky Framework for both
flaming and smoldering processes (i.e., SCCs 281XXXX002).
Smoldering is forced into layer 1 (by adjusting heat flux). Daily
resolution.

Other North America
fires:

ptfire othna

Alpha
Platform

Point source day-specific wildfires and prescribed fires for 2016.
Canada fires provided by Environment Canada, with data for missing
months filled in using fires from the Fire INventory (FINN) from
National Center for Atmospheric Research (NCAR) fires (NCAR,
2016 and Wiedinmyer, C., 2011). Fires in Mexico and Central
America are from FINN. Daily resolution.

Canada dust sources:

othafdust

Alpha
Platform

Fugitive dust sources from Canada's 2013 and 2025 inventories
(interpolated to 2016). A transport fraction adjustment is applied
along with a meteorology-based (precipitation and snow/ice cover)
zero-out. Also includes afdust emissions in Alaska, Hawaii, Puerto
Rico, and Virgin Islands from 2014NEIv2. County and annual
resolution.

Canada and Mexico
point sources:
othpt

Alpha
Platform

Point sources from Canada's 2013 and 2025 inventories (interpolated
to 2016) and Mexico's 2014 and 2018 inventories (interpolated to
2016), annual resolution.

Canada and Mexico
nonpoint and
nonroad:

othar

Alpha
Platform

Year 2016 Canada (province or sub-province resolution) emissions,
interpolated from 2013 and 2025: monthly for agricultural ammonia,
and nonroad sources; annual for rail, CMV and other nonpoint Canada
sectors. Year 2016 Mexico (municipio resolution), interpolated from
2014 and 2018: annual nonpoint and nonroad mobile inventories.

Other non-NEI
onroad sources:
onroad can

Alpha
Platform

Monthly year 2016 Canada (province resolution or sub-province
resolution, depending on the province) onroad mobile inventory,
interpolated from 2013 and 2025. Also includes onroad emissions in
Alaska, Hawaii, Puerto Rico, and Virgin Islands from 2014NEIv2.

Other non-NEI
onroad sources:

onroad mex

Alpha
Platform

Monthly year 2016 Mexico (municipio resolution) onroad mobile
inventory, interpolated from 2014 and 2018.

Point source wildland
fires outside North
America: g_ptfire

FINN

Point source day-specific wildfires and prescribed fires for 2016 from
the Fire INventory (FINN) from National Center for Atmospheric
Research (NCAR) fires (NCAR, 2016 and Wiedinmyer, C., 2011).
Includes all fires outside North America. Daily resolution.

Point source
agricultural fires
outside North
America: g ptagfire

FINN

Point source day-specific agricultural fires for 2016 from the Fire
INventory (FINN) from National Center for Atmospheric Research
(NCAR) fires (NCAR, 2016 and Wiedinmyer, C., 2011). Includes all
fires outside North America. Daily resolution.

Agricultural
emissions outside
North America and
China: g ag

HTAP v2
+ CEDS

Agricultural emissions outside North America and China, from the
gridded HTAP version 2 dataset, agriculture sector. Year 2010
projected to 2014 using factors from CEDS dataset. Monthly
resolution.

6


-------
Platform Sector:

abbreviation

Data
Source

Description and resolution of the data input to SMOKE

Aircraft
landing/takeoff
emissions: g air Ito

HTAP v2

Landing and takeoff emissions from aircraft, from the gridded HTAP
version 2 dataset, air sector. Annual resolution.

Aircraft climbing and
descent emissions:

g air cds

HTAP v2

Climbing and descent emissions from aircraft, from the gridded HTAP
version 2 dataset, air sector. Annual resolution.

Aircraft cruising
emissions: g air crs

HTAP v2

Cruising emissions from aircraft, from the gridded HTAP version 2
dataset, air sector. Annual resolution.

Energy sector
emissions outside
North America and
China: g energy

HTAP v2
+ CEDS

Energy sector emissions from the gridded HTAP version 2 dataset,
excluding North America and China. Year 2010 projected to 2014
using factors from CEDS dataset. Monthly resolution.

Industry sector
emissions outside
North America and
China: g industry

HTAP v2
+ CEDS

Industry sector emissions from the gridded HTAP version 2 dataset,
excluding North America and China. Year 2010 projected to 2014
using factors from CEDS dataset. Monthly resolution.

Residential sector
emissions outside
North America and
China: g residential

HTAP v2
+ CEDS

Residential sector emissions from the gridded HTAP version 2 dataset,
excluding North America and China. Year 2010 projected to 2014
using factors from CEDS dataset. Monthly resolution.

Transport sector
emissions outside
North America and
China: g transport

HTAP v2
+ CEDS

Transport sector emissions from the gridded HTAP version 2 dataset,
excluding North America and China. Year 2010 projected to 2014
using factors from CEDS dataset. Monthly resolution.

Shipping emissions
outside North
America: g ships

HTAP v2
+ CEDS

Shipping emissions from the gridded HTAP version 2 dataset, ships
sector. Excludes emissions in U.S. State and Federal Waters. Year
2010 projected to 2014 using factors from CEDS dataset. Annual
resolution.

China fertilizer
application:
china agrf

THU
China

Fertilizer application emissions in China from the THU dataset.
Annual resolution with hourly temporal profiles.

China livestock:
china agrl

THU
China

Livestock emissions in China from the THU dataset. Annual
resolution with hourly temporal profiles.

China domestic
biofuel: china dobi

THU
China

Domestic biofuel emissions in China from the THU dataset. Annual
resolution with hourly temporal profiles.

China domestic
combustion:
china docb

THU
China

Domestic combustion emissions in China from the THU dataset.
Annual resolution with hourly temporal profiles.

China domestic fossil
fuel: china dofu

THU
China

Domestic fossil fuel emissions in China from the THU dataset. Annual
resolution with hourly temporal profiles.

China domestic
solvent use:
china doso

THU
China

Domestic solvent use emissions in China from the THU dataset.
Annual resolution with hourly temporal profiles.

China other domestic:
china doth

THU
China

Other domestic emissions in China from the THU dataset. Annual
resolution with hourly temporal profiles.

China industry
combustion:
china inch

THU
China

Industrial combustion emissions in China from the THU dataset.
Annual resolution with hourly temporal profiles.

China power plant:
china ppcb

THU
China

Power plant emissions in China from the THU dataset. Annual
resolution with hourly temporal profiles.

7


-------
Platform Sector:

abbreviation

Data
Source

Description and resolution of the data input to SMOKE

China cement:
china prce

THU
China

Cement process emissions in China from the THU dataset. Annual
resolution with hourly temporal profiles.

China steel:
china prir

THU
China

Steel process emissions in China from the THU dataset. Annual
resolution with hourly temporal profiles.

China industrial
solvent use:
china prso

THU
China

Industrial solvent use emissions in China from the THU dataset.
Annual resolution with hourly temporal profiles.

China other
industrial: china prot

THU
China

Other industrial process emissions in China from the THU dataset.
Annual resolution with hourly temporal profiles.

China transport off-
road: china trof

THU
China

Off-road mobile emissions in China from the THU dataset. Annual
resolution with hourly temporal profiles.

China transport on-
road: china tron

THU
China

On-road mobile emissions in China from the THU dataset. Annual
resolution with hourly temporal profiles.

Lightning NOx
emissions: lightning

GEIA

Global NOx emissions from lightning strikes.

Table 2-2. Summary of emissions differences between 2015 spinup and 2016 base case emissions by

sector

Platform Sector

2016 base case

2015 spinup case

afdustadj

2014NEIv2, adjusted with 2016
meteorology

2014NEIv2, adjusted with 2015
meteorology

onroad

2016 activity / emission factors /
meteorology

2015 activity / emission factors /
meteorology

onroad ca adj

CARB inventory interpolated to 2016

CARB inventory interpolated to 2015

ag

2016 fertilizer + 2014NEIv2 livestock +
2016 met-based temporalization

2014NEIv2 fertilizer and livestock + 2015
met-based temporalization

cmv clc2

2014NEIv2 with SO2 reduced by 90%

Same as 2016

cmv c3

2014NEIv2 with SO2 reduced by 90%

Same as 2016

nonpt

2014NEIv2

Same as 2016

nonroad

2016 inventory from MOVES2014a

Interpolation of 2014 and 2016 inventories
from MOVES2014a

np oilgas

2014NEIv2 projected to 2016

2014NEIv2 projected to 2015

rail

2014NEIv2

Same as 2016

rwc

2014NEIv2 + 2016 met-based
temporalization

2014NEIv2 + 2015 met-based
temporalization

othafdust

2013 and 2025 Canada, interpolated to 2016

2013 and 2025 Canada, interpolated to 2015

onroad can

2013 and 2025 Canada, interpolated to 2016

2013 and 2025 Canada, interpolated to 2015

onroad mex

2014 and 2018 Mexico, interpolated to 2016

2014 and 2018 Mexico, interpolated to 2015

othar

2013	and 2025 Canada, interpolated to 2016

2014	and 2018 Mexico, interpolated to 2016

2013	and 2025 Canada, interpolated to 2015

2014	and 2018 Mexico, interpolated to 2015

othpt

2013	and 2025 Canada, interpolated to 2016

2014	and 2018 Mexico, interpolated to 2016

2013	and 2025 Canada, interpolated to 2015

2014	and 2018 Mexico, interpolated to 2015

ptagfire

2016 point ag fire dataset

2015 point ag fire dataset

ptegu

2016 point inventory

2015 point inventory

ptnonipm

2016 point inventory

2015 point inventory

pt_oilgas

2016 point inventory, with remaining 2014
sources projected to 2016

2015 point inventory, with remaining 2014
sources projected to 2015

ptfire

2016 SMARTFIRE2 dataset

2015 SMARTFIRE2 dataset

g ptfire

2016 FINN dataset

2015 FINN dataset

8


-------
Platform Sector

2016 base case

2015 spinup case

ptfireothna

2016 Environment Canada fires + FINN
fires

2015 STI fires

beis

2016 biogenics

2015 biogenics

g biog

2016 biogenics

2015 biogenics

B ag

2010 HTAPv2 + 2014 CEDS projection

Same as 2016

g ptagfire

2016 FINN dataset

2015 FINN dataset

g air cds

2010 HTAPv2

Same as 2016

g air crs

2010 HTAPv2

Same as 2016

g air lto

2010 HTAPv2

Same as 2016

g energy

2010 HTAPv2 + 2014 CEDS projection

Same as 2016

g industry

2010 HTAPv2 + 2014 CEDS projection

Same as 2016

g residential

2010 HTAPv2 + 2014 CEDS projection

Same as 2016

g ships

2010 HTAPv2 + 2014 CEDS projection

Same as 2016

g transport

2010 HTAPv2 + 2014 CEDS projection

Same as 2016

lightning

Gridded lightning NOx dataset

Same as 2016

All China sectors
(china *)

2015 China THU

Same as 2016

Documentation for the North America emissions sectors is provided in the 2016 alpha platform for
regional modeling TSD. The following sectors concern the emissions inventories for the portions of the
hemispheric platform that are not part of the regional modeling platform.

9


-------
2.1 Global HTAP version 2

The Hemispheric Transport of Air Pollution Version 2 inventory, or HTAPv2 inventory, is the basis for
most emissions in the platform outside of North America and China. HTAP inventories are available here:
http: //edgar. i rc. ec. europa. eu/htap v2/

Each HTAP sector is mapped to one or more platform sectors as outlined in Table 2-3. The representative
dates column indicates whether emissions are processed for every day of the year ("All"); seven days per
month, one for each day of the week ("Week"); four days per month, corresponding to Mondays, other
WeekDays, Saturdays, and Sundays ("MWDSS"); or for a single average day per month ("Aveday").

Table 2-3. HTAP sector mappings to platform sectors

HTAP sector

Platform sector(s)

Inventory
resolution

Pollutants

Representative
dates

Air

g_air_cds
g_air_crs
g air lto

Annual

CO, NOx, PMio, PM2.5,
BC (PEC), OC (POC),
S02, NMVOC

All

Ships

g_ships

Annual

CO, NOx, PM10, PM2.5,
BC (PEC), OC (POC),
S02, NMVOC

All

Energy

g_energy

Monthly

CO, NH3, NOx, PMio,
PM2.5, BC (PEC), OC
(POC), S02, NMVOC

All

Industry

g_industry

Monthly

CO, NH3, NOx, PMio,
PM2.5, BC (PEC), OC
(POC), S02, NMVOC

All

Transport

gtransport

Monthly

CO, NH3, NOx, PMio,
PM2.5, BC (PEC), OC
(POC), S02, NMVOC

Week

Residential

gresidential

Monthly

CO, NH3, NOx, PMio,
PM2.5, BC (PEC), OC
(POC), S02, NMVOC

MWDSS

Ag

g ag

Monthly

nh3

Aveday

Hemispheric emissions totals from each HTAP sector are provided in Table 2-4. These emissions totals
include CEDS projections (see section 2.1.5), do not include emissions in North America or China that
were zeroed out (see section Error! Reference source not found.), and are domain-wide for the 108km
hemispheric modeling domain.

10


-------
Table 2-4. Emissions totals by HTAP sector, 108km hemispheric domain, tons/year for 2016

sector

CO

NH3

NOX

PM10

PM2.5

S02

voc

g_ag



27,053,284











g_air_c;ds

134,140



1,089,169

18,439

18,439

92,194

31,552

g_air_c;rs

179,798



1,457,442

24,682

24,682

123,425

42,107

g_air_lt°

116,928



282,723

2,350

2,350

26,785

4,415

g_energy

8,597,034

71,485

14,168,795

4,951,791

2,077,552

26,978,678

884,267

g_industry

132,095,290

992,589

6,997,163

6,873,446

3,563,763

16,333,374

38,620,852

g_residential

232,141,408

4,546,822

4,344,817

15,661,354

11,074,589

3,478,108

33,272,813

g_ships

1,175,509



12,078,998

1,207,008

1,207,008

6,633,296

628,816

g_transport

99,498,142

189,347

24,530,368

1,827,155

1,678,990

1,621,702

23,083,023

2.1.1 Spatial Allocation

The HTAP dataset includes gridded annual and monthly datasets on a global 0.1° latitude by 0.1°
longitude grid. Spatial allocation of these inventories in SMOKE consists of two steps. First, in the
SMOKE program Smkinven, each 0.1° by 0.1° point on the global grid is mapped to a country code
(GEOCODE) and a time zone using a file called the GRIDMASK. Then, the SMOKE program Grdmat
spatially reallocates emissions from the global input grid to the 108km polar stereographic grid used for
air quality modeling.

The GRIDMASK file is a gridded file on the same 0.1° by 0.1° global grid and maps each point on the
grid to a GEOCODE and a time zone. Smkinven maps each point in the gridded inventory data to a
GEOCODE, enabling calculation of national emissions totals, and application of country-specific
projections or controls. GEOCODEs are defined not only for each country, but also for distinct bodies of
water, e.g. the North Atlantic Ocean, Gulf of Mexico, Caribbean Sea, and so on. The time zone
information is used by the SMOKE program Temporal to apply diurnal profiles with the correct time zone
offsets. Time zones with a 30-minute offset (e.g. Newfoundland, India) are rounded to the nearest hour in
SMOKE modeling.

A plot of the GEOCODEs in the GRIDMASK file is shown in Figure 2-1, in which distinct GEOCODEs
appear as different colors.

11


-------
Figure 2-1. Map of GEOCODE Country Codes

2.1.2 Chemical Speciation

HTAP inventories include the following pollutants: "no methane VOC" (NMVOC); black carbon (BC,
mapped to model species PEC); organic carbon (OC, mapped to model species POC); total PM2.5; and
other standard CAPs (CO, NH3, NOx, PM10, S02). HTAP inventories do not include any HAP
emissions.

Unlike regional modeling which uses emissions inventories with SCCs, HTAP inventories do not include
more detailed information beyond the sector total for each point on the grid by month and pollutant.
Therefore, we must apply sector-average speciation profiles to each HTAP sector. For VOC, average
speciation profiles for HTAP sectors for the CB6 chemical mechanism were developed from the 2011
emissions modeling platform. In that platform, U.S. SCCs were each mapped to one of the HTAP
sectors, and then speciated emissions were summed by HTAP sector group to develop an average VOC
profile for each HTAP sector. Since the HTAP inventories specify "NMVOC" instead of "VOC", these
average profiles do not include methane (CH4), but do include other species which may be considered
part of TOG but not VOC, such as ethane (ETHA). There is no VOC-to-TOG conversion prior to
speciation like there is in traditional emissions modeling, since these average profiles are computed on the
basis of (NM)VOC.

For PM2.5 speciation, the HTAP dataset includes emissions for black carbon, organic carbon, and total
PM2.5. In SMOKE modeling, we map black carbon to the model species PEC, and organic carbon to the
model species POC. Since PEC and POC are also part of total PM2.5, we must subtract PEC and POC
from total PM2.5 in order to prevent a double count. Prior to SMOKE modeling, an additional set of
gridded inventory files is generated for PM2 5 0TH (other PM2.5), which is equal to total PM2.5 minus
PEC minus POC. Then, PM2 5 0TH is speciated using speciation profiles which map to the remaining

12


-------
PM species needed for CMAQ modeling. It is still necessary for SMOKE to use total PM2.5 for the sole
purpose of calculating PMC (PM10 minus total PM2.5), but only PM2 5 0TH is mapped to PM model
species other than PEC, POC, and PMC. Similar to the VOC speciation profiles, speciation of
PM2 5 0TH uses sector-specific average profiles.

For all HTAP sectors, NOx is speciated to NO and N02 using a 90/10 split. Emissions for the HONO
species are not created outside of North America. The SULF species is calculated as a percentage of S02
in the HTAP energy, industry, and residential sectors.

2.1.3 Temporal Allocation

Similar to with chemical speciation, each HTAP sector has a single set of temporal profiles (weekly and
diurnal, and also monthly for sectors without monthly inventories) that is applied to the entire sector.
These temporal profiles were estimated using North American source-specific examples and best
judgement. Because different countries celebrate different holidays, there are no considerations for
holiday temporalization in the HTAP sectors.

Similar to some North America sectors, some HTAP sectors are processed with representative dates, as
specified in Table 2-3 (above). For example, the HTAP ag sector uses flat day-of-week temporalization,
and so for the g ag sector, we process emissions for one representative day per month. The aircraft,
energy, industry, and ships sectors are processed for every day to support plume rise, as discussed in the
next section.

The weekly and diurnal temporal profiles for HTAP sectors are shown in Figures 2-2 and 2-3. The HTAP
air and ships sectors each use a flat monthly temporal profile, while the other HTAP sectors have monthly
inventories and do not use monthly temporal profiles.

Figure 2-2. Day-of-week temporal profiles for HTAP sectors

M	Tu	W	Th	F

— ii3/ AiR/SHIF;- EI.ERGV —INDUSTRY/ RESIDENT

13


-------
Figure 2-3. Hour-of-day temporal profiles for HTAP sectors

o.i
0.09
0.08
0.07
0.06
0.05
0.04
0.03
0.02
0.01
0

1 2 3 4 5 6 7 B 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

AG		AIR CDS/LTO	AIRCRS ^—ENERGY

^—INDUSTRY ^—RESIDENTIAL-^—SHIPS	TRANSPORT

2.1.4 Vertical Allocation

Unlike in North America sectors, HTAP inventories do not include stack parameters or other information
which can be used to estimate plume rise for point sources. This means we cannot run CMAQ with inline
plume rise, and must apply vertical allocation in SMOKE to all sectors. In North America, we can use the
Laypoint program to calculate plume rise using stack parameters and meteorology. Without stack
parameters, we cannot do that for HTAP sectors; instead, we apply vertical profiles on a sector-wide
basis.

Figure 2-4 shows the vertical profiles applied to HTAP sectors. Among HTAP sectors, only the aircraft,
energy, industry, and ships sectors have a vertical profile applied. For ag, residential, and transport, all
emissions remain in Layer 1.

Figure 2-4. Layer fractions for HTAP sectors

0.35

layer

ships industry energy air LTO air CDS airCRS

Vertical profiles are applied using the SMOKE program Layalloc. First, the SMOKE program Smkmerge
generates daily emissions with everything in a single layer, and then the Layalloc program splits the
emissions from Smkmerge into vertical layers. Layer fractions are input to Layalloc in the following
format, where "bottom" and "top" are layer heights in meters:

1 ay er,b ottom,top, fracti on

14


-------
1,0,20,0.06667
2,20,44,0.06667
3,44,70,0.06667
4,70,101,0.06667

41,14388,15345,0
42,15345,16394,0
43,16394,17565,0
44,17565,18874,0

Layalloc requires that the bottom and top of each layer be specified in the layer fraction file, and applies
layer fractions according to those layer heights, not according to the layer number in the first column.
Layer heights are not constant, but vary according to meteorology. Our layer fraction files use layer
heights averaged across the year. Because layer heights vary temporally and spatially, the actual layer
fractions as applied by Layalloc are not constant. This means that occasionally, a small amount of
emissions may appear in a layer with a 0% layer fraction. For example, the layer fractions for aircraft
cruising emissions allocate 20% of the emissions to each of five layers from 34 to 38; but because of
varying layer heights, Layalloc does allocate a small amount of emissions to layers 33 and 39.

2.1.5 Projection to 2014 via CEDS

The HTAP version 2 inventory represents emissions for the year 2010, which is six years prior to the
model year for this project. A projection of the HTAP emissions to a year as close to 2016 as possible was
desired.

Data from the Community Emissions Data System, or CEDS, includes emissions totals for the years 2010
and 2014, among other years, by country and emissions sector. Emissions totals for years after 2014 were
not available at the time of this study, so 2014 was used as a reasonable approximation for 2016.

Table 2-5 shows how CEDS sectors were mapped to HTAP sectors. CEDS data is available at a much
finer resolution than the aggregate sectors listed in this table, but because adjustment factors can only be
applied on a sector-wide basis in SMOKE, we could not apply projection factors at a finer resolution than
as shown in this table.

Table 2-5. Mapping of CEDS sectors to HTAP sectors

CEDS aggregate sector(s)

HTAP sector (SMOKE sector)

Agriculture

Agriculture (g ag)

Power

Energy (g energy)

Industry, Solvents

Industry (g industry)

Residential, Waste

Residential (g residential)

Ship

Ships (g ships)

Transportation

Transport (g transport)

HTAP projection factors for each country, HTAP/SMOKE sector, and pollutant were calculated using
CEDS emissions totals for 2010 and 2014 (examples in Figure 2-5), and then converted into GCNTL
format for application in SMOKE. Separate projection factors were calculated for PEC (BC), POC (OC),
and total PM2.5, as well as for CO, NH3, NOX, S02, and NMVOC. In China, we applied CEDS
projection factors to HTAP emissions for carbon monoxide (CO) only, because the THU dataset does not
include CO emissions. In the EdgarCHN sensitivity (see Section 23.1.9) CEDS factors were applied to

15


-------
HTAP emissions for all pollutants in China. For the ships sector, because most shipping emissions take
place over open water and not in a specific country, and because S02 is the primary pollutant of interest
with the ships sector, a global projection factor was applied to the entire domain for S02 only. The CEDS
dataset does not include emissions for aircraft, and so HTAP aircraft emissions were not projected.

Country-specific projection factors can be applied in SMOKE according to the GEOCODE for each
country, with GEOCODEs mapped to a set of points in the gridded HTAP inventories using the
GRIDMASK file. To limit the influence of unrealistically high or low projection factors from the CEDS
dataset, projection factors were limited to a range of 0.5 (50% decrease) to 1.5 (50% increase).

Figure 2-5. Projection factors from year 2010 to 2014 for POWER(left) and TRANSPORT(right)

sectors for NOx emissions derived from CEDS.

To prevent double counting of emissions with other datasets, a portion of the emissions from the HTAP
datasets must be zeroed out. In North America, all HTAP emissions except for in-flight aircraft are zeroed
out, to prevent a double count with the U.S., Canada, and Mexico inventories. In China, all HTAP
emissions except for CO (all sectors), and except for aircraft and shipping (all pollutants), are zeroed out,
to prevent a double count with the THU datasets. Zero-outs are implemented with a GCNTL file, which is
read by the SMOKE program Cntlmat and applied to model-ready emissions by pollutant and country
GEOCODE. The procedure for applying zero-outs is the same as the procedure for applying CEDS-based
projection factors, except in the case of zero-outs, the "projection factor" is zero.

In most HTAP sectors, including ag, residential, transport, energy, and industry, we zeroed out emissions
from the following GEOCODEs: United States, Canada, Mexico, and China (except CO). Puerto Rico,
Virgin Islands, Hong Kong, and Macau have separate GEOCODEs from the United States and China, and
were also zeroed out, as these areas were determined to overlap the U.S. and China datasets, respectively.
Taiwan is not covered by the THU dataset and is not zeroed out.

Bodies of water which border the United States, Canada, and Mexico were also zeroed out in most HTAP
sectors. This is because emissions from the HTAP dataset extend slightly into points that are assigned
water GEOCODEs instead of land GEOCODEs in the GRIDMASK file, and zeroing out bodies of water
adjacent to North America as well as the land areas prevents a "ring" of emissions from appearing along
the North America coast. In China, to prevent THU China and HTAP emissions from occuring in the

16


-------
same grid cells, the GRIDMASK file was modified to ensure that any point covered by the THU China
datasets is mapped to either the original China GEOCODE, or to a new GEOCODE called "China
Coastline". In SMOKE, we then zeroed out both China and China Coastline emissions (except CO).

The HTAP ships and air sectors are handled differently with respect to zero-outs. For shipping, the U.S.
CMV sectors (cmv_clc2, cmv_c3) include emissions in U.S. state and federal waters, extending 200
nautical miles off the U.S. coastline. The original GRIDMASK file did not distinguish U.S. federal waters
from non-federal waters, and so it was originally not possible to zero out HTAP ships emissions in the
exact area covered by the cmv_clc2 and cmv_c3 sectors. Previous hemispheric emissions modeling
applications resolved this by dropping U.S. federal water emissions from the cmv_clc2 and cmv_c3
sectors, but for this platform, use of U.S.-based emissions in federal waters was preferred over use of
HTAP ships emissions. To facilitate this, additional GEOCODEs were added to the GRIDMASK file in
U.S. state and federal waters. First, points in U.S. state waters that were originally assigned to a water
GEOCODE (e.g. 034000 for Gulf of Mexico) were reassigned to the code 3100XX, where XX = the U.S.
state FIPS code (e.g. 310012 for Florida). Second, points in U.S. federal waters that were originally
assigned to a water GEOCODE were reassigned to the code 85X000, where X = a number from 1 through
7 corresponding to the FIPS code associated with each offshore zone (1 = North Pacific, 2 = South
Pacific, 3 = Gulf Coast, 4 = East Coast, 5 = Alaska, 6 = Hawaii, 7 = Puerto Rico). When running the
HTAP ships sector, we zero out all emissions from the U.S., Canada, and Mexico GEOCODEs, and also
from all U.S. state and federal water GEOCODEs, but not from any bodies of water outside of U.S.
federal waters. We do not zero out shipping emissions in or around China because the THU dataset does
not include shipping.

The U.S. and Canada inventories include emissions from aircraft that are landing or taking off, which
overlaps the HTAP air LTO sector. Therefore, in the g air lto sector, we zero out emissions over land in
the United States and Canada. We do not zero out LTO emissions in Mexico, China, or over any bodies of
water including U.S. federal waters. We do not zero out any emissions in the HTAP g air cds
(climbing/descent) or g air crs (cruising) sectors, as the U.S. and Canada aircraft inventories do not
include those categories of aircraft emissions.

2.2 THU China

Gridded annual emissions for China for the year 2015 were provided by Tsinghua University (THU)

(Zhao et al., 2018,). The THU dataset includes emissions for 16 sectors, 15 of which are included in the
platform. The THU dataset includes annual gridded emissions on a Lambert projection grid with 27km
resolution; emissions forNH3, NOX, PM10, PM2.5, S02, and VOC, including partial speciation; and
temporal profiles which are used to generate hourly emissions.

The THU dataset emissions sectors, along with a mapping to the equivalent HTAP sectors, are listed in
Table 2-6. We are not using the open burning sector emissions from THU. Fire emissions from the FINN
dataset are preferred over the THU open burning emissions, primarily because the FINN data has better
daily resolution. Note that the THU dataset does not overlap the HTAP aircraft or ships sectors. Annual
emissions totals by sector for China are in Table 2-7.

The emissions files provided by Tsinghua University were annual files on a Lambert grid with 27km
resolution with partial speciation. Conversion of these emissions to the format needed for air quality
modeling was performed using standalone tools and not with the SMOKE model. First, the emissions
were temporalized from annual to hourly; then, additional speciation and unit conversion was performed

17


-------
as needed; then, emissions were reallocated to the 108km hemispheric grid; and finally, the emissions
were allocated vertically, depending on the sector.

Table 2-6. List of THU China sectors and equivalent HTAP sectors

THU China sector

Equivalent HTAP sector

AGRF (fertilizer application)

Agriculture

AGRL (livestock)

Agriculture

DOBI (domestic bio-fuel)

Residential

DOCB (domestic combustion)

Residential

DOFU (domestic fossil fuel)

Residential

DOSO (domestic solvent use)

Residential

DOTH (other domestic)

Residential

INCB (industry combustion)

Industry

PPCB (power plant)

Energy

PRCE (cement)

Industry

PRIR (steel)

Industry

PRSO (industry solvent use)

Industry

PROT (other industry process)

Industry

TROF (off-road transport)

Transport

TRON (on-road transport)

Transport

OPEN (open burning)

None; using FINN fires instead of THU fires

Table 2-7. Annual emissions totals by THU China sector, tons/year

China sector

NH3

NOX

PM10

PM2.5

S02

VOC1

AGRF (fertilizer application)

4,105,865











AGRL (livestock)

6,164,346











DOBI (domestic bio-fuel)



247,159

1,535,996

1,487,994

37,311

247,159

DOCB (domestic combustion)











2,254,033

DOFU (domestic fossil fuel)



684,444

2,030,505

977,311

3,894,725

684,444

DOSO (domestic solvent use)











1,863,873

DOTH (other domestic)

741,314









213,655

INCB (industry combustion)



2,520,744

1,453,406

965,475

5,920,805

2,653,154

PPCB (power plant)



4,390,020

1,175,701

702,486

4,992,558

4,390,020

PRCE (cement)



2,100,046

1,370,862

757,175

1,005,509

2,100,046

PRIR (steel)



603,922

1,300,269

939,834

836,222

603,922

PRSO (industry solvent use)











5,792,690

PROT (other industry process)

353,494

2,389,083

1,591,475

1,014,461

1,881,593

7,557,787

TROF (off-road transport)



2,684,613

219,988

208,409

159,214

3,354,114

TRON (on-road transport)



8,385,993

405,461

384,119

22,164

15,460,059

'VOC totals in this table are estimated from a sum of individual VOC species.

18


-------
2.2.1 Temporal Allocation

Included in the THU data release were annual emissions files, and temporal profiles on a month-of-year,
day-of-week, and hour-of-day basis. Temporal profiles vary by broad sector (AG/agriculture,
DO/domestic, IN/industry, PP/power plant, PR/industry process, and TR/transport) and by Chinese
province. Because daily temporalization depends only on the day of the week, emissions for seven
representative days per month, one for each day of the week, are sufficient.

A province-to-grid cell cross-reference for the China 27km grid was used in combination with the
temporal factors. Figure 2-6. shows the averages of province-level monthly, daily, and hourly temporal
profiles. These temporal profiles were gridded for each sector and combined to create an annual hour-of-
year basis file. Then, the gridded hourly profiles were applied to the annual gridded emissions for each
sector to create emissions files with 25 hourly timesteps for each representative day of the year (one file
for each day of the week for each month). Application of gridded hourly profiles was applied using a
Python utility called apply fracs.py which multiplies the annual China emissions files by the hourly
fraction matrix to generate the representative day, hourly emissions files.

For both 2016 modeling (base year) and 2015 modeling (spinup year), the same set of representative day
emissions are used in China.

19


-------
Figure 2-6. Time allocation factors for the China Inventory. Averages of province-specific

allocation factors are shown.

	 Agricultural 	 Industry		 Power Plant 	 Processing	Transport

	 Domestic 	 Open Burning

	 Agricultural 	 Industry		 Power Plant 	 Processing 	 Transport

	 Domestic 	 Open Burning

0.20 -

0.15

0.10 -

tu
>
<

16

6

'5i

£ 0.05

0.00 -

	1	1	1	1	1	r~

1	3	5	7	9	11

month

0.16 -

u 0.14-

QJ
CD

rc

<

0.12

a o.io -


-------
2.2.2	Chemical Speciation

The THU dataset included pre-speciated emissions for VOC and PM2.5 species, but additional processing
related to speciation was necessary to prepare these emissions for air quality modeling. The following
steps were all performed using the same Python utility (apply fracs.py) as the temporalization. For VOC
species, only a time conversion was necessary; emissions units from the previous step were in moles per
hour, which we converted to moles per second for CMAQ modeling. For PM2.5 species, emissions units
were in metric tons per hour (not short tons as in most North America emissions inventories), and were
converted to grams per second. The THU dataset included emissions for NOX, but not individual species
NO or N02; we split NOX into NO (90%) and N02 (10%), and also converted from metric tons per hour
to moles per second. NH3 and S02 only required a unit conversion from metric tons per hour to moles
per second.

Finally, the THU dataset included PM10 and individual PM2.5 species, but not the PMC species. We
added the PMC species to the emissions files by subtracting the sum of PM2.5 from PM10. Occasionally,
the sum of PM2.5 emissions slightly exceeded the PM10 value, resulting in a small negative value for
PMC. This was most common in the transport sector. In order to prevent errors in emissions processing or
air quality modeling, any instance of negative PMC was reset to a value of zero.

2.2.3	Spatial Allocation

The next step to prepate the THU China emissions for air quality modeling is to regrid the emissions from
the Lambert 27km grid to the 108km hemispheric grid. For this, the I/O API utility mtxcalc was run to
create a conversion matrix from the 27km China grid to the 108km hemispheric grid. Then, a Python
script was used to re-project the hourly speciated emissions from the 27km China grid to the 108km
hemispheric grid using the conversion matrix from mtxcalc. Early versions of I/O API v3.2 are known to
produce incorrect conversion matrices from mtxcalc, but this has been fixed in the latest build of v3.2; I/O
API v3.1 is also valid.

2.2.4	Vertical Allocation

Vertical allocation was applied to the hourly, speciated, and gridded emissions, consistent with how the
HTAP sectors are processed. China sectors that are mapped to the HTAP industry sector (see Table 2-6)
were vertically allocated using the same layer fractions that are used to allocate the g industry sector.
Similarly, the China power plant sector was vertically allocated using the layer fractions from the
g energy sector. All Chinese domestic, transport, and agriculture emissions were retained as single layer,
low-level emissions.

Layer fractions were applied using the SMOKE program Layalloc using a similar procedure as for the
HTAP sectors g industry and g energy. As with the HTAP sectors, because layer heights vary
temporally, layered emissions for the China industry and energy sectors are processed for each day of the
year, separately for 2016 (base year) and 2015 (spinup year).

The resulting China emissions files, which include hourly temporalization, full speciation, vertical
allocation, and spatial allocation on the hemispheric 108km grid, are read directly by the SMOKE
program Mrggrid and merged with the other sectors.

2.3 Global FINN fires (g_ptfire, g_ptagfire)

Annual 2015 and 2016 fire emissions outside of North America were developed from FINN (Fire
Inventory from NCAR) vl.5 daily fire emissions. For FINN fires, listed vegetation type codes of 1 and 9

21


-------
are defined as agricultural burning, all other fire detections and assumed to be wildfires. Wildfires are
processed in the g_ptfire sector separately from ag burning in the g_ptagfire sector. All wildland fires that
are not defined as agricultural are assumed to be wild fires rather than prescribed. FINN fire detects less
than 50 square meters (0.012 acres) are removed from the inventory because they are smaller than the
pixel size of the remote sensing data and are therefore assumed to be false detects. Fires in North
America, Central America, and the Caribbean were excluded from the g_ptfire and g_ptagfire sectors
because fires in these areas are included in North America portion of the hemispheric platform (ptfire,
ptagfire, and ptfire othna sectors). In China, we include fire emissions from the g_ptfire and g_ptagfire
sectors and do not include open burning emissions from the THU inventory.

Global fire processing uses a unique set of country codes that are similar to, but not the same as, standard
FIPS codes used in North America. The locations of FINN fires are geocoded from latitude and longitude
to a country code and a time zone code. The 6-digit FlPS-like codes used for global fires follow the
format 9TTCCC, where TT indicates the time zone (01 through 24), and CCC is the country-level
GEOCODE that is used for HTAP sector processing. This allows for the generation of emissions
summaries by country, and supports diurnal temporalization of fires in local time. Diurnal profiles for
global FINN wild fires and ag fires are shown in Figure 2-7.

Figure 2-7. Diurnal temporal profiles for global wild fires and ag fires

Global fire diurnal profiles

18.0056
16.00%

14.00%

12.00%

10.00%

8.00%

6.00%

4.00%

2.00%

0.00%

°_ptf r e	g_ptagfire

The FINN fire inventory includes heat content and location information specific to individual fires, which
can be used to calculate fire-specific plume rise as is done for North America fires. However, given the
large number of fires that may exist around the globe at any given time, processing all global fires
individually through the SMOKE model is more computationally intensive than can be accomplished on
EPA computing systems in a reasonable amount of time. To reduce the computational complexity, we
reduce the number of unique fires in the global wildfire and ag fire datasets. After fires are geocoded to
time zone and country, latitude and longitude coordinates for each fire are rounded to the nearest tenth of
a degree (0.1°). Then, all fires that share the same time zone, country code, and rounded coordinates, are
summed to a single inventory source. This reduces the size of the annual fire inventory by a factor of
approximately 15, and the size of the daily fire inventory by a factor of approximately 10. While this does
reduce the spatial accuracy of the inventory, this new resolution is equivalent to the resolution of gridded
HTAP emissions inventories and is still more than 10 times more precise than the resolution of the 108km
hemispheric modeling domain, and so the impact of coordinate rounding on 108km air quality modeling
is minimal.

hour of day (local time)

22


-------
When summing fire sources that share a common time zone, country code, and rounded coordinates, we
lose the fire-specific heat content information that is needed to calculate plume rise. So for the g_ptfire
and g_ptagfire sectors, we apply sector-wide layer fractions with the SMOKE program Layalloc, similar
to how HTAP and THU China sectors are layered. Different layer fractions are applied to g_ptfire and
g_ptagfire; this is the main reason wild fires and ag fires are processed separately. The layer fractions for
the two global fire sectors are shown in Figure 2-8.

Figure 2-8. Layer fractions for global wild fires (g_ptfire) and ag fires (g_ptagfire)

Global fire layerfractions

g_ptagfire	g_ptf"re

Annual emissions totals for the g_ptfire and g_ptagfire sectors are in Table 2-8.

Table 2-8. Emissions totals for global fire sectors, 108km hemispheric domain, year 2016, tons/year

Sector

CO

NH3

NOX

PM10

PM2.5

S02

voc

g ptfire

121,285,767

2,005,131

6,058,488

21,767,734

13,499,258

780,250

34,654,568

g ptagfire

25,959,983

382,301

1,576,886

2,362,013

1,847,549

150,067

9,932,375

2.4 Other global emissions

In addition to emissions from the North America regional modeling platform, HTAP, THU (China), and
global fires from FINN, the hemispheric modeling platform also includes biogenic emissions, lightning
NOx emissions, and ocean chlorine emissions.

2.4.1 Biogenic emissions

Biogenic emissions for the 108km hemispheric modeling domain consist of two components: emissions
from the Biogenic Emission Inventory System (BEIS) version 3.61 in and near the United States, and
emissions from the Model of Emissions of Gases and Aerosols from Nature (MEGAN) elsewhere in the
domain. BEIS emissions from 12km regional modeling were elected for the region surrounding the
United States to be as consistent with the regional modeling platform as possible. However, the BEIS
model cannot be used globally because the land use data upon which BEIS relies, specifically the BELD
dataset, does not extend beyond North America. Therefore, we use biogenics from the MEGAN model to
fill in the rest of the hemispheric domain.

BEIS 3.61 was run using SMOKE for the 12US1 regional modeling domain. The gridded emissions from
12US1 were then spatially allocated to the 108km hemispheric domain using a Python re-gridding utility.

23


-------
Figure 2-9 shows the spatial extent of 12US1 BEIS emissions on the 108km hemispheric domain. This is
a plot of emissions on the hemispheric domain, zoomed in on North America; the red region represents
nonzero emissions from 12US1 BEIS. To simplify the process of preventing double-counted emissions
along the edges of the red area in Figure 9, rather than identify each individual cell as either a "BEIS" cell
or a "MEGAN" cell, we identified a rectangular region which has 100% 12US1 BEIS coverage -
specifically, (73,31) to (124,64), as shown in blue on Figure 2-9. For BEIS, we keep all emissions inside
the blue rectangle and zero out all emissions outside the blue rectangle. Conversely, for MEGAN, we
keep all emissions outside the blue rectangle and zero out all emissions inside the blue rectangle. All zero-
outs were performed using a Python-based netCDF utility. Sector names for these two sectors are "beis"
for the BEIS component and "g biog" (global biogenics) for the MEGAN component.

Figure 2-9. Spatial extent of 12US1 BEIS emissions on the 108km hemispheric domain

Biogenic emissions from the MEGAN model were processed for a 2° x 2.5° global GEOS-Chem domain.
The MEGAN emissions were reallocated to the 108km hemispheric domain. Model output mass flux
emissions values were converted to emissions rates using the area of each grid cell. Gas phase species
values were then converted from mass rates to mole emissions rates. Volatile organic species were
combined to match the CB6 species needed for CMAQ. To avoid the double counting of biogenic
emissions, the region in the blue rectangle zeroed out as described above. Biogenic emissions from
MEGAN are processed for a single representative day per month for both 2015 and 2016. Formaldehyde
emissions from MEGAN were lower than expected compared to alternative datasets, and so FORM
emissions were increased by 70% compared to MEGAN outputs.

Both BEIS and MEGAN include NOx emissions from soil. The total Co, NO and VOC emissions are
shown in Table 2-9.

24


-------
Table 2-9. Annual biogenic emissions totals, 108km hemispheric domain, year 2016, tons/year

Sector

CO

NO

VOC1

beis (US + vicinity)

9,956,463

1,961,185

58,179,072

megan (rest of domain)

34,386,775

16,252,008

260,825,447

1 VOC emissions are approximated using a sum of individual VOC species.

2.4.2 Lightning NOx emissions

NOx emissions from lightning strikes, known as Lightning NOx, were developed. Monthly lightning NOx
emissions including vertical allocation for all 44 layers were created from a global inventory (Price,
1997). Lightning NOx was speciated to model species NO and N02 using a standard 90/10 split.

Diurnal temporalization for lightning NOx was applied using lightning strike flash rate data from
Blakeslee, et al., 2014 (http://www.sciencedirect.eom/science/article/pii/S0169809512003250#t0010).
This lightning data includes diurnal variation in lightning strikes in UTC by season (summer, fall, winter,
spring) and by continent. Diurnal temporal profiles were calculated from this data for each season and
continent and then applied to the monthly total lightning NOx emissions, using the GRIDMASK file to
map cells to continents. Lightning strikes over open water use a flat diurnal profile.

There is no day-of-month temporalization, and so there is a single hourly lightning NOx emissions file for
each month. The same lightning NOx emissions are used for both 2015 (spinup year) and 2016 (base
year).

Figure 2-10 shows a plot of lightning NOx annual emissions for the 108km hemispheric domain. There
are approximately 30 million tons per year of lightning NOx emissions.

25


-------
Figure 2-10. Lightning NOx Annual emissions
Annual total lightning NOx

2.4.3 Ocean chlorine emissions

The ocean chlorine gas emission estimates are based on the build-up of molecular chlorine (C12)
concentrations in oceanic air masses (Bullock and Brehme, 2002). Ocean chlorine emissions for the
108km hemispheric domain were processed using an ocean mask file for the hemispheric domain,
representing the amount of salt water coverage in each grid cell in the domain.

2.5 Changes to North America sectors to support hemispheric modeling

Documentation regarding emissions from the United States, Canada, and Mexico, is available in the 2016
alpha regional platform TSD. This section covers differences between the regional platform and the
hemispheric platform concerning emissions in North America.

2.5.1 Emissions inventory differences

As discussed in the next section, hemispheric emissions in the onroad and afdust_adj sectors were
reallocated from 12US1 to the 108km hemispheric domain. Therefore, those two sectors exclude
emissions in Alaska, Hawaii, Puerto Rico, and Virgin Islands, collectively referred to as AK/HI/PR/VI. In
order to include AK/HI/PR/VI onroad and afdust emissions in the hemispheric modeling platform without
rerunning the onroad or afdust emissions for the entire U.S., the AK/HI/PR/VI onroad inventory from

26


-------
2014NEIv2 was included in the onroad can (Canada onroad) sector, and the AK/HI/PR/VI afdust
inventory from 2014NEIv2 was included in the othafdust (Canada afdust) sector.

In the regional modeling platform, the cmv_c3 sector includes emissions in U.S. state waters, U.S. federal
waters, and "non-US" emissions beyond federal waters. In the hemispheric modeling platform, the
cmv_c3 sector includes emissions in U.S. state and federal waters, but does not include non-US emissions
beyond federal waters. As discussed in the HTAP section of this document, non-US C3 emissions are
excluded from the hemispheric cmv_c3 sector to prevent a double count with emissions from the HTAP
ships sector.

Biogenics, as discussed in Section 2.4.1, are based on a combination of the BEIS and MEGAN models in
the hemispheric platform, compared to the regional platform which only uses BEIS.

Otherwise, the 2016 hemispheric modeling platform uses the same North America emissions inventories
as the 2016 alpha regional modeling platform.

2.5.2 Spatial Allocation

For most North America sectors, emissions were processed for the 108km hemispheric domain in
SMOKE using spatial surrogates for the hemispheric domain. For some North America sectors,
emissions were processed for the 12US1 regional modeling domain and then reallocated to the 108km
hemispheric domain, in order to ensure consistency between regional modeling and hemispheric modeling
for sectors that are sensitive to meteorology or other grid-specific inputs. Table 2-10 specifies the
approach for all North America emissions sectors.

Table 2-10. Hemispheric spatial allocation approach for all North America sectors

Platform Sector

Spatial allocation approach

afdust adj

Regridded from 12US1; excludes AK/HI/PR/VI

onroad

Regridded from 12US1; excludes AK/HI/PR/VI

onroad ca adj

Regridded from 12US1

ag

Processed with 108km surrogates; temporalization based on 12US1 meteorology

cmv clc2

Processed with 108km surrogates

cmv c3

Point sources allocated to 108km grid; excludes non-US C3

nonpt

Processed with 108km surrogates

nonroad

Processed with 108km surrogates

np oilgas

Processed with 108km surrogates

rail

Processed with 108km surrogates

rwc

Processed with 108km surrogates; temporalization based on 12US1 meteorology

othafdust

Processed with 108km surrogates and adjusted with 108km meteorology; includes AK/HI/PR/VI

onroad can

Processed with 108km surrogates; includes AK/HI/PR/VI

onroad mex

Processed with 108km surrogates

othar

Processed with 108km surrogates

othpt

Point sources allocated to 108km grid

ptagfire

Point sources allocated to 108km grid

ptegu

Point sources allocated to 108km grid

ptnonipm

Point sources allocated to 108km grid

pt oilgas

Point sources allocated to 108km grid

Ptfire

Point sources allocated to 108km grid

ptfire othna

Point sources allocated to 108km grid

Beis

Regridded from 12US1 and zeroed out outside of (73,31) to (124,64); see Section 2.4.1

27


-------
2.5.3 Vertical Allocation

For regional modeling, point source emissions are processed as "inline" point sources, and plume rise is
calculated not within SMOKE, but within CMAQ. Inline CMAQ point source layering requires stack
information for individual point sources. Because the gridded HTAP inventories do not have stack
information from which plume rise can be calculated within CMAQ, we must process layering and create
3-D layered emissions in SMOKE for all sectors, including North America sectors for which we do have
stack parameters.

For North America sectors, plume rise is calculated using the SMOKE program Laypoint, which uses heat
content for fires and stack parameters for other point sources to estimate plume rise on a source-by-source
basis. This is for the following sectors: cmv_c3, othpt, ptegu, ptnonipm, pt_oilgas, ptagfire, ptfire, and
ptfire_othna.

2.6 Final CM A Q-Rea dy Hemispheric Emissions

The first two sections of this technical support document has highlighted the various inventory and
ancillary datasets that were used to generate CMAQ-ready emissions for modeling air quality for the
Northern Hemispheric modeling domain. Annual NOx and S02 emissions are shown in Figure 2-11.
These annual emissions can be used to compare versus the various sensitivity CMAQ runs discussed in
next section.

Figure 2-11. Annual NOx(left) and S02(right) emissions for base case run.

28


-------
3 Sensitivity Runs

In addition to the "base case" modeling, CMAQ modeling was performed for several sensitivity cases.
Most of the sensitivity cases involve zero outs to emissions in a particular region, or a particular category
of emissions, or a combination. For each sensitivity case, we processed new emissions for both year 2015
and year 2016. The following sensitivity cases were processed:

-	ZANTH: Zero out all anthropogenic emissions in the entire domain.

-	ZROW: Zero out all anthropogenic emissions outside the United States.

-	ZUSA: Zero out all anthropogenic emissions inside the United States.

ZSHIP: Zero out all CMV C3 emissions in the entire domain.

-	ZCHN: Zero out all anthropogenic emissions in China.

-	ZIND: Zero out all anthropogenic emissions in India.

-	ZCANMEX: Zero out all anthropogenic emissions in Canada and Mexico.

ZFIRE: Zero out all fire emissions in the entire domain.

-	EdgarCHN: Use HTAPv2 emissions in China instead of the THU China dataset.

The specifications for each sector in each sensitivity run are summarized in Table 3-1, with additional
detail in Section 3.1.

Table 3-1. Hemispheric sensitivity run sector table summary

Sector

Includes

ZANTH

ZROW

ZUSA

ZSHIP

ZCHN

ZIND

ZCANMEX

ZFIRE

EdgarCHN

afdust adj

Continental U.S.

zero out



zero out













ag

U.S.

zero out



zero out















U.S. state+federal



















cmv clc2

waters

zero out



zero out















U.S. state+federal



















cmv c3

waters

zero out



zero out

zero out











nonpt

U.S.

zero out



zero out













nonroad

U.S.

zero out



zero out













np oilgas

U.S.

zero out



zero out















Continental U.S.



















onroad

except California

zero out



zero out













onroad ca adj

California

zero out



zero out













pt oilgas

U.S.

zero out



zero out













ptagfire

U.S.

zero out



zero out









zero out



ptegu

U.S.

zero out



zero out













ptnonipm

U.S.

zero out



zero out













rail

U.S.

zero out



zero out













rwc

U.S.

zero out



zero out





















zero out















Canada +



zero out

AK HI.







zero out





onroad can

AKHI/PRVI

zero out



PR VI













onroad mex

Mexico

zero out

zero out









zero out













zero out















Canada +



zero out

AK. HI.







zero out





othafdust

AK/HI/PR/VI

zero out



PR VI























zero out





















Canada
C3











othar

Canada + Mexico

zero out

zero out



CMV





zero out





othpt

Canada + Mexico

zero out

zero out









zero out









zero out



zero out













ptfire

U.S.

prescribed



prescribed









zero out





Canada. Mexico.

zero out

zero out









zero out





ptfire othna

Central America,
Caribbean

prescribed

prescribed









prescribed

zero out





Outside North



















g ptfire

America















zero out



29


-------
Sector

Includes

ZANTH

ZROW

ZUSA

7. SHIP

ZCHN

ZIND

ZCANMEX

ZFIRE

EdgarCHN

g ptagfire

Outside North
America

zero out

zero out





zero
out

zero
out



zero out



g ag

Outside

US/CAN/MEX/China

zero out

zero out







zero
out





rerun with
China

g air cds

Entire domain

zero out

zero out
all except
US

zero out
US



zero
out

zero
out

zero out
Canada,





g air crs

Entire domain

zero out

zero out
all except
US

zero out
US



zero
out

zero
out

zero out
Canada,





g air lto

Outside US/CAN

zero out

zero out





zero
out

zero
out

zero out





g energy

Outside

US/CAN/MEX/China
(includes CO in
China)

zero out

zero out





zero
out
China
CO

zero
out





rerun with
China

g industry

Outside

US/CAN/MEX/China
(includes CO in
China)

zero out

zero out





zero
out
China
CO

zero
out





rerun with
China

g residential

Outside

US/CAN/MEX/China
(includes CO in
China)

zero out

zero out





zero
out
China
CO

zero
out





rerun with
China

g ships

Outside US Federal
Waters

zero out

zero out



zero out

zero
out

zero
out







g transport

Outside

US/CAN/MEX/China
(includes CO in
China)

zero out

zero out





zero
out
China
CO

zero
out





rerun with
China

all THU China

China

zero out

zero out





zero
out







zero out

beis

Continental U.S. and
vicinity



















g biog

Entire domain except
Continental U.S.
vicinity



















lightning

Entire domain



















ocean cl2

Entire domain



















3.1 Sensitivity Run Specifications

Zero-out sensitivity cases are processed by excluding emissions sectors from the final sector merge,
reprocessing sectors with a partial zero-out, and reusing remaining sectors from the base run, or from
other previously completed sensitivity runs.

3.1.1 ZANTH

The ZANTH run excludes all anthropogenic emissions in the entire hemispheric modeling domain, in
order to isolate the impact of non-anthropogenic emissions on air quality modeling results. This means
almost all sectors from the base case are excluded from the ZANTH case. The only emissions included in
the ZANTH run are:

-	Biogenics (beis and g_biog sectors)

Lightning NOx

Ocean chlorine

-	Wildfires from the ptfire, ptfire othna, and g_ptfire sectors

30


-------
For the purposes of the ZANTH and other sensitivity cases concerning anthropogenic emissions,
prescribed fires and ag fires are considered anthropogenic, while wildfires are considered non-
anthropogenic.

U.S. fires are represented in two sectors: ptfire, which includes wildfires and prescribed fires (labeled with
different SCCs), and ptagfire, which includes ag fires. For the ZANTH run, we exclude the ptagfire
sector, and reprocess the ptfire sector with all prescribed fire emissions removed from the inventory.

Table 3-2 shows the fraction of emissions from wildfires and prescribed burns in 2016.

Table 3-2. Fraction of 2016 emissions from wildfire and prescribed burns in USA

Pollutant

Wild

Prescribed

CO

60.10%

39.90%

NH3

59.90%

40.10%

NOX

49.20%

50.80%

PM10

58.90%

41.10%

PM25

58.90%

41.40%

S02

54.50%

45.50%

VOC

59.90%

40.10%

Other North America fires are in the ptfireothna sector, which includes both wildfires and ag fires. The
ptfireothna inventories do not have prescribed fires broken out separately; that is, all fires that are not ag
fires share the same SCC. Therefore, all ptfire othna fires that are not ag fires are considered to be
wildfires and are included in ZANTH. For the ZANTH run, we reprocess the ptfire othna sector with all
ag fire emissions removed from the inventory.

Fires outside of North America are in the g_ptfire (wildfires) and g_ptagfire (ag fires) sectors. Like in the
ptfire othna sector, the g_ptfire inventories do not have prescribed fires broken out separately, and so
100% of the g_ptfire sector emissions are included in the ZANTH run. The g_ptagfire sector is excluded
from ZANTH.

31


-------
Figure 3-1. Annual NOx(left) and S02(right) emissions for ZANTH sensitivity run.

emissions total: mrggrid S02

3.1.2 ZROW

The ZROW ran excludes all anthropogenic emissions outside the United States, in order to isolate the
impact of all non-U.S. emissions on air quality modeling results. "ROW" is an acronym which means
"Rest Of World". The ZROW run includes:

-	Everything that is included in ZANTH

All U.S. anthropogenic sectors from the base case

The AK/HI/PR/VI portion of onroadcan and othafdust adj

-	HTAP aircraft sector emissions within the United States

For fire sectors, the ZROW case includes the full ptfire and ptagfire sectors (all U.S. fires, including
prescribed and ag); reuses the ptfire_othna emissions from the ZANTH run (with ag fires removed but
wildfires retained); and includes 100% of g_ptfire from the base run and none of g_ptagfire (same as
ZANTH).

All U.S. sector emissions from the base case are included in full in ZROW. In the base case, the othafdust
and onroad can sectors include a combination of Canada emissions and emissions from Alaska, Hawaii,
Puerto Rico and Virgin Islands; for ZROW, these sectors were reprocessed so that they include
AK/HI/PR/VI only and do not include any Canadian emissions. All sectors which only include
anthropogenic emissions in Canada and Mexico were excluded from ZROW.

Offshore CMV emissions within U.S. federal waters are considered to be United States emissions for the
purposes of these sensitivities. Therefore, ZROW includes the 100% of the emissions from the cmv_clc2
and cmv_c3 sectors, and does not include any emissions from the g_ships sector.

All HTAP sector emissions are excluded from ZROW, with the exception of the g_air_cds and g_air_crs
sectors. The CDS and CRS aircraft sectors are the only HTAP sectors which include emissions within the
United States, or in this case, United States airspace. For ZROW, the g_air_cds and g_air_crs sectors were
reprocessed with all emissions outside the boundaries of the United States zeroed out. For aircraft
emissions, only emissions over land areas of the United States are retained; this includes Alaska, Hawaii,

32


-------
and PR/VI. The g_air_lto sector from the base case does not include any United States emissions, since
they would have double counted emissions from the North America inventories; therefore, we can
exclude g air I to from ZROW entirely.

Figure 3-2. Annual NOx(left) and S02(right) emissions for ZROW sensitivity run.

2016fe zrow hemi annual emissions total: mrggrid NOX





2016fe zrow hemi annual emissions total: mrggrid S02





























&L i





*





>18689
16990



\ r k

, Vv n jh.

¦	>1725

¦	1568



'¦ ' V'"*





15291
13592



' ' ~"H

Hl254



{£si?





11893
10194 I

2
S

8495
6796
5097
3398



i ."V£J \ /"		 1-t .-li

V .rv-v-f

< . £

v ,£ -C'*

H 1097

I

¦	784

I

¦	470
<156















3.1.3 ZUSA

The ZUSA run excludes all anthropogenic emissions inside the United States, in order to isolate the
impact of U.S. emissions on air quality modeling results. ZUSA includes:

-	Everything that is included in ZANTH

All anthropogenic sectors from the base case except those from U.S. sectors
The Canada portion of onroad can and othafdust adj

-	HTAP aircraft sector emissions outside the United States

For fire sectors, the ZUSA case includes the full ptfire_othna, g_ptfire, and g_ptagfire sector emissions
from the base case (all non-U.S. fires, including ag); reuses the ptfire emissions from the ZANTH run
(with prescribed fires removed but wildfires retained); and excludes the ptagfire sector.

All sector emissions from the base case with no U.S. component are included in full in ZROW. In the
base case, the othafdust and onroad_can sectors include a combination of Canada emissions and
emissions from Alaska, Hawaii, Puerto Rico and Virgin Islands; for ZUSA, these sectors were
reprocessed so that they include Canada only and do not include AK/HI/PR/VI only. All sectors which
only include anthropogenic emissions in Canada and Mexico were included in full in ZUSA.

Offshore CMV emissions within U.S. federal waters are considered to be United States emissions for the
purposes of these sensitivities. Therefore, ZUSA excludes the entirety of the cmv_clc2 and cmv_c3
sectors, and includes 100% of emissions from the g_ships sector.

All HTAP sector emissions from the base case are included in full in ZUSA, with the exception of the
g_air_cds and g_air_crs sectors. The CDS and CRS aircraft sectors are the only HTAP sectors which
include emissions within the United States, or in this case, United States airspace. For ZUSA, the

33


-------
g_air_cds and g_air_crs sectors were reprocessed with all emissions inside the boundaries of the United
States zeroed out. For aircraft emissions, only emissions over land areas of the United States, including
AK/HI/PR/VI, are excluded. The g_air_lto sector from the base case does not include any United States
emissions, since they would have double counted emissions from the North America inventories;
therefore, we can include 100% of the g_air_lto sector in ZUSA.

Figure 3-3. Annual NOx(left) and S02(right) emissions for ZUSA sensitivity run.

34


-------
3.1.4 ZSHIP

In the ZSHIP sensitivity run, we exclude all emissions from CMV Category 3 (C3) emissions for the
entire domain, in order to isolate the impact of shipping emissions on model results.

The only differences between ZSHIP and the base case concern the cmv_c3, othar, and g_ships sectors. In
the United States, ZSHIP includes 100% of the cmv_clc2 sector but excludes the cmv_c3 sector. In
Canada, C3 emissions are included in the othar sector, and so we reprocess the othar sector for ZSHIP
with Canada C3 sources removed from the inventory. For the rest of the domain, because we cannot
isolate C3 ships from other types of ships in the HTAP shipping sector, we exclude 100% of the g ships
sector emissions from ZSHIP.

Figure 3-4. Percent difference from basecase NOx(left) and S02(right) weekday-summer emissions

for ZSHIP sensitivity run.

3.1.5 2CHN

The ZCHN ran excludes all anthropogenic emissions inside China, in order to isolate the impact of China
emissions on air quality modeling results. All emissions in ZCHN are the same as in the base case, except
for the HTAP, THU China, and global FINN fire sectors.

Recall that emissions sectors based on HTAP inventories, such as g_residential and g_transport, include
carbon monoxide emissions in China, since the THU China dataset does not include CO emissions
Therefore, it was necessary to reprocess all HTAP sectors for ZCHN, where instead of retaining CO
emissions in China, we zero out emissions for all pollutants in China. Emissions were zeroed out in both
China land areas and along the China coastline, as defined by the extent of the THU China dataset (see
Section 2.2). Aircraft emissions over China were zeroed out from the g_air_cds, g_air_crs, and g_air_lto
sectors. Emissions in Taiwan were not zeroed out in ZCHN.

All emissions from the THU China dataset are excluded from ZCHN for all sectors.

For fire sectors, emissions in the g_ptfire sector are retained in full, since that sector's emissions are
treated as wildfires and as non-anthropogenic. We reprocessed the g_ptagfire sector for ZCHN with all
fires that have a China GEOCODE removed from the inventory.

35


-------
Figure 3-5. Annual NOx(left) and S02(right) emissions for ZCHN sensitivity run.

2016fe zchn hemi annua! emissions total: mrggrid NOX





2016fe zchn hemi annual emissions total: mrggrid S02





X A







V

p) v 7k





^ / "S- > -

y . *, ',>*¦ > ¦

r





>41144
37404

29923
26183



% ' -;r- f ¦

' t'.',



>29737
27033
24330
21626
18923

Iv ?

f,

< set'W Mlrv 0.1«« ( .i"





22442 |

18702

14961

11221

7480

<3740



-A I .

iv ¦

" yS - If;
*.*• ,/v

* $sL

"j >-•'



16220 |

1

13516
10813
8110
5406
<2703

3.1.6 ZIND

The ZIND run excludes all anthropogenic emissions inside India, in order to isolate the impact of India
emissions on air quality modeling results. All emissions in ZIND are the same as in the base case, except
for the HTAP and global FINN fire sectors.

For ZIND, we reprocess all HTAP sectors with all emissions in India, according to the India GEOCODE,
zeroed out. This includes aircraft emissions as well as other types of emissions in the I-ITAP dataset.

For fire sectors, emissions in the g_ptfire sector are retained in full, since that sector's emissions are
treated as wildfires and as non-anthropogenic. We reprocess the g_ptagfire sector for ZIND with all fires
that have an India GEOCODE removed from the inventory.

Figure 3-6. Annual NOx(left) and S02(right) emissions for ZIND sensitivity run.

2016fe zind hemi annual emissions total: mrggrid NOX





2016fe zind hemi annual emissions total: mrggrid S02







? X- u















>-V .SUkA M t

- ¦< *< V ¦

%* . -$l





>46566
42332

33866

a

— j



' .'*»& j J\ yn f





>35967
32697
29428
26158
22888
19618 |f



ivr

* "VH* JSP

Kj * , ^





12699
8466
<4233



*: cj 'r

. - . • *?





16348
13079
9809
6539
<3269

























Hfi1, l.CUc.tK Mm: 8.0	





36


-------
3.1.7 ZCANMEX

The ZCANMEX am excludes all anthropogenic emissions inside Canada and Mexico, in order to isolate
the impact of Canadian and Mexican emissions on air quality modeling results. ZCANMEX includes:

-	Everything that is included in ZANTH

All anthropogenic sectors from the base case except those from Canada and Mexico sectors
The AK/HI/PR/VI portion of onroad can and othafdust adj

-	HTAP aircraft sector emissions outside Canada and Mexico

For fire sectors, the ZCANMEX case includes the full ptfire, ptagfire, g_ptfire, and g_ptagfire sector
emissions from the base case. For the ptfire_othna sector, we reprocess the emissions with all Canada and
Mexico ag fires removed, but we retain all Canada and Mexico wildfires. The ZCANMEX ptfire_othna
sector is equivalent to the ZANTH ptfire_othna sector, except that ZCANMEX ptfire othna includes ag
fires in Central America and the Caribbean, whereas Z ANTH ptfire othna does not include any ag fires.

All sector emissions from the base case with no Canada or Mexico component are included in full in
ZROW. In the base case, the othafdust and onroad_can sectors include a combination of Canada
emissions and emissions from Alaska, Hawaii, Puerto Rico and Virgin Islands. For ZCANMEX, we reuse
othafdust_adj and onroad_can emissions from the ZROW run, which include AK/HI/PR/VI only and do
not include Canada. All sectors which only include anthropogenic emissions in Canada and Mexico were
excluded from ZCANMEX.

All HTAP sector emissions from the base case are included in full in ZCANMEX, with the exception of
the g_air_cds, g_air_crs, and g air lto sectors. The aircraft sectors are the only HTAP sectors which
include emissions within Canada or Mexico. For ZCANMEX, the three HTAP aircraft sectors were
reprocessed with all emissions inside the boundaries of Canada and Mexico zeroed out. Note that in the
base case, Canada emissions were already removed from g_air_lto to prevent a double count with the
Canadian aircraft inventory, but Mexico was not already removed from g air lto.

Figure 3-7. Annual NOx(left) and S02(right) emissions for ZCANMEX sensitivity run.

2016fe zcanmex hemi annual emissions total: mrgqrid NOX

7 '

. • tS

H.. 4	M-vOMU		f	'		

2016fe zcanmex hemi annual emissions total: mrgqrid S02

? V

• -••• 4 .W

¦¦ . v , -¦

v:









if',

iTV

:

I 36555
32899

37


-------
3.1.8 2FIRE

In the ZFIRE sensitivity simulation, all fire emissions in the entire domain are excluded, in order to
isolate the impact of fire emissions in air quality model results. Wildfires, prescribed fires, and ag fires are
all excluded from ZFIRE.

All emissions from the base case are included in ZFIRE, except for the ptfire, ptagfire, ptfire_othna,
g_ptfire, and g_ptagfire sectors, each of which is excluded from ZFIRE.

Figure 3-8. Annual NOx(left) and SG2(right) emissions for ZFIRE sensitivity run.

2016fe zfire hemi annual emissions total: mrggrid NOX



2016fe zfire hemi annual emissions total: mrggrid S02







; k





> ...







•> f'/f ?% /VH. -r-

" ,/fi >>. k

S •" \i{r

Vr" ' ¦



>51533



- M % .. f





>42755





46948
42164



m ^





38868
34981



'

¦it,

i'



8

28109 |
3

18739
1 14054

'<4684



- S, ' S- 'J

'' < ' C ( Ss •

. *r~*. » ~

i-'-m

WW- CT X

\.:A .. . - . ¦ ¦¦ >¦ Yt

* "Kr . • V





27207

|

23321 |
S

19434
15547
11660

<3886



01S18 	(- - '	





tK mWilli mhjul	* - *		





3.1.9 EdgarCHN

The EdgarCHN sensitivity is not a zero-out simulation, but instead uses alternative data in China. Instead
of using the THU China emissions, for EdgarCHN we use the HTAP emissions projected to 2014 using
CEDS factors. In other words, in EdgarCHN, China is treated the same way as other countries outside
North America. "Edgar" refers to the Emissions Database for Global Atmospheric Research, which
includes the HTAP v2 dataset used for this study.

For the EdgarCHN case, we exclude all THU China sector emissions. We then reprocess all HTAP
sectors which were affected by China zero-outs in the base case: g_ag, g_energy, g_industry,
g residential, and g_transport. Instead of zeroing out China emissions in those sectors, for EdgarCHN we
keep all China emissions, and also apply 2010-to-2014 growth factors from the CEDS dataset (see
Section 2.1.5). In the base case, we did not zero out China emissions from the HTAP aircraft or shipping
sectors, so those sectors' emissions can be reused from the base case without reprocessing.

Other than the HTAP and THU China sectors, the EdgarCHN case is equivalent to the base case,
including fires and biogenics in China.

38


-------
Figure 3-9. Percent difference NOx(left) and S02(right) weekday-summer emissions for

EdgarCHN sensitivity run.

2016fe edchn hem

minus 2016fe_hemi 20160712 daily emissions percent difference

: total merged NO>

2016fe edchn hemi minus 2016fe_hemi 20160712 daily emissions percent difference

: total merged SOI



'/\ )' i-









I .=- •









'' * J \









> .. S*' V \ 1









'¦. (\ ?). , , V







|
1





"" a!'>: -

fi .



























rlSr

m















sp

' 1





1

;





Wv.":-''





1













1





< 'x. />•v

1 •>



¦- Or





•I



» T

- -J fr
/















. s*'

. tJ



1







f " "•

- >rf
y ">?



1

'

4 Emission summaries

Table 4-1 summarizes emissions for the 2016 hemispheric modeling platform by sector for the entire
108km hemispheric modeling domain. The afdust and othafdust sector emissions represent the summaries
after application of both the land use (transport fraction) and meteorological adjustments; therefore, these
sectors are called "afdust_adj" and "othafdust adj" in these summaries. The onroad sector totals are post-
SMOKE-MOVES totals, representing air quality model-ready emission totals. CARB emissions for
California are reported under the onroad_ca_adj sector. For CMV, the cmv_clc2 and cmv_c3 sectors
include emissions in U.S. state and federal waters, the othar sector includes CMV emissions in Canadian
waters, and the g_ships sector includes CMV emissions elsewhere. Table 4-2 shows annual domain total
emissions for each of the sensitivity cases.

Table 4-1. Domain total emissions by sector for 2016 hemispheric platform, tons/year

Sector

Includes

CO

NH3

NOX

PM10

PM2.5

S02

voc

afdust adj

Continental U.S.







6,216,650

874,142





ag

U.S.



2,778,508









180,293



U.S. state+federal















cmv clc2

waters

116,380

335

611,167

17,366

16,713

580

10,842



U.S. state+federal















cmv c3

waters

57,469

97

596,900

19,212

17,797

15,919

26,072

nonpt

U.S.

2,754,695

122,447

767,493

618,831

504,676

164,401

3,734,585

noiiroad

U.S.

12,425,503

2,282

1,209,937

123,000

116,208

2,448

1,513,284

np oilgas

U.S.

646,802

15

680,416

17,847

17,580

39,132

3,020,375



Continental U.S.















onroad

except California

19,646,578

87,562

3,815,719

251,155

120,018

25,654

1,857,060

onroad ca adj

California

799,749

13,667

230,117

21,700

10,245

1.702

104,935



Canada +















onroad can

AK/HI/PR/VI

2,305,624

9,519

481,221

29,091

20,552

1.746

203,278

onroad mex

Mexico

6,297,363

10,351

1,501,683

74,477

57,028

26,521

554,751



Canada +















othafdust adi

AK/HI/PR/VI







2,461,411

476,986





othar

Canada + Mexico

6,068,987

1,393,414

1,414,212

1,059,151

680,593

67,747

5,050,225

othpt

Canada + Mexico

2,353,692

57,497

1,844,920

399,581

263,139

3,402,750

1,338,420

pt oilgas

U.S.

238,554

4,385

450,675

13,724

12,618

43,909

183,275

ptagfire

U.S.

593,082

80,365

18,298

96,340

68,103

5,636

36,121

ptegu

U.S.

672,384

25,018

1,289,229

171,237

140.845

1,544,799

33,467

ptfire

U.S.

37,941,136

621,241

441,988

3,792,193

3,213,737

261,975

8,930,350

39


-------
Sector

Includes

CO

NH3

NOX

PM10

PM2.5

S02

voc



Canada, Mexico,

















Central America,















ptfire othna

Caribbean

18,814,941

308,314

725,893

2,595,963

1,841,790

137,980

5,410,402

ptnonipm

U.S.

1,879,574

61,825

1,136,271

414,564

268,625

723,900

817,149

rail

U.S.

118,925

364

675,285

20,806

19,227

702

34,874

rwc

U.S.

2,121,291

15,514

30,856

317,792

317,266

7,754

341,934



Continental U.S. and















beis

vicinity

9,956,463



1,961,185







58,179,072



Entire domain except
Continental U.S.















g biog

vicinity

34,386,775



16,252,008







260,825,447



Outside















g ag

US/CAN/MEX/China



27,053,284













Outside North















g ptfire

America

121,285,767

2,005,131

6,058,488

21,767,734

13,499,258

780,250

34,654,568



Outside North















g ptagfire

America

25,959,983

382,301

1,576,886

2,362,013

1,847,549

150,067

9,932,375

g air cds

Entire domain

134,140



1,089,169

18,439

18,439

92,194

31,552

g air crs

Entire domain

179,798



1,457,442

24,682

24,682

123,425

42,107

g air lto

Outside US/CAN

116,928



282,723

2,350

2,350

26,785

4,415



Outside

















US/CAN/MEX/China















g energy

(includes CO in
China)

8,597,034

71,485

14,168,795

4,951,791

2,077,552

26,978,678

884,267



Outside

















US/CAN/MEX/China















g industry

(includes CO in
China)

132,095,290

992,589

6,997,163

6,873,446

3,563,763

16,333,374

38,620,852



Outside

















US/CAN/MEX/China















g residential

(includes CO in
China)

232,141,408

4,546,822

4,344,817

15,661,354

11,074,589

3,478,108

33,272,813



Outside US Federal















g ships

Waters

1,175,509



12,078,998

1,207,008

1,207,008

6,633,296

628,816



Outside

















US/CAN/MEX/China















g transport

(includes CO in
China)

99,498,142

189,347

24,530,368

1,827,155

1,678,990

1,621,702

23,083,023

china agrf

China



4,105,865











china agrl

China



6,164,346











china dobi

China





247,159

1,535,996

1,487,994

37,311

247,159

china docb

China













2,254,033

china dofu

China





684,444

2,030,505

977,311

3,894,725

684,444

china do so

China













1,863,873

china doth

China



741,314









213,655

china incb

China





2,520,744

1,453,406

965,475

5,920,805

2,653,154

china ppcb

China





4,390,020

1,175,701

702,486

4,992,558

4,390,020

china prce

China





2,100,046

1,370,862

757,175

1,005,509

2,100,046

china prir

China





603,922

1,300,269

939,834

836,222

603,922

china prso

China













5,792,690

china prot

China



353,494

2,389,083

1,591,475

1,014,461

1,881,593

7,557,787

china trof

China





2,684,613

219,988

208,409

159,214

3,354,114

china tron

China





8,385,993

405,461

384,119

22,164

15,460,059

lightning

Entire domain





30,860,577









TOTAL

Entire domain

781,379,966

52,198,696

163,586,924

84,511,726

51,489,332

81,443,235

540,715,955

40


-------
Table 4-2. Domain total emissions for 2016 hemispheric sensitivity cases, tons/year



CO

NH3

NOX

PM10

PM2.5

S02

voc

Base case

781,379,966

52,198,696

163,586,924

84,511,726

51,489,332

81,443,235

540,715,955

ZANTH

205,815,303

2,661,361

55,994,576

26,485,224

17,144,907

1,053,965

363,772,982

ZROW

263,425,215

6,103,894

68,276,616

36,393,852

20,982,070

3,797,678

379,286,276

ZUSA

723,767,414

48,756,163

151,287,174

74,602,782

47,651,854

78,697,944

525,202,275

/ship

780,131,216

52,198,364

150,753,319

83,277,710

50,257,391

74,788,669

540,053,998

ZCHN

591,973,607

40,781,009

139,226,575

73,209,207

43,875,407

62,661,852

492,117,370

ZIND

701,221,898

40,811,982

151,620,782

74,049,474

43,936,446

69,579,876

520,341,737

ZCANMEX

763,590,698

50,710,585

158,220,989

80,426,511

49,928,210

77,930,685

533,105,441

ZFIRE

576,785,057

48,801,345

154,765,370

53,897,483

31,018,895

80,107,326

481,752,138

EdgarCHN

781,379,966

51,173,214

173,025,698

92,178,616

57,909,585

96,907,344

521,767,906

41


-------
5 References

Adelman, Z. 2012. Memorandum: Fugitive Dust Modeling for the 2008 Emissions Modeling Platform.
UNC Institute for the Environment, Chapel Hill, NC. September, 28, 2012.

Adelman, Z. 2016. 2014 Emissions Modeling Platform Spatial Surrogate Documentation. UNC Institute
for the Environment, Chapel Hill, NC. October 1, 2016.

Adelman, Z., M. Omary, Q. He, J. Zhao and D. Yang, J. Boylan, 2012. "A Detailed Approach for
Improving Continuous Emissions Monitoring Data for Regulatory Air Quality Modeling."
Presented at the 2012 International Emission Inventory Conference, Tampa, Florida. Available
from http://www.epa.gOv/ttn/chief/conference/ei20/index.html#ses-5.

Anderson, G.K.; Sandberg, D.V; Norheim, R.A., 2004. Fire Emission Production Simulator (FEPS)

User's Guide. Available at http://www.fs.fed.us/pnw/fera/feps/FEPS users guide.pdf.

Appel, K.W., Napelenok, S., Hogrefe, C., Pouliot, G., Foley, K.M., Roselle, S.J., Pleim, J.E., Bash, J.,
Pye, H.O.T., Heath, N., Murphy, B., Mathur, R., 2018. Overview and evaluation of the
Community Multiscale Air Quality Model (CMAQ) modeling system version 5.2. In Mensink C.,
Kallos G. (eds), Air Pollution Modeling and its Application XXV. ITM 2016. Springer
Proceedings in Complexity. Springer, Cham. Available at https://doi.org/10.1007/978-3-319-
57645-9 11.

Bash, J.O., Baker, K.R., Beaver, M.R., Park, J.-H., Goldstein, A.H., 2016. Evaluation of improved land
use and canopy representation in BEIS with biogenic VOC measurements in California. Available
from http J/www, geosci-model-dev. net/9/2191/2016/.

Bullock Jr., R, and K. A. Brehme (2002) "Atmospheric mercury simulation using the CMAQ model:

formulation description and analysis of wet deposition results." Atmospheric Environment 36, pp
2135-2146.

Coordinating Research Council (CRC). Report A-100. Improvement of Default Inputs for MOVES and
SMOKE-MOVES. Final Report. February 2017. Available at http://crcsite.wpengine.com/wp-
content/uploads/2019/05/ERG FinalReport CRCA100 28Feb2017.pdf

Drillinginfo, Inc. 2015. "DI Desktop Database powered by HPDI." Internet address:
http://www.enverus.com/

England et al., 2007, Dilution-Based Emissions Sampling from Stationary Sources: Part 2... J. Air &
Waste Manage. Assoc. 57.

England, G., Watson, J., Chow, J., Zielenska, B., Chang, M., Loos, K., Hidy, G. 2007. "Dilution-Based
Emissions Sampling from Stationary Sources: Part 2— Gas-Fired Combustors Compared with
Other Fuel-Fired Systems," Journal of the Air & Waste Management Association, 57:1, 65-78,
DOI: 10.1080/10473289.2007.1046529.

Environ Corp. 2008. Emission Profiles for EPA SPECIATE Database, Part 2: EPAct Fuels (Evaporative
Emissions). Prepared for U. S. EPA, Office of Transportation and Air Quality, September 30,
2008.

42


-------
EPA, 1994. Onboard Refueling Vapor Recovery for Motor Vehicles, Fact Sheet. April, 1994. Available
at: http://www.epa.gov/oms/regs/ld-hwv/onboard/orvrfact.txt.

EPA, 1998. AP-42, Fifth Edition, Compilation of Air Pollutant Emissions Factors, Volume 1: Stationary
Point and Area Sources. September 1998. Available at https://www.epa.gov/air-emissions-
factors-and-quantification/ap-42-compilation-air-emissions-factors.

EPA, 2000. Light-Duty Vehicle, Light-Duty Truck, and Medium-Duty Passenger Vehicle Tier 2 Exhaust
Emission Standards. Office of Transportation and Air Quality, Ann Arbor, MI 48105. Available
at: HYPERLINK "http://www.epa.gov/otaq/standards/light-
duty/tier2stds.htm"http://www. epa.gov/otaq/standards/light-duty/tier2stds.htm.

EPA. 2007a. National Scale Modeling for the Final Mobile Source Air Toxics Rule, Office of Air

Quality Planning and Standards, Emissions Analysis and Monitoring Division, Research Triangle
Park, NC 27711, EPA 454/R-07-002, February 2007. Available at
http://www.regulations.gov/#!documentDetail;D=EPA-HO-OAR-2005-0036-1134.

EPA, 2015a. Population and Activity of On-road Vehicles in MOVES2014 - Draft Report. EPA-420-D-
15-001. Available at http://www.epa.gov/otaq/models/moves/documents/420dl5001.pdf.

EPA, 2015b. Draft Report Speciation Profiles and Toxic Emission Factors for Nonroad Engines. EPA-
420-R-14-028. Available at

https://cfpub.epa.gov/si/si public record Report.cfm?dirEntryId=309339&CFID=83476290&CF
TOKEN=35281617.

EPA,2015c. Speciation of Total Organic Gas and Particulate Matter Emissions from On-road Vehicles in
MOVES2014. EPA-420-R-15-022. Available at
https://nepis.epa.gov/Exe/ZyPDF.cgi?Dockev=Pl OONOJG.pdf

EPA, 2016. SPECIATE Version 4.5 Database Development Documentation, U.S. Environmental

Protection Agency, Office of Research and Development, National Risk Management Research
Laboratory, Research Triangle Park, NC 27711, EPA/600/R-16/294, September 2016. Available
at https://www.epa.gov/sites/production/files/2016-Q9/documents/speciate 4.5.pdf

EPA, 2018. AERMOD Model Formulation and Evaluation Document. EPA-454/R-18-003. U.S.
Environmental Protection Agency, Research Triangle Park, North Carolina 27711.

ERG, 2014a. Develop Mexico Future Year Emissions Final Report. Available at

ftp://newftp.epa.gov/Air/emismod/2016/vl/reports/Mexico Emissions WA%204-
09 final report 121814.pdf

ERG, 2014b. "Technical Memorandum: Modeling Allocation Factors for the 2011 NEI."

ERG, 2016b. "Technical Memorandum: Modeling Allocation Factors for the 2014 Oil and Gas Nonpoint
Tool."

ERG, 2017. "Technical Report: Development of Mexico Emission Inventories for the 2014 Modeling

Platform.". Available at ftp://newftp.epa.gov/Air/emismod/2014/v2/2014fd/emissions/EPA%205-
18%20Report Clean%20Final 01042017.pdf

Eyth, 2016. "Development of 2011 Hemispheric Emissions for CMAQ". Presented at the 15th CMAS
Conference, Chapel Hill, NC. Presentation available at:

https://www.cmascenter.Org/conference//2016/slides/evth development hemispheric 2016.pptx

43


-------
Hildebrandt Ruiz, L. and Yarwood, G., 2013. Interactions between Organic Aerosol and NOy: Influence
on Oxidant Production, Final report for AQRP project 12-012. Available at
http://aqrp.ceer.utexas.edu/projectinfoFY12_13%5C12-012%5C12-012%20Final%20Report.pdf.

Joint Fire Science Program, 2009. Consume 3.0—a software tool for computing fuel consumption. Fire
Science Brief. 66, June 2009. Consume 3.0 is available at:
http://www.fs.fed.us/pnw/fera/research/smoke/consume/index.shtml.

LADCO, 2012. "Regional Air Quality Analyses for Ozone, PM2.5, and Regional Haze: Base C
Emissions Inventory (September 12, 2011)". Lake Michigan Air Directors Consortium,
Rosemont, IL 60018. Available at:

http://www.ladco.org/tech/emis/basecv8/Base C Emissions Documentation Sept 12.pdf.

McCarty, J.L., Korontzi, S., Jutice, C.O., and T. Loboda. 2009. The spatial and temporal distribution of
crop residue burning in the contiguous United States. Science of the Total Environment, 407 (21):
5701-5712.

McQuilling, A. M. & Adams, P. J. Semi-empirical process-based models for ammonia emissions from
beef, swine, and poultry operations in the United States. Atmos. Environ. 120, 127-136 (2015).

MDNR, 2008; "A Minnesota 2008 Residential Fuelwood Assessment Survey of individual household
responses". Minnesota Department of Natural Resources. Available from
http://files.dnr.state.mn.us/forestry/um/residentialfuelwoodassessment07 08.pdf.

NCAR, 2016. FIRE EMISSION FACTORS AND EMISSION INVENTORIES, FINN Data, downloaded
2014 SAPRC99 version from http://bai.acom.ucar.edu/Data/fire/.

NESCAUM, 2006; "Assessment of Outdoor Wood-fired Boilers" (NESCAUM, 2006). Northeast States
for Coordinated Air Use Management (NESCAUM) report. Available from
http://www.nescaum.org/documents/assessment-of-outdoor-wood-fired-boilers/2006-1031-owb-
report revised-iune2006-appendix.pdf.

NYSERDA, 2012; "Environmental, Energy Market, and Health Characterization of Wood-Fired
Hydronic Heater Technologies, Final Report". New York State Energy Research and
Development Authority (NYSERDA). Available from:
http://www.nvserda.ny.gov/Publications/Case-Studies/-

/media/Files/Publications/Research/Environmental/Wood-Fired-Hvdronic-Heater-Tech.ashx.

Pinder, R., Strader, R., Davidson, C. & Adams, P. A temporally and spatially resolved ammonia emission
inventory for dairy cows in the United States. Atmos. Environ. 38.23, 3747-3756 (2004). 2.
Pinder, R., Pekney, N., Davidson, C. & Adams, P. A process-based model of ammonia emissions
from dairy cows: improved temporal and spatial resolution. Atmos. Environ. 38.9, 1357-1365
(2004).

Pinder, R., Pekney, N., Davidson, C. & Adams, P. A process-based model of ammonia emissions from
dairy cows: improved temporal and spatial resolution. Atmos. Environ. 38.9, 1357-1365 (2004).

Pouliot, G., H. Simon, P. Bhave, D. Tong, D. Mobley, T. Pace, and T. Pierce. (2010) "Assessing the
Anthropogenic Fugitive Dust Emission Inventory and Temporal Allocation Using an Updated
Speciation of Particulate Matter." International Emission Inventory Conference, San Antonio, TX.
Available at http://www.epa.gov/ttn/chief/conference/eil9/session9/pouliot.pdf.

44


-------
Pouliot, G. and J. Bash, 2015. Updates to Version 3.61 of the Biogenic Emission Inventory System
(BEIS). Presented at Air and Waste Management Association conference, Raleigh, NC, 2015.

Price, C.G., J.E. Penner and M.J. Prather NOx from lightning, Part I: Global distribution based on
lightning physics Journal of Geophysical Research 102, 5229-5241, No editor, 1997

Pye, H. and Pouliot, G. 2012. Modeling the Role of Alkanes, Polycyclic Aromatic Hydrocarbons, and
Their Oligomers in Secondary Organic Aerosol Formation. Environmental Science &

Technology, 46, 11, 6041-6047.

Raffuse, S., D. Sullivan, L. Chinkin, S. Larkin, R. Solomon, A. Soja, 2007. Integration of Satellite-
Detected and Incident Command Reported Wildfire Information into BlueSky, June 27, 2007.
Available at: http://getbluesky.org/smartfire/docs.cfm.

Reichle, L.,R. Cook, C. Yanca, D. Sonntag, 2015. "Development of organic gas exhaust speciation

profiles for nonroad spark-ignition and compression-ignition engines and equipment", Journal of
the Air & Waste Management Association, 65:10, 1185-1193, DOI:
10.1080/10962247.2015.1020118.

Reff, A., Bhave, P., Simon, H., Pace, T., Pouliot, G., Mobley, J., Houyoux. M. "Emissions Inventory of
PM2.5 Trace Elements across the United States", Environmental Science & Technology 2009 43
(15), 5790-5796, DOI: 10.1021/es802930x

RIGDATA, 2015. U.S. Well Starts By Depth Range, January 2014 through December 2014. Used by

Permission and Approved for Publication by by Lori Spence at RIGDATA (www.rigdata.com) in
e-mail Regi Oommen, Eastern Research Group, Inc. July 9, 2015.

Russell, A.G. and G.R. Cass, 1986. Verification of a Mathematical Model for Aerosol Nitrate andNitric
Acid Formation and Its Use for Control Measure Evaluation, Atmospheric Environment, 20:
2011-2025.

Sarwar, G., S. Roselle, R. Mathur, W. Apel, R. Dennis, "A Comparison of CMAQ HONO predictions
with observations from the Northeast Oxidant and Particle Study", Atmospheric Environment 42
(2008)5760-5770)

Schauer, J., G. Lough, M. Shafer, W. Christensen, M. Arndt, J. DeMinter, J. Park, "Characterization of

Metals Emitted from Motor Vehicles," Health Effects Institute, Research Report 133, March 2006.

Seigneur, C., K . Vigjayaraghavan, K. Lohman, P. Karamchandani, C. Scott. "Global Source Attribution
for Mercury Deposition in the United States." Environ. Sci. Technol. 2004, 38, 555-569.

Seigneur, C., P. Karamchandani, K. Lohman, K . Vigj ayaraghavan. "Multiscale modeling of the

atmospheric fate and transport of mercury." Journal of Geophysical Research, Vol. 106, NO. D21,
Pages 27,795-27,809, November 16, 2001.

Skamarock, W., J. Klemp, J. Dudhia, D. Gill, D. Barker, M. Duda, X. Huang, W. Wang, J. Powers, 2008.
A Description of the Advanced Research WRF Version 3. NCAR Technical Note. National
Center for Atmospheric Research, Mesoscale and Microscale Meteorology Division, Boulder, CO.

45


-------
June 2008. Available at:

https://opensky.ucar.edu/islandora/obiect/technotes%3A500/datastream/PDF/view.

Sullivan D.C., Raffuse S.M., Pryden D.A., Craig K.J., Reid S.B., Wheeler N.J.M., Chinkin L.R., Larkin
N.K., Solomon R., and Strand T. (2008) Development and applications of systems for modeling
emissions and smoke from fires: the BlueSky smoke modeling framework and SMARTFIRE: 17th
International Emissions Inventory Conference, Portland, OR, June 2-5. Available at:
http://www.epa.gov/ttn/chief/conferences.html.

U.S. Census, 2012; "2012 Statistical Abstract, Construction and Housing: Housing Units and
Characteristics". U.S. Census Bureau. Available from:

http://www.census.gov/compendia/statab/cats/construction housing/housing units and characteri
stics.html.

Wang, Y., P. Hopke, O. V. Rattigan, X. Xia, D. C. Chalupa, M. J. Utell. (2011) "Characterization of

Residential Wood Combustion Particles Using the Two-Wavelength Aethalometer", Environ. Sci.
Technol., 45 (17), pp 7387-7393.

Wiedinmyer, C., S.K. Akagi, R.J. Yokelson, L.K. Emmons, J.A. Al-Saadi3, J. J. Orlando1, and A. J. Soja.
(2011) "The Fire INventory from NCAR (FINN): a high resolution global model to estimate the
emissions from open burning", Geosci. Model Dev., 4, 625-641. http://www.geosci-model-
dev.net/4/625/2011/ doi: 10.5194/gmd-4-625-2011

Yarwood, G., J. Jung,, G. Whitten, G. Heo, J. Mellberg, and M. Estes,2010: Updates to the Carbon Bond
Chemical Mechanismfor Version 6 (CB6). Presented at the 9th Annual CMAS Conference, Chapel
Hill, NC. Available at

https://www.cmascenter.org/conference/2010/abstracts/emery updates carbon 2010.pdf.

Zhao, B., Zheng, H., Wang, S., Smith, K.R., Lu, X., Aunan, K., Gu, Y., Wang, Y., Ding, D., Xing, J., Fu,
X., Yang, X., Liou, K.-N., Hao, J., 2018. Change in household fuels dominates the decrease in
PM2.5 exposure and premature mortality in China in 2005-2015. Proceedings of the National
Academy of Sciences 115, 12401-12406. https://doi.org/10.1073/pnas.1812955115

Zue, Henze, et al, 2013. "Constraining U.S. Ammonia Emissions using TES Remote Sensing
Observations and the GEOS-Chem adjoint model", Journal of Geophysical Research:
Atmospheres, 118: 1-14.

46


-------
47


-------
48


-------