* A vo
l ®!
PRO"^
2017 National Emission Inventory Based
Photochemical Modeling for Sector Specific Air
Quality Assessments
-------
EPA-454/R-21-005
July 2021
2017 National Emission Inventory Based Photochemical Modeling for Sector Specific Air
Quality Assessments
U.S. Environmental Protection Agency
Office of Air Quality Planning and Standards
Air Quality Assessment Division
Research Triangle Park, NC
-------
BACKGROUND
This project is intended to update sector-specific air quality surfaces developed from the 2005 National
Emission Inventory (NEI) that were used as input to the Benefits Mapping Analysis Program -
Community Edition (BenMAP-CE) (U.S. Environmental Protection Agency, 2018a) to generate benefit-
per-ton values (Fann et al., 2012) to support regulatory assessments in situations where photochemical
modeling of specific rules was not feasible. The EPA periodically updates these benefit-per-ton values to
account for changes in population (e.g. size and distribution) and economic factors (e.g., inflation,
personal income) that each influence the size of the benefit per-ton value (U.S. Environmental
Protection Agency, 2018b). However, these periodic updates do not account for changes overtime in
the number of facilities, distribution of facilities, and facility specific emissions of each emissions sector;
these are each factors that influence the magnitude of the BPT value. The results of recently performed
sector-specific modeling suggests that these factors should be updated to account for changes since
2005 (Industrial Economics Incorporated, 2019). Mobile source sector benefit estimates have been
updated recently and described elsewhere (Wolfe et al., 2019).
This document presents an overview of EPA photochemical modeling of 2017 NEI based sector specific
emissions on downwind 03 and secondary PM2.5. Ozone contributions were estimated using Ozone
Source Apportionment Technology and PM2.5 contributions using Particulate Source Apportionment
Technology as implemented in the CAMx photochemical model (Ramboll Environ, 2020). The
contribution from each of these emissions sectors to model predicted 03 and PM2.5 sulfate and nitrate
ions were tracked using reactive tracers which track impacts of chemistry, atmospheric transport and
deposition in the photochemical model (Kwok et al., 2015; Kwok et al., 2013; Ramboll Environ, 2020).
Primary emitted PM2.5 was tracked with inert tracers which track impacts of atmospheric transport and
deposition in the photochemical model. All precursor impacts on PM2.5 and 03 are tracked separately
(e.g. NOx to 03, VOC to 03, etc.).
MODEL CONFIGURATION & APPLICATION
Emissions Sectors Tracked for Contribution
The emissions sectors tracked for contribution are shown in Table 1. These emissions sectors were
selected based on a review of sectors that were the subject of past or planned future regulation that
included a health benefits assessment with monetized impacts related to PM2.5 exposure. Sectors with
very small amounts of precursor emissions were not included as part of this assessment.
Table 1. List of emissions sectors tracked for contribution as part of this project.
brick
cement kilns
coke ovens
electric arc furnaces and argon
ferroalloys
gasoline distribution bulk terminals
industrial boilers
integrated iron and steel
internal combusion engines
iron and steel foundries
oil and natural gas
oil and natural gas transmissions
paint stripping / misc surface coating
primary copper smelting
pulp and paper
refineries
residential woodstoves
secondary lead smelters
synthetic organic chemical
taconite mining
1
-------
The industrial boilers and internal combustion engine sectors were tracked by state. All other sector
contribution were tracked with a broader geographic source region delineation based on EPA's U.S.
Nine-region MARKAL Database (LeNOx et al., 2013). A total of 3 regions will be used to track contribution
and were defined to match up to MARKAL regions: 1) WEST = Pacific and Mountain, 2) NORTH = New
England, Middle Atlantic, East North Central, and West North Central, and 3) SOUTH = South Atlantic,
East South Central, and West South Central (Figure 1).
Figure 1. EPA USA9r Regions used to geographically group emission sectors in the U.S. Nine-region
MARKAL database.
R1
New England
R2
Middle Atlantic
NORTH
R3
East North Central
R4
West North Central
R5
South Atlantic
R6
East South Central
SOUTH
R7
West South Central
R8
Mountain
WEST
R9
Pacific
Annual emission totals in tons per year (tpy) for each of the sectors tracked for contribution by region is
shown in Table 2. Emissions are provided for total primarily emitted PM2.5 (PM25PRI), primarily emitted
organic aerosol PM2.s (POA), primarily emitted elemental carbon (PEC), volatile organic compounds
(VOC), ammonia (NH3), sulfur dioxide (S02), and nitrogen oxides (NOx). Emissions are shown for
industrial boilers by State, offshore, and Tribal lands in Table 3 and internal combustion engines similarly
in Table 4. Offshore emissions include emissions from state boundaries to the edge of the economic
exclusion zone (EEZ). Sectors were defined based on North American Industrial Classification System
(NAICS) and Source Classification Code (SCC). Some industrial boilers and internal combustion engine
emissions may also be part of other 18 sectors tracked for contribution. These are the only 2 sectors
that may have sector overlap in this assessment.
2
-------
Table 2. Annual total emissions (tpy) for each sector tracked for contribution.
Sector Name - Region
NOX
S02
NH3
voc
PEC
POA
brick - NORTH
3,458
3,849
79
218
44
383
brick-SOUTH
2,539
5,552
22
702
71
569
brick - WEST
301
292
1
151
5
48
cement kilns - NORTH
36,224
16,432
820
1,755
54
320
cement kilns - SOUTH
45,537
9,378
325
2,464
45
274
cement kilns - WEST
22,818
1,929
554
659
14
84
coke ovens - NORTH
7,635
9,641
24
437
51
142
coke ovens - SOUTH
1,280
1,733
12
170
18
19
coke ovens - WEST
-
-
-
-
-
-
electric arc furnaces and argon - WEST
195
117
-
43
-
3
electric arc furnaces and argon- NORTH
843
591
-
523
1
8
electric arc furnaces and argon- SOUTH
1,752
1,041
-
775
3
36
ferroalloys - NORTH
1,021
2,027
-
175
120
91
ferroalloys - SOUTH
3,245
4,299
3
194
156
111
ferroalloys - SOUTH
-
-
-
-
-
-
gasoline distribution bulk terminals -
NORTH
4
-
-
60,481
-
-
gasoline distribution bulk terminals -
SOUTH
34
-
-
117,591
-
1
gasoline distribution bulk terminals -
WEST
-
-
-
28,186
-
-
integrated iron and steel - NORTH
7,374
7,217
94
2,639
18
887
integrated iron and steel - SOUTH
4,677
1,967
33
865
4
263
integrated iron and steel - WEST
343
60
-
55
1
80
iron and steel foundries - NORTH
1,138
524
57
4,570
45
495
iron and steel foundries - SOUTH
675
246
26
2,058
26
167
iron and steel foundries - WEST
175
163
2
184
1
8
oil and natural gas - NORTH
158,632
13,750
1,790
687,867
191
1,850
oil and natural gas - SOUTH
488,506
45,686
63
1,357,120
699
4,948
oil and natural gas - WEST
149,057
12,933
222
499,533
271
2,448
oil and natural gas transmissions - NORTH
66,288
361
76
6,365
125
1,026
oil and natural gas transmissions - SOUTH
101,291
660
29
20,152
153
1,409
oil and natural gas transmissions - WEST
17,337
321
17
6,271
38
288
paint stripping / misc surface coating
- NORTH
-
-
-
291,043
-
-
paint stripping / misc surface coating
-SOUTH
-
-
-
261,702
-
-
paint stripping /misc surface coating
-WEST
-
-
51
108,850
19
52
primary copper smelting - NORTH
5
141
-
15
-
1
primary copper smelting - SOUTH
3
-
-
-
-
-
primary copper smelting - WEST
245
25,012
3
8
-
7
pulp and paper - NORTH
7,282
1,687
550
4,933
27
170
pulp and paper - SOUTH
49,777
18,649
3,807
56,112
185
942
pulp and paper - WEST
5,388
2,003
466
3,579
28
129
refineries - NORTH
18,221
5,840
434
11,922
266
951
refineries - SOUTH
40,855
14,869
1,486
37,056
620
2,495
refineries - WEST
17,879
7,576
1,620
16,126
240
1,046
residential woodstoves - NORTH
8,195
1,257
3,094
60,605
2,661
42,824
residential woodstoves - SOUTH
5,443
847
2,421
55,948
2,162
34,792
residential woodstoves - WEST
3,663
564
1,582
40,536
1,597
25,702
secondary lead smelters - NORTH
429
3,425
-
65
-
-
secondary lead smelters - SOUTH
298
8,332
-
128
-
-
secondary lead smelters - WEST
-
-
-
-
-
-
synthetic organic chemical - NORTH
10,365
1,960
332
10,814
151
1,269
synthetic organic chemical - SOUTH
21,324
18,390
2,410
13,122
149
822
synthetic organic chemical - WEST
608
51
6
348
7
26
taconite mining - NORTH
38,192
5,090
24
402
124
1,146
taconite mining - SOUTH
-
-
-
-
-
-
taconite mining - WEST
3
-------
Table 3. Industrial
boiler annual total
emissions
(tpy) by group
State
NOX
S02
NH3
voc
POA
Alabama
9,508
6,464
14
584
820
Arizona
132
1
1
8
6
Arkansas
10,198
15,207
40
354
524
California
5,706
845
312
581
390
Colorado
2,228
144
22
149
88
Connecticut
165
2
4
8
7
Delaware
833
150
23
27
12
District of Columbia
24
0
-
1
1
Florida
7,347
4,841
297
3,251
815
Georgia
10,397
6,511
7
703
762
Idaho
4,576
702
0
121
135
Illinois
6,386
13,482
194
241
270
Indiana
7,599
10,104
157
398
1,751
Iowa
3,583
3,122
154
239
142
Kansas
2,410
82
40
145
145
Kentucky
3,068
948
71
128
237
Louisiana
25,407
5,587
248
1,792
1,574
Maine
4,208
2,027
113
122
268
Maryland
1,695
8,595
1
25
24
Massachusetts
976
115
24
59
46
Michigan
5,115
3,545
98
212
243
Minnesota
7,601
3,336
175
510
230
Mississippi
4,595
257
0
1,305
575
Missouri
1,437
307
48
103
95
Montana
2,450
295
63
389
48
Nebraska
1,481
434
15
116
153
Nevada
212
16
-
38
46
New Hampshire
137
123
-
4
8
New Jersey
891
65
19
47
70
New Mexico
290
5
0
24
23
New York
3,818
9,005
39
143
145
North Carolina
9,250
4,140
38
733
644
North Dakota
4,915
6,478
122
197
309
Offshore to EEZ
243
4
8
13
8
Ohio
5,803
3,203
156
307
576
Oklahoma
3,221
1,424
15
104
196
Oregon
4,314
244
-
388
364
Pennsylvania
7,309
6,263
398
456
602
Puerto Rico
231
210
-
7
3
Rhode Island
114
12
3
7
5
South Carolina
7,418
5,540
263
497
644
South Dakota
242
83
29
43
43
Tennessee
10,330
13,107
97
319
274
Texas
13,458
7,686
239
1,074
1,072
Tribal Data
285
50
-
24
34
Utah
1,068
1,413
16
35
37
Vermont
45
4
-
7
12
Virginia
5,397
479
113
290
320
Washington
5,162
971
23
349
189
West Virginia
2,444
2,175
14
91
64
Wisconsin
7,826
9,816
33
358
311
Wyoming
5,614
4,628
3
115
58
4
-------
Table 4. Internal combustion endings annual total emissions (tpy) by group.
State
NOX
S02
NH3
voc
POA
Alabama
8,514
79
11
570
194
Arizona
2,109
20
59
51
49
Arkansas
5,444
13
2
302
59
California
4,121
211
599
598
365
Colorado
14,999
143
-
4,813
236
Connecticut
296
8
9
64
24
Delaware
65
19
0
13
14
Florida
7,999
136
26
522
141
Georgia
4,427
121
245
560
382
Idaho
1,140
7
-
34
11
Illinois
5,712
45
10
323
134
Indiana
3,810
22
0
210
40
Iowa
5,885
38
1
344
87
Kansas
23,675
35
58
1,403
190
Kentucky
9,468
15
0
632
174
Louisiana
28,027
235
757
2,118
754
Maine
86
1
0
4
1
Maryland
118
3
19
9
6
Massachusetts
669
71
22
109
23
Michigan
11,179
29
45
931
112
Minnesota
3,895
66
65
176
19
Mississippi
11,784
32
-
734
150
Missouri
3,882
20
0
156
40
Montana
1,682
17
0
544
54
Nebraska
3,507
4
0
277
42
Nevada
652
35
0
39
17
New Hampshire
27
1
-
1
1
New Jersey
424
13
5
101
21
New Mexico
14,348
106
103
2,118
189
New York
2,802
42
54
349
71
North Carolina
1,086
33
51
167
51
North Dakota
1,335
12
-
114
36
Offshore to EEZ
48,915
435
-
1,387
304
Ohio
10,072
32
8
630
201
Oklahoma
47,977
66
-
11,959
769
Oregon
325
4
-
4
4
Pennsylvania
5,155
47
47
830
232
Puerto Rico
4
3
-
0
0
Rhode Island
150
4
0
26
3
South Carolina
728
35
9
53
22
South Dakota
251
8
-
10
63
Tennessee
3,991
17
-
553
50
Texas
57,551
864
928
9,039
2,163
Tribal Data
5,358
22
-
801
37
Utah
2,329
13
14
110
46
Virginia
1,936
41
3
134
40
Washington
770
38
10
54
30
West Virginia
4,914
5
-
322
60
Wisconsin
773
1
9
45
10
Wyoming
11,212
406
15
7,959
556
-------
Mode! Configuration
Annual 2017 photochemical mode! simulations were performed for a domain covering the contiguous
United States with 12 km sized grid cells (Figure 2). Each simulation tracked a different combination of
sectors and pollutants. All simulations were conducted using version 7.00 of the Comprehensive Air
Quality Model with Extensions (CAMx) photochemical grid model (www.camxxom). This CAMx
application includes ISORROPIA inorganic chemistry (Nenes et a!.; 1998), gas phase reactions based on
the Carbon Bond (CB6r4) mechanism (Ramboll Environ, 2016), and aqueous phase reactions (Ramboll
Environ, 2020). Chemical boundary inflow is extracted from a photochemical mode! simulation for 2017
with a larger geographic domain covering the northern hemisphere (U.S. Environmental Protection
Agency, 2019). A total of 35 layers resolve the vertical atmosphere to 50 mb with thinner layers nearer
the surface (the height of the layer closest to the surface is approximately 20 m). More details about the
meteorological model simulation used to supply inputs to the emissions and CAMx model are available
elsewhere (ref 2017 WRF TSD). Baseline emissions include anthropogenic sources based on the
"2017gb" version of the 2017 National Emission Inventory (U.S. Environmental Protection Agency, 2020)
and biogenic sources estimated with the Biogenic Emission Inventory System version 3.6.1 (Bash et al.,
2016). Wildland fire emissions are day specific for 2017 (U.S. Environmental Protection Agency, 2020).
Figure 2. Photochemical model domain.
Model Application
The photochemical model was applied for the entire year of 2017 at 12 km grid resolution. Multiple
annual model simulations were performed to capture precursor/pollutant impacts on ozone and PM2.s
for each sector and geographic source region. Each sector/source region combination is called a source
apportionment "tag".
Table 5 shows the relationship between precursor emissions and contribution to modeled primary and
secondary pollutants. The model was applied so that primary and secondary precursors were tracked for
6
-------
contribution to modeled PM2.5 components. NOx emissions were tracked for contribution to PM2.5
nitrate ion, NH3 emissions were tracked for contribution to PM2.5 ammonium ion, and S02 emissions
were tracked for contribution to PM2.5 sulfate ion. Primarily emitted elemental carbon was tracked to
model predicted PM2.5 elemental carbon. The contribution to PM2.5 nitrate does not include primarily
emitted PM2.5 nitrate and the contribution to PM2.5 sulfate does not include primarily emitted PM2.5
sulfate.
Table 5. Relationship between emissions species and tracked primary and secondarily formed PM2.5
and also ozone in the modeling system.
Precursor Tagged Pollutant
so2
Secondarily formed PM2.5 sulfate ion
NOx
Secondarily formed PM2.5 nitrate ion
nh3
Secondarily formed PM2.5 ammonium ion
Primary PM2.5 elemental carbon
Primary PM2.5 elemental carbon
Primary PM2.5 organic aerosol
Primary PM2.5 organic aerosol
VOC
Ozone
NOx
Ozone
MODEL PERFORMANCE EVALUATION
Particulate Matter
An operational model performance evaluation for the speciated components of PM2.5 (e.g., sulfate,
nitrate, elemental carbon, organic carbon, etc.) was conducted using 2017 monitoring data in order to
estimate the ability of the modeling system to predict ambient concentrations. The evaluation of PM2.5
component species includes comparisons of predicted and observed concentrations of sulfate (S04),
nitrate (N03), ammonium (NH4), elemental carbon (EC), and organic carbon (OC). Chemically speciated
PM2.5 ambient measurements for 2017 were obtained from the Chemical Speciation Network (CSN) and
the Interagency Monitoring of PROtected Visual Environments (IMPROVE). The CSN sites are generally
located within urban areas and the IMPROVE sites are typically in rural/remote areas. The
measurements at CSN and IMPROVE sites represent 24-hour average concentrations. In calculating the
model performance metrics, the modeled hourly species predictions were aggregated to the averaging
times of the measurements.
Model performance statistics were calculated for observed/predicted pairs of all daily concentrations
measured in 2017 (Simon et al., 2012). All daily average chemically speciated PM2.5 prediction-
observation pairs are shown in Figure 3. Aggregated metrics and number (N) of prediction-observation
pairs are shown by chemical specie in Table 6. PM2.5 ammonium ion is not measured at most IMPROVE
monitors so metrics were not generated for that network. Model performance was compared to the
performance found in recent regional PM2.5 model applications for other assessments. Overall, the mean
bias (bias) and mean error (error) statistics are within the range or close to that found by other groups in
recent applications (Kelly et al., 2019; Simon et al., 2012). The model performance results provide
confidence that this application of CAMx provides a scientifically credible approach for estimating PM2.5
concentrations for the purposes of this assessment.
7
-------
Table 6. Aggregated model performance metrics for speciated components of PM2.5 for the
IMPROVE and CSN monitor networks.
Mean Bias Mean Error Normalized Normalized
Specie Network N (ng/mB) (ng/mB) Mean Bias (%) Mean Error (%) r*
PM2.5 sulfate ion
CSN
13,523
0.37
0.56
39
59
0.32
PM2.5 sulfate ion
IMPROVE
16,542
0.28
0.40
44
63
0.47
PM2.5 nitrate ion
CSN
13,593
0.43
0.87
49
98
0.28
PM2.5 nitrate ion
IMPROVE
16,529
0.17
0.36
53
111
0.34
PM2.5 elemental carbon
CSN
13,041
-0.01
0.27
-2
47
0.29
PM2.5 elemental carbon
IMPROVE
16,666
-0.03
0.11
-16
55
0.17
PM2.5 organic carbon
CSN
13,066
0.64
1.15
31
56
0.29
PM2.5 organic carbon
IMPROVE
16,894
0.15
0.86
11
64
0.28
PM2.5 ammonium ion
CSN
11,419
0.49
0.60
126
155
0.17
Ozone
An operational model performance evaluation for eight-hour daily maximum (MDA8) ozone was
conducted in order to estimate the ability of the modeling system to replicate the 2017 base year
concentrations. Ozone measurements were taken from 2017 monitoring site data in the Air Quality
System (AQS). The ozone metrics covered in this evaluation include eight-hour average daily maximum
ozone bias and error (Simon et al., 2012). The evaluation principally consists of statistical assessments of
model versus observed pairs that are paired in time and space. Aggregated metrics and number (N) of
prediction-observation pairs are shown in Table 7. All MDA8 ozone prediction-observation pairs are
shown in Figure 3 (bottom right panel).
Table 7. Aggregated model performance metrics for MDA8 03. Metrics are shown for all prediction-
observation pairs, pairs where the model predictions exceed 60 ppb, and pairs where the
observations exceed 60 ppb.
Mean Bias
Mean Error
Normalized
Normalized
Specie
Network
N
(ppb)
(ppb)
Mean Bias (%)
Mean Error (%)
r2
MDA803
AIRS-ALL
215,245
-0.08
7.44
0
16
0.52
MDA803
AIRS - model > 60 ppb
24,834
5.57
9.17
9
15
0.22
MDA803
AIRS - obs > 60 ppb
23,783
-6.30
8.89
-9
13
0.29
Only prediction-observation pairs from April through September were included in the aggregated
metrics. This ozone model performance includes all prediction-observation pairs, a subset of prediction-
observation pairs where observed ozone exceeded 60 ppb, and a subset of prediction-observation pairs
where predicted ozone exceeded 60 ppb. This cutoff was applied to evaluate the model on days of
elevated ozone which are more policy relevant. Overall, the mean bias (bias) and mean error (error)
statistics are within the range or close to that found by other groups in recent applications (Simon et al.,
2012). The model performance results provide confidence that this application of CAMx provides a
scientifically credible approach for estimating ozone mixing ratios for the purposes of this assessment.
PRODUCTS
The photochemical model output was processed to be in the format for input to the BenMAP-CE tool.
This requires temporal aggregation to convert hourly model output to annual average PM2.5 and ozone
season (April through September) average of MDA8 ozone. BenMAP-CE input files were generated for
the baseline simulation where all sources are included. Additionally, a set of BenMAP-CE input files
8
-------
were generated representing air quality surfaces for which the contribution from specific sectors or
state/sector combinations are removed from the baseline. The difference in air quality in the baseline
and sector specific BenMAP-CE input files represent the health burden associated with that sector. A set
of BenMAP-CE inputs were generated for annual average PM2.5 (sum of all chemical constituents of
PM2.5), and annual average PM2.5 chemical components: elemental carbon, sulfate ion, nitrate ion, and
ammonium ion. BenMAP-CE inputs were also generated for multiple 03 related metrics: 1) April-
September MDA8, 2) June-August MDA8, 3) April-September MDA1, 4) May-October MDA1, and 5)
April-October MDA8. The photochemical model output species were temporally aggregated but were
not adjusted toward ambient measurement data in order to preserve the simulated emissions to
ambient concentration relationship.
PM2.5 nitrate impacts were linked to secondary formation attributed to N0X emissions only and
therefore these outputs do not include the impacts from primarily emitted PM2.5 nitrate. PM2.5 sulfate
impacts were linked to secondary formation attributed to S02 emissions only and similarly do not
include the impacts from primarily emitted PM2.5 sulfate. PM2.5 ammonium impacts were linked to
secondary formation attributed to NH3 emissions only and do not include the impacts from primarily
emitted PM2.5 ammonium.
Figure 3. Contribution to PM2.5 from N0X, S02, NH3, and primarily emitted elemental carbon from
industrial boilers in Louisiana.
PM25 from boilers (19) NH3
PM25 from boilers (19) NOX
PM25 from boilers (19) PEC
PM25 from boilers (19) S02
9
-------
The sum of PM2.5 elemental carbon and primarily emitted organic aerosol could be used as a surrogate
for all primarily emitted PM2.s emissions because those emissions and model concentrations do not have
any secondary component and the emissions are considered more robust than many of the metals and
ions that are highly variable spatially and generally more uncertain with respect to the underlying
emission inventory. This approach is consistent with other approaches to represent primary PM2 5
impacts from photochemical model output (Heo et al., 2016). Separate attribution was done for
primarily emitted PM2.5 sulfate ion and PM2.5 nitrate ion to differentiate primary and secondary forms of
these pollutants.
Figure 3 shows the contribution to annual average PM2.5 from emissions of NOx, S02, NH3, and elemental
carbon in Louisiana. The model shows very localized impacts of primarily emitted pollutants and NH3 on
annual PM2.5. The PM2 5 impacts from NOx and S02 extend further downwind due to chemical
transformation over time.
Figure 4. Average ozone season MDA8 for IC engines (top left), state subset (Wyoming, Kentucky,
California, and Florida) of IC engines (top right), industrial boilers (bottom left), and state subset
(California, Texas, North Carolina, and Wisconsin) of industrial boilers (bottom right).
-------
Precursor emissions of NOx and VOC were separately tracked for contribution to MDA8 03 predictions.
Figure 4 shows average 03 season MDA8 NOx contribution from the entirety of the internal combustion
engines and boilers sectors tracked here and an example of state-level impacts for each. The left panels
show the entirety of impacts from internal combustion engines (top left) and industrial boilers (bottom
left) on seasonal average MDA8 03. The right panels show the same sectors but only impacts from a
subset of states.
REFERENCES
Bash, J.O., Baker, K.R., Beaver, M.R., 2016. Evaluation of improved land use and canopy representation
in BEIS v3. 61 with biogenic VOC measurements in California. Geoscientific Model Development 9, 2191.
Fann, N., Baker, K.R., Fulcher, C.M., 2012. Characterizing the PM2.5-related health benefits of emission
reductions for 17 industrial, area and mobile emission sectors across the U.S. Environment international
49, 141-151.
Heo, J., Adams, P.J., Gao, H.O., 2016. Reduced-form modeling of public health impacts of inorganic PM2.
5 and precursor emissions. Atmospheric Environment 137, 80-89.
Industrial Economics Incorporated, 2019. Evaluating Reduced-Form Tools for Estimating Air Quality
Benefits, https://www.epa.gov/sites/production/files/2019-
11/documents/rft combined report 10.31.19 final.pdf.
Kelly, J.T., Koplitz, S.N., Baker, K.R., Holder, A.L., Pye, H.O., Murphy, B.N., Bash, J.O., Henderson, B.H.,
Possiel, N.C., Simon, H., 2019. Assessing PM2. 5 model performance for the conterminous US with
comparison to model performance statistics from 2007-2015. Atmospheric Environment 214, 116872.
Kwok, R., Baker, K., Napelenok, S., Tonnesen, G., 2015. Photochemical grid model implementation of
VOC, NO x, and O 3 source apportionment. Geoscientific Model Development 8, 99-114.
Kwok, R., Napelenok, S., Baker, K., 2013. Implementation and evaluation of PM2.5 source contribution
analysis in a photochemical model. Atmospheric Environment 80, 398-407.
LeNOx, C., Dodder, R., C., G., Kaplan, O., Loughlin, D., Yelverton, W., 2013. EPA U.S. Nine-region MARKAL
Database. EPA 600/B-13/203. .
Nenes, A., Pandis, S.N., Pilinis, C., 1998. ISORROPIA: A new thermodynamic equilibrium model for
multiphase multicomponent inorganic aerosols. Aquatic geochemistry 4, 123-152.
Ramboll Environ, 2016. Technical Memorandum: EPA Contract EPD12044; WA4-07; Meterologogical,
Photochemical, and Dispersion Modeling Support Task 7: Update Carbon Bon Chemical Mechanism.
http://www.camx.com/files/emaq4-07 task? techmemo rl laugl6.pdf.
Ramboll Environ, 2020. User's Guide Comprehensive Air Quality Model with Extensions Version 7.00.
http://www.camx.com/files/camxusersguidt .pdf.
Simon, H., Baker, K.R., Phillips, S., 2012. Compilation and interpretation of photochemical model
performance statistics published between 2006 and 2012. Atmospheric Environment 61, 124-139.
U.S. Environmental Protection Agency, 2018a. Environmental Benefits Mapping and Analysis Program -
Community Edition (BenMAP-CE). https://www.epa.gov/benmap.
11
-------
U.S. Environmental Protection Agency, 2018b. Technical Support Document Estimating the Benefit per
Ton of Reducting PM2.5 Precursors from 17 sectors, https://www.epa.gov/sites/production/files/2018"
02/documents/sourceapportionmentbpttsd 2018.pdf.
U.S. Environmental Protection Agency, 2019. 2016 Hemispheric Modeling Platform Version 1:
Implementation, Evaluation, and Attribution. Research Triangle Park, NC. U.S. Environmental Protection
Agency. U.S. EPA.
U.S. Environmental Protection Agency, 2020. 2017 National Emission Inventory Technical Support
Document, https://www.epa.eov/air-emissions-inventories/2017-national-emissions-inventorv-nei"
technical-support-document-tsd.
Wolfe, P., Davidson, K., Fulcher, C., Fann, N., Zawacki, M., Baker, K.R., 2019. Monetized health benefits
attributable to mobile source emission reductions across the United States in 2025. Science of The Total
Environment 650, 2490-2498.
12
-------
United States Office of Air Quality Planning and Standards Publication No. EPA-454/R-21-005
Environmental Protection Air Quality Assessment Division July 2021
Agency Research Triangle Park, NC
------- |