&EPA
United SUos
Envirwinmlfll Protection
Air Quality Modeling Technical Support Document:
Source Sector Assessments
-------
EPA-454/R-11-006
August 2011
Air Quality Modeling Technical Support Document:
Source Sector Assessments
U.S. Environmental Protection Agency
Office of Air Quality Planning and Standards
Air Quality Assessment Division
Research Triangle Park, NC 27711
-------
I Introduction
This document describes the air quality modeling performed by EPA in support or air quality and
deposition assessments related source sector groups. A national scale air quality modeling
analysis was performed to estimate the impact of current and future year sector emissions on
annual and 24-hour PM2.5 concentrations, 8-hr maximum ozone, as well as visibility
impairment. Air quality benefits are estimated with the Comprehensive Air Quality Model with
Extensions (CAMx) model. CAMx simulates the numerous physical and chemical processes
involved in the formation, transport, and destruction of ozone, particulate matter and air toxics.
In addition to the CAMx model, the modeling platform includes the emissions, meteorology, and
initial and boundary condition data which are inputs to this model. It is important to note that the
inventories used in the air quality modeling may be different than a final sector inventory
presented in a Regulatory Impact Analyses.
II. Photochemical Model Version, Inputs and Configuration
Photochemical grid models use state of the science numerical algorithms to estimate pollutant
formation, transport, and deposition over a variety of spatial scales that range from urban to
continental. Emissions of precursor species are injected into the model where they react to form
secondary species such as ozone and then transport around the modeling domain before
ultimately being removed by deposition or chemical reaction.
The 2005-based CAMx modeling platform was used as the basis for the air quality modeling.
This platform represents a structured system of connected modeling-related tools and data that
provide a consistent and transparent basis for assessing the air quality response to projected
changes in emissions. The base year of data used to construct this platform includes emissions
and meteorology for 2005. The modeling system treats the emissions, transport, and fate of
criteria pollutants.
The modeling system was used to calculate daily and annual PM2.5 concentrations, 8-hr
maximum ozone, and visibility impairment. Model predictions are used to estimate future-year
design values of PM2.5 and ozone using a 2016 reference scenario compared to a 2005 baseline.
This is done by calculating the simulated air quality ratios between any particular future year
simulation and the 2005 base. These predicted ratios are then applied to ambient base year design
values. The design value projection methodology used here followed EPA guidance for such
analyses (USEPA, 2007).
A. Model version
CAMx version 5.30 is a freely available computer model that simulates the formation and fate of
photochemical oxidants, ozone, primary and secondary PM concentrations, and air toxics, over
regional and urban spatial scales for given input sets of meteorological conditions and emissions.
CAMx includes numerous science modules that simulate the emission, production, decay,
deposition and transport of organic and inorganic gas-phase and particle-phase pollutants in the
atmosphere (Baker and Scheff, 2007; Nobel et al., 2001; Russell, 2008). CAMx is applied with
ISORROPIA inorganic chemistry (Nenes et al., 1999), a semi-volatile equilibrium scheme to
1
-------
partition condensable organic gases between gas and particle phase (Strader et al., 1999),
Regional Acid Deposition Model (RADM) aqueous phase chemistry (Chang et al., 1987), and
Carbon Bond 05 (CB05) gas-phase chemistry module (ENVIRON, 2008; Gery et al., 1989).
CAMx contains a variety of ozone source apportionment tools, including the original ozone
source apportionment tool (OSAT) and the anthropogenic pre-cursor culpability assessment
(APCA) tool (ENVIRON, 2008). Ozone source apportionment tracers are treated using the
standard model algorithms for vertical advection, vertical diffusion, and horizontal diffusion.
Horizontal advective fluxes for each of the regular model species that make up nitrogen oxides
(NOx) and volatile organic compounds (VOC) are combined and normalized by a concentration
based weighted mean. Separate ozone tracers are used in CAMx to track ozone formation that
happens under NOX and VOC limited conditions.
Particulate matter source apportionment technology (PSAT) implemented in CAMx estimates
the contribution from specific emissions source groups to PM2.5 and all forms of mercury using
reactive tracers (ENVIRON, 2008; Wagstrom et al., 2008). The tracer species are estimated with
source apportionment algorithms rather than by the host model routines. PSAT tracks
contribution to PM2.5 sulfate, nitrate, ammonium, secondary organic aerosol, and inert primarily
emitted species. Non-linear processes like gas and aqueous phase chemistry are solved for bulk
species and then apportioned to the tagged species. Emissions of nitrogen oxides are tracked
through all intermediate nitrogen species to particulate nitrate ion. Ammonia emissions are
tracked to particulate ammonium ion. This modeling assessment used the PSAT approach to
estimate source contribution to PM2.5 species and the APCA method to estimate source
contribution to modeled ozone.
B. Model domain and grid resolution
The modeling analyses were performed for a domain covering the continental United States as
shown in Figure II-l. This domain has a parent horizontal grid of 36 km with two finer-scale 12
km grids over portions of the eastern and western U.S. The model extends vertically from the
surface to 100 millibars (approximately 15 km) using a sigma-pressure coordinate system. Air
quality conditions at the outer boundary of the 36 km domain were taken from a global model
and vary in time and space. The 36 km grid was only used to establish the incoming air quality
concentrations along the boundaries of the 12 km grids. Only the finer grid data were used in
determining the impacts of the emissions changes. Table II-l provides geographic information
about the photochemical model domains.
-------
Table II-l. Geograj
Map Projection
Grid Resolution
Coordinate Center
True Latitudes
Dimensions
Vertical extent
}hic elements of domains used in photochemical modeling.
Photochemical Modeling Configuration
National Grid
Western U.S. Fine Grid
Eastern U.S. Fine Grid
Lambert Conformal Projection
36km
12km
12km
97 deg W, 40 deg N
33 deg N and 45 deg N
148 x 112x 14
213 x 192 x 14
279 x 240 x 14
14 Layers: Surface to 100 millibar level (see Table II-3)
Figure II-l. Map of the photochemical modeling domains. The black outer box denotes the 36
km national modeling domain; the red inner box is the 12 km western U.S. grid; and the blue
inner box is the 12 km eastern U.S. grid.
C. Modeling Time-period
The 36 km and both 12 km modeling domains were modeled for the entire year of 2005. Data
from the entire year were utilized when looking at the estimation of PM2.5 and visibility
impacts. Data from April through October is used to estimate ozone impacts.
D. Model Inputs: Emissions, Meteorology and Boundary Conditions
The 2005-based modeling platform was used for the air quality modeling of future emissions
scenarios. In addition to the photochemical model, the modeling platform also consists of the
-------
base- and future-year emissions estimates, meteorological fields, as well as initial and boundary
condition data which are all inputs to the air quality model.
1. Emissions Input Data
The emissions data used in the base year and future reference and future emissions adjustment
case are based on the 2005 v4.3 platform. Emissions are processed to photochemical model
inputs with the SMOKE emissions modeling system (Houyoux et al., 2000). The 2016 reference
case is intended to represent the emissions associated with growth and controls in that year
projected from the 2005 simulation year. The United States EGU point source emissions
estimates for the future year reference and control case are based on an Integrated Planning
Model (IPM) run for criteria pollutants. Both control and growth factors were applied to a subset
of the 2005 non-EGU point and non-point emissions to create the 2016 reference case. The 2005
v4 platform projection factors were the starting point for most of the 2016 SMOKE-based
projections.
Table II.2 Domain total estimated sector emissions
Sector Year VOC
Biogenics
Cement Kilns
Pulp& Paper
Refineries
Coke Ovens
Iron/Steel Foundries
Integrated Iron/Steel
Electric Arc Furnaces
Taconite Mining
Ferroalloy Production
Residential Wood
Non-point Other
Non-EGU point Other
Average Fires
Air/Locomotives/Marine
Nonroad mobile
Onroad mobile
Canada + Mexico
Point EGU
Commercial Marine (SECA-C3)
Area Fugutive Primary PM2.5
2016
2016
2016
2016
2016
2016
2016
2016
2016
2016
2016
2016
2016
2016
2016
2016
2016
2016
2016
2016
2016
59,388,365
3,059
121,597
111,391
7,821
14,384
9,620
3,560
606
150
538,466
9,380,925
1,259,745
1,874,035
43,547
1,953,067
2,357,108
5,615,200
63,198
66,093
0
(TPY) for 2016 modeling scenario.
NOX Primary PM2.5 SO2
2,218,207
130,536
240,139
118,206
16,110
5,867
31,925
15,707
41,350
3,412
33,786
1,633,261
1,263,276
189,428
1,342,849
1,259,578
4,239,971
2,841,702
1,826,582
1,534,234
0
0
1,106
10,067
7,379
368
1,366
2,856
622
884
201
192,492
325,820
67,614
459,641
35,604
106,975
118,986
242,412
30,078
7,407
47,438
0
48,737
170,393
132,337
27,952
3,590
29,045
6,088
8,823
4,580
4,720
1,243,154
877,620
49,094
9,087
2,879
26,786
2,855,974
3,793,362
439,987
0
NH3
0
679
10,859
3,556
1,084
166
167
119
4
510
6,586
126,802
140,948
36,777
940
2,345
82,094
1,077,333
36,706
0
0
The future year projection includes emissions reductions related to the NOx SIP Call, Boiler
MACT, RICE, and the proposed Transport Rule (TR1). The residential wood combustion
category is defined based on SCC codes 2104008 and 2104009. Certain sectors are based on a
list of facilities that would likely be subject to a NESHAP related to that sector: cement kilns,
pulp & paper, refineries, coke ovens, iron/steel foundries, integrated ion/steel, electric arc
-------
furnaces, taconite mining, and ferroalloy production. The model domain total estimated
emissions by source sector used in this modeling assessment are shown in Table II.2 for 2016
and Table II.3 for 2005.
Table II.3 Domain total estimated sector emissions
Sector Year VOC
Biogenics
Cement Kilns
Pulp& Paper
Refineries
Coke Ovens
Iron/Steel Foundries
Integrated Iron/Steel
Electric Arc Furnaces
Taconite Mining
Ferroalloy Production
Residential Wood
Non-point Other
Non-EGU point Other
Average Fires
Air/Locomotives/Marine
Nonroad mobile
Onroad mobile
Canada + Mexico
Point ECU
Commercial Marine (SECA-C3)
Area Fugutive Primary PM2.5
2005
2005
2005
2005
2005
2005
2005
2005
2005
2005
2005
2005
2005
2005
2005
2005
2005
2005
2005
2005
2005
59,388,365
9,044
128,390
112,622
7,914
15,522
9,687
3,630
607
150
647,141
12,453,787
1,747,491
1,874,035
58,890
3,353,497
4,734,097
5,615,200
63,404
42,438
0
(TPY) for 2005 modeling scenario
NOX Primary PM2.5 SO2
2,218,207
217,948
246,714
150,924
16,145
5,985
31,961
15,707
41,350
3,412
38,190
1,659,651
1,297,668
189,428
1,909,526
2,083,093
9,006,389
2,841,702
3,726,455
1,169,907
0
0
2,342
11,979
7,750
416
1,435
2,862
622
1,035
201
222,081
326,140
71,899
459,641
53,671
181,038
258,251
242,412
39,544
10,501
47,315
0
157,163
345,917
246,351
29,239
3,611
32,646
6,088
11,285
4,580
5,280
1,251,930
1,215,203
49,094
154,016
195,597
176,525
2,855,974
10,380,773
740,998
0
NH3
0
861
10,990
3,556
1,089
166
167
119
4
510
7,238
126,802
141,230
36,777
773
1,969
156,276
1,077,333
21,684
0
0
Other North American emissions of criteria and toxic pollutants (including mercury) are based
on a 2006 Canadian inventory and 1999 Mexican inventory. Both inventories are not grown or
controlled when used as part of future year baseline inventories. Global emissions of criteria are
included in the modeling system through boundary condition inflow.
2. Meteorological Input Data
The gridded meteorological input data for the entire year of 2005 were derived from simulations
of the Pennsylvania State University / National Center for Atmospheric Research Mesoscale
Model. This model, commonly referred to as MM5, is a limited-area, nonhydrostatic, terrain-
following system that solves for the full set of physical and thermodynamic equations which
govern atmospheric motions. Meteorological model input fields were prepared separately for
each of the three domains shown in Figure II-l using MM5 version 3.7.4. The MM5 simulations
were run on the same map projection as shown in Figure II-l.
All three meteorological model runs were configured similarly. The selections for key MM5
physics options are shown below:
-------
• Pleim-Xiu PEL and land surface schemes
• Kain-Fritsh 2 cumulus parameterization
• Reisner 2 mixed phase moisture scheme
• RRTM longwave radiation scheme
• Dudhia shortwave radiation scheme
Three dimensional analysis nudging for temperature and moisture was applied above the
boundary layer only. Analysis nudging for the wind field was applied above and below the
boundary layer. The 36 km domain nudging weighting factors were 3.0 x 104 for wind fields and
temperatures and 1.0 x 105 for moisture fields. The 12 km domain nudging weighting factors
were 1.0 x 104 for wind fields and temperatures and 1.0 x 105 for moisture fields.
Table II-4. Vertical layer structure (heights are layer top).
CAMX Layers
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
MM5 Layers
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
Sigma P
1.000
0.995
0.990
0.985
0.980
0.970
0.960
0.950
0.940
0.930
0.920
0.910
0.900
0.880
0.860
0.840
0.820
0.800
0.770
0.740
0.700
0.650
0.600
0.550
0.500
0.450
0.400
0.350
0.300
0.250
0.200
0.150
0.100
0.050
0.000
Approximate Height
(m)
0
38
77
115
154
232
310
389
469
550
631
712
794
961
1,130
1,303
1,478
1,657
1,930
2,212
2,600
3,108
3,644
4,212
4,816
5,461
6,153
6,903
7,720
8,621
9,625
10,764
12,085
13,670
15,674
Approximate Pressure
(mb)
1000
995
991
987
982
973
964
955
946
937
928
919
910
892
874
856
838
820
793
766
730
685
640
595
550
505
460
415
370
325
280
235
190
145
100
All three sets of model runs were conducted in 5.5 day segments with 12 hours of overlap for
spin-up purposes. All three domains contained 34 vertical layers with an approximately 38m
deep surface layer and a 100 millibar top. The MM5 and CAMx vertical structures are shown in
Table II-4 and do not vary by horizontal grid resolution. The meteorological outputs from all
-------
three MM5 sets were processed to create model-ready inputs for CAMx using the MCIP
processor.
Before initiating the air quality simulations, it is important to identify the biases and errors
associated with the meteorological modeling inputs. The 2005 MM5 model performance
evaluations used an approach which included a combination of qualitative and quantitative
analyses to assess the adequacy of the MM5 simulated fields. The qualitative aspects involved
comparisons of the model-estimated synoptic patterns against observed patterns from historical
weather chart archives. Additionally, the evaluations compared spatial patterns of estimated to
observed monthly average rainfall and checked maximum planetary boundary layer (PEL)
heights for reasonableness.
Qualitatively, the model fields closely matched the observed synoptic patterns, which is not
unexpected given the use of nudging. The operational evaluation included statistical
comparisons of model/observed pairs (e.g., mean normalized bias, mean normalized error, index
of agreement, root mean square errors, etc.) for multiple meteorological parameters. For this
portion of the evaluation, five meteorological parameters were investigated: temperature,
humidity, shortwave downward radiation, wind speed, and wind direction. The three individual
MM5 evaluations are described elsewhere (Baker, 2009a, b, c). It was ultimately determined that
the bias and error values associated with all three sets of 2005 meteorological data were
generally within the range of past meteorological modeling results that have been used for air
quality applications.
3. Initial and Boundary Conditions
The lateral boundary and initial species concentrations are provided by a three-dimensional
global atmospheric chemistry model, the GEOS-CHEM model (standard version 7-04-11). The
global GEOS-CHEM model simulates atmospheric chemical and physical processes driven by
assimilated meteorological observations from the NASA's Goddard Earth Observing System
(GEOS). This model was run for 2005 with a grid resolution of 2.0 degree x 2.5 degree
(latitude-longitude) and 30 vertical layers up to 100 mb. The predictions were used to provide
one-way dynamic boundary conditions at three-hour intervals and an initial concentration field
for the 36 km CAMx simulations. The 36 km photochemical model simulation is used to supply
initial and hourly boundary concentrations to the 12 km domains. The 36 km domain simulation
includes 10 days of spin-up before the start of each calendar quarter that are not used in the
analysis. The 12 km domain simulations include 3 days of spin-up before each calendar quarter.
The future base conditions from the 36 km coarse grid modeling were used as the
initial/boundary state for all subsequent future year 12 km finer grid modeling scenarios.
-------
III. Base Case Model Performance Evaluation
A. PM2.5
An operational model performance evaluation for the speciated components of PM2.5 (e.g.,
sulfate, nitrate, elemental carbon, organic carbon, etc.) was conducted using 2005 monitoring
data in order to estimate the ability of the modeling system to replicate base year concentrations.
The evaluation of PM2.5 component species includes comparisons of predicted and observed
concentrations of sulfate (SO4), nitrate (NO3), ammonium (NH4), elemental carbon (EC), and
organic carbon (OC). PM2.5 ambient measurements for 2005 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 daily concentrations
(Boylan and Russell, 2006). The aggregated metrics and number (N) of prediction-observation
pairs are shown by chemical specie and quarter in Table III-l. Model performance was
compared to the performance found in recent regional PM2.5 model applications for other, non-
EPA studies. 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 (Doraiswamy, 2010; Tesche et al.,
2006). The model performance results give us confidence that our application of CAMx using
this modeling platform provides a scientifically credible approach for estimating PM2.5
concentrations for the purposes of this assessment.
-------
TABLE III-l. Model performance metrics for speciated PM2.5 averaged by quarter.
Specie Quarter N Observed Predicted Bias Error
PM2.5Sulfatelon
PM2.5 Nitrate Ion
PM2.5 Ammonium Ion
PM2.5 Organic Carbon
PM2.5 Elemental Carbon
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
6,218
6,537
6,108
5,894
5,870
6,146
5,828
5,757
4,045
4,075
3,869
3,763
5,882
6,180
5,836
5,600
6,368
6,552
6,061
5,937
2.7
3.8
5.4
2.5
2.4
0.8
0.5
1.3
1.7
1.5
1.8
1.3
1.7
2.1
2.4
2.0
0.5
0.5
0.5
0.6
4.2
4.4
5.2
3.7
1.9
0.6
0.2
1.1
1.8
1.5
1.5
1.4
2.2
1.7
2.3
2.0
1.0
0.7
0.8
0.9
1.6
0.7
0.0
1.3
-0.4
-0.1
-0.2
-0.1
0.4
0.3
0.0
0.4
0.6
-0.4
-0.1
0.1
0.5
0.3
0.3
0.4
2.0
1.6
1.9
1.5
1.3
0.6
0.4
0.8
0.8
0.7
0.8
0.7
1.1
0.9
0.9
1.0
0.6
0.4
0.4
0.5
B. Ozone
An operational model performance evaluation for hourly and eight-hour daily maximum ozone
was conducted in order to estimate the ability of the modeling system to replicate the base year
concentrations. Ozone measurements were taken from the 2005 State/local monitoring site data
in the Air Quality System (AQS) Aerometric Information Retrieval System (AIRS). The ozone
metrics covered in this evaluation include one-hour and eight-hour average daily maximum
ozone bias and error. The evaluation principally consists of statistical assessments of model
versus observed pairs that are paired in time and space. This ozone model performance was
limited to the prediction-observation pairs where observed ozone exceeded or equaled 60 ppb.
This cutoff was applied to evaluate the model on days of elevated ozone which are more policy
relevant. Aggregated performance metrics by quarter are shown in Table III-2 for daily
maximum 1-hr average ozone and Table III-3 for daily maximum 8-hr average ozone.
-------
TABLE III-2. Model performance metrics for daily maximum 1-hr avg ozone by quarter.
Fractional Fractional
Quarter N Observed Predicted Bias Error Bias Error
2
3
19,821
22,699
76
80
65
71
-11
-9
12
13
-16
-13
18
17
TABLE III-3. Model performance metrics for daily maximum 8-hr avg ozone by quarter.
Fractional Fractional
Quarter N Observed Predicted Bias Error Bias Error
2 19,833 69 60 -9 11 -15 17
3 22,714 70 64 -6 10 -10 16
This model performance is consistent with photochemical modeling used to support other
national regulations (USEPA, 2010).
IV. References
Baker, K., Dolwick, P., 2009a. Meteorological Modeling Performance Evaluation for the Annual 2005 Continental
U.S. 36-km Domain Simulation. US Environmental Protection Agency OAQPS.
Baker, K., Dolwick, P., 2009b. Meteorological Modeling Performance Evaluation for the Annual 2005 Eastern U.S.
12-km Domain Simulation. US Environmental Protection Agency OAQPS, RTF.
Baker, K., Dolwick, P., 2009c. Meteorological Modeling Performance Evaluation for the Annual 2005 Western U.S.
12-km Domain Simulation, in: EPA, U. (Ed.). US Environmental Protection Agency OAQPS.
Baker, K., Scheff, P., 2007. Photochemical model performance for PM2.5 sulfate, nitrate, ammonium, and precursor
species SO2, HNO3, and NH3 at background monitor locations in the central and eastern United States.
Atmospheric Environment 41, 6185-6195.
Boylan, J.W., Russell, A.G., 2006. PM and light extinction model performance metrics, goals, and criteria for three-
dimensional air quality models. Atmospheric Environment 40, 4946-4959.
Chang, J.S., Brost, R.A., Isaksen, I.S.A., Madronich, S., Middleton, P., Stockwell, W.R., Walcek, C.J., 1987. A 3-
Dimensional Eulerian Acid Deposition Model - Physical Concepts and Formulation. J. Geophys. Res.-Atmos. 92,
14681-14700.
Doraiswamy, P., Hogrefe, C., Hao, W., Civerolo, K., Ku, J., Sistla, G., 2010. A Retrospective Comparison of
Model-Based Forecasted PM2.5 Concentrations with Measurements. Journal of Air & Waste Management
Association 60, 1293-1308.
ENVIRON, 2008. User's Guide Comprehensive Air Quality Model with Extensions. ENVIRON International
Corporation, Novato.
Gery, M.W., Whitten, G.Z., Killus, J.P., Dodge, M.C., 1989. A Photochemical Kinetics Mechanism for Urban and
Regional Scale Computer Modeling. J. Geophys. Res.-Atmos. 94, 12925-12956.
Houyoux, M.R., Vukovich, J.M., Coats, C.J., Wheeler, N.J.M., Kasibhatla, P.S., 2000. Emission inventory
development and processing for the Seasonal Model for Regional Air Quality (SMRAQ) project. Journal of
Geophysical Research-Atmospheres 105, 9079-9090.
10
-------
Nenes, A., Pandis, S.N., Pilinis, C., 1999. Continued development and testing of a new thermodynamic aerosol
module for urban and regional air quality models. Atmospheric Environment 33, 1553-1560.
Nobel, C.E., McDonald-Buller, E.G., Kimura, Y., Allen, D.T., 2001. Accounting for spatial variation of ozone
productivity inNOx emission trading. Environmental Science & Technology 35, 4397-4407.
Russell, A.G., 2008. EPA Supersites Program-related emissions-based paniculate matter modeling: Initial
applications and advances. J. Air Waste Manage. Assoc. 58, 289-302.
Strader, R., Lurmann, F., Pandis, S.N., 1999. Evaluation of secondary organic aerosol formation in winter.
Atmospheric Environment 33, 4849-4863.
Tesche, T.W., Morris, R., Tonnesen, G., McNally, D., Boylan, J., Brewer, P., 2006. CMAQ/CAMx annual 2002
performance evaluation over the eastern US. Atmospheric Environment 40, 4906-4919.
USEPA, 2007. Guidance on the Use of Models and Other Analyses for Demonstrating Attainment of Air Quality
Goals for Ozone, PM2.5, and Regional Haze, RTF.
USEPA, 2010. Air Quality Modeling Technical Support Document: Boiler Source Sector Rules (EPA-454/R-10-
006), Research Triangle Park, North Carolina.
Wagstrom, K.M., Pandis, S.N., Yarwood, G., Wilson, G.M., Morris, R.E., 2008. Development and application of a
computationally efficient paniculate matter apportionment algorithm in a three-dimensional chemical transport
model. Atmospheric Environment 42, 5650-5659.
11
-------
United States Office of Air Quality Planning and Standards Publication No. EP A-454/R-11 -006
Environmental Protection Air Quality Assessment Division August 2011
Agency Research Triangle Park, NC
------- |