EPA/600/A-97/078
SCIENCE FEATURES IN MODELS-3 COMMUNITY MULTISCALE AIR QUALITY SYSTEM
Jason Ching1', Daewon Byun1', Jeff Young1', Francis S. Binkowski1', Jonathan Pleim1', Shawn Roselle1', James Godowitch1',
William Benjey1*, and Gerald Gipson2
'Atmospheric Sciences Modeling Division, Air Resources Laboratory,
National Oceanic and Atmospheric Administration,
Research Triangle Park, NC 27711
2 Human Exposure and Atmospheric Sciences Division, National Exposure Research Laboratory,
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
1. INTRODUCTION
Air quality simulation models are important tools for
use by regulatory, policy and research communities. The Clean
Air Act provides a societal mandate to assess and to manage air
pollution levels to protect human health and the environment.
The U.S. Environmental Protection Agency (USEPA) has
established National Ambient Air Quality Standards (NAAQS),
requiring the development of effective emissions control
strategies for such pollutants as ozone, paniculate matter and
nitrogen species. National and regional policies are needed for
reducing and managing the amount and type of emissions that
cause acid, nutrient and toxic pollutant deposition to ecosystems
at risk and for enhancing the visual quality of the environment.
Air quality models are used to develop emission control
strategies that achieve these objectives. Control strategies must
be both environmentally protective and cost effective.
However, for effectiveness, one must recognize that air
pollution problems and strategies for their mitigation are very
complex and the linkages between sources, meteorology and
natural sources and landscapes are highly varied, complex and
not very well understood. The goal of developing cost-effective
control strategies is challenging and is best considered
holistically. The effectiveness of any control strategy is very
limiting when air pollution issues are handled in isolation.
Emissions from chemical, manufacturing, and other industrial
activities, power generation, transportation and waste treatment
activities contribute to a variety of air pollution issues,
including ozone, particulate matter (PM), acid, nutrient and
toxic deposition, and visibility in complex ways, and at' a
variety of spatial and temporal scales. The residence times of
pollutants in the atmosphere can extend to multi day, thus,
transport consideration must be at least regional in scale.
NAAQS requirements and other goals for a cleaner environment
vary over a large range of time scales, from peak hourly to
annual averages.
Corresponding author address: Jason K.S. Ching, MD-80. 'On
assignment to National Exposure Research Laboratory, U.S.
Environmental Protection Agency, Research Triangle Park, NC
27711; email: ching.jason@epamail.epa.gov.
Development of air quality simulation models started
in the late seventies. The Urban Airshed Model (UAM)
followed by the Regional Oxidant Model (ROM) provided
Eulerian-based models for ozone, the former for urban and the
latter for regional scale. Strategies for State Implementation
Plans (SIPS) used ROM to provide boundary conditions for
UAM simulations. Attention to acid deposition issues was
addressed in the eighties with the development and evaluation
of regional acid deposition models such as the Regional Acid
Deposition Model (RADM), the Acid Deposition and Oxidant
Model (ADOM), and the Sulfur Transport and Emissions Model
(STEM). Other major modeling systems included the Regional
Lagrangian Modeling of Air Pollution model (RELMAP), a
Lagrangian framework system, and semi-empirical and
statistical models. Models of this period were designed to
address specific air pollution issues, such as ozone or acid
deposition. Thus, flexibility to deal with other issues such as
particulate matter or toxics was very limited. With the passage
of the Clean Air Act Amendments of 1990 (CAAA-90), a wide
range of additional issues was identified including visibility,
and fine and coarse particles, as well as indirect exposure to
toxic pollutants such as heavy metals, semi-volatile organic
species, and nutrient deposition to water bodies. These latter
issues will require multi-media models.
To meet the challenges posed by the CAAA-90, the
USEPA embarked upon the development of an advanced
modeling framework, Models-3, designed for holistic
environmental modeling utilizing state of science representation
of atmospheric processes in a high performance computing
environment. Descriptions of Models-3 can be found in Novak
et al. (1998) and Byun et al. (1998). The science components
in Models-3 are called the Community Multiscale Air Quality
(CMAQ) system. The Models-3/CMAQ system is designed as
a multi-pollutant, multi-scale Eulerian framework air quality
and atmospheric deposition modeling system. It contains state-
of-science parameterizations of atmospheric processes affecting
transport, transformation and deposition of such pollutants as
ozone, particulate matter, airborne toxics, and acidic and
nutrient pollutant species. With science in a continuing state of
advancement and review, it is an important design feature that
the Models-3 framework and CMAQ have the capability to
integrate and test future formulations in an efficient manner,
without needing to develop a completely new modeling system.
In June 1998, the first release version (CMAQ-98)
-------
will become available. It will contain options representing
different model descriptions of the major science processes.
The science options available to the user include the gas phase
chemistry mechanisms, RADM2 and CB-IV, a set of numerical
solvers for the mechanisms, options for horizontal and vertical
advection schemes, algorithms for fine and coarse particulate
matter predictions, photolysis rates and a plume-in-grid
approach. The purpose of this paper is to briefly list the key
science processes in general and the specific modeling options
included throughout CMAQ-98, including the emissions and
meteorological processors and their interface processors. Also
discussed are methods to determine and test the relative
contribution of different atmospheric processes to the air quality
predictions, important functions built into the CMAQ system.
2. SCIENCE AND MODELING FEATURES OF CMAQ
In this section, the science components and related
features of the CMAQ-98 system is briefly described. The
multi-scale capabilities of Models-3/CMAQ are handled by a set
of nested domains, each with successively finer resolution. The
selection of grid resolution, domain size and area of interest is
dependent on the application, and on computational resources,
A configuration of three levels of nests with grid resolutions of
36-12-4 km will be provided in CMAQ-98. With such a
configuration, the 36km grid resolution resolves the regional
scale concentration fields and provides boundary conditions for
an intermediate scale at 12 km resolutions to set up the
transition to the 4km scale to handle features on urban scales.
2.1. CMAO Preprocessors:
2.1.1 Meteorology
The CMAQ's generalized coordinate system
accommodates meteorological data produced by any
comprehensive meteorology model. Initial implementation uses
the Fifth Generation PSU/NCAR Mesoscale Model (MM5),
(Grell et al., 1993). The configuration for preliminary
evaluation includes non-hydrostatic dynamics, four-dimensional
data-assimilation (FDDA) of winds, temperature, and humidity,
the Kain-Fritsch convective parameterization scheme, and the
high resolution boundary layer. A set of four nested grids
(108/36/12/4)km is used to achieve urban scale resolution which
is consistent with continental scale dynamics.
In addition to the standard community version of
MM5, a more sophisticated land-surface model has been
developed and coupled to the MM5 to improve simulation of
surface fluxes, surface temperature and humidity, and planetary
boundary-layer (PEL) development (Pleim and Xiu, 1995).
The surface scheme includes explicit simulation of soil moisture
and vegetative evapotranspiration. The CMAQ has been
designed to take advantage of the additional surface and PBL
parameters produced by this scheme including bulk stomatal
resistance which is then used to compute chemical dry
deposition velocities. The surface model is being added to the
community MM5 and will be available as an option in future
releases from NCAR.
CMAQ uses an interface processor to link the
meteorology and the chemistry-transport model (CTM) called
MCEP. As part of its functionality, MOT contains a generalized
coordinate capability. It also contains algorithms for treating dry
deposition,
2.1.2 Emissions
CMAQ-98 will use an emission processing system
called Models-3 Emissions Processing and Projection System
(MEPPS). The MEPPS is a crucial part of the Models-3
modeling framework. The MEPPS will processes emission
inventory data, performs future projections (including control
scenarios run using the Models-3 study planner) as well as
pre-processing of data for use by the Models-3 CMAQ model.
The MEPPS' input Data PROcessor (MEDPRO) will be capable
of importing, performing QA and converting emission data,
including but not limited to the 1985 National Acid
Precipitation Assessment Program Inventory, 1990 Interim
National Inventory, and 1990 National Emissions Trends
Inventory. MEPPS estimates mobile sources using the Mobile
5a model, and biogenic emissions using the BEIS-2 model.
Point and county level area source emissions data are taken
from inventories. Emission data are tracked and reported by
source category code. Spatial gridding is performed using
Arc/Info G1S software. Meteorological data from MM5 via
MCIP are used in modeling mobile sources and biogenic
emissions. CMAQ-98 MEPPS will have the capability to
provide speciated emissions needed for using CB-IV and
RADM2 chemistry mechanisms or user-defined modifications,
Future versions will be able to support other mechanisms.
2.1.3 Boundary and Initial Conditions
CMAQ-98 will establish the boundary conditions for the nests
by first creating modeled fields at the coarse mode, either at the
108 or 36km resolution. Alternatively, future capabilities can
incorporate observed data fields such as from remote sensors
and satellite platforms. Initial conditions will be obtained from
the end of spinup runs of several days.
2.2 The Chemistry-Transport Model (CTM).
The science core of the CMAQ system is the CTM.
Science processors in CTM will include the following
capabilities:
2.2,1 Chemistry.
CMAQ includes both the RADM2 and the CB-IV gas phase
mechanisms. In addition the CMAQ provides the capability to
edit the RADM2 and the CB-IV mechanisms or to import a
completely new chemical mechanism by means of a generalized
chemical mechanism reader. This reader is external to the
CTM, and can significantly simplify the task of altering the
chemical representation in a CTM. However, any change to the
-------
fundamental representation of the organic species in these two
mechanisms will correspondingly require substantial changes to
MEPPS. The CMAQ also accounts for the formation of
secondary aerosols and the reactions of pollutants in the
aqueous phase. Secondary aerosol formation is parameterized
on the basis of important gas-phase reaction rates and aqueous
reactions are simulated by means of the aqueous chemical
mechanism incorporated in RADM. Both the CB-IV and
RADM2 gas-phase mechanisms are linked to these processes,
however, providing the capability to simulate multi-phase
interactions using either gas-phase mechanism. Finally, two
chemistry solvers are available — the Sparse Matrix Vectorized
Gear(SMVGEAR) algorithm developed by Jacobson and Turco
(1994) and the Quasi-Steady State Approximation (QSSA)
method used in the Regional Oxidant Model. SMVGEAR is
generally recognized as the more accurate of the two, but it is
much slower than QSSA on non-vector computers.
2.2.2 Cloud processes:
Modeling clouds are essential in air quality modeling
due to their critical role in atmospheric pollutant transport and
chemistry processes. Clouds have both direct and indirect
effects on the pollutant concentrations: they directly modify
concentrations via aqueous chemical reactions, vertical mixing,
and wet deposition removal processes, and indirectly affect
concentrations by altering radiative transmittances which affect
photolysis rates and biogenic fluxes. CMAQ-98 will model
deep and shallow clouds using the RADM-type (Walcek-
Taylor) algorithms for 36 and 12 km; an explicit type cloud
scheme is being tested for the 4 km simulations.
2.2.3 Diffusion and Advection
Options for computing subgrid vertical transport
include eddy diffusion, and the Asymmetric Convective Model
(ACM) (Pleim and Chang, 1992), the latter applicable to
convective conditions. Horizontal diffusion is limited to use of
a constant eddy diffusion coefficient. Numerical methods differ
in the handling of advection of concentration fields. Several
methods are implemented in the CMAQ; these include the
method by Smolarkewicz (1983), a scheme by Sort (1989), and
a piecewise parabolic method (PPM) (Collela and Woodward,
1984). The Smolarkiewicz technique is mass conservative, and
it is based on the first order upstream method. The Bott scheme
CPU costs are comparable to Smolarkiewicz's, negative
concentration values are limited, and it is highly accurate and
mass conservative. The PPM subgrid distributions of advective
quantities are represented by a parabola in each grid. PPM
provides local fit to data, is monotonic, and contains special
steepening procedures where sharp gradients exist.
2.2.4 Particle Modeling and Visibility:
One of the major advancements in CMAQ is the
modeling of fine and coarse mode particles. The fine fraction
model described in Binkowski and Shankar (1995) is
incorporated into CMAQ-98. CMAQ will predict hourly
gridded concentration of fine particle mass whose size is equal
to or less that 2.5 microns, speciated to sulfate, nitrate,
ammonium, organics and aerosol water. Secondary sulfate is
produced by chemical reactions of hydroxyl radicals with sulfur
dioxide producing sulfuric acid that either condenses to existing
particles or nucleates to form new particles. Anthropogenic and
biogenic hydrocarbon precursors also react with hydroxyl
radicals, and with ozone and nitrates to produce condensible
material. Aerosol water is equilibrated with relative humidity
and the ammonium to sulfur molar ratios. The fine range is
modeled with two log-normal size distributions; the coarse
fraction uses one log-normal distribution. Nuclei and Aitken
particles are modeled in the smaller of the two fine particle
modes. The second mode is called the accumulation mode. At
the point of overlap, particles from the nuclei mode are shifted
to the accumulation mode when the Aitken-mode growth rate
exceeds that of the accumulation mode. CMAQ model outputs
also include number densities for both fine modes and for the
coarse modes. CMAQ-98 will provide methods for computing
particle nucleation, aerosol dry deposition, and cloud processes
including the oxidation of Sulfur(IV) to Sulfur(VI) by hydrogen
peroxide. Primary particles enter from MEPPS.
The modeling of aerosols in CMAQ provides the
capability to handle other environmental issues including
visibility and semi-volatile air toxics. Visibility (in Deciview
units) is output for each grid and model time step. The visibility
formulation in CMAQ-98 is based on integrating the Mie
scattering over the distribution of the predicted particles. In
another potential application, CMAQ can provide the basis for
modeling the atmospheric transport and deposition of semi-
volatile organic compounds (SVOC) with parameterizations for
their rates of condensation to and/or volatilization from the
modeled particles.
2.2.5 Plume-in-Grid (PinG) modeling:
CMAQ-98 will include algorithms to provide a more
realistic treatment of the subgrid scale physical and chemical
processes impacting pollutant species in plumes released from
selected Major Elevated Point Source Emitters (MEPSEs). The
key PinG components developed to treat the relevant processes
at the proper spatial and temporal scales for pollutant plumes
include a Plume Dynamics Model (PDM) processor, which
simulates plume rise, vertical/horizontal growth, and plume
position and a Lagrangian Reactive Plume Module (LRPM),
which simulates the relevant dynamic and chemical reaction
processes of subgrid plumes until certain criteria are met
triggering the handover of plume material to intercepted grid
cells. PinG will be used for the 36km and the 12 km
simulations while at 4 km resolutions, PinG is not invoked as
the MEPSE emissions are directly released into the CTM 3-D
grid cells. A detailed description of the PinG features is given
in Gillani et al. (1998).
2.2.6 Photolysis Modeling:
The photochemistry of pollution is initiated by photo
-------
dissociation of smog precursors driven by solar radiation. The
amount of radiation available for photochemistry is clearly
dependent on sun angle (time of day), season, latitude, and land
surface characteristics, and greatly affected by atmospheric
scatterers and absorbers, as well as cloud fields. Additionally,
photolytic rates are wavelength and temperature dependent,
thus, making their calculation very complex, and requiring great
accuracy. An advanced photolysis rate model will be installed
in CMAQ-98, with capabilities to predict temporally resolved
3-D gridded photolyses rates using input data from the MM5-
FDDA and other available sources. It will have flexibility in the
specification of wavelength bands, extraterrestrial irradiances,
vertical profiles of ozone and aerosols, cloud distribution and
adsorption cross section and quantum yield data. The model
will have the capability of computing photolysis rates for any
chemical mechanism with a user specified absorption cross
section and quantum yields data. CMAQ-98 defaults have been
set up for CB-IV and RADM2 mechanisms. Studies have
shown great sensitivities to the modeled cloud fields, and
clearly photolysis rates will be limited by the accuracy of the
modeled cloud fields.
2.3. Process Analysis
Process analysis involves examining the effects that
different physical and chemical atmospheric processes have on
pollutant concentrations. It is accomplished in a model by
quantifying the contributions of individual processes to the
overall change- in a species concentration. This information
explains how individual model predictions come about and
reveals the relative importance of each process. It can be used
diagnostically to identify potential sources of error in either the
model formulation or the model input data. Further, it is
particularly useful for gaining understanding of the effects of
making changes to the model or to its inputs. CMAQ-98
provides the capability to perform process analyses using two
different pieces of information: Integrated Process Rates (IPRs)
and Integrated Reaction Rates (IRRs).
The IPRs are obtained during a model simulation by
computing and saving the change in concentration of each
species caused by physical processes including advection,
diffusion, emissions, etc. IPRs are also calculated for chemical
reaction, aerosol production, and aqueous chemistry, but values
provide no information on the particulars occurring within each
process (i.e., they only give the net effect of each process).
These IPRs can be output for each species at varying times to
show how the effects of each process vary both in time and in
space. Thus, they are particularly useful for identifying
unexpectedly low or high process contributions which could be
indicative of model errors.
The IRR analysis deals with the details of the
chemical transformations. CMAQ-98 currently provides the
capability to conduct IRR analyses for gas-phase chemical
species only, but extensions to aqueous chemistry and aerosol
formation are planned. For gas-phase chemistry, the CTM has
been instrumented to compute not only the concentration of
each species, but also to compute the integral of the individual
chemical reaction rates. The computation and output of the
IRRs are synchronized with the standard concentration field
calculations and outputs to permit one to study details of the
atmospheric transformations and their effects on pollutant
concentrations. IRR analyses have been primarily applied to aid
in the understanding of ozone formation (Jeffries and Tonnesen,
1994). For example, IRRs have been used to quantify such
chemical characteristics as OH and NO chain lengths, the
amount of new radical production and termination, the amount
of ozone produced by and the yield of radical species from each
VOC species. These quantities have typically been used to
understand the reasons for differences in model predictions
obtained with different chemical mechanisms.
2.4. CMAQ Code Integration:
The codes for CMAQ system integrated into the
Models-3 framework are in FORTRAN. The integration
procedure was facilitated by following a small set of design,
coding and implementation standards that include: (1)
modularity, which employs the concept of interchangeable
modules within process classes, allowing easy exchange of
time-splitting, science process solvers such as Bott and PPM;
(2) a standard subroutine interface at the module level. (Each
science module is conceived as operating on the gas and aerosol
species concentrations within the domain grid space for its
specified time period (model synchronization time step) in
sequence throughout the modeling scenario.); (3) restriction of
coding practices, which conceal data dependencies, hinder
maintenance and foster hidden bugs, e.g., eliminating common
blocks, at least across modules; and (4) the Models-3 I/O-API
(ref 9), which contains standardized file I/O functions and a
modeler-friendly interface built on top of self-describing
netCDF (ref 10) files that are portable across most, if not all
Unix platforms,
The CMAQ codes have been generalized to
accommodate a range of different computational coordinate
systems. The essential solution in making the code sufficiently
general is to relegate the problem to the MCIP which produces
data in the correct coordinate form for CMAQ. Once a user has
selected the modules and options, building that version of a
CMAQ is made relatively easy. The selected domain and
chemistry information is encapsulated in parameter or data
statements that are automatically written to FORTRAN include
files, which in turn, are linked into the code during the build
process to produce the CMAQ executable. Execution of this
CMAQ requires the correct linkage with the meteorological and
emissions interface processor data to ensure the correct
generalized coordinate implementation.
3.
DISCUSSION
A typical simulation of Models-3/CMAQ provides
hourly air quality fields for regional to urban scales for multi-
day episodes, typically of up to five days in duration. The new
PM-fme standard will be based on annual averages; thus,
utilization of CMAQ in this application will require aggregation
-------
techniques. One such technique, initially developed for RADM
wet deposition applications, was recently modified and
successfully applied to fine paniculate matter by Eder and
LeDuc (1996). This approach, which utilizes visibility data as
a surrogate for fine PM, will be applied to CMAQ on a
continental scale (i.e., contiguous United States, southern
Canada and northern Mexico). Future efforts will be needed to
validate this approach when a network of fine particulate
samplers is deployed; also, aggregation approaches for
mesoscale domains will need to be developed perhaps utilizing
the method by Eder et al., (1994).
The design basis for the CMAQ is its ability to adapt
to science advances in atmospheric process modeling. Thus
there will be a continuing need for evaluating the performance
and the veracity of the process modules as the CMAQ advances
its science basis. The model evaluation activity for the CMAQ
will be staged with the initial efforts to show relative
performance against the RADM, which itself has undergone
extensive model evaluation efforts. Diagnostic evaluation will
continue using-data bases from different regional studies such
as the 1995 Southern Oxidant Study conducted in the vicinity
of Nashville, TN and the NARSTO-NE study.
CMAQ can be configured for a wide range of
applications from science studies and investigations to
regulatory applications. An evaluated CMAQ can provide the
benchmark from which more operational configurations can be
referenced. It is anticipated that as science advances are
provided to CMAQ, future configurations of a more operational
nature can be periodically rebenched as appropriate.
Finally, we encourage full participation by the
scientific and modeling communities to engage in the growth
and use of Models-3/CMAQ. Models-3/CMAQ has flexibility
for incorporating scientific and modeling advances in its air
quality science process modules, for testing of alternative
science descriptions of processes, and for extending its current
capability to handle multimedia environmental issues.
Additionally, the community of users should be vigilant in
performing evaluations and testing against improved databases
and measurement technology to insure model comparability to
the real world.
Comput.Phys. 54: 114-201.
(5) Eder, B.K., J.M. Davis and P. Bloomfeld, 1994: An
automatic classification scheme designed to better elucidate the
dependence of ozone on meteorology. J. Appl. Meteor, 33:
1182-1199.
(6) Eder, B.K. and S. LeDuc, 1996: Can selected RADM
simulations be aggregated to estimate annual concentrations of
fine particulate matter? Reprint from the llth Annual
International Symposium on the Measurement of Toxic and
Related Pollutants, May 7-10,1996, RTF, NC 732-739
(7) Gillani, N.V., A. Biazar, Y. Wu, J. Godowitch, J. Ching, R.
Imhoff, 1998: The Plume-in-Grid treatment of major elevated
point source emissions in Models-3. 10th Joint Conference on
Applications of Air Pollution Meteorology with AWMA,
January 11-16, 1998, Phoenix, AZ, Amer. Meteorol. Soc.,
Boston, MA.
(8) Grell, G.A., J. Dudhia, and D.R. Stauffer, 1993: A
description of the Fifth-Generation Penn State/NCAR
Mesoscale Model(MM5). NCAR Technical Note, NCAR/TN-
398+IA
(9) http://www.iceis.mcnc.org/EDSS/ioapi/index.htm I/
(10) http://www.unidataucar.edu/packages/netcdf/
(11) Jacobson M. and R.P. Turco, 1994: SMVGEAR: A
Sparse-Matrix, vectorized Gear code for atmospheric models.
Atmos. Environ. 28: 273-284.
(12) Jeffries, H. E. and S. Tonnesen, 1994. "A comparison of
two photochemical reaction mechanisms using mass balance
and process analysis", Atm. Em, 28(18): 2991-3003.
(13) Novak, Joan, et al., 1998: Models-3: A unifying framework
for environmental modeling and assessments. Preprint Volume,
10th Joint AMS and AW&MA Conference on the Applications
of Air Pollution Meteorology, Phoenix, AZ, Jan 11-16,1998.
(14) Pleim J.E., and J.S. Chang, 1992: A non-local closure
model in the convective boundary layer. Atm Environ., 26A:
965-981.
(15) Pleim, J.E., and A. Xiu, 1995: Development and testing of
a surface flux and planetary boundary layer model for
application in mesoscale models, J. Appl. Meteor., 34:16-32.
(16) Smolarkiewicz, P.K., 1983: A simple positive definition
advection scheme with small implicit diffusion. Man Wea Rev.,
Ill: 479-486.
4. REFERENCES:
(1) Binkowski; F.S., and U. Shankar, 1995: The Regional
Particulate Model: Part I. Models description and preliminary
results. J.Geophys. Res., 100(D12): 26,191-26,209.
(2) Byun, Daewon, et al., 1998: Description of the Models-3
Community Multi-scale Air Quality(CMAQ) Modeling
System. Preprint Volume, 10th Joint AMS and AW&MA
Conference on the Applications of Air Pollution Meteorology,
Phoenix, AZ, Jan 11-16, 1998.
(3) Bott, A., 1989:, A positive definite advection scheme
obtained by nonlinear renormalization of the advective fluxes,
Man. Wea. Rev. 117: 1006-1015.
(4) Colella, P., and P. L. Woodward, 1984: The Piecewise
Parabolic Method (PPM) for gas-dynamical simulations, J.
DISCLAIMER
This paper has been reviewed in accordance with the
U.S. Environmental Protection Agency's peer and
administrative review policies and approved for presentation
and publication. Mention of trade names or commercial
products does not constitute endorsement or recommendation
for use.
-------
TECHNICAL REPORT DATA
1. REPORT NO.
EPA/600/A-97/078
2,
4. TITLE AND SUBTITLE
Science Features in Models-3 Community Multiscale
Air Quality System
5.REPORT DATE
6.PERFORMING ORGANIZATION CODE
7, AUTHOR(S)
Jason Ching, et al.
8.PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Same as Block 12
10.PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME AND ADDRESS
U.S. Environmental Protection Agency
Office of Research and Development
National Exposure Research Laboratory
Research Triangle Park, NC 27711
13.TYPE OF REPORT AND PERIOD COVERED
Preprint, FY-98
14. SPONSORING AGENCY CODS
EPA/600/9
15. SUPPLEMENTARY NOTES
16. ABSTRACT
Models-3 is framework for environmental modeling utilizing state of science representation of
atmospheric processes in a high performance computing environment. The Community-based Multi
scale Mr Quality (CMAQ) system is Models-3 air quality modeling system. Models-3/CMAQ model
is designed as a multi-pollutant Eulerian grid regional to urban scale air quality and
atmospheric deposition modeling system. It contains state-of-science parameterizations and
algorithms of the relevant and contributing atmospheric processes affecting transport,
transformation and deposition of sources contributing to such pollutant issues as ozone,
particulate matter, airborne toxics and deposition of acidic and nutrient pollutant species.
The release version of Models-3/CMAQ will feature science options available to the user
including the gas phase chemistry mechanisms, RADM2, and CB-IV a set of chemistry solvers,
options for horizontal and vertical advection schemes, algorithms for fine and coarse
particulate matter, photolysis rates and a plume-in-grid approach. These current science
options are discussed in the context of the overall structure of Models-3/CMAQ model which
includes an emission and meteorological preprocessors linked to the CMAQ by interface
processors. Also discussed are the methods to- determine and test the relative contribution
of different atmospheric processes to the air quality predictions, important functions built
into the CMAQ system.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/ OPEN ENDED
TERMS
C. COSATI
18. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (This
Report)
UNCLASSIFIED
21.NO. OF PAGES
20. SECURITY CLASS (This
Page)
UNCLASSIFIED
22. PRICE
------- |