Member Price:
$5.00
Nonmember Price:
$15.00
|
Your Price: $15.00
Order Code:
Add to Cart |
USEPA’s Air Quality Modeling Group (AQMG) has recently engaged in a series of modeling
applications and evaluation undertaken over multiple (urban, regional, and continental) scales
using the Models-3/Community Multi-scale Air Quality (CMAQ) modeling system developed by
USEPA. The objective of this series of on-going modeling efforts is to establish the feasibility of
using the Models-3/CMAQ system for applications involving criteria pollutants such as ozone
and PM 2.5, as well as related regional haze issues. As part of this series of modeling
applications, an annual model simulation was conducted for the entire continental United States
using a 36-km resolution grid throughout the whole year of 1996. This paper gives a preliminary
evaluation of the model results against the ambient observations such as the IMPROVE and
CASTNet network data as well as against existing modeling work such as the REMSAD model.
The focus of this evaluation work was mainly on PM, especially fine PM (PM2.5) and its
constituents. A number of model sensitivity studies were also conducted using two common gas-phase
chemistry mechanisms (CB4 and RADM2) and ammonia (NH3) emission scenarios. The
preliminary results showed that over the summer months when sulfate PM was the dominant
PM2.5 species, the model simulated PM2.5 fairly well with a slight overestimation. However,
over the winter months when nitrate PM became a key component of PM2.5, the model
significantly overestimated PM2.5, mainly because of the overestimation of nitrate PM. This may
be partially due to the overestimate of ammonia emissions in the winter months. Another
plausible cause of the nitrate PM overestimation is the overestimation of nitric acids (HNO3) in
the model. This nitrate PM overestimation issue is being further investigated. The sensitivity
runs showed that the model gave very similar PM 2.5 predictions using the CB4 and RADM2
mechanisms. The ammonia sensitivity runs showed that the PM predictions were highly sensitive
to the emission rates of ammonia. In general, REMSAD predicted higher sulfate PM but lower
nitrate PM than the Models-3/CMAQ. The next step in this series of modeling efforts is to
engage in an annual simulation through the entire year of 2001 since a relatively large set of ambient PM speciation data over the urban areas will be available for the evaluation of model
results. Initial testing of model simulation also showed that the PC/Linux platform provided a
great cost-effective option for conducting long-term and large-scale photochemical air quality
simulations.
|
CNumber:
ACE 2002 Baltimore, Maryland June 23-27, 2002
|
|
Publication Type:
Proceedings
|
|
Publication Date:
June 2002
|
|
|
|
