Evaluation of 2002 Multi-pollutant Platform:  Air Toxics, Ozone, and
Particulate Matter

Sharon Phillips, Kai Wang, Carey Jang, Norm Possiel, Madeleine Strum,
and Tyler Fox 

Main Contact:  Sharon Phillips, Air Quality Modeling Group, Office of
Air Quality Planning and Standards, USEPA, MD: D243-01, 109 T.W.
Alexander Drive, RTP, NC  27711

E-mail:   HYPERLINK "mailto:phillips.sharon@epa.gov" 
phillips.sharon@epa.gov 

ABSTRACT

An annual 2002 multi-pollutant model application and evaluation study
was performed for the continental United States using a multi-pollutant
version of the U.S. EPA’s Community Multi-scale Air Quality (CMAQ)
modeling system at grid resolutions of 36-km continental U.S. and 12-km
Eastern U.S.  The CMAQ multi-pollutant v4.6.1 was developed to predict
ozone, particulate matter, mercury, and 38 other hazardous air
pollutants (HAPs) within one model simulation.  The focus of this effort
is the evaluation of model predictions of toxics species (including
metals) and mercury deposition using 2002 measured data.  Model
evaluation results for ozone, PM2.5 component species and precursor
gases, and deposition of sulfate and nitrate will also be presented. 
Based on the 36-km and 12-km 2002 simulations, sulfate predictions
compare very well with measurements over the summer months in the
eastern U.S. where sulfate is the dominant PM2.5 species.  The model
appears to simulate the total nitrate (nitrate PM and nitric acid, HNO3)
fairly well, but the partitioning between nitrate PM and HNO3 leads to
modest scatter in predictions of nitrate PM.  Hourly and eight-hour
maximum ozone model performance shows good agreement with ambient
measurements during the ozone season (May, June, July, August, and
September).  Model performance for some non-ubiquitous HAPs shows a
general under-prediction tendency.  This paper will discuss HAP model
performance in terms of:  (1) uncertainties in monitoring methods, (2)
limited measurements in time and space to characterize ambient
concentrations, (3) commensurability issues between measurements and
model predictions, (4) HAP monitoring issues, (3) HAP inventory issues,
and/or (4) boundary conditions, emissions, and science uncertainty
issues.  This paper will also examine model predictions to gain a better
understanding of the chemical and physical interactions between
concentration and deposition of ozone, PM2.5 component species, toxic
species, and precursor gases (including gaseous toxics) as well as their
temporal and spatial relationships and the extent that model performance
varies between the 36-km versus 12-km simulations.

