Model assessment and verification - Air Pollution Modeling and Its Application - page 296

2. Study Description
2.1. Model overview and scenarios
The Community Multiscale Air Quality (CMAQv4.5) chemical transport model
(Byun and Schere, 2006) was applied in this study using the CB-IV photochemical
mechanism. Simulations were performed for the same 3-month summer period
(June 1-August 31) in 2002, 2004, and 2005. The modeling domain covered the
eastern United States (US) and southeastern Canada with a 12 km horizontal grid
cell resolution. There were 14 vertical layers with a layer 1 thickness of 40 m.
Gilliland et al. (2008) provides details about these model simulations and results
of a prototype application of the dynamic evaluation approach involving an
assessment of the impact on modeled maximum 8-h O 3 concentrations in the
eastern US due to major point source NO x emission reductions over these periods.
Meteorological fields were generated by the Penn State/NCAR fifth-generation
mesoscale model (MM5v3.6.3) with a four-dimensional data assimilation (FDDA)
technique. The CMAQ Meteorology-Chemistry Interface Processor (MCIP v3.1)
was exercised to post-process MM5 output into compatible input data sets with
hourly 2-D and 3-D meteorological parameter fields for the CMAQ simulations.
Specific details about the MM5 simulations are given in Gilliland et al. (2008).
The 3-D emissions were generated by the Sparse Matrix Operator Kernel
Emissions (SMOKE v2.2) processing system. Anthropogenic emissions from the
EPA 2001 National Emissions Inventory (NEIv3) were used to generate surface
and elevated point source emissions with complete details in Gilliland et al.
(2008). The MOBILE6 model generated gridded on-road vehicle emissions based
on projections of vehicle-miles-traveled (VMT) and fleet factors for a reference
county in each state for each summer period. The typical temporal variation in
average weekday mobile emissions (Fig. 1) shows a rapid early morning ramp up
due to increasing traffic with the peak rush period emissions from 7:00 to 8:00
AM. Figure 1 also indicates mobile NO x emissions declined between these two
summer seasons in this grid cell encompassing Washington, DC. In fact, mobile
NO x emissions in the modeling domain decreased by close to 18% between
summers 2002 and 2005 from the MOBILE6 model (Gilliland et al., 2008).
2.2. Measurements
Hourly NO x measurements at numerous sites in the eastern US were analyzed in
this study. Data for different years were obtained from US EPA's Air Quality
System data base (http://www.epa.gov/ttn/airs/airsaqs/detaildata/).