Environmental Engineering Reference
In-Depth Information
from EPA's 2002 Multi-pollutant Modeling Platform, updated for fine-scale
modeling (Tooly and Wesson, 2009). These emissions were processed using
SMOKE, and the Surrogate Tool was used to provide spatial surrogates for grid-
ding the emissions for the 4 and 1 km domain. More information about the
meteorological data and emissions data can be found at EPA, 2008.
CMAQ v4.6.1i was run in a one-way nest. The meteorological inputs were pro-
cessed through MCIP for the 12 km domain. For the 4 and 1 km domains,
meteorological data were provided by disaggregating the 12 km MCIP output
meteorology to 4 and 1 km modeling domains.
CAMx v4.5 was run for the same three domains as CMAQ, using the two-way
nested grid structure and flexi-nesting features. The 1 km domain was run nested
within the 4 km domain, which was nested within the 12 km domain. The 12 km
MM5 meteorological data were processed using MM5CAMx and provided to the
4 and 1 km domains using flexi-nesting (Environ, 2008). The emissions, however,
were provided for each of the three domains at the appropriate resolution.
As part of this study, the EPA's AERMOD dispersion model (EPA, 2004a) was
exercised over a 36 by 48 km grid of receptors spaced 1 km apart centered in the
urban Detroit core. SMOKE was used to generate temporalized and speciated
emissions, which were then formatted for input into AERMOD. Meteorological data
for AERMOD was extracted from the 2002 MM5 12 km data for the grid cell over
the Detroit Metropolitan Airport and were processed using the MM5AERMOD tool.
To provide subgrid cell texture for multiple pollutants, AERMOD estimated
concentrations were then combined with CMAQ 12 km concentrations using a
post-processing technique called the Multiplicative Approach to the Hybrid Method
(MAHM).
1
Model outputs are compared with ambient measurements for a variety of
pollutants. Total PM
2.5
monitoring data was extracted from the Federal Reference
Monitor (FRM) Network, speciated PM data were taken from the Speciation
Trends Network, O
3
data was used from AIRS, and toxics data were used from the
state toxic monitoring sites and the Detroit Air Toxics Initiative Study (DATI).
2. Model Evaluation and Results
The CMAQ and CAMx modeled concentrations from the 12, 4, and 1 km domains
were compared against monitoring results for O
3
and total PM
2.5
. All data were
extracted from the 1 km photochemical domain for comparison.
Table 1
shows the
The MAHM equation is C=CMAQ_primary* (AERMOD_rec/AERMOD_gridavg) + CMAQ_
secondary where CMAQ_primary and CMAQ_secondary are the primary and secondary CMAQ
emissions of a pollutant within the relevant CMAQ grid cell; AERMOD_rec is the concentration
of a pollutant at an AERMOD receptor; and AERMOD_gridavg is the average concentration of a
pollutant for all the AERMOD receptors located within the relevant CMAQ grid cell.
1
