Environmental Engineering Reference
In-Depth Information
4. Impacts of Meteorological Mismatch on Air Quality
Simulation
The MM5 modeled winds were input to CMAQ using the SAPRC99-AE3aq
chemical mechanism, a 4-km horizontal grid resolution and 16 vertical layers.
Time series for 24-h observed and simulated PM levels are shown in
Fig. 2
.
Fig. 2.
Simulated and observed 24-h PM
2.5
levels for 2 SFBA PM monitors. Episodes
highlighted: persistent R2-type (diagonal hatch), other (vertical hatch)
For three persistent R2-type episodes
(Fig. 2
), the R2-R1 mismatch occurred
and PM levels were underestimated. Simulated winds were too strong in southern
SV. Light overnight drainage (downslope) flows off the SV rims were observed,
but MM5 winds were instead persistently downvalley. Simulated winds into the
Bay Area from the east and in the northern SJV from the south agreed with the
observations during the R2-R1 mismatch. Winds for another episode with rapidly
evolving large-scale conditions (17-24 December) were realistically simulated,
and modeled PM levels were reasonable.
5. Conclusions
The novel model validation technique indicates days for which a meteorological
simulation cannot capture localized circulation patterns. It also suggests how air
quality model performance may be adversely affected. This categorical model
evaluation is useful for seasonal PM simulations that must distinguish between a
variety of atmospheric conditions. The diagnosed MM5 shortcomings are consistent
with MM5's documented problems of providing excessive synoptic push during
periods of weak large-scale pressure gradients and light terrain-induced winds.
References
Seaman, N.L., 2000.
Atmospheric Environment
,
34,
2231-2259
.
Beaver, S. et al., 2008.
A&WMA Ann. Conf. and Exh.
, #623, Portland, OR.
