Improved air quality simulations through the use of GOES-derived cloud data for the TexAQS-II intensive study period

Inaccurate cloud predictions in meteorological model can cause incorrect representations of photolysis rates used in the air quality simulations. For modeling ozone concentration, cloud information is one of the key meteorological inputs affecting photochemical reactions in the atmosphere. However, current meteorological models have difficulties in simulating convective clouds within a 4-km grid, which is a preferred resolution used for studying air quality in metropolitan areas. Large areas of clouds associated with frontal system can be misplaced due to the inaccurate large scale inputs that are used to initialize the mesoscale meteorological model. The Second Texas Air Quality Study (TexAQS-II) in summer 2006 was conducted to understand air quality problems in Eastern Texas. Meteorological and chemical simulation results from the fifth-generation Penn State/NCAR Mesoscale Model (MM5) and the EPA Models-3 Community Multiscale Air Quality modeling system (CMAQ) were analyzed using measurements for the TexAQS-II intensive period. There were cases of over- and under-prediction of ozone levels due to incorrect clouds in the meteorological inputs. During 23rd – 25th August 2006, MM5 inaccurately predicted clouds and precipitations resulting in over-prediction of ozone peaks in the Houston/Galveston area while under-prediction of ozone in the Dallas area was shown on 2nd September 2006 due to overcast condition simulated by MM5. With the GOES-satellite derived clouds generated by University of Alabama in Huntsville (UAH), the cloud related parameters in the meteorological inputs were modified and the updated inputs were used in CMAQ simulations in 12-km & 4-km resolutions to examine how the satellite-observed cloud data can improve air quality simulations. The work shows the results of this approach and its evaluation with the observations during the TexAQS-II campaign.