One of the keys for the air quality predictions of ozone, particulate matter (PM) and air toxics pollutants to assess human exposure and human risk at neighborhood scales (~1 km horizontal grid spacing and at the canopy level) is the meteorological fields which are required to compute the transport, dispersion and removal of pollutants. The roughness approach used to model the dynamics by most mesoscale atmospheric models requires an implicit assumption of statistical stationarity and spatial homogeneity of the surface roughness elements. That assumption is not satisfied for vertical turbulent fields in the canopy and the urban roughness sub-layer (RSL) where pollutants are emitted, and thus that should have an impact on the pollutant dispersion simulation.

To address this problem, a new urban canopy parameterization (UCP) has been developed inside MM5 using the drag-force approach to represent the dynamic effect of the canopy following the previous works of Brown & Williams (1998), Lacser & Otte (2001), and Martilli et al. (2002). To represent the thermodynamic effects of the canopy, this UCP has been coupled to the soil model SM2-U (Dupont et al., 2002) which considers in detail both rural and urban surfaces. This overall approach is called DA-SM2U. To support this approach, computer-based analysis techniques (e.g., ArcView GIS tools and methodologies) are applied to various datasets, including digitized buildings, land use/land cover, and other essential datasets for the Houston area to assess morphological parameters required by DA-SM2U. Results will be shown for a Houston study including gridded UCP from the digitized databases, applications of DA-SM2U inside MM5. Additionally, results of air quality prediction for Houston using the USEPA Community Multiscale Air Quality (CMAQ) modeling system driven by results of both DA-SM2U/MM5 as well as the standard form of MM5.