Effects of sub-grid scale terrain in meteorology and air quality modeling

Due to the smoothing employed, a meteorological model's view of the terrain elevation in each grid cell may be significantly different from the actual mean terrain elevation for that grid cell. Moreover, there may well be considerable terrain variation within each grid cell. This leads to two kinds of modeling errors: those due to the grid scale terrain height errors, and those due to sub-grid scale terrain variability and the consequent interaction between land surface and the modeled atmosphere. Because of the computational expense of increasing resolution, it is not computationally feasible to resolve the terrain adequately within the meteorological models. Parameterizations and corrections are therefore necessary. There are significant direct effects of sub-grid scale terrain height effects upon emissions, depositions, and other surface fluxes; these effects then interact particularly with the transport and the vertical distribution of pollutants in air quality models under stable night time conditions. These differences in pollutant distribution and transport then affect the chemical response of air quality models. Here, we investigate the scope of these effects for air quality modeling results, for test domains in the California-Nevada and the Alabama-Georgia regions of the US, based on modeling with MM5, SMOKE, MAQSIP-RT, and CMAQ.