Development and Testing of Improved Techniques for Modeling the Hydrologic Cycle in a Mesoscale Weather Prediction System.

This project addresses the need to improve the surface hydrology component in mesoscale atmospheric prediction models and specifically the temperature and humidity forecasts. One way we did this was to initialize the mesoscale model with continuously updated and reasonable values of soil moisture content. To accomplish this task two approaches were taken one in which a soil hydrology model SHM Capehart and Carlson, 1994a was used to update the initial values of soil water content in the BATS land surface component of the Penn StateNCAR mesoscale model MM, specifically its most recent, non-hydrostatic version Smith et al., 1994. The other approach was to use two remote sensing techniques to make better estimates of initial conditions. One of these techniques involved using radar reflectivity data to initialize the humidity field on the basis of existing convection ion which is specified from radar observations during a pre-forecast period. The other technique is to estimate the surface soil moisture availability using remote measurement of surface radiant temperature and vegetation index obtained via satellite and then to nudge the values of initial soil water content determined from the hydrology toward those values of initial soil water content determined from the hydrology toward those values estimated from the satellite measurements Capehart and Carlson, 1994b. -BKA