The Influence of Variations in Surface Treatment on 24-Hour Forecasts with a Limited Area Model, Including a Comparison of Modeled and Satellite-Measured Surface Temperatures

Abstract

The effect of variations in surface parameters on 24-hour limited area forecasts has been examined on a day in July 1981. The vehicle for the study is a ten-level primitive equation model covering most of the continental United States. Variations in moisture availability, surface roughness and soil flux treatment generally do not produce large differences in 24-hour forecasts of primary variables, except in the extreme cases. Precipitation totals, however, are surprisingly sensitive to the surface treatment in several areas where significant amounts of precipitation are evidenced in a control forecast. In these areas, modulations of 10–30% of the control amount due to surface changes are common. In the areas exhibiting principally large-scale precipitation, evidence points to modulations in evaporation as the cause for the precipitation differences. Areas of principally convective precipitation exhibit differences which can be attributed to model-calculated changes in low-level moisture convergence patterns between forecasts. Diurnal surface temperature range measured by the VAS instrument on GOUES-3 is compared to those generated in model experiments over the model grid. Good agreement is found in arm with moderate to large moisture availability and roughness heights. Poor agreement is evidenced in the western United States where the daytime surface temperature and flux balance is shown to be critically sensitive to small errors in the moisture availability.