Tuesday, 11 February 2003: 4:30 PM
An investigation of the sensitivity of model predictions of precipitation to vertical grid spacing
We are investigating the applicability of atmospheric limited area models, as part of a system of models that also includes surface and subsurface hydrology models, to study regional water cycles and their variabilities. This paper will present the results of sensitivity tests to understand the influence of model vertical grid spacing on the predictions of moisture fluxes and precipitation. Relatively fine vertical grid spacing can allow for better resolution of the vertical structure of the moisture and wind fields, but has the undesirable effects of higher computational costs due to the smaller timestep required to satisfy the CFL condition. We will present the results of simulations that apply the Regional Atmospheric Modeling System (RAMS) to the Walnut River watershed in Kansas. The simulations employ four nested grids, allowing for horizontal grid spacing of 48 km, 12 km, 4 km, and 1 km. Simulations of two precipitation events during 1 -10 March 2000 are used for this study. Vertical grid spacings that stretch with height, starting with 50 m or 200 m near the surface are tested.