Simulations of regional high-resolution climate scenarios are becoming more and more important for local resource management and environmental impact assessment. All the regional climate models (RCMs) are developed based on regional numerical weather prediction (NWP) models. Therefore, RCMs are sensitive to the uncertainties in the lateral boundary conditions and the physics parameterizations just as the NWP models are. Since regional climate over land is the result of the interaction between the large-scale flow and the local scale forcing exerted by the topography and the inhomogeneity of the land surface, it is important to illustrate and eventually measure the relative importance of the sensitivity of an RCM to the uncertainties in the specification of the large-scale flow and those in the physics parameterization.

In this presentation, we illustrate and provide a measure of the relative importance of the sensitivity of an RCM to the uncertainties in the lateral boundary forcing and those in the physics parameterizations. As an example, simulations of the low-level winds in Central California are performed using the Weather Forecasting and Research (WRF) model, and the results of the simulations are compared with the wind profiler and surface observations during an IOP of the 2000 Central California Ozone Study (CCOS). These comparisons are intended to reveal the sensitivity of the simulated low-level winds in the San Joaquin Valley to the large-scale atmospheric forcing and soil initial conditions that are derived from the NCEP and ECMWF analyses. This work is motivated by the fact that it has become very appealing to develop an RCM based on the WRF model due to the availability in the model of the high-order numerical scheme and the extensive options of the physics parameterizations.