Using temporal modes of rainfall to evaluate the performance of a numerical weather prediction model

The authors demonstrate that much can be learned about the performance of a numerical weather prediction (NWP) model by examining the temporal modes of its simulated rainfall. Observations from the Weather Surveillance Radar-1988 Doppler (WSR-88D) network are used to evaluate the rainfall frequency, and its diurnal and semidiurnal modes, in simulations made by a preliminary version of the Weather Research and Forecasting (WRF) model for the conterminous United States during the summer of 2003. Simulations and observations were broadly similar in the normalized amplitudes of their diurnal and semidiurnal modes, but not in the modes' phases, and not in overall frequency of rain. Simulated rain fell too early, and light rain was too frequent. The model also did not produce the distinct, nocturnal maximum in rainfall frequency that is integral to the hydrologic cycle of the Great Plains. The authors provide evidence that there were regional and phenomenological dependencies to the WRF model's performance.