Precipitation Verification Studies of the Weather Research and Forecasting (WRF) Model along the East Coast of the US

During the period of June – August 2004, a precipitation monitoring experiment was conducted along the Eastern shores of the US. The experiment was conducted in a region centered on the Chesapeake Bay. A regional scale and micro scale rain gauge network was deployed to observe the regional to small-scale variability of precipitation. The NASA 10-cm polarimetric Doppler weather radar (NPOL) was deployed on the Eastern shore of the Chesapeake Bay (near Oyster, VA) to provide high-quality polarimetric rainfall estimates. These data were merged the surrounding WSR-88D observations and evaluated with the deployed rain gauge network. These precipitation products were the basis for WRF model verification of the prediction of precipitation on temporal scales of 0 – 6 hr.

During the study period, the WRF model was run over the domain centered at Oyster, VA (37.3°N, -75.9°W), with 91 x 82 grid at a horizontal discritization of 10 km and 31 vertical levels. The model was initialized using 40 km ETA analysis. A variety of precipitation events were observed including synoptic scale cold front passages, a tropical storm, mesoscale convective systems, and smaller scale summertime convection. This study performs case study analysis of the WRF model ability to predict the range of precipitation events mentioned above. A variety of statistical measures including RMSE, mean absolute error, probability of detection, and false alarm ratio will be used for evaluation of the WRF model precipitation forecasts. An overview of the study and a discussion of the results will be given during the presentation.