Impact of Configurations of Rapid Intermittent Assimilation of WSR-88D Radar Data for the 8 May 2003 Oklahoma City Tornadic Thunderstorm Case

Various configurations of the intermittent assimilation procedure for incorporating Level-II WSR-88D radar data are examined for the analysis and prediction of a tornadic thunderstorm that occurred on the 8 May 2003 near Oklahoma City. Several strong tornadoes were produced by this thunderstorm, causing extensive damages in south Oklahoma City area.

In the assimilation, the ARPS 3DVAR is employed to analyze radar radial velocity data while the ARPS complex cloud analysis procedure is used to adjust in-cloud temperature, moisture and hydrometer fields according to the reflectivity. Forecasts up to 2.5 hours are made from the assimilated initial conditions. Two one-way nested grids are used with grid spaces of 9 and 3 km respectively. The assimilation configuration experiments are conducted for the 3-km grid using Oklahoma City radar data. The different combinations of the assimilation interval and the length and coverage of the assimilation window are tested, and these factors are viewed with respect to the length and stage of the thunderstorm life cycle. Two different schemes for adjusting in-cloud temperature in the cloud analysis are experimented with different lengths of assimilation cycle and window and their behaviors and impact on the analysis and forecast are analyzed.

The results of these experiments indicate that, starting from the initial conditions in which WSR-88D data are properly assimilated, the model is able to capture well the evolution of the tornadic thunderstorm of that day for up to two and half hours into the prediction. It is also found that good predictions can be obtained through judiciary combinations of several different assimilation settings, for reasons that can be understood. The implications of the choices of assimilation settings for realtime operational applications are also discussed.