Couple Ensemble Kalman Filter System and Variational Algorithm - the Impact of Assimilating Radar Observations for Very Short-Term Forecast : Part Ii

Weather radars provide very high-resolution wind and rainfall information in time and space.Recently, Radar data is assimilated to numerical model, and expect to improve the capability ofshort-term weather forecast. By Observing System Simulation Experiments (OSSEs) with perfectmodel assumption, this study continues the work in part I, investigate the impact of assimilatingdirect (radial wind and reflectivity) and indirect (retrieved pressure, temperature and humidity) radarobservations. The purpose of this project is to: (1) how to shorten the cycling process of assimilatingradar data, so that one can obtain optimal analysis in time and improve quantitative precipitationforecast (QPF)；(2) analyze and examine the impact of assimilating radar data for short-term weatherforecast at mesoscale / convective scale；(3) verify the performance of microphysics process in thenumerical model by using dual-Pol radar observations, and understand what is the model errors. Bycoupling the ensemble Kalman Filter data assimilation system and retrieval technique from radarobservation (Variational algorithm), one would like to assimilate radar observations efficiently andimprove the very short-term forecast (0-6h).