Assimilation of GMI level-1C radiance and DPR level-2A reflectivity in the Goddard WRF Ensemble Data Assimilation System

Currently, millions of observations are incorporated in operational data assimilation for numerical weather prediction (NWP) and re-analyses for climate studies. However, precipitation observations indirectly measured by satellite instruments are not routinely used. Because precipitation is a non-linear microphysics process, when radiances are affected by precipitation, it remains a scientific challenge to connect the observed signals to model physical and dynamic states to make effective corrections to forecasts and analysis. The Goddard cloud-resolving WRF ensemble data assimilation system has been developed with a focus on utilizing satellite observed precipitation information and incorporating microwave and radar techniques, particularly in estimation of precipitation properties. We present the bias correction and the ensemble assimilation algorithms developed for GPM observations, and results from experiments carried out in the GPM field campaign domain and intense observing period. The data impact to 4D dynamical precipitation estimation will be examined by using level-1 observations from multi-sensors at field-of-view resolutions and at varying overpass times during the assimilation window. Data from ground validation campaign are utilized in estimation problem as additional constraint or served as independent verification.