AIRS impact on precipitation analysis and forecast of Tropical Cyclone Nargis in a global data assimilation and forecast system

The impact of assimilating quality-controlled Atmospheric Infrared Sounder (AIRS) temperature retrievals obtained from partially cloudy regions is assessed, with focus on precipitation analysis produced by the GEOS-5 Data Assimilation System (DAS), and precipitation forecast produced by the global model. The first event chosen is tropical cyclone Nargis (April 27 - May 03, 2008), which hit Myanmar causing immense devastation. It is found that the precipitation analysis obtained when assimilating AIRS cloudy retrievals (AIRS experiment) can capture regions of heavy precipitation associated with Nargis much better than without AIRS data (CONTRL) or when using AIRS clear-sky radiances (RAD). The precipitation along the storm track shows that the AIRS assimilation produces larger mean values and more intense rain rate than the CONTRL and RAD assimilations. The corresponding precipitation forecasts initialized from AIRS analyses show reasonable prediction skill and better performance than forecasts initialized from RAD analysis up to day-2. Additional results for other tropical cyclones will be presented.