Assimilation of Precipitation-affected Radiance in NCEP FV3 Hybrid Data Assimilation System

In current NCEP operational FV3 hybrid data assimilation (FV3DA) system, both clear and cloudy radiances without precipitation from AMSU-A and ATMS sensors are assimilated under all-sky condition using the Community Radiative Transfer Model (CRTM). CRTM provides the simulated radiances and sensitivities to atmospheric, hydrometeor, and surface parameters. The cloudy scenes are assumed to be overcast because CRTM used in operational FV3DA cannot handle fractional cloud coverage. To assimilate cloud and precipitation affected radiance appropriately, the capability of handling fractional cloud coverage had been developed in CRTM and the performance of CRTM under scattering conditions has also been validated through collocated observations and against Radiative Transfer for TOVS (RTTOV) as a reference. It is found that CRTM has systematic cold biases under scattering condition.

The development of the two-column radiance calculation to handle fractional coverage and enhancement made in CRTM to remove the systematic cold biases under scattering condition will be described in detail in the presentation. The changes in FV3DA system such as analysis variables, quality control procedure, background error, observation error assignment for assimilating precipitation affected radiances will be summarized. Finally, the impact of the precipitation affected radiance to analysis and forecast will be discussed.