Physical initialization in the NMC Global Data Assimilation System

The impact of assimilating Goes Precipitation Index (GPI) rain rates in the National Meteorological Center Global Data Assimilation System (NMC GDAS) via physical initialization is explored. Reverse algorithms of the simplified Arakawa-Schubert cumulus parameterization and the parameterized surface fluxes are developed in order to assimilate the GPI data over the tropics during the six hour forecasts linking successive GDAS analyses. Results from two experiments show improvements to both analyses and forecasts when GPI rain rates are assimilated.