The impact of TRMM Microwave Imager (TMI) derived surface rainfall data on the numerical simulation of Super Typhoon Paka (1997) is assessed. Numerical simulations using the Penn State University/National Center for Atmospheric Research mesoscale model MM5 are conducted for the mature stage of Paka. Two sets of initial conditions are derived from global analyses from the NASA Goddard Earth Observing System (GEOS): one without rainfall data and one including the rainfall data. Simulation results clearly demonstrate that the GEOS analysis with TMI rainfall data leads to an improved forecast of Paka in terms of its intensity and kinematic and precipitation structures.

The previous results involved assimilation of the rainfall data into a global analysis system with relative coarse horizontal grid resolution and then simulation with a higher-resolution mesoscale model. At the conference, preliminary results of direct assimilation of the rainfall data into the mesoscale model will be presented.