Vertical Structure of Rainfall from TRMM PR Data for Assimilation in GCM

1.Introduction Use of proxy satellite observations for improving the model states of divergence, humidity and temperature has gained considerable impetus in the past one decade. Various researchers (Donner, 1998,Heckley et al, 1990, Krishnamurti, 1990) have used the satellite derived rainfall and moisture fields ameliorating the spin up problem in the tropics. Physical initialization methods (Pal et al, 1999) have shown the feasibility of correcting latent heating locations in the model by assimilation of observed rainfall during initial phase of model integration. The rainfall being a representative of integrated column heating cannot be directly assimilated into the model, because the model requires the heating profiles for different levels. In earlier studies in the absence of observed vertical profiles, modelers have derived the profiles from the total rainfall by creating a background profile from the model statistics over a large area. Separate profiles have been generated for land and oceans, because rain has different characteristics over different surfaces. At present the vertical profile of rainfall has started becoming available from the precipitation radar (PR) in space onboard TRMM satellite. This gives an opportunity to study the vertical structure of rain corresponding to a particular rain-rate. In future Global Precipitation Mission (GPM) and Indo-French Megha-Tropiques satellite missions will provide rainfall observations. Microwave analysis and detection of Rain and Atmospheric Structure (MADRAS) sensor will provide the measurement of precipitation over land as well as over the sea. Here we have attempted to derive the vertical structure of convective and stratiform rain using TRMM Microwave Imager (TMI) and PR combined data set. The combined data derives vertical hydrometeor profiles using the data from PR and 10 GHz channels of TMI. The vertical structure derived from these data sets can be used to determine the vertical profile of latent heating using the rain measurements from MADRAS sensor of Megha-Tropiques. The heating profile can directly be assimilated in general circulation model for dynamic physical initialization.