The Effect of Assimilating Clear-Sky INSAT 3D Radiances on Heavy Rainfall Prediction over Indian Region

Early mitigation of natural hazards like flash floods, cyclones etc. has been the focus of much of mainstream meteorological research. With the advent of satellite based remote sensing, forecasting and the attendant mitigation has become more prudent and efficient. This study looks at the problem of heavy rainfall over the Indian ocean and ways and means to improve the forecast skill of such events. Currently, clear-sky infrared radiances from INSAT 3D, a geo-stationary sounder, is being assimilated to access its impact on heavy rainfall prediction by a few researchers. In order to eliminate the cloudy radiance observations, an adaptive threshold based cloud screening algorithm which can vary the threshold values both temporally and spatially, is used in this study. It is inevitable for any assimilaion study to have a fast forward radiative transfer model. In view of this an ANN based fast radiative transfer model was developed and tested on various clear-sky scenarios. Initial ensembles required for the assimilation are generated using an iterative EOF (Emprical Orthogonal Function) based perturbation technique wherein the perturbations were made 12 hours prior to the proposed analysis time in order to allow for the model spin-up. The Local Ensemble Transform Kalman Filter (LETKF) assimilation strategy is adopted. Assimilations are performed for a series of precipiation events over the Indian region in order to bring out the robust view of the assimilation strategy. Assimilation experiments are being performed to evaluate the influence of temperature and humidity update on the precipitation forecast. The precipitation forecast will be validated with observed rainfall estimates from satellites.