Preliminary Steps in Assimilating SSM/I Brightness Temperatures in a Hurricane Prediction Scheme

Different aspects of assimilating satellite-observed microwave radiances (brightness temperatures) into the initial vortex of a hurricane prediction model are discussed. The tangent linear and adjoint observation operators were developed from a computationally inexpensive and reasonably accurate radiative transfer model. These models have the advantage of being able to perform in all types of weather, including rain. The adjoint radiative transfer model was used to conduct a sensitivity analysis of brightness temperatures to different atmospheric and surface variables. The sensitivities computed by the model compare favorably with physical understandings of how brightness temperatures are affected by the atmosphere and the surface. The errors associated with some of the approximations in the radiative transfer model were estimated from comparisons with a more accurate model. These errors were found to be smaller than estimates from previous studies. The random errors associated with brightness temperature observations were also estimated from statistical structure function calculations and were found to be in line with estimates previously used. The models developed and the errors calculated for this study will be used in future work to assimilate brightness temperatures in hurricane initializations and to evaluate the performance of different microphysical schemes in hurricane prediction.