Forecast model applications of retrieved three dimensional liquid water fields

Forecasts are made for tropical storm Emily using heating rates derived from the SSM/I physical retrievals described in chapters 2 and 3. Average values of the latent heating rates from the convective and stratiform cloud simulations, used in the physical retrieval, are obtained for individual 1.1 km thick vertical layers. Then, the layer-mean latent heating rates are regressed against the slant path-integrated liquid and ice precipitation water contents to determine the best fit two parameter regression coefficients for each layer. The regression formulae and retrieved precipitation water contents are utilized to infer the vertical distribution of heating rates for forecast model applications. In the forecast model, diabatic temperature contributions are calculated and used in a diabatic initialization, or in a diabatic initialization combined with a diabatic forcing procedure. Our forecasts show that the time needed to spin-up precipitation processes in tropical storm Emily is greatly accelerated through the application of the data