A model for the effects of rain on scatterometer data is proposed. Data from the Advanced Microwave Scanning Radiometer (AMSR) and the SeaWinds scatterometer, both on the Midori-H satellite, are used. The model includes three basic rain effects: attenuation, rain roughening of the sea surface, and volumetric backscatter. Attenuation is calculated directly from the radiometer data and beam filling is explicitly addressed. The model simultaneously solves for both the rain roughening and volumetric backscatter. Fitting the coefficients of the model requires an estimate of the radar cross section because of wind roughening, and NCEP Global Data Assimilation System (GDAS) wind vectors are used for this purpose. Both the derived rain roughening and volumetric backscatter are similar to results in published work, but the values are slightly smaller, especially for vertical polarization. Drop size distribution variability is accounted for by formulating the radar equation in terms of the parameters of the radiative transfer equation and using additional radiometric information. Explicit inclusion of vertical profile variability results in an underdetermined problem, but it is implicitly included in fitting the model to the data. The correction makes large improvements in wind speeds and modest improvements in wind directions. Wind statistics and specific examples are shown to illustrate the nature of the improvements. The correction is limited, however, by measurement mismatch issues and the nonlinear nature of the wind retrieval.
