Wind retrievals under rain for passive satellite microwave radiometers and its application to hurricane tracking

We have developed an algorithm that retrieves wind speed under rain using C-band and X-band channels of passive microwave satellite radiometers. The spectral difference of the brightness temperature signals due to wind or rain allows to find channel combinations that are sufficiently sensitive to wind speed but little or not sensitive to rain. We have trained a statistical algorithm that applies under hurricane conditions and is able to measure wind speeds in hurricanes to an estimated accuracy of about 2 m/s. We have also developed a global algorithm, that is less accurate but can be applied under all conditions. Its estimated accuracy is between 2 and 5 m/s, depending on wind speed and rain rate. We also extend the wind speed region in our model for the wind induced sea surface emissivity from currently 20 m/s to 40 m/s. The data indicate that the signal starts to saturate above 30 m/s. Finally, we make an assessment of the performance of wind direction retrievals from polarimetric radiometers as function of wind speed and rain rate.