An Improved Model for L-Band Brightness Temperature Estimation Over Foam-Covered Seas Under Low and Moderate Winds

Abstract

In this study, an improved T B model is proposed on the basis of the SSA model and Monahan foam coverage model. A 22-month matched dataset from the Aquarius observation and Argo buoys was collected to build the improved model. However, it was determined that the wind-induced emissivity calculated from the SSA model at low wind speed (WS) displays obvious deviation compared with the satellite observations, which can be corrected using the systematical error correct coefficient (SECC). To evaluate the effects of sea wave spectra on T B , different sea wave spectra are used and compared in the SSA model and the foam coverage model. A new set of parameters for the foam coverage model for different spectra is obtained by fitting the wind-induced emissivity calculated from a 16-month matching dataset of the Aquarius and Argo buoys. The SSA model with a Kudryavtsev spectrum, as well as the improved foam coverage model, is selected for simulating the satellite T B under low and moderate winds. Finally, a 6-month match dataset obtained in 2013 is used for model validation. The root mean square error (RMSE) of the estimated brightness temperature is approximately 0.5 K in V polarization and less than 0.7 K in H polarization. The correlation coefficient is approximately 0.9, except for the case with 45.6° in H polarization. The bias between simulation and observation in V polarization is lower than that in H polarization, and the biases in both polarizations are acceptable when the WS is less than 20 m/s.