An Efficient Two-Scale Model for the Computation of Thermal Emission and Atmospheric Reflection From.pdf

560 IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, VOL. 44, NO. 3, MARCH 2006 An Efficient Two-Scale Model for the Computation of Thermal Emission and Atmospheric Reflection From the Sea Surface Joel T. Johnson, Senior Member, IEEE Abstract—An efficient implementation of the two-scale model of sea surface thermal emission and atmospheric reflection is described. The model is applied in a study of the reflection of downwelling atmospheric radiation. Results show that reflected downwelling radiation can increase azimuthal variations of total observed brightnesses. Index Terms—Microwave radiometry, rough surface scattering, sea remote sensing. I. INTRODUCTION DATA from the WindSat mission [1] are providing the firstlarge archive of polarimetric sea brightness data. Phys- ical models of this dataset are of interest in wind vector re- trieval applications, as well as for improving understanding of the emission physics involved. The two-scale model is an ap- proximate theory of sea emission that has been applied in several studies [2]–[12] to analyze polarimetric brightnesses, primarily from ground-based or aircraft missions prior to the WindSat launch. Although other approximate theories are available for comparing with data [13]–[18], previous studies have yet to demonstrate any conclusive advantages over or general inac- curacies in the two-scale theory. Numerical methods for com- puting rough surface thermal emission [19], [20] remain too computationally expensive to be practical for general use at present. For this reason, use of the two-scale model for com- parison with the WindSat dataset remains highly relevant. As described in [3], the two-scale model is based on a sepa- ration of the sea-surface into “long” and “short” wave regions, with the choice of the separation point a free parameter. Sea-sur- face waves in the “long” region contribute to the long wave slope variance of the surface; the short waves are then tilted over the distribution of these long wa