Inferring wind direction from polarimetric passive microwave measurements: the role of anisotropic contributions from reflected sky radiance

The retrieval of wind direction from passive microwave measurements is usually addressed as an inverse problem; a radiative transport model (RTM) is solved for candidate directions. The discrepancy of the fit of the computed top-of-the-atmosphere (TOA) brightness temperatures with the measurements is then used as a measure of the likelihood that the candidate direction is the true direction. This approach depends on having a realistic description of the directional dependence of the component models of the RTM. In previously proposed retrieval algorithms, it has been assumed that the directionality in the RTM is dominated by that of the ocean surface emissivity. Here, we examine the role of the dependence of the reflected sky radiance on wind direction through the change in atmospheric path length as the surface facet is tilted by waves. The effect has been recognized but assumed negligible. Our study shows that this source of directionality in the RTM can be neglected only under a limited range of environmental conditions. Under more general conditions, an accurate modeling of the directionality of the reflected sky radiance through this mechanism is needed to deduce the wind direction.