Modeling the Anisotropic Reflectance of Snow in a Kernel-Driven BRDF Model Framework Using a Snow Kernel

Abstract

The linear kernel-driven RossThick-LiSparseReciprocal (RTLSR) bidirectional reflectance distribution function (BRDF) model was originally developed for modeling the simplified scenarios of the continuous and discreet vegetation canopies, and has been widely used to fit the multiangle observations for the vegetation-soil system of the land surface in many fields. However, there is a need to develop this model to characterize the light scattering properties of snow, which tends to exhibit strongly forward scattering behaviors. This study proposes a snow kernel to describe the reflectance anisotropy of snow, mainly based on the asymptotic radiative transfer theory (ART) for a semi-infinite weakly absorbing layer of snow, and then applies this kernel to the framework of kernel-driven BRDF model. This snow kernel adopts the analytic form of the ART model with an improved ability in forward scattering direction, particularly in a case of a large viewing zenith angle (> 60°) where the simulation accuracy of the ART model somewhat decreases in the principal plane (PP). Validation of this method was implemented using observed multiangle data. Pure snow targets were selected from the entire archive of the POLDER BRDF data. This validation demonstrates that this proposed snow kernel in the framework of the kernel-driven RTLSR model show potentials for many potential applications, particularly in the field of Earth’s water cycle and radiation budget where snow cover plays an important role.