Bidirectional reflectance of Earth targets: evaluation of analytical models using a large set of spaceborne measurements with emphasis on the Hot Spot

Multidirectional observation from the spaceborne POLDER (Polarization and Directionality of the Earth Reflectance) instrument makes it possible to measure the bidirectional reflectance of a large variety of Earth targets. A careful selection of cloud-free measurements with a large directional coverage lead to about 22,000 sets of measured Bidirectional Reflectance Distribution Functions (BRDFs). This data set is used to evaluate the ability of analytical models to reproduce the observed directional signatures. Among those evaluated, the best models appear to be the three-parameter linear Ross–Li model, and the nonlinear Rahman–Pinty–Verstraete (RPV) model. On the other hand, all models fail to accurately reproduce the sharp reflectance increase (hot spot) close to the backscattering direction. Based on physical considerations, we suggest a modification of the Ross–Li model, without adding a free parameter, to account for the complex radiative transfer within the canopy that leads to the hot spot signature. The modified linear model performs better than all others, including the RPV nonlinear model. Although the correction modifies the retrieved directional signature parameters, it does not change significantly the surface albedo estimates.