To derive a prior database of archetypal BRDF shapes from ground measurements using anisotropic flat index (AFX)

In this study, we develop a new technique to derive a prior database of archetypal BRDF shapes from accumulated ground measurements based on a newly developed angular index named anisotropic flat index (AFX). Through the analysis of characteristics of the semi-empirical, kernel-driven, linear BRDF models, we find that the anisotropic reflectance patterns of land surface at different wavelengths are actually determined by a net scattering magnitude that is dependent of two basic spectral scattering types, volume scattering and geometric-optical surface-scattering. A further normalization of this net magnitude by isotropic scattering results in a new anisotropic flat index (AFX) that can indicate basic dome-bowl anisotropic reflectance patterns of terrestrial surface. This trait makes it possible to support a novel method to acquire some archetypal BRDF shapes, rather than directly based on conventional land cover classification schemes. The sensitivity of the derived BRDF archetypes is initially examined by cross-comparison of the AFX and other variables including model parameters, white sky albedo (WSA) and the NDVI.