Advances in Land Remote Sensing

Abstract

The availability of a diverse set of directional observations sampling the viewing hemisphere over a particular land surface can effectively capture its surface anisotropy and thus be used to accurately compute the surface albedo of that surface. While numerous samples may be possible in the field or laboratory, remotely sensed retrieval methods based on data from individual satellites usually must suffice with a limited number of directional reflectances of the surface, and the producers of such data sets must acknowledge that these observations may not necessarily represent a well-distributed sampling (Privette, 1997). Therefore a model is usually adopted to characterize the surface anisotropy ¿ a model which can be inverted with a finite set of angular samples and then be used to predict surface reflectance in any sun-view geometry and derive surface albedo (Roujean et al., 1992; Walthall et al., 1985; Rahman et al., 1993; Engelsen et al., 1996; Wanner et al., 1997, Pinty et al., 2000a, Bréon et al., 2002; Maignan et al., 2004).