Physical Mechanisms in Hyperspectral BRDF Data of Grass and Watercress

A hyperspectral bidirectional reflectance data set of grass lawn and watercress canopies acquired under controlled laboratory conditions at the JRC European Goniometric Facility, at Ispra, Italy was analyzed with respect to basic physical reflectance mechanisms and factors influencing reflectance anisotropy. By normalizing reflectance data with the nadir reflectance as a reference, a strong influence of reflectance intensity on observed reflectance anisotropy is demonstrated for both vegetation canopies. This is explained by multiple scattering effects inside vegetation canopies as a function of canopy spectral absorbance characteristics. Two quantities based on canopy reflectances are utilized for characterizing the spectrally variable dynamics of BRDF effects: the anisotropy factor (ANIF), which is simply a normalization with nadir reflectance, and the anisotropy index (ANIX) defined as the ratio between the maximum and minimum reflectance values in the principal plane (or defined azimuth plane) per spectral band. Using these two quantities the study revealed that the basic physical reflectance mechanisms described by Kimes (1983) for broad spectral bands are extendable to hyperspectral bidirectional reflectance data. In addition, the influences of canopy geometry and multiple scattering on BRDF are clearly demonstrated for the erectophile grass lawn and the planophile watercress canopy. Since multiple scattering effects and canopy geometry are dominant factors regulating BRDF effects, vegetation canopy architecture parameters such as the leaf area index might be derived from hyperspectral BRDF data.