Effects of shadows on vegetation indices

An important methodology in the remote sensing of biophysical variables is the use of Spectral Vegetation Indices (SVI). An SVI is an algebraic combination of reflectance values from different wavelength bands to produce a single value. Two SVI are evaluated for their usefulness in determining the Fraction of Absorbed Photosynthetically Active Radiation (FAPAR). Evaluating an SVI requires a comparison to a perfect metric for the desired variable. For remote sensing that means the SVI being tested should be sensitive to the variable to be measured, and should provide consistent results for different vegetation types, minimizing the effect of varying canopy characteristics, such as soil background reflectance. The SVI evaluated in this study are the Normalized Difference Vegetation Index (NDVI) and the Soil Adjusted Vegetation Index (SAVI). Most work on the remote sensing of biophysical variables has been performed on crop and grassland canopies. This study broadens the landscape types to include forest and woodland canopies. Forest canopy spectral reflectance is simulated using the GeoSail model. The GeoSail model combines a geometric model, which calculates the amount of shadows, with the Scattering from Arbitrarily Inclined Leaves (SAIL) model, which calculates the reflectance of the trees. Scene reflectance is modeled by calculating an area weighted average of three landscape components: illuminated canopy, illuminated background, and shadowed background. The reflectance of these components are obtained from the SAIL model