Use of laboratory hyperspectral reflectance data of soils for predicting their diurnal albedo dynamics accomodating their roughness

The objective of this study was to assess the relationship between the hyperspectral reflectance of soils and its albedo, measured under various roughness conditions. 108 soil surfaces measurements were conducted in Poland and Israel. Each surface was characterized by its diurnal albedo variation in the field as well as its reflectance spectra that was obtained in the laboratory. The best fit to the model was achieved by postprocessing manipulation of the spectra, namely second derivate transformation. Using stepwise elimination process, four spectral wavelengths, as well as roughness index, were selected for modeling. The resulted models allow predicting the albedo of a soil at specific roughness for any solar zenithal angle, provided that hyperspectral reflectance data is available.