Abstract
Color is an important parameter in Soil Science, widely used since the very first soil investigations. Color visually estimated in the field with the Munsell Atlas is used for soil identification and classification, for estimating various soil properties and in soil survey and mapping. The difficulties encountered in the first attempts in establishing empirical relationships between surface color and remote sensing data led to use a more scientific approach. In this paper the significance of Munsell soil color versus spectral signature is investigated and discussed with the help of colorimetric concepts and techniques. Different experiments based on field and laboratory spectral reflectance measurements are summarized. Among the results, the transformation of Munsell data into Red, Green, Blue color coordinates (R,G,B) is the basis for a pragmatic, physically based model relating surface color to satellite data in visible bands. Applications of this model to satellite imagery for soil mapping, soil mineralogy studies and arid lands vegetation monitoring are presented, with emphasis on TM data. In conclusion, the promises of remote sensing of surface color for environmental changes monitoring are discussed in anticipation of the new instruments that will fly on polar orbiting platforms at the end of this decade.
