Relationships between Satellite-Based Radiometric Indices Simulated Using Laboratory Reflectance Data and Typic Soil Color of an Arid Environment

Abstract

By definition, the color of an object such as soil is highly dependent on its reflectance properties in the visible spectrum. In this study, the relationships between soil color and simulated reflectance values for the Landsat TM and SPOT HRV satellites are examined from a laboratory standpoint. Visible reflectance spectra were acquired for 124 soil samples originated from an arid environment, and selected radiometric indices were worked out for both sensors. All the earlier studies relative to soil color and remote sensing have considered the widely known Munsell method as a reference for soil color quantification. Some characteristics of this system based on a visual comparison of a soil sample with painted color chips may complicate the establishment of simple relationships between reflectance data and soil color. We have applied the CIE 1931 standard method of color measurement which consists in computing color parameters directly from reflectance spectra using colorimetric equations. Color data are expressed according to two polar coordinates called Helmholtz coordinates (dominant wavelength and purity of excitation) and a luminance variable having a similar meaning to the Munsell hue, chroma, and value, respectively. The Munsell system is also employed to estimate soil color. Linear regression analysis between soil color and radiometric indices show a systematic improvement of correlations (r) from about 0.7 to more than 0.9 using Munsell data and Helmoltz data, respectively. Simple radiometric indices (band combinations) calculated from broad blue, green, and red bands are found to be good predictors of each of the soil color components. The increasing availability of spectroradiometers, including in the field, should stimulate the use of Helmholtz coordinates, as a beneficial alternative to the Munsell chart to obtain a precise and reproducible color quantification which may be useful for remote sensing applications. The radiometric indices utilized in this study are potentially helpful to contribute to soil resource and soil degradation cartography using visible satellite data in vast arid regions where soil data are not readily available.