Mapping the distribution of wet soils through the use of reflectance modeling : Dragoon Mountains, Cochise County, Arizona

Abstract

Soils darken upon wetting due to changes in the scattering properties of the individual soil particles. The objective of this research was to develop a procedure to map the distribution of wet soils using the radiance measurements acquired by a spaceborne imaging scanner. The soil-mapping procedure was designed for use in the regional exploration for ground water resources. The soil-mapping procedure was based upon the detection of reflectance changes in a comparison of Landsat 5 Thematic Mapper (TM) scenes acquired before and after a rain. The Stronghold watershed, which is situated on the western slopes of the Dragoon Mountains, Cochise County, Arizona, was chosen as the test site for the soil-mapping procedure. TM scenes depicting the watershed on 7 June 1985 and 14 November 1985 were used in the change-detection analysis. The region was dry at the time of the June overpass, the November overpass occurred two days after a rain. The recovery of reflectance from radiance requires knowledge of 1) the orientation of the surface relative to the sun and the satellite, 2) the exoatmospheric solar irradiance, 3) the atmospheric optical depth, and 4) the atmospheric path radiance. The orientation of the surface elements were defined through the use of a digital elevation model of the Stronghold watershed. The solar irradiance and atmospheric optical depth were obtained from the literature; the atmospheric path radiance was estimated from shadowed areas depicted in the images. Temporal changes in reflectance were detected by subtracting the November reflectance estimates from those recovered from the June radiance measurements. Changes significant at the 0.05 level were identified through use of the Student-t test. The identical significance level was used to identify temporal changes in the Perpendicular Vegetation Index, or PVI. A surface element was classified as an anomaly if there was a significant temporal change in reflectance with no attendant change in PVI. Field checks of the anomalies proved that wet soils could be mapped via the remote detection of changes in their reflectance. The majority of the false anomalies could be attributed to the disparity between the spatial resolutions of the radiance measurements and the topographic data.