Abstract
Successful detection of boundaries of soil mapping units (SMU's) using differences in spectral characteristics of surface soils requires that any given SMU appears homogeneous but different from all other SMU's as viewed by the recording device. The purpose of this study was to assess surface characteristics affecting aerial images of soils and the degree to which they differ among SMU's. Soil surface samples were collected along four transects (crossing 11 SMU's) at 10-m intervals. Munsell color coordinates (dry soils) and organic matter, oxalate iron, sand, silt and clay contents were subjected to a statistical analysis. The results showed that except for Munsell hue, more than 50% of the total variability occurred among SMU's. Distinct groups of SMU's were denned on the basis of Munsell value, organic matter, sand, and silt content, but in no case were all SMU s distinguished from one another. This and related studies suggest that these soils could not be mapped accurately on the basis of their spectral response alone but spectral differences would provide additional information on soil distribution Since soil reflectance characteristics are described in terms of Munsell color coordinates and other properties measured by pedologists, these results may be extrapolated into various areas with similar soils.