Soil Property Relationships with SPOT Satellite Digital Data in East Central Illinois
1990
Agbu, Patrick A. | Fehrenbacher, Donald J. | Jansen, Ivan J.
The second-generation remote sensing satellites with improved spatial and spectral resolution offer additional possibilities for conducting soil surveys. This study was undertaken to determine the relationships that exist among high-resolution satellite digital data and soil properties useful in map-unit delineation and classification. A second objective was to develop models (linear equations) based on soil properties to predict satellite spectral response. Site, morphological, physical, and chemical soil properties were determined to a depth of 1.2 m for 442 observation sites in two rectangular grids, at two contrasting areas of 3108 ha each. Multiple-regression analyses between SPOT spectral data and soil properties showed that many surface and some subsurface soil properties were significantly correlated to spectral data. Landscape position and percent slope were not important as site characteristics that predict satellite spectral variables. Color of dominant mottles and depth to loess did not significantly affect satellite data. Measured soil properties were tested in terms of their usefulness in predicting SPOT spectral response. The results showed that most surface soil properties and some subsurface properties pertinent to soil classification and map-unit separation can be used to predict satellite data and vice versa. Brightness index proved to be a more useful spectral parameter if surface soil properties are to be extracted from spectral data, but ratioing of the original data resulted in more association of spectral data with subsurface soil properties. Joint contribution from the Dep. of Agronomy, Univ. of Illinois and USDA-SCS.
显示更多 [+] 显示较少 [-]