A Remote Sensing Technique for Estimating Watershed Runoff

摘要

Research was carried out to derive runoff coefficients for watersheds of Mill Creek in Coshocton County, Ohio, from LANDSAT multispectral data by including the more detailed (level II) surface conditions and soil types along with land cover (level I) types in the analysis techniques. Regression models were developed from treating the runoff coefficient of each sample site as the dependent variable, and the mean radiance measurements in the four LANDSAT channels as the independent variables. The models were then extended over the area, and the classification results consisted of digital color-coded maps illustrating spatial variability of runoff potential for the surface.