Abstract
Eleven environmental variables, elevation, slope angle, slope aspect, surface roughness, soil type, geology, percentage of vegetation, percentage of trees, percentage of shrubs, percentage of herbs, and percentage of bare ground, are evaluated as to their impact on LANDSAT thematic mapper (TM) spectral signatures. Multiple regression models indicate that a combination of percentage of trees and shrubs have the most significant impact on TM spectral response, and, in fact, mask the effects of other tested environmental variables. Regression models are also developed for major soil types which significantly predict the amount of trees and shrubs present on a site.