In recent decades, increasing numbers of Lesser Snow Geese have damaged vegetation areas of their northern breeding grounds. Quantitative data on the timing and extent of this damage are needed to set management goals, and Landsat imagery is one of the few consistent, long-term, spatially extensive data sources available at these remote sites. Ground information is needed for detailed analysis of the imagery. With ground-based radiometry and vegetation data collected at a long-term goose study site, a relationship between reflectance and vegetation amount is established for the coastal salt marsh, a preferred foraging area for the geese. For natural vegetation, relatively little attention has been paid to the choice of radiometric index and vegetation measure. Three vegetation measures (biomass, LAI, and cover) were collected, along with radiometric information at 0.5m resolution in the first five Landsat bands. Various published indices and bandwise regression were evaluated for their sensitivity to vegetation measures and their robustness across species composition. With autocorrelation measures, the results will be applied to Landsat images to map habitat change over time.
展开▼
