PREDICTING SOIL DRAINAGE CLASS USING REMOTELY-SENSED AND DIGITAL ELEVATION DATA

Soil drainage strongly affects the patterns and processes of ecosystem s, including biomass production, vegetation community distribution, so il development, aeration, hydrologic processes, and trace gas fluxes. To obviate the need for extensive field surveys, we present a techniqu e to use a remotely sensed optical image and digital elevation data to predict soil drainage class at a 6- by 4-km research site in a mixed conifer forest in Howland, Maine. Elevation, detrended elevation, slop e, aspect, and flow accumulation were determined from a 10-m resolutio n digital elevation model (DEM) of the site. Normalized Difference Veg etation Index (NDVI) data derived from the Advanced Visible and Infrar ed Imaging Spectrometer (AVIRIS) were used to represent differences in vegetation cover. Classification tree analysis predicted soil drainag e class with an average of 78 percent accuracy.