Regional land cover characterization using multiple sources of intermediate-scale data

Abstract

Many United States federal agencies need accurate, intermediate scaled, land cover information. While many techniques and approaches have been successfully used to classify land cover in relatively small regions, there are substantial problems in applying these techniques to large multi-scene regions. An evaluation was conducted of the multiple layer land characteristics data base approach for generating large area land cover information. Mosaicked leaves-on Landsat thematic mapper scenes were used in conjunction with leaves-off thematic mapper data, digital elevation (and derived slope, aspect and shaded relief) data, population census information, defense meteorological satellite program "city lights" data, land use and land cover data, digital line graph data, and national wetlands inventory data to derive land cover information. This approach was evaluated for Region III of the United States Environmental Protection Agency (middle Atlantic states).