Conterminous U.S. and Alaska Forest Type Mapping Using Forest Inventory and Analysis Data

Abstract

Classification-trees were used to model forest type groups and forest types for the conterminous United States and Alaska. The predictor data were a geospatial data set with a spatial resolution of 250 m developed by the U.S. Department of Agriculture Forest Service (USFS). The response data were plot data from the USFS Forest Inventory and Analysis program. Overall accuracies for the conterminous U.S. for the forest type group and forest type were 69 percent (Kappa � 0.66) and 50 percent (Kappa � 0.57), respectively. The overall accuracies for Alaska for the forest type group and forest type were 78 percent (Kappa � 0.69) and 67 percent (Kappa � 0.61), respectively. This is the first forest type map produced for the U.S. The forest type group map is an update of a previous forest type group map created by Zhu and Evans (1994).