Improved forest classification in the northern Lake States using multi-temporal Landsat imagery
PE and RS, Photogrammetric Engineering and Remote Sensing
Forest classifications using single date Landsat data have been only moderately successful in separating forest cover types in the northern Lake States region. Few regional forest classifications have been presented that achieve genus or species level accuracy. We developed a more specific forest cover classification using data from early summer in conjunction with four MSS dates to capture phenological changes of different tree species. Among the 22 forest types classified, multi-temporal image analysis aided in separating 13 types. Of greatest significance, trembling aspen, sugar maple, northern red oak, northern pin oak, black ash, and tamarack were successfully classified. The overall classification accuracy was s83.2 percent and the forest classification accuracy was 80.1 percent. This approach may be useful for broad-scale forest cover monitoring in other areas, particularly where ancillary data layers are not available.
Wolter, P.T.; Mladenoff, D.J.; Host, G.E.; and Crow, T.R., "Improved forest classification in the northern Lake States using multi-temporal Landsat imagery" (1995). Aspen Bibliography. Paper 1839.