Date of Award
Master of Arts (MA)
Geography and Geology
Dr. John F. Shroder, Jr.
The present study represents an attempt to classify and compare features of selected glaciers and their surroundings in Afghanistan and Pakistan for two different time periods. The classification utilizes digital data imaged by the MSS (multi-spectral scanner) carried by the LANDSAT satellite. This digital data permits fairly reliable measurements of the many different features including glacial ice, snow, moraine, water, vegetation and different rock types in the adjacent valleys. The LANDSAT digital data were evaluated using both supervised classification and unsupervised classification algorithms. The supervised classifications include Parallelepiped, Minimum Distance to Means and Maximum Likelihood classifications; the unsupervised classification incorporated was a CLUSTER algorithm. The study indicates that the LANDSAT multispectral scanner data are still very useful and contain more information about the larger features than maps. In many cases the snow and ice boundaries can be clearly identified. A comparison of images from two different dates showed significant variations. Overall the classification is about 50 to 80 percent accurate.
Khan, Mohammad Saqib, "Digital Analysis of Selected Glaciers of Afghanistan and Pakistan" (1989). Student Work. 2041.
Files over 3MB may be slow to open. For best results, right-click and select "save as..."