Comparison of different methods for monitoring glacier changes observed by Landsat images

With the acceleration of global warming, it has been increasingly important to investigate the roles of glaciers as freshwater sources and sensitive indicators of climate change. Thus, it is of great significance to acquire accurate information on glacier changes. However, few papers have focused on the comparison of glacier monitoring methods. The objectives of this paper are to (1) present three methods for classifying glacier boundaries, including visual interpretation, ratio between TM channels 4 and 5 as well as Normalized Difference Snow Index (NDSI); (2) compare the tree methods to give users some advice on how to choose an appropriate method; (3) analyze the relationship between glacier change and the trends of precipitation and temperature. Current distribution and glacier changes since the 1980s were mapped using multi-temporal optical remote sensing data from the Landsat series. Thematic maps were then generated using three classification methods. Furthermore, GIS-supported investigation was also conducted to get information of glacier changes. Finally, the results were compared. The results indicated that: (1) the visual interpretation method is accurate but time-consuming and operator-dependent; (2) the ratioing method using channel 4 and 5 of Landsat image is fast, accurate but need too much follow-up work; (3) NDSI cannot classify snow and glacier very well, and it sometimes misclassifies snow into glaciers; (4) analyses of precipitation and temperature indicate that global warming is a major factor affecting changes of glaciers.