Alpine wetland landscape changes in the Eastern Tibetan Plateau based on coupling analysis of multispectral factor decision tree

In the past 15 years (1995-2010), remotely sensed landscape index analysis, indoor and outdoor experimental analysis and statistical analysis were utilized to study a typical alpine wetland at Maqu County in eastern Qinghai-Tibet Plateau of China for the relation between landscape change and climate change. We investigated: 1) The decision tree module for wetland information extraction was built through analyzing data with spatial character, spectrum character and statistic. 2) changes of alpine wetland landcover and the main climate characteristics (temperature, precipitation) of the alpine wetland; and 3) dynamic changes and causes of alpine wetland landscape. The results indicated the method of decision tree was suitable for information extraction of alpine land, and the local climate change was the primary cause for resulting in reduction of wetland areas and degradation of wetlands in the last many years.