Monitoring and evaluation of desertification in Shihezi area using Landsat TM imagery

Desertification is one of the most serious ecological and environmental problems in the arid and semi-arid areas of western China. This study demonstrated a cell-based modeling approach to monitor and evaluate the land degradation using remotely sensed data in Shihezi area, Xinjiang, China. Two-date Landsat TM imagery of 2000 and 2008 was used to derive factors such as land surface temperature, NDVI and soil moisture etc. The preprocessing of the images was conducted and the DN values were converted into albedo. The mono-window algorithm was applied to the TM band 6 to compute land surface temperature, and the regressive relationship between Temperature Vegetation Dryness Index (TVDI) and field-surveyed soil moisture was modeled and then used to derive soil moisture factor. After that a weighted linear combination of those factors was applied to create an integrated index to characterize the desertification and degradation of land in the study area. The resultant index was then categorized into four classes: non-desertificated, slightly-desertificated, moderate-desertificated, and heavily-desertificated, and finally a map of desertification was created and used to analyze the land degradation in the Shihezi area. The map shows that the degree of desertification diminished gradually from north to south. Due to the graze control policy and land rehabilitation, the threatening of desertification in 2008 is smaller than that in 2000, and especially in the northern and middle areas. Field verification also supports the results positively and thus the map of desertification can be used as a reference for land management and regional environmental protection.