干旱区生态环境知识资源中心

Based on the climatic, surface and human factors of desertification, an index system and an intelligent model of desertification forewarning are established for the downstream of Tarim River. Based on the remote sensing, climate, land surface, and human data in 1990, 2000, and 2006, the ArcGIS software is used to quantify and rasterize the data of each factor. The forewarning model reveals the distribution of desertification degree in the three study periods on a scale of 30 x 30 m and the results are used to conduct the correction and verification of model parameters. The development trend of desertification in the study area in 2015 is then forecasted under the conditions of "intermittent water conveyance" and "no water conveyance" respectively. Through the parameter correction, the accuracy of the simulated distribution of desertification degree can be more than 90%, showing a good performance of the model. The intermittent water conveyance plays a certain role in the reversal of desertification in the downstream of Tarim River, which is a necessary measure to prevent a complete desertification. However, the reversal effect is only confined to a limited extent along the river channel and cannot thoroughly change the current situation of desertification in the study area.