干旱区生态环境知识资源中心

Due to the sparse and irregular distribution of vegetation in desertification area, the Low inversion precision of vegetation coverage and its change using single satellite remote sensing data become the bottleneck of further exploration of ecological evolution in this region. In order to improve the retrieval accuracy of vegetation coverage in desertification area, this article inverses and dynamic monitors the vegetation coverage of Mu Us Sandy Land in north of Yanchi County combining multi-source data include digital image, Landsat TM and MODIS-NDVI using scale conversion and two pixel model. The results showed that: (1) Vegetation information based on NDVIDC (normalized difference vegetation index based on digital camera) was accurately extracted with the classification accuracy up to 94.3%, which provided a convenient and accurate method for ground survey and remote sensing revise of the vegetation coverage. (2) Vegetation coverage significantly increased after grain to green program and grazing prohibition measures in Yanchi county since the beginning of 2000-2002. (3) The method of vegetation coverage inversion based on multi-source remote sensing information provides a new reference for rapid and efficient vegetation monitoring in desertification areas.