Monitoring grassland degradation in yiminhe mine of china using TM remotely sensed data

For mining would destroy grassland ecosystem seriously, monitoring grassland degradation and analyzing its influence mechanism in mine areas can play an important role in grassland ecosystem restoration and reconstruction. This paper was mostly focused on the remotely-sensed monitoring of grassland degradation. We first computed NDVI (Normalized Difference Vegetation Index) in different time using TM images of 5 years (1987, 1992, 1995, 1998 and 2005) in Yiminhe Mine of China. Using sub-pixel decomposed model based on NDVI, we established distinct vegetation fractional coverage models for different sub-pixel structure in every pixel, to estimate vegetation fractional coverage in the five years. Then, grassland degradation system was constructed taldng declining proportion of vegetation fractional coverage in four stages, dominant species, prophyta and land surface conditions as references. We subdivided degradation degrees into three grades: slight, moderate and serious. This study showed that grassland degradation was more and more serious as a result of common influence of human activities and climate drought.