Research on dynamic estimation models of grassland based on MODISNDVI in the middle section of northern hillside of Tianshan Mountain

Taking the grassland in middle section of the northern hillside of Tianshan Mountain as a typical study area, this research utilized EOS/MODIS satellite remote sensing data to process the dynamic monitoring of biomass change of grassland. The article analyzed the spatiotemporal change characteristic of vegetation index, as well as the correlativity between NDVI and aboveground biomass with MODIS vegetation index, consequently established the MODISNDVI of biomass dynamic estimation model in different seasons at mountain meadow grassland, mountain grassland and mountain desert grassland. The result indicated that the optimal dynamic yield estimation models of three types of grassland were the linear regression model, quadratic curve regression model, and the power function curve model, and their yield estimation accuracy reached 83.06%, 90.85%, 88.06%, respectively.