Modeling Aboveground Biomass Of Alpine Grassland In The Three-River Headwaters Region Based On Remote Sensing Data

Three-River Headwaters Region is not only an important livestock base in our country, but also an important water source conservation area and alpine grassland ecological function area, timely and accurately estimation of the spatio-temporal variation of alpine grassland aboveground biomass is significant for determining regional carrying capacity and utilizing grassland resources. In this paper, we developed a remote sensing inversion model of alpine grassland AGB through various regression analyses based on AGB observations and corresponding MODIS-NDVI and MODIS-EVI data in the Three-River Headwaters region. The temporal and spatial pattern of AGB of alpine grassland in the past decade was then estimated. The main conclusions are as follows:(1) Through regression analyses of AGB and EVI or NDVI, it was found that the two vegetation index and alpine grassland aboveground biomass was positively related. Among them, the correlation coefficient of the EVI and AGB higher than NDVI, reached0.791. Thus it demonstrated the feasibility of inversion alpine grassland aboveground biomass by MODIS vegetation index method.(2) According to the model fit and precision, comparative analyses of AGB and EVI or NDVI between five different regression models, it showed that the power model of EVI had the best performance in estimation of AGB. To improve the simulation results of the estimation model, based on the model established structures and variables, further parameters can be determined by70measured samples AGB and EVI, and then the alpine grassland aboveground biomass estimation model of the Three-River Headwaters region is obtained.(3) According to the estimation model, the analysis of the temporal and spatial pattern of AGB in the last ten years shows that, the spatial patterns of AGB from2005to2014were similar, which reduced gradually from southeast to northwest. the interannual variability of AGB had significant change, which decreased from2005to2008and increased in fluctuation from2008to2014. The results show that the alpine grassland ecosystem function is increasing year by year, and further explains the effectiveness of the nature reserve project of the Three-River Headwaters region in the first phase was significantly and timely launch of the second phase.