Remote Sensing Monitoring Model Contrast And Analysis Of Northern Xinjiang Aboveground Biomass

Northern Xinjiang is the extremely important livestock product supply area in Xin Jiang Uighur Autonomous Region. Timely and accurate understanding of the northern Xinjiang grassland biomass distribution on the ground could provide reference for grassland resources monitoring, protection and rational utilization in North Xinjiang. It has established the different monitoring model of vegetation index and aboveground biomass by using the field survey data, and SPOT VGT, MODIS data from 2011 to 2013 in this region in this paper. It uses the inversion results of optimization model to analysis the spatial distribution of aboveground biomass in northern Xinjiang. On this basis, it takes further compared the effects of different scale modeling on model foundation and estimation accuracy. The main results are as follows:1. Grassland above-ground biomass in northern Xinjiang was significantly correlated with MODIS NDVI, EVI and SPOT VGT NDVI data, this result indicated that it is feasible to combine two kinds of remote sensing data with ground survey data to establish yield estimation model. But it is better to inverse aboveground biomass in northern Xinjiang for quadratic polynomial function which established on the basis of MODIS EVI. And the yield estimation model is: y = 17089x2- 3825.5x + 1001.8, R2 = 0.7268.2. Compared the annual dynamic characteristics of grassland aboveground biomass in northern Xinjiang for 3 years, we can find that: The time of maximum aboveground biomass of grassland appeared earlier in 2011 than in 2012 and 2013. During the growing season, aboveground biomass in 2012 was lower than in 2011 and 2013.3. From northern Xinjiang aboveground biomass spatial distribution, we can found: in northern Xinjiang, the grassland area that their average aboveground biomass was below 3000 kg ? hm-2 accounts for more than 80 percent of the total grassland area. The grassland which their average above-ground biomass was greater than 3000 kg ? hm-2 mainly focused on mountainous area.4. It finds that inversion model of grassland aboveground biomass which based on Ili region was better than that of based on northern Xinjiang both in goodness of model fitting and average yield estimation accuracy. This shows that due to the differences of climate and environment condition, the estimation research and partition modeling for grassland aboveground biomass in Ili region were better than the overall modeling. Both partition modeling and whole model can well reflect the variation characteristic in different types of grassland. This paper provides a new idea for improvement of the accuracy and precision of Grassland Resources Remote Sensing for the further study.