Grassland Degradation And Remote Sensing Inversion About Aboveground Biomass In Yanchi County

The paper taked the grassland in Yanchi County as the topic and analyses the change of soil nutrients and soil enzyme activities in the different desertificated grassland. In the meantime, it maked use of TM image to extract eight vegetations as indexes, and built the regression model with six vegetation indexes and aboveground biomass. The aboveground biomass inversion can timely and accurately get the grass growth, and provided certain theoretical basis for the aboveground biomass remote sensing estimat and grassland sustainable development, also provided effective method to carry out large-scale grassland estimation and dynamic monitor. The main research findings of this paper were as follows:Firstly, with the increase of degradation, the soil nutrients decreases at the same time, the effective phosphorus taked on desertification grassland> light desertification grassland>serious desertification grassland>moderate desertification grassland> extremely serious desertification grassland, total P、total N, total K、organic carbon、alkali-hydrolyzable nitrogen、soil potassium are potential desertification grassland> light desertification grassland> moderate desertification grassland> serious desertification grassland> extremely serious desertification grassland. Four kinds of soil enzyme activity contents decreased with the exacerbation of the desertification grassland. The correlation analyses were conducted for the rations between the soil enzyme activities and soil nutrients. Soil catalase content had no significant correlation between organic carbon and total potassium(P>0.05).however have significant or extremely significant correlation with other soil nutrients (P＜0.05 or P＜0.01). Protease、phosphatase、invertase contents have significant or extremely significant correlation with other soil nutrients (P＜0.05 or P＜0.01). The change of soil enzyme activity and soil fertility factors were consistent in the process of grassland desertification.Secondly, the paper used the TM image of Yanchi County in August 2013, taked advantage of the software of ENVI and ArcGis, extracted eight vegetation indexes of RVI、NDVI、TVI、MSAVI、SAVI、 DVI、WVI and BVI, and makes correlaton analysis on the aboveground biomass in the same period. The results showed that apart from WVI and BVI, the rest six vegetation indexes have extremely significant correlation with aboveground biomass. The correlation coefficient between RVI and grassland bimass reached 0.956. It maked regression analysis on six vegetation indexes and aboveground biomass respectively and 36 regression models were established.Thirdly, the paper analysed vegetation indexes and aboveground biomass and built regression model, and finally found that RVI has the highest correlation coefficient, and the second was NDVI. The optimum model is cubic polynomial regression model, followed by the quadratic polynomial regression model, the worst is exponential regression model, and the cubic polynomial regression model is y=4.7539RVI3-38.708RVI2+213.04RVI-161.71.Fourthly, It taked precise testing of RVI-Cubic polynomial regression model, and it showed that the average error coefficient of above-ground biomass and prediction measurement is 16.56%,83.44% of the regression rating precision, which could be easily seen that the regression model obtained from the remote sensing vegetation index could be used to monitor the grassland aboveground biomass.