Study on Models for Monitoring of Grassland Biomass around Qinghai Lake Assisted by Remote Sensing
青海湖环湖地区草地植被生物量遥感监测模型

Taking the region around the Qinghai Lake as the study area and using the Landsat Thematic Mapper data and the measured grass yield data, the monadic linear regression models and the non-linear regression models were established, respectively, to express the relations between grassland biomass and the vegetation indices. There are two types of sampling site, i.e., the larger one is 30 m×30 m and the smaller one is 1 m×1 m. Each larger sampling site includes one smaller one which was randomly selected. The major conclusions from this study are: 1) the fitting accuracies of the non-linear regression models are much higher than those of the non-linear regression models, namely, the results obtained from the non-linear regression models are more accordant with the measured grassland biomass data in comparison with those from the monadic linear regression models; 2) the comparison of different forms of the non-linear regression analysis on the relations between the vegetation indices and the measured grassland biomass data indicates that the cubic equation is the best one in terms of the suitability of use in the study area; 3) the results from the non-linear regression analysis show that the order is RVI, NDVI, SAVI, MSAVI and DVI in terms of the fitting accuracy between these vegetation indices and grassland biomass data; and 4) the non-linear model Y = -18.626RVI3+220.317RVI2-648.271RVI+691.093 is the best model which can be used in monitoring grassland biomass based on the vegetation indices in the region around the Qinghai Lake.