Qinghai Lake Valley Grassland Biomass Multi-source Remote Sensing Inversion Study Together

Biomass is used to evaluate the grassland ecosystem structure and basic functions of an important indicator, is also used to study the parameters of the basis of the first productive force of the grass.Grassland biomass is the energy of the grassland ecological system and the source of material, is the study of productivity(the productivity), net primary productivity(net primary the productivity, NPP).Vegetation biomass is a ecological system, it can obtain the ability of a potential energy, it is the basic function of the ecosystem and ecosystem structure and so on all has certain significance, is also used to reflect the characteristics of a study area of ecological environment of an important indicator. A research area of vegetation growth, learned this study can be used to regional ecological environment situation of the good and the bad. But, a research area of soil water content, to some extent restricts the growth status and characteristics of the vegetation. Soil moisture content, however, as an important parameter of the land surface, it profoundly affects the ability of a research area of weather and climate, balance, the change of the global water cycle. Based on optical remote sensing image data of vegetation index and grass biomass of remote sensing for estimating model is established on the ground, through correlation analysis method, fitting out the suitable for this study of the statistical regression model of grassland biomass on the ground and in the model as the basis, used to inverse estimating grassland biomass on the ground, and then realize the real-time, dynamic monitoring of large areas of grassland resources. With the continuous development of remote sensing platform, the use of synthetic aperture radar(SAR) data inversion estimates of biomass, has in the grassland ecological research in the field to become a relatively new research subject. However, as a new remote sensing platform, the advantages of day, all-weather SAR data, and won’t be affected by factors such as climate, also has the very strong ability to penetrate the surface, is extremely important in the area of grassland ecosystem research a remote sensing data sources, can accurately and quickly reflect the grassland ecological environment. Therefore, this article focuses on the grassland ecological environment element of qinghai lake basin of passive remote sensing inversion technique of Lord for qinghai lake basin is established the inversion method of grassland biomass for more time spatial scales inversion of factors of ecological environment to provide the support theory and method.The Qinghai lake basin as study area, this study will be of qinghai lake basin in the grass on the ground biomass as an important object of study, and all kinds of grass for the qinghai lake basin types of grassland samples, on the basis of field observation data, respectively using passive optical remote sensing data with active radar remote sensing data to inverse estimation of grassland biomass on the ground in the study area, and select some measured data for model validation and precision evaluation. Through the study, the following preliminary conclusions:1. In the study of soil condition of vegetation index(SAVI), when L = 0.5, S/N reached the maximum value, then decreases gradually, so the qinghai lake valley grassland best L value of 0.5.2. The biomass of inversion model based on passive optical remote sensing dataFor different vegetation index, based on the determination of the statistical regression model of SAVIL = 0.5 coefficient were higher than the statistical regression model based on NDVI coefficient of determination.For different statistical regression model, based on the Landsat 8 OLI, MODIS NDVI of optical remote sensing image, SAVIL = 0.5 cubic polynomial model is set up with the measured biomass determination coefficient were higher than a yuan linear model, logarithm model and exponential model and the power function model.3. Based on the primary biomass inversion model of passive remote sensing data togetherBased on the the passive remote sensing data improved estimates of the grassland vegetation scattering model of aboveground biomass and the measured values high coefficient of determination(R2 = 0.7186), scatter distribution around the 1:1 line, the precision of the model inversion is higher(RMSE = 46.40 g/m2).For different remote sensing platform, based on the passive remote sensing data(Radarsat SAR, Landsat 8 OLI) of the improved waterclouds model inversion grass aboveground biomass based on coefficient of determination is higher than passive optical remote sensing data(Landsat 8 OLI, MODIS) statistical regression model of the coefficient of determination.