青海湖流域草地产草量与MODIS植被指数的相关性分析

Correlation Analysis of MODIS Vegetation Index and Grass Production in Qinghai Lake Basin

借鉴国内外基于遥感手段监测植被的方法,利用MODIS-NDVI遥感影像数据,探讨如何对草地变化进行监测。提取青海湖流域典型草原遥感影像植被指数,并与野外调查获得的79个样方生物量数据进行对比分析,建立了多个回归模型用于研究植被指数与草地产草量的相关关系。结果表明,植被指数与草地产草量之间存在较好的相关性,但是不同的回归模型与草地产草量相关性程度存在一定的差别,非线性回归模型优于一元线性回归模型,该方法的提出有助于监测青海湖流域草地变化情况。

By referring to domestic and international methods based on remote sensing to monitor vegeta- tion,we focus on how to monitor grassland change using the MODIS - NDVI remote sensing data. Vegeta- tion index of typical grasslands in the Qinghai Lake watershed has been extracted from remote sensing im- ages and compared with grass production data of 79 plots. Multiple regression models are established to study the correlation between MODIS vegetation index and grass production. The results show that these two factors have good correlation, even if there are certain differences in the degree of correlation by dif- ferent regression models. The nonlinear regression model has obvious superiority in comparison with the linear regression model. This method is conductive to monitor grassland change in the Qinghai lake watershed.