Responses of NDVI to climate factors in Inner Mongolia using geographically weighted regression

Abstract

Using the MODIS normalized difference vegetation index (NDVI) datasets along with the climate data (precipitation and temperature), geographically weighted regression (GWR) was constructed to explore the spatial nonstationarity and heterogeneity relationships between NDVI and climate factors in Inner Mongolia, China. Our work compared the results of GWR model accuracy with ordinary least squares (OLS) model. The results indicated that GWR method yielded better goodness of fit and higher model accuracy than OLS. Moreover, the GWR model could deeply reveal the complex relationship between NDVI and climate factors. At the same time, the research results could also provide scientific basis for vegetation modeling in Inner Mongolia and similar areas.