In this study, we compared the in-site ground-based climate variables (1982-2016) with a multiyear normalized difference vegetation index (NDVI) dataset to characterize climate change and vegetation-climate interactions at Ergune, Inner Mongolia, China, using the time series analysis, the correlation analysis and the principal component analysis (PCA). To reveal the time lag effects in climate-vegetation relationship, vector auto regression (VAR) model was constructed and the impulse-response analysis, the causality analysis were conducted. We found that the regional climate change over the past decades could be summarized as climate warming and drying. And the regional climate warming was mostly contributed by summer warming, rather than the widely reported winter warming in the north hemisphere. Climate variables were highly correlated. The PCA analysis revealed that the 1st principal component represented the temperature related variables, and the 2nd principal component represented the humidity related variables. At seasonal scale, however, the humidity and temperature was the 1st principal component for summer and winter respectively. VAR analysis revealed that, the precipitation has higher impact on NDVI than the temperature. The feedback of NDVI to humidity was significant, but feedback of NDVI to Temperature was non-significant. VAR model had better performance in prediction of NDVI than the multi-linear regression approach. This study investigated the climate-vegetation relationship with full considerations on the co-linearity and the time lag effect in climate system. The cause-effect in climate-vegetation relationship indicated the feedback of NDVI to climate variability and the ecological function of vegetation in mitigating and regulating the regional climate.
You et al. (Mon,) studied this question.