Vegetation‐Climate Interactions on the Loess Plateau: A Nonlinear Granger Causality Analysis

Abstract

Abstract In this study, we used a multiyear normalized difference vegetation index data set (1981–2015), and corresponding data sets for observed climatic factors, to investigate vegetation‐climate interactions on the Loess Plateau using an innovative nonlinear Granger causality approach. From eight selected local near‐surface climate factors, we found that the diurnal temperature range, precipitation, and relative humidity were the dominant climate factors influencing the variability of vegetation on the plateau, Granger causing changes in vegetation over 26%, 49%, and 23% of the plateau, respectively. The reverse causality analysis showed that all of the climate factors except precipitation and sunshine hours were sensitive to normalized difference vegetation index on the Loess Plateau. Taken together, these results suggest that there is indeed an interaction between the local climate and changes in vegetation on the Loess Plateau.