Response time and lag time of vegetation drought to meteorological, soil, and groundwater droughts: Non-stationary, linear and nonlinear perspectives

Study region: The Heihe River Basin, China. Study focus: In the study of drought propagation, understanding the temporal characteristics is crucial for drought resistance assessment and drought occurrence prediction. Due to the influence of climate change and human activities, drought index time series may exhibit non-stationarity, and the propagation process tends to be nonlinear, which complicates the temporal characteristic assessment of drought propagation. This study combined the non-stationary drought indices with both linear (correlation coefficient method) and nonlinear methods (convergent cross mapping, CCM) to evaluate the temporal characteristics of meteorological, soil, and groundwater drought propagation to vegetation drought in the study area. New hydrological insights for the region: The response time and lag time of vegetation drought to the three types of drought exhibited spatial heterogeneity from both linear and nonlinear methods, indicating spatial differences in the resistance and sensitivity of vegetation drought to different types of drought. Compared to the linear method, the nonlinear method generally resulted in longer response time and lag time. The nonlinear method generated higher CCM correlation coefficient for soil and groundwater droughts, indicating that the propagation of these two types of droughts to vegetation drought exhibits more pronounced nonlinear characteristics. This highlights the necessity of using nonlinear methods in drought propagation analysis, especially in exploring the propagation between soil/groundwater drought and vegetation drought. It is expected that the results of this study can contribute to vegetation drought monitoring and early warning in the region, and provide support for ecosystem management and agricultural production.