Spatiotemporal dynamics of vegetation drought and causal cascade effects of drivers in the Yellow River Basin's water conservation areas

Abstract

Vegetation drought and the cascading effects of its drivers are critical to ecological security and water supply in river water conservation areas. Using the Temperature–Vegetation–Precipitation Drought Index (TVPDI), this study examined vegetation drought in the Yellow River Water Conservation Area (YRWC) during 2001–2022 and investigated its nonlinear causality, cascading pathways, and spatial heterogeneity. Vegetation drought showed a slight alleviating trend (0.003/10a), with clear monthly periodicity and a stepwise spatial pattern. Precipitation (PRE) and the Normalized Difference Vegetation Index (NDVI) increased significantly, whereas Land Surface Temperature (LST) decreased slightly. PRE was the dominant internal factor affecting TVPDI, with a total effect coefficient of 0.92. Among external drivers, Relative Humidity (RH), Temperature (TEM), Available Water Content (AWC), Actual Evapotranspiration (ET), and Elevation showed strong causal relationships with TVPDI. Structural equation modeling indicated positive total effects of RH and ET on TVPDI (0.20 and 0.14, respectively), but a negative effect of TEM (−0.11). Multiscale geographically weighted regression further revealed pronounced spatial heterogeneity in driver effects. These findings support vegetation drought monitoring, ecological restoration planning, and ecological water management in the YRWC.