Vapor pressure deficit dominates vegetation productivity during compound drought and heatwave events in China's arid and semi-arid regions: Evidence from multiple vegetation parameters

Compound drought and heatwave (CDHWs) events, characterized by low soil moisture (SM) and high vapor pressure deficit (VPD), pose significant threats to the stability of vegetation ecosystems. However, regional-scale assessments on vegetation responses to CDHW events and their driving factors remain limited. This study integrates multiple satellite remote sensing indices and meteorological data to analyze the spatiotemporal dynamics of climate extremes and vegetation responses across China's Arid and Semi-Arid Regions (CASR) from 2001 to 2020. To address the debate over the dominant factors influencing vegetation productivity, we employed the XGBoost model to quantify the independent contributions of factors such as SM and VPD to multiple vegetation parameters. We also examined the interactions between these factors using a percentile-based heatmap approach. The results show that the three major CDHWs during the study period significantly suppressed vegetation growth, as evidenced by pronounced negative SM anomalies and positive anomalies in temperature (TEM) and VPD. We identified a potential critical VPD threshold, below which vegetation productivity declines rapidly. Moreover, the observed negative VPD-SM coupling (e.g., low SM and high VPD) is primarily driven by land-atmosphere feedbacks. Finally, Model results based on XGBoost demonstrate that VPD predominantly drives the decline in vegetation productivity during CDHWs in this region. These findings provide new insights into vegetation responses to compound climate extremes in dryland ecosystems, with implications for forecasting vegetation dynamics and informing adaptive management under future climate change scenarios.