Drought thresholds that impact vegetation reveal the divergent responses of vegetation growth to drought across China.

Identifying droughts and accurately evaluating drought impacts on vegetation growth are crucial to understanding the terrestrial carbon balance across China. However, few studies have identified the critical drought thresholds that impact China's vegetation growth, leading to large uncertainty in assessing the ecological consequences of droughts. In this study, we utilize gridded surface soil moisture data and satellite-observed normalized difference vegetation index (NDVI) to assess vegetation response to droughts in China during 2001-2018. Based on the nonlinear relationship between changing drought stress and the coincident anomalies of NDVI during the growing season, we derive the spatial patterns of satellite-based drought thresholds (T SM ) that impact vegetation growth in China via a framework for detecting drought thresholds combining the methods of feature extraction, coincidence analysis, and piecewise linear regression. The T SM values represent percentile-based drought threshold levels, with smaller T SM values corresponding to more negative anomalies of soil moisture. On average, T SM is at the 8.7th percentile and detectable in 64.4% of China's vegetated lands, with lower values in North China and Jianghan Plain and higher values in the Inner Mongolia Plateau. Furthermore, T SM for forests is commonly lower than that for grasslands. We also find that agricultural irrigation modifies the drought thresholds for croplands in the Sichuan Basin. For future projections, Earth System Models predict that more regions in China will face an increasing risk for ecological drought, and the Hexi Corridor-Hetao Plain and Shandong Peninsula will become hotspots of ecological drought. This study has important implications for accurately evaluating the impacts of drought on vegetation growth in China and provides a scientific reference for the effective ecomanagement of China's terrestrial ecosystems.