Abstract: Global climate change has markedly increased compound drought frequency and intensity worldwide, severely threatening ecosystem function and sustainable socio-economic development. As a climate-vulnerable region with high drought susceptibility, China requires a deeper understanding of the spatiotemporal evolution of compound droughts, along with their of deviation of compound drought characteristics from their historical state. Based on national‐ scale meteorological observations and remote‐sensing data, methods were developed for identifying compound drought events and analyzing their novelty. Ecosystem resilience was then computed, and the mechanisms by which compound drought influences resilience were examined. Results indicate pronounced regional differences in compound drought across China, the spatial centroid of compound drought frequency showed an overall westward shift. Regarding ecosystem resilience, 42.5% of the area experienced a significant decline, 32.6% showed a significant increase. Partial Least Squares Structural Equation Modeling (PLS-SEM) further showed that compound drought was overall significantly negatively correlated with AR1, whereas compound drought novelty was significantly positively correlated with AR1. These results suggest that regions with long-term exposure to stronger compound drought characteristics may possess higher resilience, whereas when combinations of drought characteristics deviate from their historically experienced range, ecosystem recovery tends to slow and resilience correspondingly declines. The proposed concept of compound drought novelty and its mapped spatial patterns reveal new risk features for ecosystems under an increasingly frequent drought regime. These findings contribute scientific insights for regional drought management and strategies aimed at enhancing ecological resilience.

Key words: Compound drought, Novelty, Ecosystem resilience, SEM, Climate change