Asymmetric window detection of abrupt global drought-wetness alternations and ecological responses

Climate change-driven abrupt drought-wetness alternations (DWAEs) pose a growing threat to global ecosystem functioning and food security, undermining efforts to achieve carbon neutrality. We present a novel Asymmetric-window Drought-Wetness Abrupt Alternation Index (ADWAI), which integrates asymmetric temporal windows with antecedent dry-wet memory decay to significantly improve the detection of complex DWAEs. Applying ADWAI, we identified 12.0-12.5 million global events between 1982 and 2022, with pronounced Northern Temperate hotspots (30°-60°N, 0.12 events yr^-1) and frequent occurrences in tropical regions (AMZ, SEA, 0.08-0.10 events yr^-1). Wetness-to-drought alternations induced substantial productivity losses (signed cumulative response (SC_resp): -0.5 g m^-2 d^-1 per event), with croplands and grasslands particularly sensitive, whereas forests exhibited lagged responses. The correlation between NDVI-A and GPP-A (r = 0.82, p < 0.01) weakened under wetness-to-drought alternations, reflecting carbon cycle vulnerabilities. Our findings highlight the urgent need to strengthen the resilience of tropical croplands and other vulnerable ecosystems, providing a scientific basis for targeted climate adaptation and sustainable carbon management strategies.