Under a warmer climate, the enhancement of dry-wet abrupt alternation (DWAA) risk poses a great challenge for sustainable development. Here, we introduce a novel framework for DWAA detection based on our proposed soil moisture concentration index. By the end of this century, over humid southern China, the shift of soil moisture time series from anomalously wet to anomalously dry pattern, or the other way around, will be more abrupt. In addition, the proposed framework driven by Coupled Model Intercomparison Project Phase 6 simulations projects more widespread DWAA-affected areas over southwestern China, coastal regions of southeastern China, and the lower reaches of the Yangtze River, especially under a high emission scenario. The framework proposed in this study provides an efficient system for DWAA detection and prediction, and the findings of this study provide a reference for upgrading hydraulic infrastructure and mitigating future DWAA events.

Under a global warming scenario, extreme events, such as heavy rainfall, flooding, and drought, are likely to increase in frequency globally 1-3 . Among these events, abrupt switches between the two extreme states of drought and flood, namely dry-wet abrupt alternation (DWAA), exert far more severe socioeconomic impacts than a single drought or flood event 4-6 . A DWAA event covers both ends of the spectrum including wet-to-dry (WD) state, dry-to-wet (DW) state and the rapid switch between these two extreme states 7,8 . Such events can cause geological disasters 9,10 , ecological stress 11,12 , and severe conflict between domestic water supply and demand 13,14 . However, the lack of adequate methods for quantifying DWAA deeply constrained our insight into the changing DWAA pattern in the coming decades.

Conventionally, DWAA events were identified using a univariate rainfall 15 or runoff 16,17 , or multivariate composite indices, such as the standardized precipitation evapotranspiration index (SPEI) 18 or Palmer drought severity index (PDSI) 19 . Among these, the univariate rainfall or runoff can indicate a rapid switch from dry to wet state for one single end of the DWAA spectrum, whereas it cannot characterize abrupt alternation