GRACE-based high-resolution propagation threshold from meteorological to groundwater drought

Groundwater drought could cause tremendous damage to the social-economy via land subsidence, seawater intrusion and permanent loss of aquifer storage capacity, and often show strong association with meteorological drought. To date, the threshold for meteorological drought triggering groundwater drought and its dominant factors have been not clarified, which inhibits the effective groundwater drought risk management based on preceding meteorological drought information. In this study, we used the Standardized precipitation index (SPI) and the drought severity index of groundwater storage anomalies (GWSA-DSI) to characterize meteorological and groundwater droughts in the Xijiang River Basin (XRB) of China, respectively. A probabilistic framework is proposed to identify the high-resolution propagation thresholds from meteorological to groundwater drought on 0.25° grid. Results show that GWSA-DSI can reliably identify groundwater drought events, and the propagation time from meteorological to groundwater drought ranges from 8 to 42 months. Although the XRB is located in a humid region with abundant precipitation, the probability of groundwater drought occurrence reached 43.8%, 54.8%, 61.2%, and 64.2% under a light, moderate, severe and extreme meteorological drought event, respectively. The propagation threshold triggering light groundwater drought is mainly dominated by moderate and severe meteorological droughts, which showed an increasing trend from central to southeast of XRB. Soil evaporation and watershed elevation are the main influencing factors on the propagation threshold. It is worth noting that anthropogenic overexploitation of groundwater not only destroy the dynamic balance of regional groundwater system, but also interfere with the propagation processes of meteorological to groundwater drought. The results have great implications for more reliably monitoring and predicting the dynamics of groundwater systems under drought stress, and our proposed framework can also be extended to other regions.