The hydrological asymmetric signature of El Niño–Southern Oscillation on global water storage

The El Niño–Southern Oscillation (ENSO) influences hydroclimatic variability worldwide, yet its asymmetric impacts on terrestrial water storage remain poorly quantified. Here, we show global wetting and drying patterns generated by ENSO from a machine-learning reconstruction of water storage anomalies (1984–2023). We identify 81 El Niño–related and 85 La Niña–related clusters covering 105.92 million km² and extending beyond tropical teleconnection zones. La Niña affects 20% more land area than El Niño, and drying occurs across nearly half of all ENSO-affected regions. Areas influenced by both phases often exhibit opposite responses, and La Niña generally dominates the combined signal, generating intense wetting that overcompensates El Niño–driven deficits. This asymmetry is associated with differences in land-surface storage and response times. Wet-prone regions retain hydrological anomalies after precipitation ceases, whereas dry-prone regions respond rapidly and amplify drying. Overall, ENSO–induced water storage variability is strongly asymmetric, with wetting and drying responses that do not compensate locally, reflecting heterogeneous freshwater responses shaped by atmospheric forcing and land-surface memory. La Niña generally exerts a stronger influence than El Niño on terrestrial water storage, producing widespread wet anomalies that exceed El Niño–related dry conditions, based on a machine learning reconstruction of global terrestrial water storage anomalies (1984–2023).