Spatiotemporal dynamics of blue-green water in a humid basin: Covariation with ENSO and underlying drivers

Study region The Xiangjiang River Basin, a typical humid basin. Study focus Blue-green water responses to climate anomalies and their underlying driving pathways remain insufficiently understood in humid basins. In this study, a jointly calibrated hydrological model and Bayesian ensemble algorithm were used to investigate spatiotemporal dynamics of blue-green water in the Xiangjiang River Basin from 1981 to 2020. On this basis, we assessed the covariation of blue-green water with ENSO, a key driver of climate anomalies. Partial Least Squares Structural Equation Modeling (PLS-SEM) was applied to further quantify the effects and pathways driving blue-green water variations. New hydrological insights for the region Abrupt changes in the seasonal and trend components of green water storage (GWS) and blue water (BW) largely coincided with the occurrence periods of typical ENSO events, whereas no credible changepoint was detected for green water flow (GWF). GWS and BW showed stronger positive responses to ENSO, with correlation coefficients of 0.52 and 0.60 (both p < 0.05), respectively, whereas the negative response of GWF to ENSO was weaker, with a correlation coefficient of −0.41 (p ≥ 0.05). Among the direct effects, precipitation dominated GWS and BW variations, whereas temperature primarily drove variation in GWF, with both exerting positive effects (the path coefficients > 0.6). Temperature indirectly affected blue-green water via vegetation dynamics and dry-wet conditions. These findings provide useful evidence for water resources management under anomalous climate conditions.