Investigating the meteorological causes of hydrological drought through the integration of spatiotemporal cubes and interpretable machine learning: A case study of the Yangtze River Basin

Yangtze River,China Climate warming and extreme weather events have intensified hydrological drought, especially at long-term and large scales, making the impact of meteorological factors particularly important. This study evaluates the impact of meteorological factors (temperature, precipitation, potential evapotranspiration) on hydrological drought in the Yangtze River basin from 1980 to 2019. The study employs a spatiotemporal cube to analyze the clustering and hotspots of meteorological factors. This study developed optimal interpretable models using XGBoost-SHAP and CNN-SHAP, quantifying the impacts of climate change and human activities on hydrological drought, and analyzing the underlying causes of hydrological drought. The climate characteristics of the Yangtze River basin show significant differences, and various meteorological factors have a notable impact on hydrological drought. Hydrological drought in the upper reaches of the Yangtze River basin is primarily driven by climate change, contributing 85 %, whereas the impact in the downstream has decreased, with a contribution rate of 55 %. The contribution of meteorological factors to hydrological drought varies between the upper and lower reaches of the Yangtze River, with the upper reaches primarily influenced by relative humidity (SHAP value 0.27) and the lower reaches primarily influenced by average temperature (SHAP value 0.25). This study explores the impact of meteorological factors on hydrological drought and provides a theoretical foundation for the prevention and control of hydrological drought in the basin. • Construct a spatiotemporal cube on the average temperature, average rainfall and potential evapotranspiration in the Yangtze River Basin. The spatiotemporal clustering and spatiotemporal hotspot analysis of meteorological elements with large spatial scale and long-term series were carried out, and the spatial and temporal distribution characteristics of meteorological elements in the Yangtze River Basin were analyzed from a deeper level. • The optimal regression model was used to restore natural runoff and quantify the impact of climate change and human activities on hydrological drought. • Based on the SHAP model, the effects of different meteorological factors on hydrological drought in climate change are explained, and the effects of the coupling of different meteorological factors on hydrological drought are analysed.