China’s Yangtze and Yellow River Basins exhibit divergent drought patterns, yet the underlying mechanisms driving these differences remain underexplored. This study compares their drought characteristics from 1961 to 2022 using the Standardized Precipitation Index, Standardized Precipitation Evapotranspiration Index, and Palmer Drought Severity Index, and identifies their drivers through attribution models and interpretable machine learning. Our results reveal two distinct paradigms: the Yangtze Basin is characterized by high-frequency, over 14% in all seasons, short-duration droughts, reflecting a rapid hydrological response, while the Yellow River Basin experiences low-frequency, long-duration events indicative of strong soil moisture memory. Quantitative attribution demonstrates that atmospheric evaporative demand (VPD) plays a significantly greater role in the Yellow River Basin, contributing over 20% to soil drought, far exceeding its 14.4% contribution in the Yangtze Basin. Furthermore, their large-scale drivers differ fundamentally: the Yangtze Basin responds primarily to the Atlantic Multidecadal Oscillation (AMO) and Arctic Oscillation (AO), whereas the Yellow River Basin is mainly influenced by solar activity and the El Niño-Southern Oscillation (ENSO). These findings reveal that Yangtze drought is primarily driven by precipitation deficits, while Yellow River drought is a composite phenomenon amplified by evaporative demand. This distinction underscores the need for basin-specific water management strategies.

Rainfall variability strongly influences both flood and drought hazards, especially in climatic transition zones where precipitation is highly seasonal and spatially heterogeneous. This study assessed long-term changes in rainfall patterns and compound flood–drought hazard in the Huaihe River Basin, China, using ERA5-Land-derived daily precipitation series at 174 spatial sampling locations during 1950–2025. Rainfall pattern indicators, flood-related rainfall extremes, and SPI-3-based drought indicators were calculated to characterize rainfall amount, frequency, intensity, dry–wet persistence, heavy rainfall events, and meteorological drought conditions. The Mann–Kendall test and Sen’s slope estimator were used to detect long-term trends, and a compound flood–drought hazard classification framework was developed based on a flood-related rainfall hazard index (FHI) and a drought-related hazard index (DHI). The results showed that annual total precipitation, wet days, and consecutive wet days decreased significantly, indicating reduced rainfall occurrence and wet spell persistence. Flood-related rainfall indicators generally showed decreasing tendencies, with more evident declines in persistent multi-day extremes than in single-day rainfall. In contrast, mean SPI-3 showed a significant drying tendency, although drought frequency, severe drought frequency, and drought intensity did not exhibit significant monotonic trends. Spatially, rainfall pattern, flood-related, and drought-related indicators showed clear heterogeneity across the basin. The compound hazard classification identified flood-dominated and drought-dominated areas as the two major hazard types, each accounting for 31.03% of the spatial sampling locations, while low compound hazard and compound flood–drought hazard areas each accounted for 18.97%. These findings indicate that flood- and drought-related hazards coexist but vary spatially across the Huaihe River Basin. The proposed framework provides preliminary rainfall-based information for differentiated flood–drought hazard assessment, climate-adaptive water resources planning, and the sustainable management of water resources in regions facing spatially heterogeneous hydroclimatic hazards.

The Anhui reach of the Huaihe River Basin features low-lying terrain, dense river networks, and high flood risk spatial heterogeneity, restricting regional socio-ecological sustainability. Traditional flood risk assessments suffer from incomplete indicators, subjective grading, and one-sided outputs when adopting single empirical or unsupervised clustering models. Based on disaster system theory, this study constructed a 14-indicator evaluation system covering hazard susceptibility, environmental sensitivity, and hazard-bearing vulnerability. Subjective-objective weights from Analytic Hierarchy Process (AHP) and entropy weight were fused via game theory equilibrium to build the Game Theory-AHP-Entropy Weight Coupled FCE Model (GAE-FCE Model) for comprehensive flood risk quantification. Meanwhile, principal component analysis (PCA) was utilized for indicator dimensionality reduction, and two unsupervised clustering models, namely, K-Means and Gaussian Mixture Model (GMM), were developed. Confusion matrix, Adjusted Rand Index, and correlation analysis were used to compare four model outputs. Individual models display weak correlation due to distinct risk characterization mechanisms; the ensemble scheme compensates single-model defects, achieving AUC = 0.9161, Recall = 0.9231, and F1 = 0.8889 under optimal weights, and a multi-model coupling framework was further proposed to obtain integrated flood risk zoning. Results show that five dominant factors (Distance to rivers, Topographic Roughness, NDVI, Distance to lakes, and Population Density) contribute over 59.6% of total weight. GAE-FCE reflects inundation characteristics yet lacks natural-factor comprehensiveness; clustering methods capture natural background differentiation but ignore socio-economic vulnerability, whereas ensemble coupling offsets individual model defects. High-risk zones are concentrated along major rivers, and low-risk areas mainly distribute in southern mountainous regions. This multi-model framework provides scientific support for flood prevention and territorial spatial optimization, advancing sustainable disaster risk reduction for the Huaihe River Basin and similar plain basins.