Flood planning needs realistic views of how floods develop, but flood size and timing depend on where rain falls and how wet the ground is beforehand. We developed a method using past rainfall and flood data to capture both effects. With the same rainfall, wetter starting conditions produced floods that were 54.4–100.0 % larger at their peak, reached the peak 19–27 hours earlier, and carried 26.6–59.9 % more water. This approach can improve reservoir safety checks and flood risk planning.

1. PINAS used a physically constrained searchable routing structure to more flexibly represent basin-scale differences and flood-wave propagation characteristics. 2. The size of the training sample had a significant impact on the performance of PINAS. However, the improvement was not entirely a monotonic trend. 3. SHAP interpretability analysis revealed significant differences in the main sources of information among different basins.

The North China Plain, home to over 300 million people, faces groundwater decline from irrigation, creating 80-meter-deep zones threatening supplies. We modeled rain and river infiltration using borehole and weather records to study restoration. Rain recharge averages 446 days, varying nearly 140 times: 10 days in sandy western foothills to 1395 days in clayey central/eastern plains. River recharge is faster at 91 days, suggesting flood basins could accelerate recovery for sustainable farming.

This study presents a model integrating attention mechanisms and physical constraints to improve flood prediction. It forecasts floods up to 6 h in advance. The model enhances accuracy by focusing on critical input features and historical patterns. Results demonstrate its superior performance compared to other models, offering improved flood prediction with greater interpretability and alignment with physical laws.

Flash drought (FD) in the North China Plain (NCP) from 1981 to 2022 was identified by three methods. It occurs more frequently in central and northern NCP than southern, with longer duration in southern and eastern NCP. FD shows seasonal variations. NCP is prone to severe FDs. There is a decreasing trend in FD, with three hotspots in northwestern, eastern and southwestern NCP. The FD frequency is affected by thresholds in the identification methods. This study may be valuable for FD mitigation.