DUALFloodGNN is a physics-informed GNN that jointly predicts water volume and flow using multi-scale loss terms and curriculum learning for efficient autoregressive flood modeling.
A spatial–temporal graph deep learning model for urban flood nowcasting leveraging heterogeneous community features.Scientific Reports, 13(1):6768, 4
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
DUALFloodGNN: Physics-informed Graph Neural Network for Operational Flood Modeling
DUALFloodGNN is a physics-informed GNN that jointly predicts water volume and flow using multi-scale loss terms and curriculum learning for efficient autoregressive flood modeling.