What question did this study set out to answer?

The central aim is to create a fast, efficient surrogate model for flood forecasting using graph-neural networks.

April 8, 2026Open Access

Toward an Operational GNN-Based Multimesh Surrogate for Fast Flood Forecasting

Puntos clave

The central aim is to create a fast, efficient surrogate model for flood forecasting using graph-neural networks.
Develop a learning-ready database of synthetic flood events based on Telemac2D model simulations.
Create a graph-neural network that utilizes projected meshes and multimesh connectivity for flood prediction.
Analyze the impact of an explicit discharge feature and pushforward training on model performance.
The developed model predicts floods in approximately 0.4 seconds, significantly faster than the traditional 180 minutes.
Conditioning on discharge feature Q(t) is critical for model accuracy in boundary-driven settings.
Multimesh connectivity enhances prediction accuracy when properly conditioned.

Resumen

Operational flood forecasting still relies on high-fidelity two-dimensional hydraulic solvers, but their runtime can be prohibitive for rapid decision support on large urban floodplains. In parallel, AI-based surrogate models have shown strong potential in several areas of computational physics for accelerating otherwise expensive high-fidelity simulations. We address this issue on the lower Têt River (France), starting from a production-grade Telemac2D model defined on a high-resolution unstructured finite-element mesh with more than 4 10⁵ nodes. From this setup, we build a learning-ready database of synthetic but operationally grounded flood events covering several representative hydrograph families and peak discharges. On top of this database, we develop a graph-neural surrogate based on projected meshes and multimesh connectivity. The projected-mesh strategy keeps training tractable while preserving high-fidelity supervision from the original Telemac simulations, and the multimesh construction enlarges the effective spatial receptive field without increasing network depth. We further study the effect of an explicit discharge feature Q (t) and of pushforward training for long autoregressive rollouts. The experiments show that conditioning on Q (t) is essential in this boundary-driven setting, that multimesh connectivity brings additional gains once the model is properly conditioned, and that pushforward further improves rollout stability. Among the tested configurations, the combination of Q (t), multimesh connectivity, and pushforward provides the best overall results. These gains are observed both on hydraulic variables over the surrogate mesh and on inundation maps interpolated onto a common 25\, m regular grid and compared against the original high-resolution Telemac solution. On the studied case, the learned surrogate produces 6-hour predictions in about 0. 4\, s on a single NVIDIA A100 GPU, compared with about 180\, min on 56 CPU cores for the reference simulation. These results support graph-based surrogates as practical complements to industrial hydraulic solvers for operational flood mapping.

Me gusta

Guardar

Ver artículo completo