Fast and accurate flood forecasting using geometry-aware neural models
A Geometry-Informed Neural Operator surrogate of LISFLOOD-FP for rapid pluvial-fluvial flood forecasting: the case of Forlì, Italy
Machine Learning
Summary
Flood forecasting using detailed physical models is often too slow for real-time warnings. The authors developed a fast neural network model that learns from simulations and predicts flood depths for a town in Italy. It uses everyday river and rain gauge data as input and produces accurate forecasts much quicker than traditional methods. This speed enables quicker decisions and the use of many forecast scenarios to better understand flood risks.
What this means in practice
- •For flood risk management teams: Generate rapid and reliable flood forecasts using limited river and rainfall measurements to support operational early warning and decision making.
- •For emergency response planners: Run large ensembles of flood predictions in minutes for better probabilistic flood scenario planning and resource allocation during flood events.
Authors
Samuele Maria Garofalo, Fabio Limardo, Simonetta Bodojra, Mohammad Amin Nabian, Mehdi Taghizadeh
Abstract
Physically based two-dimensional hydraulic models are too slow for real-time inundation forecasting and large ensembles. We present a time-dependent geometry-informed neural operator (GINO) surrogate of LISFLOOD-FP for the municipality of Forlì (Italy), severely affected by the May 2023 Emilia-Romagna floods. The surrogate is designed for an operational setting in which only river stage gauges and rain gauges are available. A reproducible pipeline turns public regional observations into a physically based training set: each river inflow is driven by the first stage gauge downstream of it through a calibrated water-surface boundary condition, and rain enters as SCS-CN effective rainfall. The surrogate is an autoregressive GINO trained on 216 simulations of 24 observed events with scaled forcing. It predicts volume-conserving polygon-mean depths on an exposure-based Voronoi mesh of 659 nodes, with a 20 min step over 72 h. The surrogate is tested on 38 independent observed events that it never saw during training. On these events it reaches a critical success index (CSI) of 0.934 at 0.05m and 0.844 at 0.3 m, a depth RMSE of 0.008m and an $R^2$ of 0.984 for maximum-depth maps. On one NVIDIA A10G GPU a 72-h forecast takes 1.7 s instead of 55 s with LISFLOOD-FP, and a 50-member ensemble, computed in a single batched rollout, takes 6.2 s instead of about 46 min, which makes probabilistic flood forecasting practical. These results show that neural-operator surrogates can deliver accurate pluvial-fluvial flood forecasts driven by river stage for operational early warning.