FIDLAR: Forecast-Informed Deep Learning Architecture for Flood Mitigation

TL;DR

FIDLAr tackles coastal flood mitigation by replacing slow physics-based optimization with a forecast-informed deep learning framework that couples a Flood Manager to generate pre-release schedules with a Flood Evaluator that forecasts their hydrological impact. The Evaluator is pre-trained and frozen, enabling gradient-based planning to train the Manager, yielding rapid, optimized releases that reduce floods while avoiding water wastage. A Graph Transformer Network underpins both components, leveraging spatiotemporal dynamics and covariate interactions; experiments on South Florida data show orders-of-magnitude speedups and superior mitigation performance over rule-based and GA-based baselines. The work demonstrates real-time applicability, interpretability through attention maps, and a practical path toward data-driven, disaster-resilient flood management.

Abstract

In coastal river systems, frequent floods, often occurring during major storms or king tides, pose a severe threat to lives and property. However, these floods can be mitigated or even prevented by strategically releasing water before extreme weather events with hydraulic structures such as dams, gates, pumps, and reservoirs. A standard approach used by local water management agencies is the "rule-based" method, which specifies predetermined pre-releases of water based on historical and time-tested human experience, but which tends to result in excess or inadequate water release. The model predictive control (MPC), a physics-based model for prediction, is an alternative approach, albeit involving computationally intensive calculations. In this paper, we propose a Forecast Informed Deep Learning Architecture, FIDLAR, to achieve rapid and optimal flood management with precise water pre-releases. FIDLAR seamlessly integrates two neural network modules: one called the Flood Manager, which is responsible for generating water pre-release schedules, and another called the Flood Evaluator, which assesses these generated schedules. The Evaluator module is pre-trained separately, and its gradient-based feedback is used to train the Manager model, ensuring optimal water pre-releases. We have conducted experiments using FIDLAR with data from a flood-prone coastal area in South Florida, particularly susceptible to frequent storms. Results show that FIDLAR is several orders of magnitude faster than currently used physics-based approaches while outperforming baseline methods with improved water pre-release schedules.

Figures (10)

• Figure 1: Forecast-Informed Deep Learning Architecture ( FIDLAr). Input data consists of five categories of variables as shown in the left table. The variables $w$ and $k$ are the lengths of the past and prediction windows, respectively. The parts colored green are provided as inputs, while the orange and blue parts (with question marks) are outputs. The Flood Manager and Flood Evaluator represent deep learning (DL) models, the former to predict control schedules of controllable hydraulic structures (e.g., gates and pumps) to pre-release water, and the latter to predict the resulting water levels for those control schedules. Loss functions, $\mathcal{L}_{1}$ and $\mathcal{L}_{2}$, penalize the flooding and water wastage beyond pre-specified thresholds, respectively.
• Figure 2: Flood Evaluator. The parts shaded green are used as inputs. Water levels (blue) are the outputs.
• Figure 3: Flood Manager. The parts shaded green (historical data) are the inputs, and the parts shaded orange are the outputs. The water levels shaded blue are not predicted.
• Figure 4: The two red bars represent a threshold of flooding and a threshold of water wastage. Shown are (a) the time spans when these thresholds are crossed, and (b) the areas between water level curves and threshold bars. Violations of the upper and lower thresholds are captured in $\mathcal{L}_1$ and $\mathcal{L}_2$.
• Figure 5: Schemetic diagram of study domain. There are three water stations with hydraulic structures (S26, S25B, S25A), one simple water station, S1 (green circle in the middle), and a station monitoring the tide level from the ocean.