DELUGE: Towards Continental-Scale Daily Pluvial Flood Damage Prediction via Interpretable Conditioning on Foundation Model Embeddings

2026-07-17

AI Summary

Existing approaches struggle to deliver nationwide, rainfall-driven flood loss predictions at daily temporal and kilometer-scale spatial resolutions due to coarse data granularity, regional limitations, or prohibitive computational costs. This work proposes DELUGE, a multimodal deep learning framework that integrates the three core components of risk—hazard, exposure, and vulnerability—to predict daily flood losses at approximately 1-kilometer resolution across the contiguous United States, focusing on the top 100 high-claim regions. DELUGE introduces novel value and temporal modulators, incorporates terrain features, and leverages embeddings from the AlphaEarth foundation model, endowing the architecture with intrinsic interpretability and establishing a generalizable embedding integration paradigm for geospatial tasks. In spatial block holdout evaluations, DELUGE significantly outperforms tuned Random Forest, XGBoost, and LightGBM baselines by 9%–30% in dollar-weighted PR-AUC, markedly enhancing predictive capability for rare, high-loss events.

Abstract

Pluvial (rainfall-driven) flooding accounts for 45% of National Flood Insurance Program (NFIP) claims in the United States and is harder to predict than its riverine and coastal counterparts, with existing approaches limited to coarse resolution, regional domains, or computationally intensive process-based models unsuitable for daily continental-scale use. We present DELUGE, a multimodal deep learning framework for daily pluvial flood damage prediction at ~1 km resolution and national scale, trained on spatially and temporally corrected NFIP claims (2017-2022) and structured around the hazard, exposure, and vulnerability components of disaster risk. Rather than blanket coverage of the Conterminous United States (CONUS), we model the top 100 highest-claim 75 km cells, distributed nationwide and accounting for ~81% of total pluvial flood claims. Our architectural novelty is a pair of parametric modules in the hydrometeorology branch, a Value Modulator and a Temporal Modulator, conditioned on terrain descriptors and AlphaEarth foundation-model embeddings, that expose directly inspectable hydrological response parameters and provide architecture-level interpretability-by-design. Under a spatial block holdout, DELUGE outperforms tuned Random Forest, XGBoost, and LightGBM baselines by 9% to 30% on a dollar-weighted area under the precision-recall curve (PR-AUC), a metric that emphasizes the rare, high-cost claims of greatest operational interest. Beyond DELUGE, we argue this interpretable conditioning scheme is a transferable pattern for integrating foundation-model embeddings into other geospatial prediction tasks.

Problem

• pluvial flood
• damage prediction
• continental-scale
• daily forecasting
• foundation model embeddings

Innovation

• foundation model embeddings
• interpretable deep learning
• pluvial flood prediction
• multimodal architecture
• hydrometeorological modeling