Answers are grounded in this paper and cite their sources.
Because they are highly interdependent, most infrastructure systems are vulnerable to disasters such as severe floods. A vast amount of data has been made available with which to better anticipate the impact of severe floods and enable areas prone to flooding to become more sustainable and resilient. However, integrating these data sources and modelling the impact of floods remains challenging. In this study, we develop a transferable, fact-calibrated, weakly supervised, physically constrained probabilistic modelling framework designed to address this issue. We estimate the facility level flood risk during a flood by integrating XGBoost and depth damage functions. Furthermore, we use Granger tests, distributed-lag models, and spatiotemporal Durbin specifications to examine cross system predictive temporal dependencies and finite-horizon lag patterns. The study offers three main findings. First, flood-induced damage exhibits a threshold-driven structure: damage across all infrastructure types is activated within a narrow depth band ( ~ 0.20–0.26 m), suggesting universal hydraulic tipping points for sustainable infrastructure management. Second, at the predictive temporal dependence level, cross system associations are more selective than the underlying physical connections suggest. Rather than being widely distributed, these associations are concentrated around a limited set of potential bridging nodes, providing a basis for targeted resilience interventions in urban contexts. Third, temporal risk dynamics exhibit systematic differences in response latency and finite-horizon risk signal persistence across infrastructure types.
No takes yet. Share an insight, caveat, or question.
Wan et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: