Stress-testing road network resilience using counterfactual flood events (1953–2024) in Great Britain

• Developed a national-scale, event-based framework to stress-testing road network resilience to flooding. • Integrated process-based approach to capture congestion, rerouting and isolation effects. • Indirect disruption losses exceed direct asset costs significantly for most flood events. • Single carriageway A roads and suburban bridges are critical for network resilience. • Rapid clearance and early speed restriction removal substantially reduce indirect flood impacts. Road infrastructure is facing increasing flooding risks, causing asset damage and disrupting traffic flows. Effective risk management requires integrated assessments that capture network vulnerability, disruption and recovery. While localised studies have simulated traffic disruptions, national-scale assessments have largely focused on flood exposure rather than systemic disruption analysis. We developed a modelling framework combining process-based flow model of passenger travel-to-work flows and applied it to stress-test Great Britain’s road networks against 17 historical flood events from 1953 to 2024. Results reveal significant variations between direct and indirect damage losses, with single carriageway A roads and suburban bridges emerging as critical points. Notably, indirect losses due to disruption and rerouting can be significantly higher than direct damages depending on hazard event. Early clearance and speed restriction removal are key to mitigating the overall indirect impacts. The model is generalisable and can be applied to stress-test other road networks and flood scenarios worldwide.