Abstract Short-duration extreme rainfall can disrupt underground metro systems and trigger rapid service degradation. Flooding often begins at a small subset of stations, but station closures and passenger rerouting can shift demand toward transfer hubs and accelerate cascading overload. Existing studies mainly emphasize physical vulnerability, while the timing of operational intervention during evolving disruptions is less explored. This study investigates performance-triggered intervention timing in an urban metro network under extreme rainfall, using the 2021 Zhengzhou event as a reference scenario. We develop a demand-weighted network model and simulate a staged disaster-chain process in which flood-induced primary failures, overload-driven secondary failures, and governance switching evolve jointly. Structural resilience is measured by the connectivity of the largest connected component, and functional resilience is measured by demand-weighted service capacity. The results show that structural resilience and functional resilience follow different trajectories during disruption. Over the main threshold range, the intervention threshold exhibits a non-monotonic effect on cumulative resilience loss, with a low-loss region at intermediate thresholds. Under the present simulation setting, a representative intermediate-threshold strategy (\: \: \: 0. 55) increases mean functional resilience from 0. 357 to 0. 379 and reduces cumulative resilience loss from 64. 29 to 62. 09 relative to the reactive benchmark. Compared with a fixed-time intervention baseline (\: t=12), the same strategy also yields lower cumulative resilience loss (62. 09 vs. 63. 71). These findings suggest that intervention timing should be treated as a case-specific design parameter rather than a universal operating rule. The proposed framework provides a tractable basis for scenario-based assessment of performance-informed metro governance under extreme rainfall.
Li et al. (Sun,) studied this question.