Revealing multi-phase resilience patterns of inter-city mobility under extreme weather

Extreme weather events pose growing threats to regional resilience and transportation network stability. To characterize spatiotemporal process and driving mechanisms of inter-city mobility under extreme weather, we investigate the Yangtze River Delta urban agglomeration during Typhoon Bebinca and identify a novel multi-phase “water-drop” resilience curve considering pre- and post- disaster fluctuations. Based on temporal and spatial heterogeneity identification, machine-learning regression is applied to determine stage-specific mobility drivers. Results reveal that: 1) mobility intensity mainly fluctuated in an M−shaped pattern throughout the research period with preventive evacuation peaks, impact-induced decline, and post-disaster recovery; 2) network redundancy exhibited opposite trends to centrality, with modular clustering reorganized after fragmentation; and 3) mobility changes were primarily driven by information awareness, disaster intensity, and mobility needs across different stages. The comprehensive analysis advances our understanding of the mobility response cycle, providing actionable insights for regional climate-resilient transportation strategies enhancing system sustainability.