Adverse Weather and Urban Transport: Travel Demand, System Per-formance, and Resilience

Weather acts on urban transport at several levels at once: on individual trip decisions, on the speed and reliability of services, on the integrity of infrastructure and on the agencies that run it. Evidence for each level comes from different disciplines and data sources, and these strands seldom cite one another. This paper synthesizes studies of metro, bus, ride-sourcing, bike-sharing, road networks and transit agencies in Asia, Australia, North America and Europe. The studies are placed in a four-level framework (travel demand; operational performance and prediction; network resilience; governance and adaptation) linked to a typology of direct and indirect disruption pathways and their weather definitions, model formula-tions and effect estimates are compared in their native metrics. Precipitation and wind reduce aggregate demand in every study that tests them, yet the sign at a single station or zone can be positive when that unit receives travelers leaving walking, cycling or open-air waiting. Weekday/weekend moderation runs in opposite directions across studies. Bus-stop shelters offset about a quarter of the weekday loss on cold and heavy-rain days and all of it on hot days. Weather inputs change average forecast accuracy very lit-tle but reduce error sharply during rare events. On Shenzhen roads the response phase lasted two to three times as long as recovery, and the usual loss–recovery relation reversed under a super typhoon. Among US transit agencies, hurricanes were the only hazard that raised risk perception. We argue that demand robustness, network performance resilience and life-safety resilience are separate properties that need separate metrics and governance responses.