Heat-driven taxi demand reveals hidden mobility stress in walkable cities

As climate extremes intensify, outdoor exposure in walkable cities becomes costly, increasing the reliance on door-to-door transportation. Using a multi-year dataset of app-based taxi trips from five Japanese metropolitan areas (2022–2024), we quantify behavioral adaptation to heat and rain. Above 27°C, ride-hailing demand increases near-linearly to 38% at 37°C, a pattern observed across all cities. This demand is strongest among infrequent users and for short-distance trips, particularly in high-density areas with poor transit access for first/last-mile connections. Hourly rainfall elicits a nearly identical response pattern, suggesting a general mobility adaptation mechanism. By translating the demand elasticity of short trips by infrequent users into a welfare metric, we map hidden “heat-mobility stress” hotspots around major rail hubs. Our findings show taxis are crucial buffers for urban mobility in extreme weather, with demand varying by user frequency, trip distance, and transit accessibility.