Published May 20, 2026 | Version v1
Conference paper Open

Is Proximity Enough? Heat-Conditioned Walkability and Accessibility Equity in the 15-Minute City

Description

The 15-minute city paradigm is commonly assessed under an implicit assumption of uniform walking conditions, overlooking how extreme heat may reduce the functional accessibility of daily services. This study develops a heat-conditioned accessibility framework to examine whether nominal proximity overstates meaningful walkability in climate-stressed urban environments. Using Shenzhen, China, as a case study, we integrate multi-source environmental, network, service, socioeconomic, and behavioural data, including ERA5-Land thermal conditions, MODIS land surface temperature, VIIRS nighttime lights, OpenStreetMap pedestrian networks and points of interest, and geotagged Weibo activity records. We compare baseline and heat-penalised accessibility across five categories of urban services, examine behavioural correspondence through high-heat versus reference-day activity patterns, and assess socioeconomic heterogeneity in heat-related accessibility loss. Preliminary results show that, under the medium penalty scenario, 13.44% of populated grid cells in 2023 and 12.65% in 2024 experience measurable service accessibility loss. Behavioural evidence indicates a consistent compression of midday activity and a corresponding increase in evening activity during high-heat conditions. Pooled equity models further reveal a significant interaction between heat exposure and socioeconomic status (β = -0.0175, p = 0.0015), indicating that heat-related accessibility degradation is not uniformly distributed across urban space. These findings suggest that proximity alone is insufficient for evaluating climate-resilient service access and provide a basis for future pedestrian-scale thermal accessibility research.

Files

GeoAI2026_Oral_3264_Guo_Tang_Hua.pdf

Files (539.4 kB)

Name Size Download all
md5:b7c90f1617696dd0136783630b8ae8cc
539.4 kB Preview Download

Additional details

Software

Programming language
Python