This study introduces a spatial methodology that models accessibility loss under extreme flood conditions, integrating a 100-year return period scenario with network-based spatial analysis. Applied to Chennai, the approach overlays flood impact on road networks, land use, and infrastructure to assess where functional accessibility collapses. A clustering analysis further identifies priority zones for safe shelter allocation and evacuation planning.

Results show that accessibility degradation is spatially uneven. Compact urban forms retain greater connectivity under stress, while dispersed or fragmented areas become structurally isolated. Key corridors and intervention areas are mapped both at local and metropolitan scales, offering actionable entry points for adaptive planning.

The method reframes resilience not as physical resistance to flooding, but as the preservation of spatial connection under disruption. It is transferable to diverse urban contexts and leverages open-access spatial and environmental data, enabling scalable diagnostics without relying on intensive modelling or proprietary inputs.

By shifting focus from exposure to accessibility, the study offers a novel framework for anticipating the lived impacts of climate hazards—and designing cities that can maintain life-saving accessibility in times of crisis.