A graph-based geospatial framework for delineating neighborhood planning units with barrier-aware isochrones

Graph-based GIS workflow is a relevant tool to urban planning and governance, yet most operational accessibility tools still rely on flat or Euclidean representations which can overestimate local dynamics in cities shaped by complex topography and fragmented urban forms. This limitation is especially critical at the neighborhood scale, where short walking trip structure everyday access to opportunities and services. This paper presents an original algorithm to delineate neighborhood scale planning units through the integration of slope-sensitive, territorial barriers, and urban morphology into isochrone modeling. The method structures a barrier-aware pedestrian graph on the block fabric, assigns edge costs from topography, and produces dual time–distance isochrones through an iterative seeding procedure to identify shortest walkable paths. A subsequent classification stage generates a typology of accessibility polygons based on area, population, radii, elevation, and administrative overlap. Applied to the municipality of São Paulo, Brazil, the framework generated 639 neighborhood scale isochrones and demonstrates that a 10-min or 800-m walking threshold does not correspond to a uniform spatial scale once topography and territorial discontinuities are explicitly incorporated. By analytically associating isochrones with transitional areas, the model reconceptualizes neighborhoods as flexible spatial perimeters rather than fixed administrative units, allowing their limits to be refined through public governance or participatory planning processes. In doing so, the study shifts isochrones from isolated accessibility outputs to a citywide system of planning-oriented units and offers a transferable geospatial computational framework for capturing local urban dynamics and supporting evidence-based active mobility and neighborhood planning in complex urban environments.