Development of the boundary permeability index of urban greenspaces and its impact on accessibility: A case study in Edinburgh

Source abstract

Urban greenspaces (UGS) deliver critical benefits, including enhanced mental and physical health, stronger social cohesion, and improved urban climate regulation, thereby supporting human well-being and environmental sustainability, yet their accessibility is often compromised by overlooked physical barriers at boundaries-a pervasive "last-mile" challenge. This study develops a novel Boundary Permeability Index (BPI) to quantify the physical barriers along UGS boundaries and their impact on accessibility assessment, using Edinburgh as a case study. Leveraging Google Street View (GSV) imagery, we classified 608 greenspace boundaries into seven types—open, restricted (three sub-types), and closed (three sub-types)—and applied the CRITIC weighting method, with entropy and subjective methods for robustness testing, to construct the BPI. Results reveal that 75.50% of Edinburgh’s greenspace boundaries are closed, with walls (32.39%) being the most prevalent. After accounting for boundary permeability, 7.3% of the population no longer met the accessible greenspace standard. A Generalized Linear Mixed Model (GLMM) confirmed that BPI significantly predicts accessibility gaps (β = 0.204, p < 0.01), with fully enclosed boundaries adding approximately 60 m of access resistance. The BPI’s contribution remains robust after controlling for greenspace type, spatial location, and socio-economic factors. This study introduces a replicable framework for integrating boundary-level barriers into UGS accessibility assessments, offering critical insights for promoting more accurate and human-centric urban green planning.