Assessing bikeability across scales: a context-specific framework for Xiamen Island, China

Understanding how urban form shapes active-travel choices is central to advancing sustainable transport. Drawing on the transport-environment nexus, this study develops and applies a context-specific, multi-scale framework to evaluate bikeability on Xiamen Island, China. We construct indicator sets at two scales: the road-level, assessing infrastructure quality, and the community-level, evaluating the built environment's propensity to foster cycling. Road-scale weights are derived from a four-level Likert questionnaire of local residents, while community-scale weights are estimated from a locally calibrated cycling-propensity model, ensuring sensitivity to the city's mobility culture. The normalised, weighted overlays yield island-wide bikeability surfaces and reveal marked intra-urban disparities: compact, mixed-use central neighbourhoods outperform peripheral urban-village areas; newly built complete-street corridors exceed legacy streets constrained by obstructions, poor greening, mixed traffic, waterlogging and recurrent congestion. The results demonstrate how differences in urban form and (built) environment manifest in cycling conditions and, by extension, cycling behaviour. The findings inform targeted policy recommendations, including the prioritized retrofitting of substandard roads in these high-potential areas. This study offers a replicable methodological framework for scholars and practitioners to develop locally attuned bikeability indices that guide effective investment in human-scale urban development. Understanding how urban form shapes active travel is essential for sustainable transport planning. This study develops a context-specific, multi-scale framework to assess bikeability on Xiamen Island, China. Two indicator sets are constructed: a road-level set measuring infrastructure quality, and a community-level set capturing how the built environment supports cycling. Road-level weights are derived from a four-level Likert survey of local residents, while community-level weights come from a locally calibrated cycling-propensity model, reflecting the city’s mobility culture. By combining normalized and weighted indicators, the study produces island-wide bikeability surfaces and identifies clear intra-urban disparities. Compact, mixed-use central neighborhoods perform better than peripheral urban-village areas, while newly built complete-street corridors outperform older streets affected by obstructions, weak greening, mixed traffic, waterlogging, and congestion. The findings show how urban form influences cycling conditions and behaviour, and provide policy guidance for prioritizing road retrofits and supporting locally grounded, human-scale urban development.