Quantifying willingness-to-walk across varying urban visual contexts using a global-local path choice model

Understanding pedestrians' preferences for network and street attributes is essential for improving walkability, but conclusions from previous studies vary. Thus, to comprehensively investigate pedestrians' preferences, this study conducts a comparative analysis of walking route choices across six areas around stations in the Tokyo metropolitan region. The six selected areas represent a range of typical urban contexts. The focus is on two types of attributes: global attributes, which pedestrians can consistently perceive (e.g., route length) in the whole network, and local attributes, which they can only perceive at local surroundings through visual observation (e.g., streetscapes). A global-local path choice model, a state-of-the-art path choice model proposed by Oyama (2024a), is used to incorporate both types of attributes. The estimation shows consistent preferences for global attributes across all areas, with pedestrians generally preferring wide sidewalks and avoiding crosswalks and long paths. However, the responses to local streetscape attributes are consistent in some areas, while varying in others, highlighting the context-specific nature of pedestrian preferences. To ensure fair and consistent comparisons across different areas, willingness-to-walk (WTW) is calculated as an indicator of the additional distance pedestrians are willing to walk in response to a unit improvement in a specific attribute. Given that pedestrians perceive environmental attributes differently, WTW is derived and analyzed separately for global and local attributes. For global attributes, widening sidewalks is identified as the most effective strategy in the Omiya area. Regarding local streetscape characteristics, enhancing the open sky view emerges as the most effective intervention across all areas. Moreover, increasing green vegetation is particularly recommended in residential areas, while it may not be necessary in other types of urban settings. These findings offer valuable insights to inform policy decisions on context-sensitive pedestrian infrastructure and streetscape design strategies. • Pedestrian route choice behavior is analyzed across diverse urban contexts • The study considers both global and local path preferences in the analysis • Street view imagery is utilized to extract visually perceived local attributes • Willingness-to-walk is calculated for global and local attributes for policy use • This study confirms pedestrians' route choice preferences are context-dependent