Shaping street rhythms: Spatial context and the temporal patterns of pedestrian activity

Urban streets provide spatial contexts in which pedestrian activity unfolds over time, yet previous research has seldom examined how these street rhythms differ across space or how they are associated with local street conditions. This study addresses this gap by examining street-level pedestrian rhythms across Seoul’s commercial streets. Using mobile phone signal–based estimates of pedestrian presence for 110,574 street segments in 2022, we extract 11 temporal features from weekday pedestrian time series and apply fuzzy C-means clustering to identify five street rhythm types. Multinomial logistic regression evaluates how these rhythms relate to 16 indicators of commercial hierarchy, physical scale, accessibility, functional mix, and surrounding populations. The findings show that street rhythm types are systematically associated with multiple dimensions of spatial context, rather than reflecting simple differences in pedestrian-presence levels or land-use intensity. Even among commercially active streets, evening-peak rhythms were distinguished by concentrations of retail, leisure, and entertainment-oriented functions. Midday- and afternoon-peak rhythms, despite their similar daytime profiles, were also differentiated by their spatial contexts: midday-peak rhythms were associated with office- and institution-based workday districts, whereas afternoon-peak rhythms were linked to production-related commercial areas. By revealing these distinctions, this underscores the value of treating time as an integral dimension of urban spatial structure.