Identification of homogenous zones using riding speed time-series to reveal the spatial structure of the urban pace

Traffic speed is essential for the measurement of urban pace. The riding speed metric is one of the most effective ways to identify the underlying dynamic patterns and structures of urban pace. However, few studies have explored the patterns of urban pace reflected by riding speed differentiation of public shared bicycle groups. To this end, this study first infers the riding speeds under unknown path conditions based on the origin–destination of large sample riding records in Suzhou, China. Second, an identification model of homogenous zones for urban pace is established by neighboring station search and time-series similarity-based neighboring station merger. Finally, the fluctuation and deviation indexes of the fitted rhythm curves are adopted to measure the heterogeneity degree of the urban pace in each homogenous zone. Results show that the urban paces based on riding speed have significant spatial differentiation characteristics in the city, and the generated homogeneous zones show a certain stable structure in urban space. The old town has the fastest urban pace and is at a consistently high level, the new district has a faster urban pace and a double-peak pattern, the suburb area has a moderate urban pace and shows a single-peak pattern, while the peripheral area has the slowest urban pace with no significant pace spikes. Meanwhile, the waveform of the urban pace in different zones have a delay of approximately one hour. This study proposes an effective approach to reflect the spatiotemporal homogeneity and heterogeneity of urban pace through the dynamic traffic flow elements, which can be universally applied to other cities and is helpful for planning more dynamic urban functional areas and detecting life stress of urban residents.