Speaker
Description
The rapid advancements in digital technologies and big data analytics have opened new avenues in urban research, enabling the exploration of urban phenomena with unprecedented spatial and temporal precision. In cities like Seoul, mobile phone signal-based data has become a powerful tool for capturing real-time pedestrian dynamics and urban mobility patterns. This type of big data makes it possible to analyze urban activity patterns and street-level pedestrian flows in ways that were previously unattainable. The pedestrian data used in this study is derived from mobile phone signal-based "active population" data, offering a strong basis for analyzing temporal rhythms in pedestrian volume.
This study examines how the physical characteristics of urban roads in Seoul shape the temporal rhythms of pedestrian activity. Unlike prior studies, which primarily focused on static, quantitative correlations between the built environment and pedestrian volume, this study emphasizes "rhythms"—temporal patterns of pedestrian fluctuations across daily and seasonal cycles. For instance, differences in pedestrian peaks and troughs between weekdays and weekends or across specific seasons may stem not only from physical attributes but also from contextual or segment-specific factors. Hence, this research explores how physical environmental factors shape pedestrian rhythms. It also examines whether variations within homogeneous environmental clusters (roads sharing similar physical characteristics) are driven by environmental attributes or latent, segment-specific traits.
The study utilizes a comprehensive dataset of 2022 pedestrian volume records in Seoul, which includes time-of-day and day-of-week across 12 months. It also incorporates detailed road environment attributes, such as sidewalk presence, road width, surrounding building height and land use, POI density and diversity, transit accessibility, and spatial configuration metrics (integration and connectivity within an 800m radius). The methodology follows a three-step approach: (1) clustering roads based on physical attributes and assessing intra-cluster rhythm consistency, (2) decomposing pedestrian rhythm variations into environmental (fixed) and segment-specific (random) effects, and (3) conducting a hierarchical analysis to measure the relative influence of different spatial and temporal factors.
This study leverages mobile phone signal-based active population data to deeply analyze the relationship between temporal pedestrian rhythms and the physical environment of urban roads, offering novel insights for data-driven urban research and planning. The findings reveal that while pedestrian rhythms strongly correlate with physical road attributes, significant variations within clusters sharing similar environmental characteristics suggest the influence of contextual factors and segment-specific latent traits. Hierarchical analysis reveals the relative influence of cluster-level physical factors, segment-specific traits, and temporal contexts (e.g., monthly and hourly patterns), emphasizing the key role of temporal dynamics in pedestrian volume changes. This research underscores the necessity of integrating temporal patterns into pedestrian analysis as a vital component of urban planning and design. By employing data-driven methods, it opens pathways to uncover hidden factors that contribute to pedestrian dynamics. These insights provide practical foundations for formulating effective policies and designs to foster sustainable and walkable urban environments.
| Keywords | Pedestrian Temporal Rhythms; Street Environments; Clustering Analysis; Mobile Phone Data |
|---|---|
| Best Congress Paper Award | No |
