Ruoxi Wang, Nan Li*, Yan Wang
Abstract: The increasing accessibility to digital traces of human whereabouts in cities has offered numerous new opportunities for exploring patterns of human mobility in urban spaces. Prior research pointed out that there exist two distinct subpopulations in cities, namely returners and explorers, whose mobility patterns differ in the extent to which their characteristic traveled distance is impacted by their recurrent mobility. However, the potential dependence of the returners and explorers dichotomy on the observation duration has been largely ignored in prior research, which may cause biased understanding of the returners and explorers dichotomy in urban mobility patterns. By analyzing the daily trajectory data of 21,240 individuals in Guangzhou over 111 weekdays, this study evidenced that the returners and explorers dichotomy is significantly dependent on the duration within which individuals’ trajectories are observed. This study further revealed that such dependence could be interpreted by three underlying explanations, which are respectively related to information accumulation, individuals’ spatial exploration behaviors and changes in individuals’ important locations. The findings provide fundamental knowledge for studying urban human mobility patterns for disease prediction, population behavioral modeling, and understanding dynamic human-environment interactions at urban scales.
Keywords: Human mobility; Returner and explorer; Observation duration; GPS trajectory; Travel behavior; Urban analytics
Figure 1 Spatial distribution of geolocations in the preprocessed dataset.
Figure 2 The trajectories of a 2-returner and a 2-explorer.
Figure 3 Returner-explorer classification results under different observation durations.
Figure 4 Average number of total visited locations of individuals in the dataset.
Link：https://www.sciencedirect.com/science/article/abs/pii/S2210670721001529
