Uncovering urban tourist mobility patterns based on large-scale mobile phone data

Amid the rapid global recovery of tourism and the increasing popularity of urban travel, cities are experiencing mounting spatial and functional pressures driven by tourists' unique mobility patterns. Although urban managers and researchers have focused on tourists' impact on urban space and transportation, systematic empirical studies at the individual level remain scarce, particularly in the detailed characterization of spatiotemporal behavior patterns. This study, based on high-frequency mobile phone data, develops an integrated analytical framework that combines tourist identification, activity inference, and multidimensional mobility indicators to systematically uncover the behavioral patterns and underlying mechanisms of tourist mobility in urban space. The results show that tourists, compared to residents, exhibit stronger spatial exploration tendencies and greater behavioral diversity, reflecting distinct non-routine mobility characteristics. Tourist activities are organized around accommodation locations, forming anchor-based activity chains. Tourist visitation is highly concentrated in certain attractions, resulting in spatially uneven “tourism corridors”. In addition, the integration of POI features and machine learning methods enables the inference of tourist activities, further enhancing the understanding of individual-level behavioral heterogeneity. This study not only extends urban mobility theories to tourism contexts but also provides data-driven insights for spatial governance, transportation planning, and service optimization in tourism-oriented cities. • Utilize large-scale mobile phone data to identify and analyze individual tourist mobility patterns. • Quantify key spatial and temporal differences between tourists and residents using multidimensional mobility indicators. • Reconstruct and classify typical tourist activity chains through POI-based semantic inference and machine learning.