Unravelling cross-boundary travel flow and its place-based explanatory factors in the Hong Kong metro system

Cross-boundary travel flows highlight urban integration and interdependence between cities with territorial boundaries. Cross-boundary travellers exhibit varying travel purposes, behaviours and demands for facilities and opportunities. Existing research rarely conducts large-scale analysis of these distinct travellers and their travel behaviours. This study develops analytical techniques to examine different types of non-local cross-boundary travellers (NCBTs). Using a large-scale smartcard dataset encompassing millions of metro trips in Hong Kong, we applied data mining to identify NCBTs travelling from Shenzhen to Hong Kong and clustered them based on their dynamic cross-boundary trip patterns using K-means clustering techniques. We then analysed these clusters’ spatiotemporal patterns and assessed how their distribution correlates with place-based characteristics at and near metro stations. The results show that NCBT trips are concentrated in four subareas of the city. A higher ratio of NCBT trips occurs at stations with specific place-based characteristics, such as proximity to boundaries, more commercial facilities and restaurants and lower percentages of low-income and highly educated populations. However, the explanatory power of these characteristics varies between NCBT clusters. Frequent NCBTs are more concentrated in station areas with a higher proportion of the population residing outside Hong Kong, while infrequent NCBTs particularly interact less with low-income populations. The findings offer valuable insights for urban planning and management in cross-boundary metropolitan areas.