Trip chaining patterns of tourists: a real-world case study

Insights into tourist travel behaviours are crucial for easing traffic congestions and creating a sustainable tourism industry. However, a significant portion of the literature analysed tourist travel behaviour by predefined tourist trip chains which result in the loss of more representative classification. Using tourist travel survey data from Nanjing, China, this paper presents an innovative methodology that combines the tourist trip chain identification and the trip chain discrete choice model to comprehensively analyse the travel behaviour of tourists. The discretized trip chains of tourists are clustered using the ordering points to identify the clustering structure (OPTICS) clustering algorithm to identify typical tourist trip chains, which will then be considered as the dependent variable in the nested logit model to estimate the significant explanatory variables. The clustering results show that there are two main categories, namely single and multiple attraction trip chains, and seven subcategories, which were named according to the characteristics of trip chains. The clustering result is analysed and three main trip chain patterns are derived. Departure city, travel cost, travel time, and travel mode show significant influence on the choice between single and multiple attraction trip chains. The urban attraction trip chain is more favoured by tourists with children, and the typical trip chain shows stronger dependence on travel intention. Visiting Lishui for the first time only affects the choice of the multiple suburban attraction trip chain. These findings are valuable for optimising tourist public transport infrastructure, promoting travel by public transport and better tourism management.