Multimodal route choice models of public transport passengers in the Greater Copenhagen Area

Understanding route choice behavior is crucial to explain travelers’ preferences and to predict traffic flows under different scenarios. A growing body of literature has concentrated on public transport users without, however, concentrating on multimodal public transport networks because of their inherent complexity and challenges. In particular, choice set generation and modeling route choice behavior while accounting for similarity across alternatives and heterogeneity across travelers are non-trivial challenges. This paper tackles these challenges by focusing on the revealed preferences of 5,641 public transport users in the Greater Copenhagen Area. A two-stage approach consisting of choice set generation and route choice model estimation allowed uncovering the preferences of the users of this multimodal large-scale public transport network. The results illustrate the rates of substitution not only of the in-vehicle times for different public transport modes, but also of the other time components (e.g., access, walking, waiting, transfer) composing the door-to-door experience of using a multimodal public transport network, differentiating by trip length and purpose, and accounting for heterogeneity across travelers.