The accuracy of a transport demand model’s predictions is inherently limited by the quality of the underlying data. This issue has been highlighted by the decline in response rates for transport surveys, which have traditionally served as the primary data source for estimating transport demand models. At the same time, mobile phone network data, not requiring active participation from subjects, have become increasingly available. However, some key trip and traveller characteristics enhancing the prediction power of the estimated models are not collected in mobile phone network data. In this paper we therefore investigate what can be gained from combining mobile phone network data with travel survey data, using the strengths of each data source, to estimate long-distance mode choice models. We propose and estimate a set of mode choice demand models on joint mobile phone network data and travel survey data. We show that combining the two data sources produces more credible estimates than models estimated on each data source separately. The travel survey should preferably include the variables: travel party size, cars per household licence, licence holding, in addition to origin, destination, mode, trip purpose, age, and gender of the respondent.
Stored only when both papers are in the corpus — 7 citations exist upstream.
Imported from OpenAlex (CC0) — nothing here is generated.