Methodology to Add Value to Ageing Travel Survey Data

Over past the several decades, most transportation organizations have gathered a large amount of data through travel surveys. The data they provide is typically used as the basis of their transportation planning and modeling processes. In the province of Quebec, regional household travel surveys take place every 5 to 10 years in the main regions. Conducting large-scale surveys more frequently is not always possible, especially for less populous areas. Relying on old data is, however, becoming increasingly problematic. In this context, it is interesting to explore methods to value historical travel surveys in innovative ways. Based on two travel surveys from the city of Sherbrooke in Quebec, a large-scale regional survey from 2012 and a smaller ad-hoc travel survey from 2019, this paper proposes a methodology for combining the two survey samples. The objective of this process is to benefit from the advantages of both samples: (i) up-to-date travel behaviors from the 2019 survey and (ii) a large and controlled sample size from the 2012 survey. The integration process relies on proportional iterative updating. The descriptive analysis of the two surveys confirmed significant changes in travel behaviors and population had occurred between 2012 and 2019. Despite these differences, the results obtained with the combined samples allow us to reproduce faithfully the trip behaviors and populations of 2019. The proposed method therefore confirms a strong potential to gain better value from historical travel survey data using innovative data combination approaches.