Two-year Worth of Smart Card Transaction Data – Extracting Longitudinal Observations for the Understanding of Travel Behaviour

This paper proposes a sampling strategy to extract longitudinal observations from a large smart card fare validation database. Internal consistency and comparability with the population are evaluated. It is revealed that operational practices, rather than theoretical life span of the card, are the determinant factor of observation duration, sample size, and the presence of spatial and temporal bias. Using mobility and location diversity indicators, the longitudinal observations are analysed individually and aggregately for understanding travel behaviour at the day-to-day, seasonal and year-to-year levels.