|
Abstract
|
Background
Quantifying physical activity accumulated through daily commuting is challenging due to the scarcity of detailed data, especially for public transport trips. Using Montreal, Canada as a case study, this paper measures and compares an individual's daily amount of walking to and from public transport in their regular commute to work using two datasets and two methods.
Methods
The first method used urban level detailed origin-destination microdata. Distances walked to and from public transport stations were measured using trip details provided from the survey. The second method used open data including commuting flows obtained from census data. Public transport trips for each flow were modeled using a fastest route algorithm applied to General Transit Feed Specification (GTFS) data obtained from public transport operators. Walking distances were then extracted from these trip paths. Multilevel mixed-effect regression modeling was used to identify the determinants of total walking when using both methods. A sensitivity analysis was then used to derive an adjustment table for those who wish to use open data to estimate walking.
Results
For commuter train users 471 m must be added to walking estimates obtained from the commuting flows data, while negative adjustments are required for subway users (122 m), city bus users (366 m), suburban bus users (516 m), and peripheral bus users (1186 m).
Conclusions
The methodology presented in this study provides researchers and professionals in other cities without access to detailed origin-destination survey data with a guide to use open data to accurately estimate total walking distances accumulated in a daily commute by public transport.
Language: en |