A station-level ridership model for the metro network in Montreal, Quebec

The development of a successful transit system must include a demand ridership analysis during the planning stage. The construction of a rail system is expensive and sufficient demand is needed for the project to be feasible. A classical and relatively simple method to predict potential ridership is to build a linear regression model that makes the linkage between the built environment, socioeconomics, transit service and ridership in existing stations. This paper develops trip production and attraction models at the station-level for the metro network in Montreal, Quebec, by exploring numerous factors that influence transit use. The results demonstrate that transit-related attributes have a significant impact on ridership. Population density, average income, bus service connectivity, distance to central station, and service frequency are linked to trip production ridership during morning peak hours. Factors such as commercial and governmental land uses, bus connectivity, and transfer stations are associated with station attraction ridership during morning peak hours. This indicates that strategies to promote densification, land use diversity as well bus transit connectivity seem to be effective means of maximizing metro station ridership.