Mode choice modeling for long-distance travel

Abstract

Investments for improving long-distance travel infrastructure are substantial and may have significant impacts on travel demand, the environment, and the economy, and therefore, deserve careful analysis before implementation. A new nested multinomial logit mode choice model is presented that is sensitive to travel costs, distance, transit station accessibility, service frequency, number of transfers, and parking costs. On the auto side, the model considers the modes drive-alone and shared-ride with 2–4 or more passengers. The transit side models regional bus, rail, and air as modal options. Special attention was given to transit station choice, as the nearest station may not provide the best connectivity for a given trip. Parameters used in this model were carefully compared with those published in the literature to ensure reasonability of model results. To explore the model sensitivities, scenarios with increased gasoline prices and improved bus service are presented.

Keywords:

• Travel demand modeling
• Long-distance travel
• Discrete choice
• Multinomial logit model
• Mode choice

Acknowledgement

The development of the long-distance mode choice model has been funded in part by the North Carolina Department of Transportation. Bill Davidson has provided valuable feedback on model design and implementation. Special thanks go to Joe Kolousek and Carlee Clymer for collecting and analyzing travel data on rail, bus, and air long-distance travel.

Notes

1 Available for download at http://nhts.ornl.gov/download.shtml.

2 http://www.transtats.bts.gov/databases.asp?Subject_ID = 3&Subject_Desc = Passenger%20Travel&Mode_ID2 = 0

3 Available at http://matrix.itasoftware.com.