The Ohio Statewide Travel Demand Model is a state-of-the-art economic, land use and transport model. The person transport component of this model is a tour-based, micro-simulation model of the entire population of the State of Ohio. A special feature of the person transport models is the inclusion of a long distance travel model designed to capture the unique behavior associated with individual’s trips longer than 50 miles. This paper will focus on the calibration and application of the long distance travel model.

The Ohio long distance travel model includes six primary components: binary choice of travel, pattern choice, scheduling, internal-external choice, destination choice, and mode choice. In application, each of these choices is simulated for the same synthetic population used by the short distance travel model. Both short and long distance trips are assigned to the same networks, competing for transportation capacity.

Previous work (1) described the design and estimation of the long distance model in further detail. This paper will describe the calibration of each of the long distance model. It compares the model’s performance to observed data, primarily from the 2002-2003 Ohio Long Distance Travel Survey. It includes discussion of changes to model parameters and model structure made as part of the calibration process.

This paper seeks to provide several benefits to future practitioners. First, by reporting the model calibration targets it illustrates the magnitude and characteristics of long distance travel. This information is not as readily available as information about short distance travel, and provides context to those considering which model components to prioritize when investing limited resources. Second, the paper compares the performance of an innovative new approach to modeling long distance travel, allowing future model developers to build upon the work with full knowledge of their starting point.