Using Linked Non-Home-Based Trips in Virginia
Corresponding Author: Hadi Sadrsadat, RSG
Presented By: Hadi Sadrsadat, RSG
Abstract
When four-step models were originally developed in the 1950’s, the process for home-based (HB) trips was simply re-applied to non-home-based (NHB) trips. Although NHB trips accounted for less than 20% of household trips in those days, by 2009 they accounted for nearly a third of all household trips. Unfortunately, the way NHB trips are represented in traditional models implies that travelers appear and disappear, making trips between locations they never traveled to in the first place, driving alone when they did not take a car from home, etc.
New methods to improve the handling of non-home-based trips in trip-based models were developed and applied to the Salt Lake City metropolitan area’s travel model as part of a study by TMIP and have now been applied and further developed for application in several locations including Tennessee, Alaska, and Virginia. This presentation will review the methodology and results from the successful application of this method to update Harrisonburg and Fredericksburg travel demand models for the Virginia Department of Transportation (VDOT).
The new methods are intended to address the fundamental problems with the accuracy and response properties of non-home-based trips in four-step models. These problems can be understood in a variety of ways, but ultimately are related to the inconsistency of the four-step model with tours. The fundamental approach to addressing these issues in the new technique is to adopt an alternative trip-based model architecture in which non-home-based generation and distribution models are run in series rather than in parallel with home-based component models. This relatively simple structural change appears to significantly improve the trip-based model’s ability to represent non-home-based trips and their response to land use changes and transportation infrastructure investments, making the mode and location of NHB trips consistent with those of HB trips.
This presentation will review both a simpler application of the method focused primarily on auto trips for Harrisonburg as well as a more multimodal application for Fredericksburg. Comparisons will also examine the transferability of parameters between various locations in which the method has now been applied.