TRB Pubsindex
Text Size:

Title:

Small Urban Area Travel Demand Modeling in Oregon

Accession Number:

01043954

Record Type:

Component

Availability:

Transportation Research Board Business Office

500 Fifth Street, NW
Washington, DC 20001 United States

Abstract:

The Oregon Transportation Planning Rule (TPR) requires that cities having a population of 2,500 or greater develop transportation system plans (TSPs) that link land use and transportation planning. While the TPR does not specifically require the use of travel demand forecasting models for the development of TSPs, it does establish principles that, in many cases, are best satisfied in this manner. Model development for many small cities is unrealistic, however, due to the significant cost involved. To address this issue, the Oregon Department of Transportation undertook the Oregon Small Urban Model (OSUM) development project. Although never attempted elsewhere, the fundamental assumption of the project was that a single, prototypical small urban area model could be estimated from joint travel behavior survey data and then calibrated to reflect local conditions within specific cities. The obvious advantage of this approach was that travel survey data collection/preparation and model estimation, the two costliest components of model development, would only need to be done once rather than for each city. The travel survey data came from approximately 3,200 two-day household activity surveys administered over eight rural counties throughout Oregon. The first calibration of OSUM was performed for the Coos Bay-North Bend area, which is comprised of two neighboring cities having a combined population of roughly 32,000. The primary model adjustments made during the calibration process were related to the estimation of external tripmaking (I-E, E-I, and E-E trips) and the representation of special generators. In small urban areas, external trips are a significant factor, comprising a much larger percentage of total trips than in large metropolitan areas. As an improvement over a simplified methodology that was initially used, a variation of the Huff probability model was developed to accurately portray these trips. Special generators were also found to have strong influence on travel patterns within small urban areas. Therefore, three special generators consisting of a shopping mall, a regional community college, and a regional hospital were explicitly represented within the model, with specific inputs and procedures developed for these. The model calibration/validation was very successful, with good validation results that were well within acceptable accuracy limits. In fact, the model’s performance was significantly better than that of most other small urban area models developed in Oregon over the past several years. Of equal importance is the fact that a functional model was developed that required much less time and cost than if the conventional process had been followed. Soon after its development, the model was applied without difficulties in the successful completion of the Coos Bay-North Bend TSP. Since then, models for six other cities have been completed with similar results, further demonstrating the effectiveness of joint model estimation as an alternative to the standard model development approach.

Monograph Accession #:

01043941

Language:

English

Corporate Authors:

Transportation Research Board

500 Fifth Street, NW
Washington, DC 20001 United States

Authors:

Schulte, Bob
Ayash, Sam

Pagination:

12p

Publication Date:

2004

Conference:

9th National Conference on Transportation Planning for Small and Medium-Sized Communities

Location: Colorado Springs Colorado, United States
Date: 2004-9-22 to 2004-9-24
Sponsors: Transportation Research Board; Federal Highway Administration

Media Type:

CD-ROM

Features:

Figures (2) ; Tables (4)

Geographic Terms:

Subject Areas:

Education and Training; Highways; Planning and Forecasting; I72: Traffic and Transport Planning

Files:

TRIS, TRB

Created Date:

Mar 9 2007 1:41PM

More Records from this Conference: