Citation

Dehghani Y, Harvey R. Transp. Res. Rec. 1994; 1452: 52-61.

Copyright

(Copyright © 1994, Transportation Research Board, National Research Council, National Academy of Sciences USA, Publisher SAGE Publishing)

DOI

unavailable

PMID

unavailable

Abstract

Traditionally, comprehensive multimodal regional models have been developed to conduct travel forecasts for both highway and transit projects in major metropolitan areas throughout the United States. These models generally have failed to provide accurate, detailed forecasts for existing and proposed facilities, and unrealistic expectations can be placed on these comprehensive (super) regional models. Most of the large regional models in the United States are based on scanty transit ridership information compared with the amount of data available for the predominating automobile users. Transit-component validation of the models usually has been for aggregate market shares and volumes at a few screenlines. Under these circumstances, it is no wonder that comprehensive regional models fail to provide accurate ridership forecasts for specific transit lines. The transit ridership modeling for the Regional Transit Authority (RTA) in Seattle overcomes usual limitations by relying on comprehensive regional models only for regional growth, highway congestion, and regional model coefficients. RTA modeling is structured so that transit ridership results are based on observed origins and destinations of transit users, observed transit line volumes, and a realistic simulation of observed transit service characteristics. External changes, in demographics and in highway costs, are staged into the process in distinct phases before estimating the impacts of incremental changes in transit service. The RTA transit ridership model is simple and fully incremental. The modeling system was validated on the basis of base year comparisons with transit ridership counts, and on a 1992 to 1985 backward "forecast" of transit demand.

Language: en