Deogratias Eustace, Graduate Teaching/Research Assistant (deu4455@ksu.edu)

E. Dean Landman, Adjunct Professor (eland@cjnetworks.net)

and

Eugene R. Russell, Sr., Professor (geno@ksu.edu)

Kansas State University

Civil Engineering Department

2118 Fiedler Hall

Manhattan, KS 66506

It has long been a concern of management that traffic forecasts are based on assumptions that may not be realized. Almost all transportation planning models have been implemented into computer software packages and thus travel demand modeling has been made simple, quick and more reliable, as long as adequate data is available. Travel demand models calculate demands for travel based on projected socioeconomic, land use and demographic characteristics of the community under study and the available transportation system to determine future traffic forecasts for a horizon year.

Transportation professional community has been quite concerned about the quality of traffic forecasting data which is used as a basis for multi-million dollar investment decisions. Although the widespread use of computerized urban transportation planning packages for forecasting traffic have increased the level of sophistication of the forecasting process, the outputs still leave much to be desired in terms of accuracy and the degree of confidence which can be placed in the results. Thus, decisions based on misleading forecasts, which are often presented to policymakers and to the general public, may lead to a misallocation of funds and to under performing projects during construction and operation.

Poor projections of demographic and socioeconomic data are usually the major source of poor traffic assignment projections and hence ill conceived planning and construction of street and highway infrastructure facilities. This paper attempts to evaluate how accurate long range projections have been by using one transportation study done in 1970s projecting for 20 years in the future. This paper compares the projected travel model inputs with what actually happened after the horizon year has come and gone and also what was projected traffic volumes versus what was the actual ground counts at the same horizon year. The results of this paper show that there is a poor correlation between what was forecasted and what actually happened in terms of socioeconomic and demographic data which are the major inputs used by travel demand models to forecast future traffic volumes on road links. The projections become poorer as the projection period increases. Equally, the projected traffic volumes were way off from the actual traffic counts.