THE URBAN TRAVEL DEMAND FORECASTING PROJECT. PHASE I FINAL REPORT SERIES, VOLUME XI. FORECASTING TRAVEL DEMAND IN SMALL AREAS USING DISAGGREGATE BEHAVIORAL MODELS: A CASE STUDY

The patronage of a new transit system proposed for a suburban city in the San Francisco Bay area is predicted using disaggregate behavioral models of transportation choice. Features of the study include: (1) detailed transportation alternatives (the models that were used estimated the probabilities of choices among seven different travel modes); (2) calculation of time and cost data needed as inputs to the forecasting models with hand measurements of walk distances for each person in the sample, and descriptions of trips both inside and outside the local area; (3) a method based on iterative proportional fitting used to correct for unrepresentative sampling of the population of potential bus users; and (4) analyses conducted to test and compare the accuracy of the probabilities estimated by different behavioral models, using data available before the bus system was running. Results suggested that the behavioral models used can be feasibly applied to travel demand forecasting in small urban areas, but that additional development and testing of the models should be done before they are used as a basis for policy decisions.