Post-Project Evaluation of the Demand Forecast for a New Transit System

Abstract

Methods of demand forecast for a new transit system, Tokadai Peachliner, are evaluated. The ridership originally predicted by the authority using 4 step method is 31,000 passengers/day in 1991, which resulted in the overprediction of 15 times. We found that the authority overestimated 1.7 times larger at the 'trip generation' stage and 7 times at the 'modal split' stage. We also found that the disaggregate modeling approach could better reach the observed value when the competing JR line and socio-economic variables are well considered.