Sensitivity of travel forecast to road pricing policies based on an impedance feedback variable

This paper evaluates a new variation to the GTA travel demand model by testing an endogenous link that implicitly associates modal constraints from the lower level mode choice model which involves nested logit modeling procedures, to modal constraints from the higher level distribution model which involves entropy gravity modeling procedures. The modal constraints developed in previous trip distribution models were ad hoc in nature and were designed based on a priori expectations to incorporate modal level of service attributes for the various zones. The revised model provides these constraints as a logsum term from model choice model and thus does away with the need to manually develop these constraints in the trip distribution phase of the model. The performance and effectiveness of the new endogenous link is investigated in the final stage of the thesis by comparing the full cost pricing policy analyses previous run with the 1991 Transportation Tomorrow Survey (TTS) travel data to those run using the 1996 TTS travel data. The analysis was divided into two categories. The first dealt with the analysis of the revised model in terms of its credibility and effectiveness in describing the trip characteristics and modal composition in the GTA. The part second dealt with a comparison of the original model to the revised model, paying particular interest to the various trip characteristics and price elasticity values. The revised model provided a theoretically sound feedback link between the mode split model and the trip distribution model and demonstrated that logsum impedance functions appear to provide better sensitivity to pricing policy scenarios.