A schedule-based mode-service choice model for the assessment of the competition between air transport and high-speed services

This paper presents a mode-service choice model to assess competition between air transport, high-speed train services and private car at regional/national scale. It has been specified through a schedule-based approach which implies that the whole modeling system (demand, supply, path choice and assignment) has to consider service timetable explicitly. For this reason scheduled services have been represented by single runs instead of traditional service lines, and attributes at single run level for each scheduled transportation service have been taken into account to estimate the mode-service choice probabilities for a given O/D pair and user target time. The mode-service choice model is defined in the framework of the Random Utility Model theory by using a nested-logit structure in which the systematic utility of scheduled services is assumed linear in both attributes and parameters. Among attributes composing systematic utility, a key role is played by the early schedule penalty or late schedule penalty, which represents the disutility that occurs because of the difference between user desired departure time and run scheduled departure time or between user desired arrival time at destination and run scheduled arrival; it can be calculated explicitly considering a schedule-based approach only. The mode-service model has been specified and calibrated on the basis of a set of available data for Italy, and its application to the actual scenario allowed us to test its capability to reproduce the actual Italian market share between regional flights and high-speed trains, to be used as starting point for assessing future scenario forecasts. The first part of the paper describes the schedule-based approach and presents the general structure of the proposed mode-service choice model; the latter part reports results of the application to the Italian case, in which the proposed model has been specified, calibrated and applied