Improved Treatment of Special Attractors
Corresponding Author: Kevin Stefan, HBA Specto, Inc.
Presented By: Kevin Stefan, HBA Specto
Special attractors such as airports, hospitals, shopping centres and stadia produce distinctive travel behaviour that is not usually well represented using typical trip distribution or destination choice models. Model systems that do handle special attractors tend to do so by making them more attractive; that is, by adding a fixed utility to the zone(s) with special attractors, or by weighting employment more highly. In some cases (particularly airports) supplemental models are used to reflect the added travel.
Using virtual added employment increases the attractiveness of these facilities to match higher observed travel volumes, but it has unrealistic distributional effects. A shopping centre with 2000 retail jobs will attract more shopping trips than the same 2000 retail jobs spread over a wide area – indeed, this is the economic rationale for building shopping centres in the first place. However, with the traditional approach of increased attractiveness, these additional shopping trips will primarily come from the immediate area around the shopping centre. In the real world, the shopping centre instead attracts trips from a larger catchment area than local stores do by offering additional unique stores and services.
To reproduce this behaviour, a novel approach was used in the BNE Generation, Distribution, Time and Mode Choice Personal Travel Model (BNE-GDTM), an advanced trip-based model developed for Brisbane, Queensland, Australia. The distribution portion of the model uses the distance as well as a logsum of time and mode choice model alternatives for travel to a destination. For special attractors including the airport, three major stadia, the hospitals and a number of major shopping destinations, additional terms were introduced in the distance component shaping the curve for a number of the travel segments represented in the model, particularly HBO and NHB trips. For zones identified as special attractors, the effects of travel are lessened, producing more trips to the attractor – but from a greater catchment area. For hospitals and “power” shopping centres, a fuzzy parameter is used identify these features based on zonal employment, identifying future facilities automatically and to weight the adjustment, reducing “cliff” effects.