1. Home
  2. TRID
  3. View Record
TRB Annual Meeting

MODELLING CONCEPTS FOR FORECASTING LONG DISTANCE TRAVEL IN THE NETHERLANDS

The Netherlands Ministry of Transport commissioned Martin & Voorhees Associates in 1980 to undertake a study of long distance travel. The first phase of this project was completed in summer 1981. The first phase consisted of a qualitative investigation of factors affecting long distance travel, and the development of a model structure. This paper describes the proposed model, which will be estimated in phase 2, drawing on data due to be collected in 1982. In view of the modelling problems encountered in previous studies, a recommendation of the initial qualitative study was that the modelling approach should concentrate on forecasting changes rather than attempting to synthesise existing travel patterns; this approach was termed the 'marginal model'. The approach involves a market segmentation system for each purpose. Analyses suggest that family life cycle and vehicle ownership will be among the key dimensions of this segmentation system. The model is divided into two major elements, the demand model and the forecasting model. The demand model would normally be used by itself in the appraisal of short run effects, preserving the full detail of the market segmentation system. For long term forecasts, it would be used in conjunction with a forecasting model which operates at a coarser level of segmentation, more compatible with the socio-demographic forecasts generally available. The model consists of three stages: - trip frequency, in which the effect of transport system changes on the frequency of visit to a particular destination is estimated. - distribution model. The distribution model is essentially very simple being used to modify an existing (observed) matrix of trips in order to reflect changes in the structure and distribution of population and employment. In the case of some recreational and social travel the model also allows for any redistribution effects attributable to changes in the transport system. - mode choice. The mode choice model will allow for the estimation of changes in mode choice due to both socio-demographic change and transport system change. There are two parts to the model the first of which is 'non-compensatory', while the second is 'compensatory'. Forecasts of future travel demand are closely related to forecasts of population and employment yet experience has shown such forecasts to be generally unreliable, especially at the level of disaggregation required for use in travel demand models. Furthermore, forecasting behaviour is itself inevitably speculative. Allowance for these inherent uncertainties will be made using risk analysis techniques, building on procedures developed by Martin and Voorhees Associates for the appraisal of major inter-urban highway schemes on behalf of the UK Department of Transport. (Author/TRRL)

Media Info

  • Features: Figures; References; Tables;
  • Pagination: p. 189-201

Subject/Index Terms

Filing Info

  • Accession Number: 00381632
  • Record Type: Publication
  • Source Agency: Transport Research Laboratory
  • ISBN: 0-86050-113-2
  • Report/Paper Numbers: REPORT
  • Files: ITRD, TRIS
  • Created Date: Apr 30 1984 12:00AM