WORLDNET: Applying Transport Modelling Techniques to Long Distance Freight Flows

WORLDNET: Applying Transport Modelling Techniques to Long Distance Freight Flows


S Newton, NEA, NL


Application of standard transport modelling techniques to estimate global multimodal freight transport chains. The work is based on the DG-TREN WORLDNET project, to be completed in 2009.


Transport models typically address patterns of vehicle movements within limited geographical territories in order to quantify the interactions between traffic patterns and a realistic representation of the transport infrastructure. In practice this generally results in local, urban, regional, and at the upper extreme, national models. Consequently, if these models address goods movements at all, they only address trip segments, and cannot accurately model realistic origin-to-destination trips.

Freight policy on the other hand cannot be limited simply to managing the ?last 100km? of a given set of deliveries. A broader scope, for example, allows policy makers to consider how different transport modes can be combined, and where to develop mode interchanges and distribution centres, since multimodal transport is typically associated with longer distance and international flows. There is also a need to consider the impacts of externalities outside the national territory. Reducing freight activity within an economy by relocating production to more distant economies masks a shift in the distribution of externalities that a national model typically will not capture.

WORLDNET (DG-TREN) sets out the objective of developing a long distance, multimodal origin-destination matrix, and a network model that will cover Europe, its neighbours, as well as intercontinental routes e.g. maritime and air cargo. These outputs will be developed for DG-TREN?s IPR-free TRANSTOOLS model, allowing greater scope for analysing the impacts of globalisation on transport networks. This is a potentially useful IPR-free resource for modellers wishing to capture long distance freight flows within their own national models.

This paper will focus on the methodology for estimating the long-distance origin-destination matrix. Previous exercises of this nature e.g. ETIS-BASE and NEAC have tended to use methods focusing upon data combination to estimate transport chains. However, as the geographical areas being considered have grown, along with the resulting data sourcing and harmonisation requirements, and as wide area representations of the transport supply side have improved, it has become more attractive to use a simplified ?four step? model as a matrix generator.

A world trade model is used to estimate traffic generation at national level; a regional (NUTS3) distribution model is applied to subdivide the trade flows; a multi-modal assignment procedure is then used to assign to transport chains. The methodology and results will be presented.


Association for European Transport