Skip to main content

Tools for transport forecasting and scenario testing

Résultats exploitables

Transtools has developed the largest and most comprehensive European transport model in terms of countries covered, population covered, and geographical scale. It covers all modes (cars, trucks, trains, canal ships, sea ships and air transport) as well as both freight and passenger transport. The model integrates existing models into one new model that comprehend the complexities of freight and passenger transport flows in the European Union in order to be able to assess large scale policy questions raised by the challenging environment of an enlarged Europe. This new modelling tool considers ETIS as the basis for assessing transport performances. The main modules of the Transtools model are: 1) The Transtools Freight Demand Module consists of the following sub-modules. - The Transtools Trade Module, which uses the ETIS O/D freight transport matrix. Its output is a forecast O/D matrix for freight including origin region, in-between transhipments and destination region, as well as transport mode at origin, in-between transhipments and at destination, commodity group and tonnes. - The Transtools Modal Split Module for freight transport based on the model in NEAC. It adjusts the stable modal split resulting from the Trade Model. Its output is the ETIS freight matrix (a forecast O/D matrix including forecast modal split. - The Transtools Logistics Module. Based on SLAM, which is a module appended to the Scenes model, it evaluates the impacts of changes in the logistic and transport systems within Europe on the spatial patterns of freight transport flows, through changes in the number and location of warehouses for the distribution of goods. 2) The Transtools Passenger Demand Module which models passenger transport at European level focusing on the transport models Scenes, Vaclav, and ASTRA. Its output for the Assignment Module are unimodal passenger O/D transport matrices at NUTS3 level in number of passenger per mode and trip purpose, as well as in number of vehicles. 3) The Transtools Economic Module, which predicts in monetary terms the effects of policy scenarios. The computed relative changes of economy by sector with respect to the baseline scenario are passed on to the NEAC model. Other outputs of the module are policy evaluation measures, in particular real GDP impacts and equivalent variation, by region, year and scenario 4) The Transtools Network Assignment Module, which produces the direct output of the Transtools Model. Transport networks are defined at unimodal level. A stochastic assignment procedure is applied being founded on probit-based models. The probability of using a certain link or route is calculated by a multinomial normal distribution resulting in the probit model Technical implications of results by theme The Commission should be aware of the capabilities of the model and the way in which it can be integrated in the policy analysis. The reliability of the model results depends on the accuracy of the information on the base year and forecasting year, which should be frequently updated. A user group with representatives of the Commission and Member States should be established. The website should play a central role for disseminating the results of the project. It should contribute to the establishment of a modelling community. Furthermore the establishment of a Helpdesk would enhance the use of the Transtools model. Data requirement of the Transtools model are linked to ETIS. Passenger flow data for ETIS have been used for calibrating the model. It is recommended to make a new calibration every five years in line with the ETIS update timing. The Transtools model is based on the implementation in the Traffic Analyst, which has helped in making the relations between the different models explicit. From now on within Transtools new modelling concepts can be easily implemented, as long as the format of the datafiles is followed. The possibility to optimise software codes and data flows in order to reduce running times has to be studied (at present it takes about three days to run the model in full details), as well as the enhancement of its user-friendliness. The model can be improved by extending the network, estimating matrices for intra-zonal traffic, providing links to other models (e.g. energy models, environmental models, etc.), including neighbouring countries (near Europe), improving the modelling of air and sea transport, including containerisation, enhancing links to the national models, including generation and attraction information, making model parameters changeable by the users. It is also important to check matrices and consistency to Eurostat data. Some of the improvements and extensions will be addressed by FP& projects I-TREN, Worldnet, and REFIT.