How to apply the four-step model for 150,000 travel zones: The HIPAT model

• Published in: Transport policy Vol. 155; pp. 150 - 160
• Main Authors: Ihrig, Jan, Jochem, Patrick, Szimba, Eckhard
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01-09-2024
• Subjects: 150,000 travel zones; European transport demand model HIPAT; Four-step model; Runtime; Transport policy assessment; Runtime; 150,000 travel zones; Transport policy assessment; European transport demand model HIPAT; Four-step model
• ISSN: 0967-070X
• DOI: 10.1016/j.tranpol.2024.06.014

Current European transport demand models following the classical four-step approach are less suitable for analysing the traffic volumes on inter-regional road infrastructure with relevant demand of regional trips, as they consider only about 10% of the trip demand with passenger cars. Mainly due to runtime problems, these models are limited to operation at the NUTS-3 level. This level comprises around 1500 regions for Europe with an average diameter of about 50km. However, more than 90% of the trips are shorter than 50km and are, therefore, not adequately considered in the trip distribution model and the trip matrix. Computing this matrix at a higher spatial resolution, such as the LAU-2 level, is not possible as the runtime of the overall model and the size of the trip matrix increase quadratically with the number of travel zones. In the HIPAT model, this obstacle is overcome by an innovative concept operating at different hierarchical levels. This reduces the complexity of the trip matrix, such that the HIPAT model can be applied at large scale for 150,000 travel zones without any runtime problems. This is demonstrated by a prototype implementation for a case study that handles 33,000 travel zones in two minutes. In addition, the limitations of applying NUTS-3-based travel zones for transport modelling are shown: 94.6% of the trips are shorter than the average diameter of the travel zones (50km) and are therefore not correctly represented by the model. Smaller travel zones decrease this number and provide a better basis for analysing infrastructure policies and, ultimately, for decision-making.
•Large-scale application of the four-step model without runtime problems.•One consistent EU transport demand model for long-distance and short distance trips.•HIPAT bases on EU transport policy assessment model HIGH-TOOL and ETISplus data.•Smaller travel zones yield to superior model results.•HIPAT was successfully applied for about 33,000 travel zones in a case study.