Guidance

Forthcoming change: modelling guidance updates

Updated 25 April 2024

Description: Updates to the existing modelling guidance to ensure the Department for Transport’s advice reflects latest evidence and best practice, including improving advice on the principles, processes, and the recommended approaches in data collection, modelling, forecasting, and the importance of reporting.

New advice related to alternative methodologies in transport modelling: an introduction to agent-based methods and activity-based demand modelling.

Existing units:

• M1.1 principles of modelling and forecasting
• M1.2 data sources and surveys
• M2.1 variable demand modelling
• M3.1 highway assignment modelling
• M3.2 public transport assignment modelling

New unit:

• M5.4 agent-based methods and activity-based demand modelling

Change announced: April 2024

Expected release date: May 2024

Description

This forthcoming change sets out updates to 5 existing TAG modelling units and one new TAG unit, to reflect the latest evidence and best practice in transport modelling.

In the drafting of technical guidance, views and feedback from stakeholders, consultancy support, and independent peer review, consisting of recognised modelling technical experts have been sought. Extensive engagements were undertaken to identify key themes and the required changes to the guidance.

Detail of changes

TAG unit M1.1

This TAG unit aims to help analysts understand why transport modelling and forecasting are required to produce evidence to support business cases for major transport schemes.

The purpose of the update is to clarify the principles and the recommended approaches in data collection, modelling and forecasting, as well as reporting requirements. The whole document is affected by the update, with new text added to aid further understanding of the concept and principles of modelling and forecasting. In particular, the main updates include the following:

• the scope of the unit (section 1.1) to cover broader purposes of transport modelling and forecasting
• clarifying the key concepts and principles of modelling, forecasting and data collection (sections 2, 3, 4, and 5)
• updating figure 1 (the standard transport model structure)
• adding explanation of multi-modal vs uni-modal approaches
• adding the use of alternative approaches, i.e. agent-based and activity-based demand modelling (section 4)
• adding a section on modelling active travel and new mobility (section 4)
• changing the term ‘mitigating risks’ into ‘quality assurance’ (sections 4 and 5) and expanding guidance to ensure quality assurance is undertaken in modelling and forecasting
• expanding guidance by outlining reporting requirements and potential checks and reviews that should be undertaken (a new section)
• minor edits throughout the document to improve understanding of current advice

Existing and updated documents: TAG unit M1.1.

TAG unit M1.2

This TAG unit identifies sources of transport data to develop transport models and addresses methodologies for gathering data.

The update aims to enhance guidance on data sources and improve clarity on how various data sources should be used. This unit has been expanded and made more coherent through a restructuring of the unit, and implementing the following main updates:

• adding new data sources and their use cases
• updating advice on existing data sources where previously information was limited
• updating the references made to specific data sources that were outdated and/or incorrect
• clarifying advice on the suitability and quality of data (for example, age of data, requirements for neutrality of data, seasonality variation in data, and data requirements for calculation of annualisation factors)
• minor edits throughout the document to improve understanding of current advice

Existing and updated documents: TAG unit M1.2.

TAG unit M2.1

This TAG unit provides advice on the implementation of a variable demand model in transport modelling, which estimates the behaviour of people making different travel choices in response to changes in transport conditions.

The update aims to reflect latest evidence and best practice, as well as improving clarity of the advice around considerations and processes required in modelling transport demand. The main changes are:

• updating the illustrative model parameters (see section 5.6 of the unit) to ensure demand responses in the model reflect observed behaviours:
  • the illustrative parameters for destination choice are unchanged, however, with the introduction of cost damping that is widely used in current models, the destination choice parameters for cost-damped models may exceed the median values by more than the indicative range of 25% during local calibration and realism testing – see table 5.1 in TAG unit M2.1 (2020)
  • the illustrative parameters for main mode choice are updated to include destination choice sensitivity parameters from PRISM and MoTiON, as shown in the table below

Table 1 illustrative main mode choice theta scaling parameters

Trip purpose and mode	Minimum	Median	Maximum	Sample
Home-based work	0.51	0.70	0.83	8
Home-based employers’ business	0.19	0.46	0.78	4
Home-based other	0.27	0.47	1.00	5
Non-home-based employers’ business	0.73	0.73	0.73	2
Non-home-based other	0.43	0.62	1.00	3

• demand elasticity values in realism testing (see section 6.4 and appendix A in the unit) to ensure the behaviour of the model reflects empirical evidence of demand changes where there are changes in travel times or costs. Acceptable ranges for elasticities in realism testing have been derived from the meta-models of cost and time elasticities reported in Wardman (2022a, 2022b)^{[footnote 1]}. The summary of recommended elasticity ranges is shown in the table below

Table 2 summary of recommended elasticity ranges

	High	Low
Average fuel cost (kms)	−0.35	−0.15
PT main mode fare (trips)	−0.9	−0.2
Bus fare (trips)	−0.9	−0.35
Rail fare (trips)	−1.5	−0.7
Light rapid transit fare (trips)	Use bus
London Underground fare (trips)	−0.6	−0.2
Car journey time (trips)	Not stronger than −0.75

• clarifying existing guidance on:
  • the assessment of the need for variable demand modelling (VDM)
  • the matrix form to use in VDM (O/D vs P/A format)
  • the appropriate % gap formula for cost averaging model
• minor edits throughout the document to improve understanding of current advice

Existing and updated documents: TAG unit M2.1.

TAG unit M3.1

This TAG unit provides guidance on developing highway assignment models.

The update adds new advice on the use of microsimulation modelling, which takes the form of a new appendix within the unit (see appendix C in the updated TAG unit M3.1). The use of microsimulation models can be a useful addition to the scheme assessment and appraisal process, as it can provide operational analysis of highway networks.

The new appendix provides an explanation on the application of the highway assignment modelling guidance to the development of microsimulation models. It is not intended to be a detailed step-by-step guidance note on how to develop microsimulation models.

The advice includes:

• a brief introduction to microsimulation modelling, explaining their common application and how the inputs/outputs differ from strategic models
• references to existing guidance and best practice for microsimulation modelling
• the potential role of microsimulation in scheme appraisal – differentiating between relying solely on microsimulation model results and using them in combination with a more traditional approach

Existing and updated documents: TAG unit M3.1

TAG unit M3.2

This unit provides detailed advice on developing public transport assignment models.

The update ensures that modelling public transport follows best practice, which includes a restructuring of the unit, aligning it with the structure in TAG unit M3.1 (2020). This unit has been significantly expanded with advice on the following aspects added or improved:

• the role of public transport assignment in the overall standard model structure
• methodologies required during model development, including modelled area, zoning system, network representation, time periods, user classes, assignment methods, and generalised cost definition
• model calibration in relation to the quality and in-vehicle parameters and network/service coding and checking principles
• model validation and its accuracy in relation to the level of aggregation and different type of model
• model convergence
• the requirements for reporting

Existing and updated document: TAG unit M3.2.

TAG unit M5.4

This new unit introduces agent-based methods and activity-based demand modelling. These alternative methods emerge from evolving requirements in transport modelling, as there is desire to understand the broader impacts of transport interventions or policies that require increasingly complex analysis.

The purpose of this unit is to illustrate where these approaches could be considered and how they may be used to form part of an evidence base to answer policy questions.

This unit explains the background to and operational elements of agent-based methods and activity-based demand models, as well as examples of their applications. The unit sets out principles of validating these models, however, it does not provide detailed guidance on how to develop these models.

This unit can be used when practitioners are considering using these approaches for strategic planning, exploring policy options, and gathering evidence for strategic cases at the very early stage of business cases. 

As there is currently limited experience in the UK, the unit and guidance in it rely heavily on overseas experience. The unit does not aim to provide guidance on their use in the economic cases of transport appraisal, or approval for that purpose. Further research on these areas of modelling approaches is required and the content of this unit will evolve accordingly.

The new document: TAG unit M5.4.

Contact

For further information on this guidance update, please contact:

Transport Appraisal and Strategic Modelling (TASM) division
Department for Transport
Zone 2/25 Great Minster House
33 Horseferry Road
London
SW1P 4DR

Email: tasm@dft.gov.uk

1. Wardman, M. (2022a) Meta-Analysis of Price Elasticities of Travel Demand in Great Britain: Update and Extension. Transportation Research A, 158, pp. 1-18 and Wardman, M. (2022b) Meta-Analysis of British Time-Related Demand Elasticity Evidence: An Update. Transportation Research A, 157, pp. 198-214. ↩