East Sussex Countywide Transport Model – Overview
East Sussex Countywide Transport Model (CWTM)
The East Sussex Countywide Transport Model (CWTM) has been developed to provide a consistent basis for the assessment of developments. It helps understand how people currently travel around the county and how this might change with future growth in housing and employment space, particularly as local and major transport schemes are developed and implemented.
The model, which is based on SATURN version 11.5.05H (highway assignment) and EMME (Public Transport assignment and Variable Demand Model), can be used to test development proposals and has other potential applications which are summarised below.
Utilising the Model:
The CWTM can be used to look at the cumulative impact of proposed Local Plan growth. It is also available for individual development proposals to provide information on trip origins / destinations, assess the impact of additional development trips on the transport network and to test the effectiveness of access proposals and mitigation measures.
Accordingly, the CWTM has multiple potential applications including:
- providing evidence for Local Plans
- assessing the impact of improvement schemes / development proposals
- informing business case submission
- informing the suitable phasing of maintenance and utilities work
- optimising the performance of the existing transport network
- informing accessibility planning for key land uses
- allowing for the assessment of new public transport schemes or priorities
Any application of the CWTM for a specific purpose should always first assess the suitability of the model for that task. The main early use of the model is likely to be for assessment of Local Plans within the County and the model is considered to be an appropriate tool for that purpose. For other applications, further local revalidation in specific areas may be required once the locality and requirements of that application is confirmed. The CWTM forms an excellent basis for the development of any further modelling that may be required.
Whilst developers are free to develop their own bespoke models, our recommendation would be that the CWTM is utilised. Any bespoke model would need to be submitted to ESCC for audit prior to results being presented as part of a planning application.
Model Area
The CWTM’s detailed modelling area extends across the whole of East Sussex, with coverage at lower levels of detail outside the County boundary. The model comprises approximately 567 zones, around 381 of which are in the area of detailed modelling covering the county and around 86 zones cover the adjacent areas. The modelled zone system was built up from Census output area boundaries so that the latest data on land use within the study area could be easily utilised. There is a single zone system used for the highway, public transport, and variable demand parts of the model, ensuring accurate transfer of demand and generalised costs between the assignment models, and variable demand model.
Model Details
The model is primarily highway based but does include a ‘light touch’ public transport capability. A Variable Demand Model, using 24-hour Production-Attraction matrices has been developed. Whilst the assignment model allows for different route choices for commuting, business, and other trip purposes, the demand model is segregated further into four home-based and two non-home-based trip purposes.
The modelled highway network includes all but the most minor of residential roads in East Sussex. The highway network was developed by building on existing transport models of the area, but also supplemented with detailed mapping from the Integrated Transport Network data.
Background growth in forecast land use was derived from:
- National Trip End Model (NTEM) v8.0 for car vehicles
- National Road Traffic Projections 2022 (NRTP22) growth factors for the South-East of England for Light Goods Vehicles (LGV) and Heavy Good Vehicles (HGV).
The trip matrices were derived by merging observed travel patterns from Mobile Network Data, with synthetic data calibrated against Census journey to work and National Travel Survey (NTS) Data. Trip matrices were subject to appropriate verification checks, as well as calibration and validation. Finally, the highway assignment was calibrated and validated and demonstrated to replicate observed traffic data to a high level of accuracy. This was achieved without overly distorting the quality of the underlying trip matrices, and the changes brought about by matrix estimation are within acceptable tolerances.
Public Transport
The public transport network was developed using Meridian shapefile data of bus and rail lines. Services on those routes were developed from rail, bus, and coach timetables. The public transport assignment model was developed following best practice to the extent that it is required for the relatively light touch nature of the model. All bus and rail services through the County were included and validation checks confirmed that trip journey times were all representative of actual observed conditions. This satisfies the primary objective of the public transport model, which is to provide suitable generalised costs to feed into the variable demand model.
Time Periods
To allow policy makers to understand both strategic and local issues/impacts and opportunities associated with developments, infrastructure improvements, and policy measures there is a need to provide assessment and forecasting capability of morning peak hour, evening peak hour, and average inter-peak traffic conditions. The highway assignment model therefore represented an average ‘neutral’ weekday (Monday-Thursday) in the following three modelled time periods (thereby covering all key time periods during which significant impacts on the transport network would occur).
- AM peak hour (08:00 to 09:00)
- Inter-peak average hour (between 10:00 and 16:00)
- PM peak hour (17:00 to 18:00)
The public transport assignment model represents an average ‘neutral’ weekday in the following three modelled time periods:
- AM peak period (07:00 to 10:00)
- Inter-peak period (10:00 to 16:00)
- PM peak period (16:00 to 19:00)
Peak periods (in contrast to peak hours) are represented within the public transport model because, over a large study area such as East Sussex, differences in public transport provision between the peak hour and the peak period are likely to be significant. This can result in certain origin-destination pairs appearing to have no public transport services in certain narrow time periods, which will cause problems with the demand model if not corrected. The 3-hour peak periods and 6-hour inter-peak period are designed to alleviate these situations. The selection of the specific hours/periods used to represent the AM, inter-peak and PM peaks was informed by an analysis of traffic counts.
The demand model, meanwhile, represents an average weekday at the 24-hour, production-attraction (PA) level, defined in terms of people, rather than vehicles. A set of factors is applied to convert the 24-hour PA demand in person trips to the public transport peak and inter-peak period passenger matrices and highway peak hour and average inter-peak hour vehicle matrices required for the assignments. This approach is consistent with the guidance set out in TAG Unit M2 and was deemed most appropriate for the demand model due to its strategic nature and the fact that it will be used to test large-scale interventions, major schemes, and development scenarios (individually and in parallel).
Base Year
The CWTM had an original base year of 2019. This was chosen because, at the time model development commenced, it was the most recent year for which the required representative data was available. The Model was developed and was calibrated against 2019 conditions. As with all strategic models, the impact of uncertainty on the model results will need to be carefully considered through a range of sensitivity tests when applying the model.
Implications of Covid-19
The Covid-19 pandemic had a profound impact on travel demand by all modes in 2020 and 2021, and it is not yet clear how it will affect longer term trends. To understand how the impacts of the pandemic should influence existing transport models the DfT published TAG Unit M4 (Forecasting and Uncertainty) in 2023. This stated that COVID-19 impacts should be accounted for in modelling and appraisal from May 2023 onwards.
The change to guidance was issued in the context that national traffic volumes are yet to return to pre-COVID levels. The implication of this is that, in the absence of COVID-19, traffic would have continued to grow such that current flows would have been higher than those seen before COVID-19. Traffic forecasts will therefore assume a level of growth which will not be achieved due to the impacts of COVID-19 and an adjustment will be required to account for this over-estimation.
Update to Base Year
In early 2024, work was commissioned to assess the impacts of COVID-19 on traffic behaviour within the CWTM study area and to consider whether changes to the base year (from 2019 to 2023) would be required. This assessment was undertaken by comparing:
- Motor vehicle traffic between the years of 2022 and 2019 in selected local authorities
- Permanent ATC traffic counts for the years 2023 and 2019 at selected sites withing the East Sussex area.
Both approaches showed a similar trend for overall reduced traffic growth. Based on these observations, the following reduction factors for cars should be applied to the 2019 CWTM base year demand matrices (there would be no change being made to LGV and HGV traffic).
- 5% in AM peak
- 3% in Inter peak
- 0% in PM peak
To understand if changes to a rebased CWTM base year (from 2019 to 2023 based on TEMPro) was needed the following was undertaken.
Step 1
Two 2023 CWTM transport models were created:
- Model 1: 2023 TEMpro v8.0 constrained model produced using information from the Uncertainty Log. Fuel and Income factors were applied in this case to take account of the growth in income and growth in fuel between 2019 and 2023.
- Model 2: Using the factors calculated using Method 2 the existing 2019 CWTM base year demand matrices were factored to produce 2023 CWTM base year demand matrices. The 2023 CWTM base demand matrices would be re-assigned to a 2023 highway model network.
Step 2
Once Step 1 was undertaken the 2023 modelled flows was compared with the 2023
observed flows (obtained from ESCC ATC counts) using the Geoffrey E. Havers
GEH) statistic which is used to compare observed and assigned flow. The statistic
uses the following formula to calculate a value for the difference between observed
survey data) ( ME ) and modelled (MG) (SATURN flow) traffic flow:
Formula [4.3 KB] [jpg]
The GEH statistic takes account of the fact that when traffic flows are low the percentage difference between the observed flow and the modelled flow may be high, but the significance of this difference is small and conversely, a small percentage difference on a large base might be important.
Step 3
If the analysis of the GEH assessment undertaken in Step 2 suggested that Model 1 calibrates and validates better, then there would be no need to “re-base” the 2019 CWTM base year models. The 2040 and 2050 CWTM forecast year models can be produced from the existing CWTM 2019 base year models.
If the analysis of the GEH assessment undertaken in Step 2 suggests that Model 2 calibrates and validates better then there would be a need to “re-base“ the 2019 CWTM base year models. The 2040 and 2050 CWTM forecast year models can be produced from the 2023 CWTM COVID-19 adjusted demand matrices.
The assessment showed that Model 2 calibrated and validated better therefore a 2023 base year model was created.
Step 4
Production of 2040 and 2050 forecast year models.
Access Protocol
Information on how the Model can be accessed can be viewed using the link below: