NTEM
The National Trip End Model (NTEM), developed by the Department for Transport, forecasts changes in trip origins and destinations, or productions and attractions, across Great Britain for use in transport modelling (Department for Transport, 2024a).
The forecasts incorporate projected changes in population, employment, housing, car ownership and trip-making behaviour.
| Developed by | Department for Transport |
| Coverage | Great Britain |
| Forecast horizon | Extending to 2051 |
| Current version | NTEM Version 8.0 |
| Data portal | National Trip End Model (NTEM) on data.gov.uk |
The six scenarios
The data portal provides NTEM Version 8.0 datasets for six scenarios:
| Scenario |
|---|
| Core |
| High |
| Low |
| Regional |
| Behaviour Change |
| Technology |
Inputs
NTEM combines demographic, socioeconomic, land-use and travel-behaviour information to generate future trip-end forecasts. Its principal inputs are:
- Zonal population and demographic characteristics
- Employment by sector
- Housing and household projections
- Car ownership
- Area and land-use characteristics
- Trip-production and attraction rates
- Mode and time-period profiles
- Scenario-specific assumptions regarding future population, employment, housing, behavioural change and other socioeconomic conditions
Two of these are themselves modelled rather than supplied:
| Input | How it is derived |
|---|---|
| Car ownership forecasts | Generated through the National Car Ownership Model (NATCOP) |
| Trip-production rates | Derived from National Travel Survey data, differentiated by traveller type, area type and journey purpose |
In practice, you rarely supply these yourself
When existing NTEM forecasts are accessed through TEMPro, the underlying variables above do not normally need to be supplied manually.
Instead, users specify the scenario, geographical area, base and forecast years, trip purpose, travel mode and time, and obtain the corresponding trip-end forecasts and growth factors directly.
How NTEM estimates trip ends
NTEM estimates trip productions and trip attractions separately, then reconciles them.
Trip productions
For trip productions, population is segmented into traveller types and combined with purpose-specific trip-production rates:
Pi,p,s = Ni,s βp,s,r(i)
| Symbol | Meaning |
|---|---|
| Pi,p,s | Trips produced in zone i for purpose p and traveller type s |
| Ni,s | The corresponding population |
| βp,s,r(i) | Trip-production rates, differentiated by traveller type and area type r(i) |
Trip attractions
Trip attractions are primarily related to the amount and type of employment and other land-use activities within each zone. In simplified form:
Ai,p ∝ Σe Ei,e αp,e,r(i)
| Symbol | Meaning |
|---|---|
| Ai,p | Trips attracted to zone i for purpose p |
| Ei,e | Employment in sector e |
| αp,e,r(i) | Purpose- and land-use-specific attraction rates |
Why productions and attractions are estimated separately
The two ends of a trip are driven by different things. How many trips a zone generates depends on who lives there; how many it attracts depends on what is there to travel to.
Modelling them independently means each can use the evidence that actually explains it, which is why the reconciliation step below is needed.
Segmentation, balancing and conversion
Three further steps complete the process:
- Segmentation. The resulting productions and attractions are segmented by travel mode and travel time.
- Balancing. A balancing procedure is applied so that total productions and attractions are consistent within defined balancing areas.
- Conversion. NTEM converts production-attraction trip ends into origin-destination trip ends using time-period and trip-chain conversion parameters, then aggregates the results to the required zone, purpose, mode and time-period classifications.
Production-attraction versus origin-destination
These are two different ways of describing the same trips. A production-attraction view anchors a trip to the traveller's home end; an origin-destination view describes the actual direction of movement in a given time period.
The conversion matters because a morning commute and an evening return are one production-attraction pair but two opposite OD flows.
Outputs
The outputs of NTEM are forecasts of trip productions and attractions by zone, together with production-attraction and origin-destination trip ends.
These forecasts can be disaggregated by:
- Journey purpose
- Travel mode
- Time period
- Traveller or car-availability category
Growth factors
NTEM also provides future-year trip-end growth factors, calculated by comparing forecast trip ends with those in a base year. These can be used to grow or constrain local and regional travel-demand models. For example:
gPi,p = Pi,pfuture ⁄ Pi,pbase
gAi,p = Ai,pfuture ⁄ Ai,pbase
That is, the growth factor for zone i and purpose p is simply the ratio of the forecast trip end to the base-year trip end, computed separately for productions and attractions.
Supporting forecasts
In addition, the NTEM/TEMPro system provides supporting forecasts of:
- Population
- Households
- Employment
- Car ownership
How NTEM is used in practice
In transport appraisal practice, these outputs are commonly used as national background growth assumptions, ensuring that local reference-case travel-demand forecasts remain consistent with DfT projections at an appropriate spatial scale.
A note on the equations above
In the source handbook these formulae are embedded as images. They have been transcribed to text here so that they scale with the page, remain readable in dark mode, and can be selected, searched and read aloud by assistive technology.