Modelling Uncertainty of Scenarios in National Demand Forecasts
Dobson, David Simmonds Consultancy, Fotheringham, David Simmonds Consultancy
An exercise to better understand the impact of uncertainty within a national land use and transport model.
Transport Scotland currently develop and maintain the national transport model, Transport Model for Scotland (TMfS12) and the national land use model (TELMoS12). Together they form a Land-Use/Transport Interaction (LUTI) model with information from one model informing the other as they forecast change over time. The land use model’s forecast of population, households and employment are inputs to the transport model’s calculation of trip patterns, whilst the calculations of accessibility based upon the transport model’s generalised costs influence the distribution of households and employment within the land use model.
Underlying the LUTI model’s forecasts are assumptions on the overall level of change in Scotland’s population and economy.
This paper reports upon an exercise, carried out by the authors’ firms on behalf of transport Scotland, to explore the impacts of changes to the underlying assumptions in order to see how they are likely to affect both the overall levels of travel and the distribution of travel across the country.
These variations to the assumptions include:
- higher and lower levels of net migration to Scotland and the impact of more (or less) population;
- higher and lower levels of economic growth within Scotland;
- higher and lower levels of fuel price; and
- a behavioural change to more working from home.
The change in fuel prices was simulated by an adjustment within the Transport Model, the other three changes are simulated by adjustment within the Land Use Model. Because of the inter-dependency of the models, the changes to one will impact upon the forecasts of the other.
This paper will describe the adjustments made and then report upon and discuss the outputs firstly in terms of their impact upon the scale and geographical distribution of population, households and employment and secondly in terms of their impact upon the strategic transport networks.
The change in employment and population is not uniform across the country. Some areas see greater or lesser change than elsewhere. The impact upon the strategic transport networks will be affected accordingly.
This exercise feeds into two areas of policy debate. Firstly it can help to inform the wider policy debate upon traffic forecasts and their sensitivity to variation in some of the underlying assumptions. Secondly by identifying areas that are most susceptible to changes in levels of economic growth and other trends, it can point to where policy initiatives may need to focus in order to develop more resilient communities.
Association for European Transport