How Backcasting is Helping Our Forecasting - a Truly Scottish Experience



How Backcasting is Helping Our Forecasting - a Truly Scottish Experience

Authors

Alison Irvine, Transport Scotland

Description

This paper looks at outcomes of the Transport Model for Scotland and considers issues for future forecasting

Abstract

Transport Scotland (TS) is the national transport agency for Scotland. Land-use And Transport Integration in Scotland (LATIS) is a service offered by Transport Scotland to assist with the appraisal of transport projects and programmes.

One of the tools developed and maintained through the LATIS service is the Transport Model for Scotland (ref. http://www.transportscotland.gov.uk/analysis/latis/models/national).

Forecasts from transport models are used extensively in the planning and decision making process and often underpin much of the analysis undertaken across a range of disciplines. Therefore, not only do they play an important role in making strategic investment decisions involving millions of pound, they are also often a key factor in making detailed decisions about design and impact mitigation. Transport models themselves are, however, reliant on the quality and robustness of their base year input data and assumptions about the future.

While calibration/validation standards exist to ensure suitable representation of base conditions, verifying the robustness of forecasts is a challenge that can only be undertaken after the event and is one that is often overlooked. Following the recent opening of certain major infrastructure schemes in Scotland (including M74 Completion and M80 Stepps to Haggs) Transport Scotland has reviewed the forecasts produced for the Transport Model for Scotland (TMfS07) 2012 Do Minimum Scenario against observed data. This paper:

• Outlines the results from the comparisons
• Explores the story behind both good and bad comparisons
• Considers issues for future forecasting

By establishing how well the forecasts compare with reality and investigating the factors that influence these comparisons a better understanding of how to interpret transport model forecasts is gained. This can be used to highlight the key factors and variables that need to be considered when looking to refine and develop forecasting methodologies that are fit for the needs of Government decision makers.

Publisher

Association for European Transport