This paper describes the development and implementation of a traffic forecasting procedure, which overcomes many of the problems arising from evaluating proposals in complex urban areas with broad land-use and transport strategies and policies. Several issues are critical to the integrity of traffic modelling for scheme appraisal in large urban areas, but conventional methods for addressing them have been increasingly criticised. Analysts are faced with the challenge of constructing a detailed very large model, attempting to include all influences, but the use of an appropriate hierarchy of separate models may be better. The paper compares the advantages and disadvantages of several alternative disadvantages of several alternative techniques, and argues that hierarchical modelling, through a cordon isolation mechanism, has several advantages over other alternatives. This approach is illustrated by a case study, where a two-tier system was developed to appraise a road improvement scheme; it used the London Transport Studies (LTS) model as the upper-level strategic model, and a CONTRAM model for the local area scheme appraisal. The paper finally discusses its findings, and suggests how to handle the fixed travel demand question.
Abstract