Using Transport Models in Spatial Planning: Issues from a Review of the London Land-Use/Transport Interaction LUTI Model
A Wenban-Smith, Urban & Regional Policy, UK
Plugging transport and LUTI models into spatial planning systems. Issues arising illustrated by Peer Review of London LUTI model
Conventional transport models rely on land-use inputs, and are unable to handle the feed-back from transport change to changed locational choices. An increasing literature (eg the so-called ?new economic geography?) suggests that within a few years changed locational choices will be sufficiently significant to risk invalidating the conclusions of such models ? both as regards transport system performance and the economic and social consequences of changed patterns of location . Land Use Transport Interaction (LUTI) models have been developed to tackle this problem, but suffer from the sheer complexity of the systems being modelled, and the lack of data for calibration.
Hence, the re is challenge faced by practitioners and policy-makers about how to interpret and use the results of LUTI modelling. Without conventional calibration, are there other ?realism tests? that can give confidence in the model and its outputs? And, in the light of the credibility of the numerical outputs, what should be the role of the LUTI model in the decision-making process?
Transport for London (TfL) commissioned David Simmonds Consultancy (DSC) to develop a London LUTI model (LonLUTI) to provide the top strategic level of analysis supporting London?s spatial planning work. A ?Peer Group? was appointed by TfL in March 2008 to provide an independent expert reference point on the ability of the evolving model to meet London?s requirements. An immediate concern was whether LonLUTI could be used to support the case for the Thames Gateway Bridge (TGB), a major new river crossing in London, and much of the review was directed to that end. However, in October 2008 the new Mayor (elected in May) decided not to proceed with TGB, and the review therefore re-focused on the wider strategic purposes of LonLUTI as part of TfL?s policy support toolkit.
The proposed paper is not about LUTI models as such (LUTI models have been discussed in previous conferences ), but about the light cast by the review on the uses made of LUTI models in:
1. the support and evaluation of major transport schemes, where key issues were the ability of a LUTI model to detect the effects of single schemes (however large) and the robustness of the interpretation of its results in an adversarial arena; and
2. the formulation of strategic spatial policy where a crucial question is whether the strategic spatial planning process that has grown up around conventional transport modelling can adapt to the broader (but superficially at least, less precise) inputs from LUTI models.
The paper will present examples of the realism tests carried out during the LonLUTI Peer Review process, and the interpretation of the results by the experts. Wider issues are discussed such as the interaction with the wider world and assumptions made in that respect, the interfacing with more traditional transport models, and the meaningful representation and interpretation of vast amounts of outputs. We will conclude with recommendations for the use of transport and LUTI models in spatial planning.
Association for European Transport