MICRO-SCALE SPACE-TIME MODELLING OF EXPOSURES AND HEALTH RISKS IN A MOBILE POPULATION

Abstract

ISEE-209 Introduction: Transport systems represent an important source of exposure and health risk, not only from air pollution but also noise and traffic accidents. In order to inform local transport management and policy, methods are needed to enable these exposures and their consequent health effects to be modelled and assessed. As part of the EU-funded HEARTS (Health Effects and Risks of Transport Systems) study, an integrated GIS-based system has been constructed to model exposures to air pollution, noise and traffic accident risks. Methods: An integrated system for micro-scale modelling of trip behaviour, and exposures to air pollution, noise and traffic accidents has been built in ArcGIS. Trip choices and behaviours (including pedestrian road crossing behaviours) are modelled using a stochastic model, informed by data on time activity patterns, modal preferences and route characteristics. Air pollution and noise are modelled either via external, loose-coupled dispersion models, or purposely-designed internal models, making use of GIS functionality. A specially designed accident risk model has also been incorporated, based on Routledge’s formula, and a user-friendly front-end provides access to the various model components. Models have been validated in Leicester, Lille and Florence, using observational data on time-activity patterns and trip behaviour, measured pollutant concentrations and personal exposure measurements. Results: This paper describes and demonstrates the system, presents results from validation studies in Leicester, UK, and illustrates applications of the modelling system to assess the impacts of different policy scenarios. Validation results show acceptable levels of performance for all the main model components. Results of applying the methodology to assess the potential effects of a local walk-to-school policy suggest conflicting costs and benefits from such interventions that are likely to be masked by more aggregate impact assessments. Discussion: GIS offer powerful tools for modelling of exposures and health risks. On the one hand, if these models are to provide realistic representations of the risks involved, and the consequences of intervention, they need to operate at the micro-scale, and to be capable of simulating complex behaviour patterns. On the other hand, GIS may also provide a dangerous pathway into fantasy!