An integrated multi-model approach for air quality assessment: Development and evaluation of the OSCAR Air Quality Assessment System

This paper reports on the development and evaluation of a new modelling system for studying air quality on local scales. A multi-model approach has been adopted to develop the OSCAR Air Quality Modelling System to conduct assessments at different levels of complexity. This work focuses on the evaluation of the performance of two of the models, CAR II (semi-empirical, annual averaged model) and CARFMI (line-source Gaussian model for hourly predictions), as integrated into the OSCAR System and employing consistent emissions and meteorological inputs. The models have been evaluated using measured NOx, NO2, PM10 and PM2.5 data for 2003 from the Vallila (Helsinki) and Cromwell Road (London) monitoring stations; both of these stations are located in urban roadside environments. The evaluation consists of comparisons of annual averaged and hourly predicted and measured data, as well as detailed statistical analysis of the model performance. Overall, closer agreement between the modelled and measured data is observed for the Vallila site, where traffic levels are medium, but the models show significant underestimation for Cromwell Road which is subject to high traffic levels. The paper also discusses possible reasons for this underestimation. The wider implications of the work are also stated for development and evaluation of environmental models.