Performance evaluation of a multiscale modelling system applied to particulate matter dispersion in a real traffic hot spot in Madrid (Spain)

Urban air pollution is one of the most important environmental problems nowadays. Understanding urban pollution is rather challenging due to different factors that produce a strongly heterogeneous pollutant distribution within streets. Observed concentrations depend on processes occurring at a wide range of spatial and temporal scales, complex wind flow and turbulence patterns induced by urban obstacles and irregular traffic emissions. The main objective of this paper is to model particulate matter dispersion at microscale while considering the effects of mesoscale processes. Computational Fluid Dynamic (CFD) PM10 simulations were performed taking into account high spatial resolution traffic emissions from a microscale traffic model and inlet vertical profiles of meteorological variables from Weather Research and Forecasting (WRF) model. This modelling system is evaluated by using meteorological and PM10 concentration data from intensive experimental campaigns carried out on 25th February and 6th July, 2015 in a real urban traffic hot-spot in Madrid. The effect of uncertainties in the inlet profiles from mesoscale input data on microscale results is assessed. Additionally, the importance of the sensible surface heat fluxes (SHF) provided by WRF and the selection of an appropriate turbulent Schmidt number in the dispersion equation are investigated. The main conclusion is that the modelling system accurately reproduces PM10 dispersion imposing appropriate inputs (meteorological variables and SHF) and a suitable turbulent Schmidt number. Better agreement is found for simulation with a low turbulent Schmidt number. This approach improves the standard microscale modelling alone because more realistic boundary conditions and mesoscale processes are considered.