Dynamics and dispersion modelling of nanoparticles from road traffic in the urban atmospheric environment-A review

Reducing exposure to atmospheric nanoparticles in urban areas is important for protecting public health. Developing new or improving the capabilities of existing dispersion models will help to design effective mitigation strategies for nanoparticle rich environments. The aims of this review are to summarise current practices of nanoparticle dispersion modelling at five local scales (i.e. vehicle wake, street canyons, neighbourhood, city and road tunnels), together with highlighting associated challenges, research gaps and priorities. The review begins with a synthesis of available information about the flow and mixing characteristics in urban environments which is followed by a brief discussion on dispersion modelling of nanoparticles. Further sections cover the effects of transformation processes in dispersion modelling of nanoparticles, and a critical discussion on associated structural and parametric uncertainties in modelling. The article concludes with a comprehensive summary of current knowledge and future research required on the topic areas covered. Appropriate treatment of transformation processes (i.e. nucleation, coagulation, deposition and condensation) in existing dispersion models is essential for extending the applicability of gaseous dispersion models to nanoparticles. Some modelling studies that consider the particles down to 1 nm size indicate importance of coagulation and condensation processes on street-scale modelling whereas others neglecting either sub-10 nm particles or Van der Waals forces along with fractal geometry suggest to discard these processes due to negligible effects on particle number and size distributions. Further, it is important to consider those transformation processes e.g. at city scale or in road tunnels because of the much longer residence time or much higher concentration levels compared to the street scale processes. Structural and parametric uncertainties affect the modelled results considerably. In particular, parametric uncertainty in the form of particle number emission factors appears to be the most significant due to considerably large variations in their estimates. A consistent approach to the use of emission factors, appropriate treatment of transformation processes in particle dispersion models and the evaluation of model performance against measured data are essential for producing reliable modelled results. (C) 2011 Elsevier Ltd. All rights reserved.