Abstract: We present a computational model for simulating the dispersion of traffic emitted particulate matter inside a road tunnel, with an emphasis on the number concentration of ultrafine particles (UFP). The model primarily calculates the size distribution of the particle number concentration at each location inside the tunnel. The proposed model differs from existing models in the sense that it uses a continuous representation of the size distribution based upon the high-order finite element method and that it solves the governing equations using the state-of-the-art discontinuous Galerkin method. Next to the traditional transport processes, the model also implements the most important aerosol transformation processes such as coagulation, condensation and dry deposition. It is shown that based upon parametrisations found in literature, the process of condensation in a traffic tunnel cannot properly be modelled. Therefore, we present a correction factor that allows for a better parametrisation. The adequate performance of the model is demonstrated by both a verification study and a validation study. For the verification we show that the discretisation error converges consistently while for the validation we compare the modelled results with a suitable set of data from a UFP measurement campaign in a Taiwanese traffic tunnel. The model is shown to correctly simulate the observed behaviour and by applying a statistical model evaluation we demonstrate that the proposed model meets widely accepted air quality model acceptance criteria. (C) 2012 Elsevier Ltd. All rights reserved.

Abstract: Environmental magnetism, and the magnetic leaf signal in particular, is amply investigated and applied as proxy for atmospheric particulate matter pollution. In this study, we investigated the magnetic signal of annual segments of tree branches, and the composition of particles deposited hereon. Branches are, contrary to leaves of deciduous trees, available during leaf-off seasons and exposed to air pollution year-round. We examined the intra- and inter-tree variation in saturation isothermal remanent magnetization (SIRM) of branch internodes of London plane (Platanus x aeerifolia Willd.) trees in an urban environment. The branch SIRM, normalized by surface area, ranged from 18 to 650 x 10(-6) angstrom; the median amounted to 106 x 10(-6) angstrom. Most of the branch magnetic signal was attributed to the epidermis or bark, and the presence of metal-containing particles on the branch surfaces was confirmed by SEM-EDX. The location of the trees and the height, the depth in the crown and the age of the branches significantly influenced the branch SIRM. The median branch SIRM was up to 135% higher near a busy ring road than in quiet environments (city park and quiet street canyon), and was linked to the presence of Fe-rich particles with co-occurrence of trace metals such as Cr, Cu, Zn and Mn on the branch surface. Within the tree crowns, the branch SIRM generally decreased with increasing height, and was 22% higher in the interior than at the periphery of the crowns. Within the branches, the SIRM increased with each year of exposure, but did not relate to year-to-year variation in particle concentrations due to branch surface changes (epidermis shedding). Our results provide indications that branches can be a valuable alternative for biomagnetic monitoring of particulate pollution, but intra-tree variability in branch SIRM can be substantial due to the branch's location in the tree and branch age.