Effects of Vegetation on Traffic Pollutant Dispersion and Air Quality on the Urban Neighborhood Scale

Traffic pollutant concentrations at the pedestrian level in a generic urban neighborhood consisting of an array of 5 x 5 square building blocks of 90 m length and 30 m height with 18 m broad street canyons were investigated (Fig. 1). Scenarios with different arrangements of avenue-trees (trees in every street canyon, trees in every other street canyon) and with a park consisting of trees replacing the central building block were studied and compared to the tree-free situation. The avenue-trees were arranged in a single row along the street center lines with medium dense crowns (leaf area density LAD = 1 m²m^-3) of 6 m width ranging from 6 to 18 m above the street level. The park was formed by a closed tree canopy of 90 m edge length with otherwise the same crown characteristics. Traffic pollutants were released homogenously from the entire street surface within the neighborhood. The investigations were done with Computational Fluid Dynamics (CFD) by employing a Reynolds Stress Model (RSM) for turbulence closure to which extra terms were added to account for the effects of the trees on air flow.

In comparison to the tree-free situation (Fig. 1a), local increases as well as decreases in traffic pollutant concentrations at the pedestrian level at 2 m above ground were found within the street network. For the scenarios with avenue-trees in every and in every other street canyon (Fig. 1b and Fig. 1c), a faster build-up of concentrations occurred in the wind-parallel streets at the upwind side of the neighborhood. Increases were also present in the second wind-perpendicular canyon row with the largest increases around the intersections. However, lower pollutant concentrations were found for large parts of the downwind side of the neighborhood in the wind-parallel street canyons. This might be in particular beneficial in terms of complying with air quality standards and threshold values since in this region the highest traffic pollutant concentrations in the tree-free situation were observed (Fig. 1a). For the park scenario (Fig, 1d), a general reduction of concentrations occurred in the immediately adjacent and downwind wind-parallel street canyons but local increases in concentrations were found in some wind-perpendicular canyons.

Fig. 1 Pollutant concentrations c [-] in the street network for the tree-free situation (a) and concentration differences Δc [-] with avenue-trees (b,c) and central park (d) relative to the tree-free situation at 2 m above ground. The vegetation setup is indicated in the upper right part of each panel. Notice that because of symmetry conditions only the upper part is shown.

The avenue-trees resulted in slightly higher average traffic pollutant concentrations within the street network at the pedestrian level. Compared to the tree-free situation with c_ave = 47.2 [-], the average concentrations were c_ave = 50.3 [-] in the scenario with avenue-trees in every street canyon (Fig. 1b) and c_ave = 53.1 [-] in the scenario with avenue-trees in every other street canyon (Fig. 1c). Lager increases were found for the maximum traffic pollutant concentrations, from c_max = 360.1 [-] in the tree-free situation to 402.2 [-] and 508.2 [-] in the scenario with avenue-trees in every and in every other street canyon, respectively. This indicates that reducing the overall number of avenue-trees (or the total vegetation volume) within the street canyons does neither necessarily result in lower average nor in lower maximum traffic pollutant concentrations on the urban neighborhood scale. In contrast, the park in the center of the neighborhood (Fig. 1d) resulted in lower average and maximum traffic pollutant concentrations (c_ave = 38.3 [-] and c_max = 237.0 [-]).