Assessment of Vegetative Measures for Mitigation of Air Temperatures in the Urban Environment during Heat Waves—A Case Study for the Dutch City of Arnhem

Evapotranspirational cooling by different types of vegetation was investigated for the city center of Arnhem/The Netherlands. A case study on the capability of locally applied avenue-trees, façade greening and roof greening on reducing air temperatures inside an urban street canyon was performed using Computational Fluid Dynamics (CFD). The reductions in air temperature relative to the status quo were analyzed for an afternoon hour of a hot summer day during a heat wave in July 2003.

The results show that evapotranspirational cooling by locally applied vegetative measures can noticeably contribute to a reduction of air temperatures inside an urban street canyon and in particular at the pedestrian level. However, the intensity of cooling in terms of air temperature reductions and their spatial extent differed distinctly among the vegetative measures. The strongest cooling effect was found in the presence of avenue-trees with mean and maximum temperature reductions of 0.6 K and 1.6 K, respectively, where the most vigorous reductions were observed inside and directly next to the tree crowns (Fig. 1 and Fig. 2, path 1). However, lower air temperatures were found throughout the entire street canyon. A distinct cooling effect of around 1 K was still present at the center of the street canyon (Fig. 2, path 2). Notice that path 1 and path 2, see Fig. 1 top, coincide with the location of the avenue-trees and with the center line of the street canyon, respectively.

Fig. 1 Air temperatures [°C] at 2 m above ground for the status quo (top) and scenario with avenue-trees indicated as dashed lines (bottom). Path 1 and path 2 coincide with the location of the avenue-trees and with the center line of the street canyon, respectively.

The façade greening resulted in comparatively small changes of air temperatures inside the street canyon with mean and maximum reductions of 0.1 K and 0.3 K, respectively. Its cooling effect was only noticeable in the vicinity of the building walls where greening was applied and faded out towards the canyon center (Fig. 2). In the center of the street canyon only marginal temperature reductions (< 0.05 K) due to the façade greening were found (Fig. 2, path 2). Finally, the locally applied roof greening at the buildings facing the street did not influence the air temperatures inside the canyon in the present situation with wind parallel to the street (Fig. 2). The air cooled by the roof greening was rather dispersed into the above-roof flow than entrained into the street canyon. However, for perpendicular or oblique winds, or under low wind conditions, the cooled roof top air may possibly get entrained into the street canyon and result in noticeable air temperature reductions.

Overall, the air temperature reductions were less pronounced at the street canyon ends. The largest cooling in all scenarios generally occurred between 100 m < x < 300 m (Fig. 2). The smaller cooling effectiveness in the eastern part is due to the fact that for the present mean wind direction the air enters the street canyon at this side and is subsequently cooled when flowing through the vegetation. The lower cooling effectiveness in the western part is attributed to the larger canyon width which is here approximately 50% larger and the smaller share of vegetation at the total street canyon volume. For all vegetation scenarios, no noticeable changes in air temperatures in the neighboring streets and squares were observed. This indicates that the cooling effect of locally applied vegetation is also restricted to the vicinity of the vegetative measures.

Fig. 2 Air temperatures [°C] at 2 m height at the location of the avenue-trees in front of the northern building walls (path 1), and in the middle of the street canyon (path 2). See Fig. 1 top for the location of the paths and x-positions. Notice that the dotted lines for the roof greening are hidden by the solid lines for the status quo.