VEG-AIR · Vegetation and urban air quality: CFD evaluation of vegetation effects on pollutant dispersion

Vegetation and urban air quality: CFD evaluation of vegetation effects on pollutant dispersion

Within the VEG-AIR research project different implications of urban vegetation on the micrometeorology and microclimatology of cities were investigated. The research was focused on two aspects, namely (i) on the effects of avenue-trees on natural ventilation and pollutant dispersion, and (ii) on the effects of various vegetative measures (avenue-trees, facade greening, roof greening) on air temperatures in the built environment. The basic objectives of the first aspect were to clarify the effects of avenue-trees in urban street canyons on air quality in a general context and to possibly figure out preferential or even beneficial avenue-tree designs and arrangements in terms of favourable pollutant dispersion. A further main goal was to validate and assess the performance of numerical modelling and in particular of vegetation models to reliably simulate flow and pollutant dispersion in the urban environment. The basic objectives of the second aspect were to assess the suitability and effectiveness of various vegetative measures locally applied at the building and street canyon scale to serve as climate change adaptation measures for the urban environment with regard to increasing global temperature. To achieve the project objectives, microscale numerical simulations employing Computational Fluid Dynamics (CFD) were performed. Reynolds-averaged Navier-Stokes (RANS) simulations as well as Large Eddy Simulations (LES) of momentum and scalar dispersion in the urban environment with emphasis on modelling the effects of vegetation on flow, turbulence and heat were done. For the first objective, simulations with various k-eps RANS turbulence models and LES of flow and pollutant dispersion in an isolated street canyon with and without avenue-trees were performed and validated against experimental data from wind tunnel measurements. Vegetation models of different complexity were analysed towards their performance in predicting the wind and concentration fields. Based on the outcomes and experience obtained from the street canyon scale studies, the natural ventilation and pollutant dispersion studies were extended to the urban neighbourhood scale. For a generic urban neighbourhood RANS simulations were performed to investigate the implications for a multitude of avenue-tree designs and arrangements on pollutant concentrations. For the second objective the transpirational cooling of different vegetative measures with regard to outdoor air temperatures was investigated for an urban street canyon located in the city centre of Arnhem in the Netherlands. Within the framework of this case study a parameterization for modelling the transpirational cooling of vegetation was developed and validated. RANS simulations for various vegetation scenarios were performed for the meteorological conditions during a heat wave by simultaneously accounting for the effects of urban vegetation on flow, turbulence and heat transport. The results for the first objective showed that among the various k-eps RANS turbulence models tested the realizable k-eps turbulence model performed best. In particular with regard to the maximum pollutant concentrations at the building facades in the central part of the street canyon, the realizable k-eps turbulence model clearly outperformed the other k-eps models. The analysis of the LES data revealed an overestimation of the maximum pollutant concentrations by the simulations and a spatially non-uniform model performance. Closer agreement with measurement data was achieved near the street canyon ends than for the central part of the canyon. Standard model acceptance criteria were satisfied more easily for the leeward than for the windward canyon wall. In general, better agreement with wind tunnel measurements was obtained with LES relative to the realizable k-eps turbulence model simulations but the immense larger computational costs question their preference to RANS simulations for large computational domains with many grid points at the moment. The performance of vegetation models improved with increasing model complexity. Here, larger performance enhancements were found in terms of pollutant concentrations than of wind velocities. For the generic urban neighbourhood, low to moderate average increases (<13.2%) in pollutant concentration at pedestrian level were obtained. An approximately 1% increase in average concentration was found with each percent of the total street canyon volume being occupied by vegetation. However, pronounced locally restricted decreases or increases in concentration (-87 to +1378%) were found. The general pattern of concentration changes relative to the tree-free situation were similar for all avenue-tree designs and arrangements studied and no particular avenue-tree arrangement beneficial for pollutant concentrations emerged. Overall, the outcomes concerning the first objective indicate the necessity to account for trees in street canyon and neighborhood scale dispersion studies and their relevance for reliable urban air quality assessment. They, though locally restricted, significantly impact flow and pollutant concentrations and hence should be considered in urban planning. The results for the second objective showed that locally applied vegetative measures at the street and building scale can contribute to a reduction of air temperatures in urban street canyons during hot summer days by transpirational cooling. However, the intensity of cooling in terms of air temperature reductions and their spatial extent differed distinctly between the various vegetative measures. Avenue-trees provided the strongest transpirational cooling within the street canyon both in terms of air temperature reductions and spatial extent with maximum reductions of 1.6°C inside and directly next to the tree crowns. The effects of facade greening were less pronounced with maximum reductions of 0.5°C and locally restricted to the vicinity of the building walls and those of roof greening on air temperatures inside the street canyon were negligible under the boundary conditions of this case study. In order to achieve a spatially extensive and noticeable transpirational cooling, a densely distributed arrangement of vegetative measures is required. This implies the arrangement of vegetative measures along many streets and squares and also in courtyards. Overall, the outcomes suggest to apply vegetative measures inside urban street canyons as a climate change adaptation measure. They reduce heat stress and related human morbidity and mortality rates encountered already nowadays during heat waves and dampen the expected increase in heat stress frequency within 50 to 100 years due to global warming. Hence, vegetative measures for cooling should become an integral part in urban planning. In a synoptic view of the outcomes of the two basic objectives addressed within this research project it becomes evident that the implementation strategies for urban vegetation are not straightforwardly to combine, they even partly counteract each other. Whereas for natural ventilation and pollutant dispersion the implementation of avenue-trees can be critical and unfavourable for air quality, the implementation of vegetative measures for the purpose of transpirational cooling is beneficial and desirable. A general advise to urban planners is to implement vegetation in areas with low or no traffic pollutant emissions, e.g. in little-used roads or in courtyards and on squares.

Data: CORDIS, © European Union