Assessing the influence of rooftop vegetation on pollution dispersion in urban canyons through Large-Eddy Simulations

Abstract

Urban air pollution is a significant hazard to human health, which is expected to worsen with the increase of urbanisation worldwide. There is a growing need to evaluate mitigation measures such as green Nature-Based Solutions. However, most studies focus on vegetation at street level, while the role of rooftop trees remains largely unexplored. Here, we investigate a simplified urban canyon where trees are placed on building rooftops, at the canyon-atmosphere interface level, which is a key area for pollutant removal mechanisms. Nine types of tree crown are analysed using highly-resolved Large-Eddy Simulations at Reynolds number $Re=2\times 10^4$. The simulation without trees is successfully validated against laboratory and numerical experiments. Results show that the key parameter is the tree crown density, while other characteristics like the leaf area index play a secondary role. Overall, rooftop trees reduce horizontal velocity and increase the turbulent kinetic energy (TKE) in the outer layer above the canyon: low-density-crown trees generate higher TKE, mimicking the presence of isolated obstacles and increasing the turbulent vertical mixing; high-density-crown trees reduce the atmospheric wind velocity, mainly acting as an extended area of momentum sink, which also reduces the wind turbulent level. Within the canyon, rooftop trees have a limited effect on the overall dynamics and the pollutant concentration distribution. Hence, unlike street trees, rooftop trees can be used to extend urban green areas and shade streets without hampering the dispersion of pollutants.