Unevenly distributed trees facilitate horizontal dispersion of traffic pollutants and redistribute pollution hotspots in a real-world street canyon

Urban street canyons often experience poor air quality because buildings and vegetation modify and obstruct airflow, thereby reducing the pollutant dispersion capacity within these built environments. While trees are widely recognised for their environmental (e.g., microclimate regulation and carbon sequestration) and social (e.g., aesthetics) benefits, their complex effects on pollutant dispersion remain debated. Existing numerical studies on pollutant dispersion in urban street canyons with tree canopies commonly adopted over-simplified geometries that neglected the heterogeneous sizes and spacing of individual trees, despite these features are prevalent in real-world settings. This simplification can limit the accuracy of simulated airflow and pollutant dispersion. Accordingly, this study develops a high-fidelity model that features realistic street geometry and explicit representation of unevenly distributed and irregularly shaped tree canopies to investigate their impact on roadside air quality in a real-world street canyon. The heterogeneity of real-world trees is considered, and two idealised scenarios are designed for comparison: treeless street and evenly-spaced trees. Results show that, in the real-world case with unevenly distributed trees, pollution hotspots are localised in mid-street leeward pedestrian areas (z/Z = 0.45-0.68), with peak NO₂ concentrations of ∼135 μg/m³, whereas the treeless case exhibits more uniform dispersion and lower peak concentrations (∼120 μg/m³). In contrast, the evenly distributed trees configuration produces the worst pollution, with near-ground hotspots extending over z/Z = 0.34-0.95 and peak concentrations exceeding 150 μg/m³. These findings highlight that while tree canopies in general reduce wind speed, uneven tree distribution can enhance horizontal airflows and relocate pollution hotspots in urban street canyons. These insights offer a conceptually novel strategy for roadside air quality management, shifting the focus from maximising vertical pollutant dispersion into the atmosphere or preventing pollutant transport toward the roadside to pollution hotspot relocation.