Numerical Investigation on the Effect of Avenue Trees on PM_2.5 Dispersion in Urban Street Canyons

The Reynolds-averaged Navier-Stokes (RANS) model and revised generalized drift flux model were used to investigate the characteristics of airflow fields and PM_2.5 dispersion in street canyons with a variety setting on tree crown morphologies (i.e., conical, spherical, and cylindrical), leaf area densities (LADs = 0.5, 1.5, and 2.5 m²/m³), and street canyon aspect ratios (H/W = 0.5, 1.0, and 2.0). Results were as follows: (1) airflow fields were reversed in the presence of trees and enhanced with higher LAD; (2) air velocity decreased negligibly when LAD increased from 1.5 to 2.5, but significantly when LAD increased from 0.5 to 1.5; (3) tree crown morphologies, building aspect ratios, and LADs were interrelated. The comparison of PM_2.5 showed that the most critical situations in H/W = 0.5, 1.0, and 2.0 corresponded to LAD = 0.5 with a conical canopy; (4) the H/W = 1.0 and LAD = 1.5 scenario was identified as the most efficient combination for PM_2.5 capture; (5) the maximum PM_2.5 reduction ratio was ordered from low to high in the sequence of conical, spherical, and cylindrical canopies. At predestinated LADs and aspect ratio, Populus tomentosa with cylindrical crown morphology exhibited the best efficiency on PM_2.5 capture with a reduction ratio of 75% to 85% at pedestrian height.