On the accuracy of idealized sources in CFD simulations of pollutant dispersion in an urban street canyon

Idealized pollutant sources are widely used by the scientific community to replicate traffic emissions in wind tunnel tests and CFD simulations. However, it is unclear to what extent such idealized sources can adequately reproduce the emission and dispersion by cars in idle (static) or moving situations in streets. This study investigates the impact of a static idealized point source (S-IPS) versus an idling or static (S-) and a moving or dynamic (D-) realistic car source (RCS) on the pollutant dispersion in a street canyon. First, 3D steady RANS and LES simulations are performed with a S-IPS and validated by means of wind-tunnel tests. Next, LES simulations are performed to analyze the impact of S-IPS versus S-RCS. Finally, three D-RCS with different car speeds are simulated and compared with S-RCS. The results show that using a S-RCS increases the plane-averaged concentration by 11%-140 % at z/H = 0.03 and by 30%-50 % at y/H = 0, with respect to S-IPS. The comparison of S-RCS and D-RCS shows that car movement can also have a large impact on pollutant dispersion along the canyon.