Analysis of urban wind conditions and wildfire smoke dispersion for downtown Montre<acute accent>al using computational fluid dynamics

Urban wind conditions and air quality have the potential to affect the majority of the world's population. Specifically, smoke from wildfires is increasingly posing risks to people living in cities as it is transported by the wind, making a more complete understanding of how atmospheric flow affects air quality in urban environments necessary. Computational fluid dynamics is used to investigate the flow and smoke dispersion characteristics in Montreal on July 17, 2023, between 16:00 and 22:00 UTC. This day exhibited moderate southwest winds carrying significant amounts of wildfire smoke into the city. Reynolds-averaged Navier-Stokes simulations using the standard k-epsilon and Shear Stress Transport k-omega turbulence models are compared against measurements of wind velocity and PM2.5 concentration from an anemometer, a Doppler lidar, and an air quality monitoring station. While the models are shown to accurately predict the urban boundary layer wind profile, only the k-epsilon model provides satisfactory predictions of wind speed in comparison with the anemometer, stressing the importance of accurately modeling dynamics on the building scale. The effect of the turbulent Schmidt number is investigated, for which the value of 0.6 most accurately reproduces the dispersion phenomena near the air quality monitoring station. Concentrations of the wildfire smoke are found to vary significantly across areas of the city, as some building morphologies are found to direct pollution to regions where it becomes trapped. Additional discussion of local wind and air quality characteristics is presented to better inform citizens of potential risks.