Modeling pollutant dispersion in urban areas with computational fluid dynamics

Abstract

Pollution is a major issue for urban areas all around the world. This report reviews the recent literature on modeling and pollution dispersion in urban areas with the application of computation fluid dynamic (CFD) models. A review of atmospheric turbulence and weather conditions is provided as pertaining to dispersion modeling. When applying CFD, the modeler must specify the closure model, which could be direct numerical simulation (DNS), Reynolds average numerical simulation (RANS), or large eddy simulation (LES). A comparison of the RANS and LES models is provided. The main advantages of CFD is that it offers advanced modeling that can account for turbulence under multiple weather conditions and 3-dimensional obstructions in the flow field. The recent research focuses on how CFD can be used to predict pollution concentrations without measurement devices or physically altering an urban landscape. Multiple researchers have explored pollution dispersion near roadways, specifically modeling pollution emission and the effect of roadway barriers on pollution dispersion. Other research focused on modeling a larger portion of a city where the buildings and street canyons play an important role in the flow patterns (and the effects on pollutant dispersion) within the city. The research explores the possibility of modifying urban areas to allow for increased pollutant dispersion. The recent research shows that CFD is a powerful tool for modeling pollutant dispersion and that more research is required to take full advantage of CFD in this area.