Neighbourhood-scale Urban Dispersion Modelling Using a Canopy Approach

It is important that urban air quality prediction models are accurate within the urban canopy since this is where humans are usually exposed to higher levels of pollution. Buildings within the urban canopy exert a drag force on the flow field. This modifies both the mean wind profile and turbulence characteristics, and consequently affects mixing of pollutants in the canopy. In current numerical weather prediction (used to provide meteorological input to offline dispersion models) and online air quality models, typically the effect of drag on the flow is accounted for in a highly simplified way, using a roughness parameter at a single height, and a standard log law is assumed for the mean wind profile.

We follow a vertically-resolved approach originally developed for vegetation canopies, in which the time- and spatially-averaged equations for momentum and scalar mass conservation are solved. It is unfeasible to directly account for the effect of individual buildings, and hence an average representation of the urban canopy is necessary. The method is adapted for urban areas using modified parametrisations for the height distributed drag and turbulence. Scalar concentration profiles both within and above the urban canopy are evaluated using direct numerical simulation data, and an investigation of the sensitivity of flow speed and scalar concentration to the turbulence and drag parametrisation is presented.