A comparison of results from scaled field and wind tunnel modelling of dispersion in arrays of obstacles

Abstract The problem of plume dispersion around urban buildings has been investigated by physical modelling using arrays of building-like obstacles at two scales—1 : 100 in a boundary layer wind tunnel and 1 : 10 at a field site. The particular effect of obstacle width-to-height ratio (W/H) was examined for a fixed obstacle packing density. At short distances from the plume source the concentration profiles in the obstacle arrays were quite variable, but at distances beyond two rows the concentration profiles were well approximated by a Gaussian distribution laterally and a reflected Gaussian vertically. A comparison of the Gaussian plume parameters as a function of nondimensional distance from the source shows the influence of the obstacles. Compared to plumes in the open terrain, the plumes in the arrays were typically 2–4 times as wide. At short to intermediate distances the plume width was closely related to the width of the obstacles. An effective vertical displacement of the plume was also found; it was typically about one building height. This resulted in a net vertical growth of the plume which was typically 2–3 times larger than that in the open terrain at similar distances from the source. In the arrays of obstacles with W/H>1, significant bodily deflection of the plume occurred for nonorthogonal wind directions. Compared to the wind tunnel, there was more scatter in the field results, and some effects of the larger scales of turbulence were observed. However, the field results were qualitatively and in most cases quantitatively the same as those from the wind tunnel.