Turbulent boundary layers developing over tall and dense urban environments

It is estimated that, by 2050, cities will host 68% of the world’s population, as compared to the current 54%, and 33% in the 1960s [1]. This increase in urban population density is accomplished by the proliferation of tall buildings that maximise the provision of housing/office per footprint area at street level. High rises are also employed by architects to iconically shape our cities and as an expression of modern art [2]. An assessment of the effect of these tall structures on the wind field within cities is, therefore, needed as this significantly affects structural loads, pedestrian comfort, and air quality at street level. This type of data is, however, currently scarce.

Wind tunnel experiments were conducted on surfaces representative of urban environments at the University of Surrey via means of two-component Laser Doppler Anemometry (LDA). Four different roughness surfaces of packing density λ_P = 0.44, two of uniform height and two with heterogeneous height, arranged in both aligned and staggered patterns, were employed in this study. The four surfaces shared the same average building height, h=80 mm, however, the height standard deviation, σ_h, and the maximum building height, h_max, were varied. These parameters were chosen to match those of the Mong Kok area in Hong Kong.

Spatial averages from the LDA measurement profiles were used to determine how the Boundary Layer Thickness (δ), Inertial Sublayer (ISL), Roughness Sublayer (RSL), and Urban Canopy Layer (UCL) depths change across the different surfaces. Measurements in the UCL demonstrated high local spatial inhomogeneities challenging the existence of a universal spatially-averaged profile representative of each surface. The RSL depth was shown to be confined, for both fixed and heterogenous height surfaces, to a region just outside the canopy layer.

In addition to LDA measurements, the averaged skin friction generated by the surfaces was evaluated directly by measuring the pressure acting onto the upstream and downstream faces of the buildings. This facilitated the estimation of the virtual origin, d, and the roughness length, z₀, via means of customarily best fitting the velocity profiles within the log-law in the ISL. These results were then compared to morphometric models previously proposed in the literature to asses the effect of tall buildings onto the aerodynamic parameters. Results indicated a satisfactory agreement with the literature, however, d and z₀ showed significant dependency on the surface morphologies, suggesting that h_max and σ_h have a non-negligible effect on the aerodynamic parameters.

Lastly, a brief analysis was performed on the strength of the spanwise and wall-normal velocities components within the TKE equation, demonstrating the complex relationship between these velocity components and the underlying wall surfaces. Further work on similar urban morphologies is currently ongoing to investigate urban dispersion processes associated with these surfaces. The results of this work will be presented at the meeting.