Estimating aerodynamic parameters over complex urban surfaces: a case study for London

Knowledge of the surface aerodynamic parameters over urban regions is of use in many applications, most commonly in the development of boundary layer wind speed and turbulence profiles used for example in wind resource assessments for energy micro-generation, and pollution dispersion modelling. Urban surfaces pose particular challenges due to the strongly heterogeneous nature of the surface canopy. Nevertheless, several studies have shown that log profiles may be extended into the roughness sub-layer of the urban boundary layer, proving spatially averaged estimations of the wind profile with height above the canopy. Defining these profiles however, is critically dependant on the parameterisation of the surface aerodynamic characteristics. In this work, detailed 3-dimensional building data based on high resolution LiDAR (light detection and ranging) data, coupled with a morphometric model, are used to estimate the aerodynamic roughness length z_o and displacement height d over a major UK city (London).

Firstly, using an adaptive grid, the city is divided into neighbourhood regions, each of a relatively consistent geometry throughout. Secondly, for each neighbourhood, a number of geometric parameters are derived from the LiDAR data. These include the mean building height h_m, the planar and frontal area densities ë_p and ë_f, b/l, D_y/D_x, and the distribution of building heights in the neighbourhood. Here, b/l is the average breadth over length ratio of the buildings in the array (with respect to the incoming wind direction), and D_y/D_x is the average breadth over length ratio of the ground area associated with each building. The height distribution of the buildings can be well represented by the vertical profile of the frontal or plan area density, ë_f(z) or ë_p(z), respectively. Finally, these are input into a morphometric model that considers the influence of height variability to predict aerodynamic roughness lengths and displacement heights.

Figure 1:  Maps of estimated ‘effective' roughness lengths (top) over London for Southerly and Northerly incoming wind directions, and profiles though the city centre (bottom) along the lines indicated upon the maps.

The variation of estimated aerodynamic parameters with incoming wind direction is explored, and the importance of modelling these effects is discussed. Figure 1 presents example maps showing the estimated roughness lengths across London for Northerly and Southerly incoming winds. The profile shown below serves to illustrate the magnitude of variation in z_o for a particular neighbourhood region depending on the incoming wind and therefore upwind roughness profile. The influence of including the effects of building height variability on the magnitude of the resulting aerodynamic parameters is also addressed. The results suggest that the influence of height variation is strongly dependent upon the density of the buildings. Changing wind direction was found to modify predicted effective roughness lengths (regional scale) more strongly than local roughness lengths, and in the city centre these effects were particularly significant. Overall, the results indicate the potential gains in accuracy that may be made by employing a morphometric model with a wider range of geometrical input parameters than those commonly used at present, and in considering the directional dependence of effective roughness lengths when modelling wind flow in central city areas. The resulting maps may be of use for various applications including wind resource prediction, forecasting and pollution dispersion modelling.