- Imperial College London, Civil and Environmental Engineering, United Kingdom of Great Britain – England, Scotland, Wales (jingzi.huang17@imperial.ac.uk)
Numerical Weather Prediction (NWP) models make weather forecasts at increasing resolution. At higher spatial resolution, urban surface representation becomes more heterogeneous, making conventional land-surface exchange parameterizations invalid. New parameterizations are needed to incorporate these inhomogeneities.
We develop a multi-scale planar-averaging framework for urban flows to address: “What are the requirements for NWP models as resolution increases?” Our computationally efficient method uses convolution filters for coarse graining. To investigate the heterogeneity, we apply the multi-scale framework to a large-eddy simulation of an idealised heterogeneous urban environment of 512 buildings based on a typical London height distribution. Figure 1 shows the plane views of streamwise velocity at various averaging filter lengths, where a decrease in heterogeneity can be observed as the length increases (the resolution lows). We conclude that for this geometry, the characteristic urban length scale is L=50 m, which is the averaging length scale at which as much variance in the flow is resolved as is unresolved. For L>400 m, the problem is approximately homogeneous, allowing non-building-resolving NWP models to be used without modification for the case under consideration.
Figure 1. The streamwise velocity field at various averaging lengths at the mean building level. (a) The original field, (b) L =6 m, (c) L=24 m, (d) L= 96 m, (e) L= 384 m, and (f) the plane-averaged field. The white boxes represent the buildings.
We then focus on the parameterisations of the distributed drag which is important in momentum transport. We show that a universal drag distribution holds reasonably well for resolutions L above 200 m, but at higher resolutions the problem becomes inhomogeneous, and the parameterisation works less well. Parameterisations of unresolved turbulence and dispersive stress are also examined, and the appropriateness of different turbulence closures is discussed.
How to cite: Huang, J. and van Reeuwijk, M.: Multi-scale analysis of flow over heterogeneous urban environments, 12th International Conference on Urban Climate, Rotterdam, The Netherlands, 7–11 Jul 2025, ICUC12-540, https://doi.org/10.5194/icuc12-540, 2025.
Supporters & sponsors
