Authors: Mathias W. Rotach, Peter De Haan
Addresses: Swiss Federal Institute of Technology (GIETH), Winterthurerstr. 190, 8057 Zurich, Switzerland. ' Swiss Federal Institute of Technology (GIETH), Winterthurerstr. 190, 8057 Zurich, Switzerland
Abstract: It was one of the findings from the model validation exercise during the third workshop 1994 in Mol that there was a common tendency for all models under consideration to underestimate the observed concentrations in the case of the Copenhagen dataset. It is argued here that this is due to the fact that this experiment was performed over a rough suburban surface. In such circumstances, a roughness sublayer covers the lower part of the surface layer wherein surface layer scaling cannot be valid owing to the presence of roughness elements and the resulting disturbances of the flow. A two-dimensional (u and w) Lagrangian stochastic dispersion model is used to demonstrate the effect of the modified turbulence structure within the urban roughness sublayer for the example of the Copenhagen dataset. If a roughness sublayer is included by modifying the turbulence and flow structure in the lowest metres of the domain according to observed (urban) roughness sublayer characteristics, it is shown that the model performance is considerably improved. The overall performance measures (such as rms difference, fractional bias, etc.) become significantly better when taking the roughness sublayer into account. Similarly, it is shown that the evolution of ground-level concentrations with distance from the source is simulated more realistically for most of the individual experiments.
Keywords: rough surfaces; roughness sublayers; urban dispersion modelling; urban turbulence; atmospheric dispersion models; Copenhagen dataset; air pollution; environmental pollution.
International Journal of Environment and Pollution, 1997 Vol.8 No.3/4/5/6, pp.279 - 286
Published online: 15 Sep 2009 *Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article