modeling
Numerical models and
wind tunnel experiments will be both utilised complementary to the observational data.
The still limited
spatial resolution of the observational data can be improved by means of the
numerical and physical model results, since a full three-dimensional picture of the urban flow and turbulence structure can only be obtained through
modeling.
Horizontal
variability of turbulence characteristics in the wind tunnel at a certain
height above the urban surface near site "Basel-Sperrstrasse" (Feddersen et al.,
2003, see publications)
|
Increasing computer power allows meso-scale meteorological models to be run at higher spatial resolution. Thus more grid points will be truly ‘urban’ grid points requiring a surface exchange parameterization that takes into account urban surface characteristics and exchange processes. Nevertheless, operational models and similarly most research models simply use some ‘urban values’ for a number of surface-characterizing variables (such as the roughness length or the surface’ thermal conductivity) without any further modification due to the high roughness elements.Air pollution models and in particular their meteorological part, are generally developed for surfaces of lesser roughness and do not take into account the dispersion conditions in the layer near the surface. Apart from the full three-dimensional numerical simulations which were focused on the meteorological part of BUBBLE the tracer experiments conducted provide an excellent opportunity to test a number of parametric dispersion models that have been developed and specifically adapted to urban conditions. Here, the focus lies on having at hand not only a tracer experiment with perhaps turbulence observations at the site and height of the tracer release, but a detailed, spatially quite well resolved field of turbulence, flow characteristics and surface data. The wind tunnel model of the central city part "Kleinbasel" at the scale 1:300 will provide an extensive reference data set for urban turbulence and pollutant dispersion particularly suited for the comparison with numerical model predictions. It is also expected to complement and help the interpretation of the full scale measurements.
|