First attempts of the validation of an LES model by acoustic tomography data
Responsible: Sonja Weinbrecht; Marcus Oliver Letzel
Project type: individual ​DFG project
Duration: 15/06/1999 - 31/05/2003

The aim of this project was to investigate turbulent convective structures near the surface by means of LES and acoustic tomography. So this was a cooperative project between the Institute of Meteorology, University of Leipzig and the Institute of Meteorology and Climatology, University of Hannover. By acoustic tomography (​Acoustic Research Group at LIM) high resolution and volume-averaged data can be measured and compared directly with the high resolution and volume-averaged simulation data (our group). All simulations were run with our LES-model PALM.

By comparing the simulation results with the measured data we tried to evaluate the LES-model PALM and wanted to test different subgrid-scale parametrization models implemented in the LES-model compare their results.

In preparation of such a quantitative comparison first the sensitivity of the LES-model against the boundary conditions and the initial parameters was observed. We found that the simulations results are highly sensitive towards the surface roughness length (if the surface temperature is prescribed). With the help of the sensitivity study results it is now possible to name the measurement accuracies of the boundary conditions and initial parameters which are needed to perform a quantitative comparison. Furthermore the influence of measurement errors of these parameters on the simulation results can now be quantified.

A first comparison with acoustic tomography data showed some differences which can be explained with the help of the sensitivity study and with inaccurate measurements of acoustic travel time and acoustic transmitter-receiver distances.

To check the method of the acoustic tomography acoustic travel time measurements have been performed with the LES-model. So the simulated data could be compared with the simulated acoustic travel times which were analysed using the algorithms of the acoustic tomography method. With the help of this comparison some problems with the algorithm to separate the mean wind influence from the sonic speed were detected. These problems will be solved by using reciprocal acoustic travel time measurement methods. (See also the follow-up study by Weinbrecht.)