This thesis aimed for the characterization of the three-dimensional turbulence structure in a narrow alpine valley. One possibility to measure turbulence in a valley volume is to use a measurement aircraft. Classical approaches, like Reynolds averaging, Reynolds decomposition or Fourier Analysis are not necessarily applicable because the flight legs are short and the conditions are very inhomogeneous. Therefore the Continuous Wavelet Transformation was used to analyze the turbulence structure in the Inn Valley based on high frequency, airborne data collected on a three-day field campaign. Furthermore, scaling approaches for scaling Mixed Layer (ML) spectra and vertical profiles of Turbulence Kinetic Energy (TKE) were evaluated.
The analysis of the three components of the TKE showed that it decreases with height in the valley. The decrease, however, is stronger over the valley center than near the slopes. Due to enhanced solar insolation during the day, there is often more TKE near the northern slope. Besides, vertical heat fluxes are negligible in the Free Atmosphere and only positive in the ML.
The measurements implied that the exchange of energy between two side valleys and the Free Atmosphere was maintained by plumes over the shared ridge. Evaluation of the horizontal heat flux showed a flux towards the ridge in the surface near atmosphere and in the opposite direction in the upper atmosphere. A positive vertical heat flux prevailed directly above the ridge and was encountered by two negative branches. The plumes acted at approximate length scales between 30m and 600m. Obtaining surface measurements from different surface stations can slightly improve the scaling performance. A major conclusion is that there is a dependence of the measurements taken in the ML and their scaling on the presence of a second ML further aloft. The scaled ML spectra, on the one hand, showed that isotropy is not maintained for the vertical velocity component. Hence, an additional variable, possibly a length scale depending on the valley geometry, influences the processes in a narrow valley. On the other hand, a pronounced gap in non-dimensional frequency between cases measured under 1 ML and 2 MLs conditions was observed. Hence, a smaller scaling height was appropriate, whereas no convenient length scale could be found. The vertically linear decrease of TKE in the Inn Valley showed a similar rate as compared to the Riviera Valley but was shifted towards higher values.