How Inhomogeneous is Turbulence in Mountainous Terrain?

In complex mountainous terrain boundary layer characteristics are inhomogeneous by definition. This inhomogeneity stems from local surface characteristics such as changes in slope angle and vegetation cover as well as inhomogeneity in the thermal forcing (angle of incoming radiation, shortwave and longwave emission of surrounding topography). On long uniform slopes or at flat valley floors one could still assume a certain level of local homogeneity. Still, mountainous terrain also causes inhomogeneity in the larger scale flow, which provides a dynamic forcing to the boundary layer. Thus the question remains to what degree the assumption of homogeneity on any scale is ever valid in mountainous terrain, as well as what the dominant sources of inhomogeneity are.

In this contribution we present results from an extended period of turbulent measurements at multiple stations within the i-Box project, located in the Inn Valley, Austria. We examine spatial inhomogeneity through turbulence statistics, scaling and turbulence spectra on length scales ranging from several tens of meters (slope scale) to several kilometers (valley scale) for different types of forcing (radiatively driven to dynamically forced flows such as foehn winds). Particular focus is placed on small-scale variability of mean flow and turbulence and its interaction with larger scale structures, on a steep slope under stable conditions. The results highlight the problems encountered when dealing with measurements in mountainous terrain, especially problems of non-ideal slope configurations, which require special treatment when evaluating turbulence. Slope angle change and dynamic forcing by larger scale flow are identified as the dominant sources of inhomogeneity and impact the relationship between wind shear and turbulence generated by it. For example, dynamically driven foehn winds cause turbulence heterogeneity even at horizontal scales of few tens of meters, however, their scaling corresponds better to ideal-terrain reference. On the other hand radiatively driven downslope flow while homogeneous shows larger deviations from ideal-terrain scaling.