Abstract
A comprehensive analysis of the turbulence structure of relatively deep midlatitude katabatic flows (with jet maxima between 20 and 50 m) developing over a gentle (1°) mesoscale slope with a long fetch upstream of the Meteor Crater in Arizona is presented. The turbulence structure of flow below the katabatic jet maximum shows many similarities with the turbulence structure of shallower katabatic flows, with decreasing turbulence fluxes with height and almost constant turbulent Prandtl number. Still stark differences occur above the jet maximum where turbulence is suppressed by strong stability, is anisotropic and there is a large sub-mesoscale contribution to the flux. Detecting the stable boundary-layer top depends on the method used (flux- vs. anisotropy-profiles) but both methods are highly correlated. The top of the stable boundary layer, however, mostly deviates from the jet maximum height or the top of the near-surface inversion. The flat-terrain formulations for the boundary-layer height correlate well with the detected top of the stable boundary layer if the near-surface and not the background stratification is used in their formulations; however, they mostly largely overestimate this boundary-layer height. The difference from flat-terrain boundary layers is also shown through the dependence of size of the dominant eddy with height. In katabatic flows the eddy size is semi-constant with height throughout the stable boundary-layer depth, whereas in flat terrain, eddy size varies significantly with height. Flux-gradient and flux-variance relationships show that turbulence data from different stable boundary-layer scaling regimes collapse on top of each other showing that the dominant dependence is not on the scaling regime but on the local stability.