Turbulent slope winds in complex terrain: from heat and moisture transport to the sampling of single plumes

<p>Local thermal circulations developing over heated valley slopes strongly influence the convective boundary layer (CBL) over mountainous terrain. For this study, large-eddy simulations are carried out both over idealized valleys and semi-idealized complex terrain. The flow is decomposed into a turbulent part, a local mean circulation capturing the slope winds, and a large-scale (upper-level) wind. This allows a detailed budget analysis for heat and moisture. The temperature distribution is horizontally fairly uniform inside each valley due to the homogenizing effect of the thermally-induced circulations. In contrast to that, the slope winds contribute strongly to the transport of moisture up to the ridges. The entrainment of dry air by the recirculation leads to a horizontally non-uniform moisture distribution. Consequently, a large-scale, upper-level wind hardly affects the horizontally homogeneous temperature distribution while it can considerably reduce the vertical moisture transport: a horizontal wind mixes the moisture from the slope-wind layer into the dryer regions of each valley. Single updrafts mark the small-scale end of coherent motions in the CBL over complex terrain. A conditional sampling method is applied in order to identify the thermal plumes using a passive tracer. In the mixed layer, the plumes are moving upslope with the slope wind. In order to quantify the contribution of the plumes to the vertical transport of heat and moisture, the joint probability density functions of the turbulent fluxes are calculated and decomposed into a local and a coherent part. From this perspective, the turbulence statistics is analyzed at different heights in the CBL and compared to the statistics over flat terrain. In general, the plumes turn out to dominate the vertical fluxes in the valleys and in the lower part of the boundary layer. Especially at ridge height, where the updrafts are few but almost stationary, the statistics are fairly different.</p>