Sheared convective boundary layers: turbulence kinetic energy and entrainment dynamics

Abstract

The atmospheric boundary layer is the lower part of the atmosphere that is directly influenced by processes occurring at the Earth’s surface. The depth of the boundary layer varies significantly during the diurnal cycle. Typically the boundary layer is rather shallow during night time (about 100 meters), but it reaches a depth of about a kilometre during daytime. In this thesis I investigated the old but still largely unsolved problem of the entrainment and growth dynamics of atmospheric boundary layers that are driven by both shear and a positive buoyancy surface flux. These growing boundary layers are relevant as they represent the typical atmospheric daytime conditions. My research goal was to systematically improve the understanding of entrainment, explain contradictions in existing theory and develop a consistent characterization of the boundary layer growth dynamics. As experimental (empirical) basis I used a series of ‘large-eddy simulations’ of sheared and convective boundary layers. These allowed a detailed analysis of the shear-dependent variation of the boundary-layer structure and the turbulence kinetic-energy budget. As a result I was able to show that over the whole stability range, from purely convective to purely shear-driven conditions, entrainment basically scales with a linear combination of integral buoyancy production and shear production of turbulence kinetic energy. Previously, such ideal behaviour had merely been assumed and discussed, but, due to inappropriate assumptions about the flow structure, never been demonstrated. For certain conditions I also observed systematic deviations from the ideal behaviour. These are caused by (A) local in-stationarity at the interface between the boundary layer and (B) the previously ignored formation of gravity waves at the upper boundary-layer interface. The influence of the latter on the boundary layer’s growth and momentum dynamics turned out to be significant in moderately sheared conditions. I argued that both interfacial effects (A and B) are physically intuitive and showed that both can be well represented using appropriate integral and local scales. As final result I obtained a boundary-layer growth model that is quite simple and at the same time physically much more coherent and accurate than previous ones.