Dynamics of the moderately stable boundary layer

Tese de doutoramento, Ciências Geofísicas e da Geoinformação (Meteorologia), Universidade de Lisboa, Faculdade de Ciências, 2018This project was divided in two main components. The first component consisted in modeling the very stable boundary layer using a large-eddy simulation (LES) model. Therefore, a novel LES model setup was developed to overcome the difficulties imposed on numerical models by the stable stratification and the associated reduction in the energetic scales in combination with the large anisotropy of turbulent motions. A series of stable boundary layers with weak geostrophic winds and strong surface cooling rates is explored. The surface heat flux is dynamically computed by resolving the surface layer since the often-used Monin–Obukhov similarity theory cannot represent very stable conditions. This requires the use of very high grid resolutions. Thus, Δ = 0.05 and 0.10 m was used. Two stability regimes can be identified based on vertical profiles of the gradient Richardson number. Nevertheless, the Richardson number exceeds the critical value for all simulations. The buoyancy effects act on the small scales of the turbulent flow from the surface to the top for the most stable regime, and only in the upper part of the SBL for the other regime. Furthermore, the analysis of the anisotropy tensor shows that the flow is in an anisotropic state and is governed by the stream-wise component of the turbulent flux u′u′. In the second component of this project, a method to decompose the vertical turbulent flux based on the corresponding joint probability density function (JPD) is introduced. Thus, the JPD is partitioned into a joint Gaussian part and its complement, which represent, respectively, the local mixing and them coherent motions, such as updrafts and downdrafts. The flow decomposition is independent of the water condensate (cloud) and can be applied to dry convection, subcloud layer, and stratiform cloud layer. Moreover, the JPD-based method can be applied to observational or model data.Fundação para a Ciência e a Tecnologia (FCT), SFRH/BD/52555/201