Turbulent Coherent Structures in the Convective Boundary Layer Over Complex Terrain

Coherent structures contribute significantly to turbulent exchange of heat, momentum and pollutants across the convective boundary layer (CBL). The typical taxonomy of CBL coherent structures includes horizontal convective rolls, open cells, alveolar polygonal structures, hairpin vortices, dust devils, near-surface streaks, and convective plumes. However, it is unknown how daytime coherent structures and complex terrain flows (such as thermally-induced and terrain-forced flows) interact and influence each other. Thermally-induced flows do not occur over flat terrain, so their possible impact on coherent structures requires elucidation. The goal of this dissertation is to investigate the coupling between spatiotemporal characteristics of coherent structures and thermally-driven flows in Owens Valley, California, the site of the T-REX experiment conducted in Spring 2006. The dissertation is organized into five chapters starting at the largest scales encompassing the entire Owens Valley and neighbouring mountain ranges, down to the smallest scales of a few tens of meters above the surface of the valley.

First, the influence of highly non-stationary macro-, meso-, and micro-scale phenomena on Monin-Obukhov similarity theory is assessed. This theory suffers from assumptions that are not valid over complex terrain. By contrasting five stationarity criteria common in boundary-layer meteorological research, we have found two criteria (one identifying the short-term signal variability, and the other identifying intermittency in the signal), to outperform the other three when determining the reduction in scatter around flux-variance similarity functions.

Next, emphasis is placed on the energy spectral gap, which often becomes elusive in complex terrain where different mesoscale phenomena may fill it completely. A better understanding of the spectral gap enables the determination of the ideal turbulence cut-off time scale necessary for the estimation of turbulence moments, which are the foundation of investigations in the remainder of the dissertation. A network of sixteen automatic weather stations is used to generate novel knowledge about the spatial variability of gap scales in a valley. Additionally, periodicity of upslope flows is observed and then validated against a simple, analytic model. Three categories depicting the dominant along-valley variability are defined, which also represent the cornerstone of two more dissertation chapters.

Large eddy simulations are performed for three climatologically-representative case study days. These simulations aim to establish the mechanisms in which a valley environment modifies structural properties of horizontal convective rolls and open cells, two main modes of mixed layer convection. The principal finding here is the valley-induced narrowing of daytime convection, whose role is to ultimately increase the occurrence of rolls and transitional roll-like features in the valley CBL.

Next, the dissertation focuses on investigating the surface energy budget closure in Owens Valley. Here, the focus is on establishing how near-surface scalar similarity breaks down under the influence of entrainment-driven drawdown of drier and potentially warmer free-tropospheric air from above the CBL all the way down to the valley bottom. Resulting scalar dissimilarity violates Bowen ratio-based closures, by introducing non-local effects into sensible and latent heat fluxes, the key variables in the budget. The spectrally decomposed temperature-humidity correlation coefficient is used to quantify the impact of rolls and open cells on the surface energy balance (SEB) closure via entrainment. Once entrainment effects weaken, an improvement in the SEB closure is observed. The independence of this observation on the category and location in the valley, points to the ubiquity of entrainment degrading the SEB closure.

Finally, the dissertation focuses on the smallest scale of CBL convection, the near-surface coherent structures collectively referred to as ramp structures. Since these have not yet been studied over complex terrain, we move beyond the T-REX and Owens Valley, and include flux tower data from CASES-99 (gently rolling topography) and MATERHORN (playa, base of an isolated mountain), to establish the pertinent mechanisms in which complex terrain modifies ramp structural properties. The role of shear and buoyancy contribution to turbulence production is found to be essential in explaining the observed differences in ramp structural properties among the sites. Different generation mechanisms (bottom-up, top-down, turbulent kinetic energy (TKE)-based) are contrasted, by determining the degree of site-wide collapse of normalized ramp structural properties.

This dissertation was typeset in LaTeX using the template provided by the Graduate School of Duke University. The template was made available at https://www.overleaf.com under a Creative Commons CC BY 4.0 license. Minor modifications were made to the original template source code to improve the visual presentation of this dissertation: reduction of page margins, reduction of line spacing, omission of the Abstract page, section header changes.