This study aims to look into the effect of small-scale surface heterogeneity on the mixing properties within the stable boundary layer (SBL) and the impact of different configurations of the heterogeneity on this mixing. The motivation for this research is to better understand the turbulent mixing phenomena in the SBL and to be able to represent the mixing more accurate in weather prediction models. In the current models, this mixing is highly uncertain or even missing. It is likely that this mixing is driven by thermal memories that are spatially distributed over the land-surface in the form of surface heterogeneity. These surface heterogeneities (e.g. ditches or roads) effectively store heat during the day and then turn into heat sources, driving turbulent mixing, during the night. A simulation with surface heat sources in an idealized SBL is setup and four case studies are performed. The simulations show that the heat sources clearly influence the mixing in the SBL. A buoyancy plume is formed above a heat source. The plumes of two heat sources are attracted towards each other. At smaller distances the plumes even merge and form a single plume. In the three-dimensional simulations, the distance between the heat sources has an influence on their effect on the SBL. At small distances, the kinetic energy in the system decreases with increasing distances between two heat sources. At larger distances this decrease appears to level off. Apparently, the bending and merging of the plumes allows turbulent length scales to increase, causing the effect of the buoyancy sources to be larger. To verify these simulation results, a field experiment is conducted. An infrared camera pointing at a vertical high-emissivity cloth was able to capture the two-dimensional (vertical and horizontal) air temperature pattern above heat sources. The merging effect between two plumes is also observed in this field experiment. At small distances a single merged heat plume was visible above two heat sources and at larger distances two separated heat plumes are formed.

Wind machines are used to prevent or mitigate the adverse effects of night frost in spring. These frost events occur during clear-sky, low-wind nights in which a thermal inversion builds up from the surface. The machines work by mixing and transporting warm air from aloft downward which consequently erodes the thermal inversion. Various studies have been conducted regarding this frost protection method, which have resulted in empirical regression models that relate affected area and temperature enhancement (i.e. the performance) with inversion strength. In this study, we assessed heightdependent temperature responses and the sensitivity of wind machine performance to various physical processes. Both of which have not been studied thoroughly in earlier works. In this regard, a large field experiment was conducted and experimental analysis was augmented with sensitivity studies using turbulent resolving Large Eddy Simulations (LES). Experimental observations showed that the temperature response strongly depends on the radial distance to the fan and the height above the surface. In agreement with previous studies, the wind machine was able to achieve rotation-averaged temperature increases of up to 50% of the inversion strength (≈3 K) in an area of 3-5 ha at 1 m height. Furthermore, it was observed that low-speed ambient winds (<1 m/s) can cause strong upwind-downwind asymmetries in the protected area, the downwind area being larger. The LES model, inspired by the field experiment, showed similar spatial temperature responses as compared to observations. Sensitivity studies using a simpli_ed case showed that the a_ected area strongly increased for slower axial rotation times (ranging from 3 to 6 min) while the temperature enhancement stayed relatively constant. Furthermore, variation of the horizontal tilt angle showed that, in our model, temperature enhancement was maximized between 8º and 16º. Presumably, those angles, corresponding to near horizontal ow, are very effective in generating strong shear layers which in turn generate Kelvin-Helmholtz like instabilities. These processes increase mixing and vertical transport of heat. Finally, analysis on the ambient wind showed that, in agreement with observations, strong upwind-downwind asymmetries in the affected area exist.