Turbulence Structure and Mixing in Strongly Stable Boundary-Layer Flows over Thermally Heterogeneous Surfaces

Abstract

AbstractDirect numerical simulations (DNS) at bulk Reynolds number Re $$=10^4$$ = 10 4 and bulk Richardson number Ri $$=0.25$$ = 0.25 of plane Couette flow are performed with the results used to analyze the structure and mixing intensity in strongly stable boundary-layer flows. The Couette flow set-up is used as a proxy for a real-world stable boundary layer flow with surface thermal heterogeneity. Along the upper and lower walls, the temperature is either homogeneous or varies sinusoidally, but the horizontal-mean surface temperature is the same in all cases. Over homogeneous surfaces, the strong stratification always quenches turbulence resulting in linear velocity and temperature profiles. However, over a heterogeneous surface turbulence survives. Molecular diffusion and turbulence contribute to down-gradient momentum transfer. The total (diffusive plus turbulent) heat flux is directed downward, but its turbulent contribution is positive, i.e., up the mean temperature gradient. Analysis of covariances of velocity and temperature, their skewness, and the flow structure suggests that counter-gradient heat transport is due to quasi-organized cell-like vortical motions generated by surface thermal heterogeneity. These motions transfer heat upwards similar to their counterparts in highly convective boundary layers. Thus, the flow over heterogeneous surface features local convective instabilities and upward eddy heat transport, although the overall stratification remains stable with downward mean heat transfer. The DNS results are compared to the results from large-eddy simulations of weakly stable boundary layers (Mironov and Sullivan in J Atmos Sci 73:449–464, 2016). The DNS findings corroborate the key role of temperature variance in setting the structure and transport properties of stably stratified flow over heterogeneous surfaces, and the importance of third-order transport of the temperature variance.