Pierre-Yves Passaggia (University of North Carolina at Chapel Hill)
New insights on turbulent horizontal convection from experiments and simulations
Horizontal convection is considered as a simple model to study the influence of heating, cooling and freshwater fluxes at the ocean surface on the Meridional Overturning Circulation (MOC). Recent work on the stability of horizontal convection and its influence on flow transitions at the onset of turbulence will be presented. The transition to turbulence is investigated using a combination of stability analyses, direct numerical simulations and laboratory experiments. Past a critical value, the baseflow becomes unstable first to three-dimensional perturbations and at higher Rayleigh numbers, to two-dimensional instability. Increasing further the Rayleigh number, the flow exhibits a supercritical bifurcation which substantially modifies the core of the flow from a deep to a shallow circulation. The results of companion DNS and laboratory experiments confirm this scenario leading to a retroflection of the plume region. Full reflection is observed whether horizontal convection is driven by heat or salt. The bottom of the domain progressively fills with stagnant heavy fluid, a process which likely explains the shutdown of the core circulation first hypothesized by Sandstrom (1908) and more recently theorized by Paparella & Young (2002). However for large Rayleigh numbers, the flow exhibits higher heat/freshwater uptakes than what was first theorized by Rossby (1965), suggesting that horizontal convection may play an important role in the MOC. The scalings of horizontal convection are compared with the recent theory of Shishkina et al. (2016) which was adapted for horizontal convection from Rayleigh-Benard theory.