Convection is a process through which changes in buoyancy drive fluid motion. Convection is a ubiquitous process and can be observed everywhere from boiling a saucepan to air conditioning. Convection is equally common in the ocean, and occurs at (extremely) small length and time scales. However, this small-scale turbulent convection can also drive large-scale circulation and dynamics in the ocean, impacting the ocean’s response to climatic changes. The interplay between buoyancy-driven convection and other sources of energy to the ocean, including winds and tides, also remains uncertain.
The primary bottleneck inhibiting research into ocean convection is the availability of computational resources. Current large-scale ocean models are able to resolve flow processes from ∼10 km to ∼104 km in scale. Processes which exist at length scales smaller than ∼10 km are approximated (parameterised) in these ocean models. Therefore, convection, which occurs at the millimetre scale, is largely approximated.
The aim of this work is to investigate the impact of convection on large-scale ocean dynamics and to interrogate the accuracy of convective parameterisations in large-scale ocean models. In order to accurately represent convection, we use a first-of-its-kind numerical model of the Southern Ocean, known as a Direct Numerical Simulation (DNS). The DNS is high-resolution and resolves all scales of flow, from small-scale turbulence to basin-scale dynamics. With a focus on the Southern Ocean, we use the DNS to address long-standing questions about the interactions between small-scale ocean processes and large-scale dynamics, as well as the interplay between various sources of energy into the ocean.