Self-stratifying turbidity currents

Abstract

Turbidity currents, submarine currents driven by the excess density of suspended particles, carry vast quantities of sediment and nutrients from the continental margins to the deep ocean. Due to their vast scale and extreme aspect ratio, simplified depth-or channel-averaged models are required to capture their dynamics. The inclusion of vertical structure (profiles of velocity, depth, and turbulent kinetic energy) and levee overspill in these models is in its infancy, and we demonstrate their importance in this study. We present a new channel-averaged model that supports an arbitrary evolving vertical structure and a compatible closure for the levee overspill. By examining this new model, connections between the vertical structure and the internal turbulent processes are revealed. Additionally, we find new requirements for models of the front of the current, demonstrating a connection between the vertical structure of the body and the mixing and erosion in the head.In our new framework we build a full 'proof of concept' model to illuminate the substantial effect that vertical structure and levee overspill have on the current. In this model, the vertical structure changes as part of the evolution of the current: it self-stratifies. Quasi equilibrium solutions are constructed, where the entrainment causes deepening. These solutions are not stable, but rather weekly unstable and connected to a slowly evolving manifold: we expect most environmental currents to evolve within such a manifold. Equilibrium solutions are not stable either, the levee overspill removing the dilute, low momentum fluid, which rejuvenates the flow, and this can cause a positive feedback loop where the fluid becomes increasingly concentrated. Finally, we present some simulations of the Congo system, for the first time capturing a current that travels out to the end of the levees in a channel-averaged model.