Dynamics of oceanic western boundary currents at high resolution (CASE)

Western boundary currents play a critical role in the Atlantic's Meridional Overturning Circulations. In the current Met Office models they are poorly resolved and poorly understood. Most of the underlying theory of western boundary currents is based on vertical sidewalls and lateral frictional boundary layers. In contrast, western boundary currents in the ocean flow along a sloping bottom boundary with a bottom boundary layer with relatively stagnant coastal water onshore.

The role of turbulent fluxes in setting the structure of western boundary currents is poorly understood. Critical processes include: suppression of mixing across strong bathymetric gradients, up-gradient momentum fluxes and down-gradient buoyancy fluxes (or downward momentum fluxes via the equivalent eddy form stresses). The western boundary currents are believed to be a region of substantial eddy energy dissipation, both due to being the graveyard of westward propagation eddies (Zhai et al., 2010, Nat. Geosci.) and due to the large eddy energy source through baroclinic instability of the western boundary currents. Improved understanding of these eddy energy sinks is important both for understanding the detailed nature of the eddy-mean flow interactions (Marshall et al., 2017, Geophys, Res., Lett.; Mak et al., 2018, submitted to J. Phys. Oceanogr.) and also the resultant diapycnal mixing within the boundary currents, important for short-circuiting of the AMOC and hence ocean heat transport in the Atlantic.

There is considerable evidence that boundary layer separation is related to deceleration of flow ("the external stream") just outside the viscous boundary layer (e.g., Marshall and Tansley, 2001, J. Phys. Oceanogr.). However, in the ocean this should be deceleration of the bottom boundary layer along sloping sea floor. A paradigm shift in our approach to the separation problem is required, moving away from separation from a vertical sidewall to separation of a surface-intensified current from a sloping bottom boundary. Also critical to the separation problem is the role of vortex stretching as the Gulf Stream leaves the North Atlantic shelf and crosses the DWBC which descends by about 800m as it passes beneath.

There are a myriad of sub-mesoscale processes at play in the surface mixed layer of the boundary currents. These include: symmetric and baroclinic instabilities of the boundary current and its mesoscale meanders/eddies, relative stress effects - the Ekman pumping associated with shear in ocean currents being of particular importance, Ekman transport fluxing cold water over warm leading to destabilising the surface mixed layer on one flank of the boundary current. The recent work of Bell (2018,Q. J. Roy. Meteor. Soc.) provides a unified framework for analysing the myriad of instabilities within the western boundary currents and eddies.

Progress in numerical models and computational resources mean that it is now feasible to run limited area western boundary current models at extremely high spatial resolution, resolving both the mesoscale and (for short time windows) the sub-mesoscale. These models should be able to represent the steep bathymetry reasonably satisfactorily and provide insight into the dynamics governing the path of the Gulf Stream around Grand Banks, which is a known source of error in SST in the Met Office models.

We propose to build on the methodology of Gelderloos et al. (2011, J. Phys. Oceanogr.) and set up a very high-resolution model of the Florida Current and separated Gulf Stream (including, potentially Grand Banks). The bathymetry will assume both idealised, semi-realistic and realistic configurations, and a return channel and sponges will be used to return the fluid equatorward and establish the inflow/outflow boundary conditions in a computationally efficient manner (see Gelderloos et al. for more details). A variety of diagnostics of vorticity tendencies, and energy conversions will be used.