Westward energy transmission in fast barotropic Rossby waves and slow baroclinic modes (nonlinear Rossby waves/mesoscale eddies/β-plumes) is the classic mechanism of boundary current production. Western energy intensification, even in nonlinear circulations, is related to basin-wide down-gradient eddy potential-vorticity flux, which is identical in simple ocean models to global production of potential enstrophy.
However, inertial recirculation gyres, large and small, interact with boundary current cores, deep and shallow. These inertial holding tanks' between boundary and mid-ocean make the dynamics more diffusive than expected, more elliptic' than hyperbolic, slower in net meridional tracer and heat-transport than expected, and more sensitive to near-western-boundary atmospheric forcing. Recirculation gyres to some extent mimic the broad boundary currents of low-resolution circulation models, with boundary-current tracer plumes rapidly expanding eastward, in observations and in our HYCOM-model simulations (Xu, Rhines, Chassignet & Schmitz, 2015).
Interaction of wind forcing with buoyancy forcing/mixing connects ocean gyre transports with water-mass transformation and meridional overturning circulations (MOCs). This interaction is particularly distorted in coarse-resolution models. Without mesoscale mixing, the sharply defined surface heat-storage fronts and deep mixing sites which guide and enhance atmospheric storm tracks, and dominate the zonally integrated MOC, are also missed. Hot-spots of mesoscale activity include the switch-yard', near the Grand Banks of Newfoundland, where the Gulf Stream extension splits into three streams: the North Atlantic Current, the southern recirculation gyre and the Atlantic subtropical gyre.
Deep western boundary currents also experience massive splitting, broadening and recirculation in deep basins, for example in the Labrador Sea, Newfoundland Basin, Iceland Basin, and Brazil Basin. In our HYCOM model simulations this is caused by a threefold combination of: PV conservation over broadening topographic slopes (as argued by Stommel and Arons), PV mixing at transitions forced by abrupt topographic features, and form-stress transmission of upper-ocean eddy/jet activity downward to the abyss.
In models, and likely in reality, diapycnal mass transport and PV stirring concentrates in western boundary currents, shallow and deep. Water-mass transformation essential to the global MOC is thus concentrated there, and at analogous sites around the Southern Ocean. PV stirring and thermodynamically active mixing work together in connecting the lateral circulation gyres with the MOC.