Thermohaline intrusions are formed by lateral interleaving motions across ocean fronts. Interleaving is thought to be driven by buoyancy forces arising from fluxes of heat and salt by double diffusion (i.e., salt fingering or diffusive convection). Most double-diffusive interleaving models apply only to fronts that are barotropic. However, many ocean fronts are baroclinic, with vertical shear and horizontal density gradients. This thesis investigates the dynamics of double-diffusive interleaving in baroclinic ocean fronts.
A new theoretical model of double-diffusive interleaving is developed. It is found that intrusions that slope in the along-front direction will be deformed by background horizontal and vertical shear. As a result, the along-front intrusion slope will be reduced in baroclinic fronts. It is found that horizontal density gradients change the stratification felt by intrusions. Importantly, if the intrusions slope between horizontal and isopycnal surfaces along the front, intrusive motions will be driven by baroclinicity as well as double diffusion.
The theory is applied to two test cases. The first is a Mediterranean salt lens, or Meddy. In the lower part of the Meddy, which is stratified appropriately for salt fingering, the observed interleaving slopes are found to be consistent with the salt-finger form of double-diffusive interleaving. In the upper part of the Meddy, which is stratified appropriately for diffusive convection, the intrusion slopes are consistent with the diffusive-convection form of double-diffusive interleaving. In both cases, it is found that the along-front slope is significantly reduced as a result of background shear.
The second test case is a front in the Arctic Ocean. Here, the background stratification is not appropriate for either type of double diffusion, so it is not clear which form of interleaving should occur. The intrusion slopes are found to be consistent with the salt-finger form of double-diffusive interleaving. The intrusions slope between horizontal and isopycnal surfaces, suggesting that they are driven by baroclinicity as well as double diffusion.
Brian's REAL web page!!