Mixing at Mooring DN

We have argued that mixing along the flanks of the Kaena Ridge, at least at the mooring sites DS and DN, is dependent on the baroclinic tide as a source of mechanical energy. The internal tide takes the form of downward propagating beams at this depth in numerical model runs, and in observations close to the ridge crest. We infer that the orientation of the beam relative to the sloping bottom is a key consideration for mixing. In the case of DN, the physical separation of the beam from the slope appears to be an issue. The resulting $M_2$ tidal currents are $50\%$ weaker at DN than at DS, where the beam appears to brush along the slope, and as a consequence mixing rates and dissipation are an order of magnitude smaller than at DS. We have identified both shear (downslope flow mixing) and convective (flow reversal mixing) instabilities as the triggers for mixing at DS. We do not find evidence for convectively driven mixing at DN, presumably due in part to the separation of the beam and slope. Tidal strain does appear to play a role in DN mixing events, consistent with a shear instability. We lack sufficient coverage of the current field with depth to make conclusive statements about the role of current shear. Whereas the maximum tidal strain occurred during the phase of strong downslope flows at DS, at DN the maximum strain occurs during maximum upward displacement of the isotherms when the flow reverses from up to downslope. We attribute the difference in phase of maximum strain between the two sites again to the details of the tidal beams. At DS, the strain phase is set by the kinematics of a downward propagating internal wave, which in turn will depend on the vertical wavenumber structure of the internal tide. At DN, we observe a decay of the tidal signal toward the boundary. This apparently causes the tidal strain to be set by the stronger tidal motions farther from the boundary; maximum strain occurs when the upper isotherms are displaced farthest from the boundary.

Returning to the issue of mixing parametrizations, the contrasts between DN and DS indicate the importance of specifying the altitude of the tidal beam above the topography. This in turn is likely to be sensitive to changes in stratification as the beam propagates downward from its generation region.

DN also differs from DS in the amount of near-inertial to diurnal band energy observed. The event-like behavior of this energy in the near-diurnal band suggests that this variability is not associated with locally generated diurnal internal tides. Moreover, diurnal internal waves generated near the ridge crest would have a shallow angle of downward propagation, making it unlikely that they would encounter the deep super-critical ridge flanks.

The smaller amount of near-inertial energy at DS is consistent with mooring DS being shadowed by the Kaena Ridge from the equatorward propagating internal near-inertial waves. These near-inertial waves appear to interact with the supercritical topography to produce vertical displacements near the bottom larger than for a freely propagating NIW. These vertical displacements near the bottom superpose with the displacements created by the semidiurnal beam further up to create periods of high strain, which presumably are prone to overturning and mixing under the action of a velocity shear. Here we observe energetic motions near the slope linked to incident, intermittent, near-inertial waves. The topographic slope around our observation site is supercritical for these waves. At other locations along the north side of the ridge, where the slope is critical for near-inertial to diurnal waves, mixing related to critical wave reflection is likely to play a role in the overall mixing budget of the ridge. On the south side of the ridge however, regardless of the slope, mixing linked to NIW is likely to be significantly lower due to sheltering effects.