Implications and Future Work

These results are not necessarily specific to Hawaii, or even ridge topographies. Internal tides are generated at many continental shelf breaks, and combined with supercritical or critical continental slopes, downgoing tidal beams are likely to interact with the topography. The relative importance of tidally-driven shear instability and convective mixing, as described by Gemmrich and van Haren (2001), remains to be determined. Given the large fraction of continental slope worldwide where downgoing tidal beams are potentially present, we speculate that these beams may account for a significant fraction of the mixing occurring at deep oceanic boundaries. This mechanism has received less attention than the critical reflections of incident internal tides. Furthermore, the exact position and orientation of the beam relative to the slope is essential to determine the associated mixing, as was illustrated by the differences between the two moorings. How much of the beam is dissipated locally over the slope may be important for closing the overall baroclinic tidal energy budget at the ridge.

We note that our study site is below the main thermocline, within a depth range that has been a focus of abyssal mixing studies (Munk and Wunsch, 1998). The preliminary HOME results from Kaena Ridge encourage speculations that tidally-driven mixing mechanisms may differ above and below the main thermocline. The two moorings DS and DN were deployed at the same depths, on opposite sides of Kaena Ridge. The water mass at the two moorings is the same as shown by the TS diagrams and the time-averaged vertical temperature profiles. However, the stratification at the south mooring DS is weaker than at the north mooring DN (Figure 2.3 and section 2.2). Likewise, the estimated turbulent mixing at mooring DS is found to be one order of magnitude larger than at DN. This leads us to suggest that the more energetic mixing observed at mooring DS leads to a more well-mixed water column, and locally weaker stratification. Exchanges with the ocean interior may result in the export of this boundary mixed water along constant density surfaces (e.g. via stirring by mesoscale variability, Munk and Wunsch (1998)), and replenishment by water with the background ocean stratification. Clearly, the exchange rate with the open ocean is a crucial factor in determining how mixing modifies the properties of the fluid at the boundary. Our experimental setup was limited in space, and the lack of lateral coverage along a density surface did not allow us to investigate this rate of exchange.

The HOME experiment involved numerous platforms, instruments and investigators with the goal of understanding the overall tidal energy budget and the associated mixing around the Hawaiian Ridge, and at Kaena Ridge in particular. Preliminary results provide a detailed picture of the internal tide structure at and away from the ridge crest. However, this study identifies the need for additional work to gain a better understanding of the deep boundary mixing at the Kaena Ridge and elsewhere along the Hawaiian Ridge. The following tasks remain :

1. Observations over a greater depth range above the slope. Our current measurements did not extend far enough from the bottom to better assess the internal tide and NIW structure away from the bottom, in relation to our temperature observations. Current shears could not be reliably measured due to uncertainties regarding the effects of side lobe reflections. More reliable current and current shear information over a greater depth range will provide more details on the mechanisms that lead to energetic mixing.
2. Observations closer to the bottom. Due to mechanical constraints on the mooring, no instrument could be deployed below $22 mab$ . We therefore missed a significant fraction of the frictional bottom boundary layer.
3. Numerical models with improved resolution. The agreement between existing numerical models of the area and our observations is poor. This is likely due to the complex bathymetry that varies over scales much smaller than the model resolution. A better model resolution may help reconcile the model with the data, and allow the extrapolation of sparse temporal and spatial observations to the entire ridge.
4. Mixing associated with NIW. NIWs have been observed throughout the ocean. At our north flank site, we observe strong temporal variability of the NIWs, and mixing associated with the superposition of internal tides and NIWs. It is unclear how much mixing can be attributed to the NIWs alone. An experiment in a region subject to high NIW and limited internal tide energy (e.g., away from the shallow ridges identified as internal tide generation sites) may answer this question.
5. We could not positively confirm the link between surface wind generation to the north and NIWs events on the slope, nor could we calculate the modal content of these waves. Long term, full depth, monitoring of the current and density field at open ocean sites like station ALOHA could offer answers to these questions.