Based on theory, laboratory experiments and field observations, canyons have long been identified as sites of intense internal waves. Turbulence measurements in Monterey Canyon (MC) and density overturns in Kaoping Canyon (Taiwan), which has similar size and shape, confirm very strong mixing. The processes generating the mixing, however, have not been determined, nor has mixing been measured in enough canyons to assess its variability from one to another. Observations and recent modeling indicate that the internal tide is a major source of energy in Monterey Canyon, coming from offshore as well as being generated locally. In addition, ridges and constrictions along the canyon make it likely that lee waves and hydraulic jumps are also major sources of mixing, as has been inferred for abyssal canyons.
The project will undertake modeling and intensive observations for Monterey Canyon and exploration of nearby Ascension Canyon which is much straighter than Monterey and apparently lacks a significant offshore source for the internal tide. The goals are to: 1) understand how the internal tide propagates in canyons, and in particular how the M2 energy flux is steered around bends and when turns are so sharp that the energy is dissipated, 2) determine whether mixing locations and levels can be predicted by convergences in the along-canyon M2 energy flux, 3) identify and quantify mixing produced by internal wave scattering, lee waves and hydraulic jumps, and 4) assess the effects that canyon shape, smoothness and alignment with offshore M2 sources have on mixing and baroclinic energy by measuring mixing and the internal tide in a relatively short, straight canyon ending on the outer shelf.
Broader Impacts U.S. coasts have dozens of canyons, and they account for almost half of the western continental slope from 45◦N to Alaska. Though they appear to be biologically rich and export both water and sediments, dynamical processes within canyons are too poorly understood to make a general assessment of the importance of canyons or whether they require special representation in large-scale or regional models. One set of measurements cannot resolve all questions about canyon dynamics, but the proposed work should provide qualitative and quantitative bases for assessing mixing in other canyons.. If fully successful, this project may unable the parameterization of some of the mixing or at least be able to point the way to parameterizations. As part of the proposed effort the investigators have also begun talks with regional modelers to incorporate intense canyon mixing into the models and to do sensitivity runs to define how accurately the mixing must be known to predict its effect on flows and structures outside the canyon. Two post-doc will be hired, one at APL/UW and another at the University of Hawaii. Both will participate in all phases of the work for their postdoctoral training. Coordination and collaboration with the regular MBARI sampling of Monterey Bay will benefit this project, and vice versa. Outreach activities will combine earlier canyon measurements and the future measurements.
This project was proposed to: 1. understand how the internal tide propagates in canyons, particularly how theM2 energy flux is steered around bends and turns 2. determine whether mixing locations and levels can be predicted by convergences in along-canyon M2 energy flux 3. identify and quantify mixing produced by internal waves scattering, lee waves and hydraulic jumps 4. assess the effects that canyon shape, smoothness and alignment with offshore M2 sources have on mixing and baroclinic energy The effort involved modeling (led by Carter) moorings (led by Alford and Lien) and intensive observations using SWIMS3, a depth-cycling towed body, and microstructure profilers (led by Gregg). In early 2009, Alford placed McLane Profilers and moored 75 kHz ADCPs in the upper 10 km of Monterey Canyon. During the following April, all of the PIs plus two post docs participated in the intensive measurements with SWIMS3 and microstructure profilers. To examine the structure of velocity, density, and turbulence in the canyon, repeated runs were made along selected cross-sections and along-canyon tracks. Each sequence lasted 25 hours. The moorings were recovered at the end of the intensive phase. Using the Princeton Ocean Model (POM), modeled the internal tide in a larger domain around Monterey Bay than had been used by previous investigators. These gave enhanced fluxes and led to the conclusion that using a large domain with a hydrostatic model gives better estimates of the internal tide than does a non-hydrostatic model over a smaller domain. Examining the internal tide in Monterey Canyon found that the along-canyon flux decreased almost monotonically from 9 MW into the mouth to 1 MW at the Gooseneck Meander, which was the focus of most of the intensive work. This rate of loss implies strong mixing along the canyon. Fluxes were most intense near the bottom over the thalweg. The intensive phase found that the internal tide followed changes in the bathymetry over along-canyon length scales of 5 km and larger but not smaller ones, even one as sharp as the Gooseneck Meander which deflected the flux but did not change its basic direction (Wain et al., 2013). The baroclinic energy flux was half as large on the eastern side of the canyon as on the western side. Measured dissipation near the Gooseneck had the same magnitude as the flux divergence, but in neither Monterey nor Ascension could we produce a credible along-canyon budget with both measurements and model. In the former, flux production is computed as a residual, and in the latter dissipation is the residual. Differences in spatial and temporal resolution of both model and measurements were too large to balance terms in both sets of estimates. Although not formally part of the Monterey measurements, during the mooring cruise Alford deployed two McLane profilers and a 50 kHz ADCP 19 km south of the Mendocino Escarpment. Time series of velocity and shear between 100 and 3,640 m and strain from 1,000 m to 3,640 m showed near-inertial features at all depths rather than the `shadowing' that had been expected from proximity to the escarpment (Alford, 2010). An upward beam of the internal tide was also found, but near-inertial shear exceeded it shear by factors of 2-4 at all depths. Dissipation rates inferred from overturns agreed with internal wave (Gregg-Henyey) scaling at mig-depths but greatly exceeded the scaling in the bottom 1,000 m and between 1,500 and 2,000 m. Overall, how well were the original questions answered? 1. Measurements and modeling did rather well in showing how the internal tide was influenced by channel bends in the upper canyon. Not surprisingly, the basic result is that the internal tide can follow bathymetric twists with horizontal scales longer than that of the tide but is only deflected by shorter features. 2. Location of much of the mixing can be predicted fairly well by the shape of the internal tide but levels only within an order of magnitude. 3. Mixing produced by lee waves and/or hydraulic jumps was identified, but any contribution from internal wave scattering was obscured by larger signals from the internal tide 4. Finding high levels of internal tide flux and mixing in Ascension Canyon leads to the tentative conclusion that smoothness and shape are relatively unimportant, as is alignment with specific offshore sources of the internal tide.
