The dissipation of surface tides is thought to be one of the major contributors to small-scale turbulence and mixing in the abyssal ocean (below $>2000~\mathrm{m}$), which is an essential component of the large-scale deep circulation. However, the energy of surface tides is first converted into internal tides and these waves can be transformed through several processes until they break and generated mixing. Over the course of two months in 2015, we made intensive in-situ observations to determine the energy fate of the incident internal tide on the continental slope off Tasmania, Australia. Over the steep slope in the southeast region, the incoming tide mostly reflects back to the deep ocean. We observed a complex interference pattern that is associated with both wave reflection and the alongshore propagation of a slope wave. During our experiment, the incoming internal tide exhibits a predictable spring-neap cycle. However, the mismatch between the observed and predicted phases of the fortnightly variability indicates that local waves are also important in the internal-tide wave field. The irregular northeastern continental slope off Tasmania leads to large scattering of the incoming internal tide. Tidal flow over a steep bump sets nonlinear lee-wave like phenomena on the offshore and onshore sides of the topography. Through these processes, the semidiurnal tide drives most of the observed energy dissipation rate. However, the temporal variability is complex on weekly timescales and may be associated with the interference between remote and local tides. Nevertheless, idealized theoretical predictions roughly agree with the observations and supports that wave scattering is responsible for the near-bottom turbulence. Despite the high reflectivity off southeastern Tasmania, small-scale bathymetric corrugations modify the near-bottom tidal flow, which is in turn coupled with the background current. Large cross-shore tidal and subtidal velocities are observed within the trough of the corrugation. The low-frequency horizontal current is $O(10^{-2})~\mathrm{m}~\mathrm{s}^{-1}$, which may be associated with large vertical velocities of up to $O(10^{-3})~\mathrm{m}~\mathrm{s}^{-1}$. The temporal variability of the near-bottom subtidal current is correlated with turbulence estimates, which do not have a clear fortnight modulation. However, spatially and temporally resolving measurements show that the sloshing of the internal tide is crucial for setting the near-bottom turbulence. The vertical divergence of the semidiurnal Reynolds stress explains the temporal variability of turbulence within the corrugation trough and suggests an internal-tide driven onshore subtidal flow. The onshore acceleration is likely balanced by an offshore pressure gradient, and the associated density field is responsible for the decrease in magnitude of the overlying alongshore flow as it approaches the top of the corrugations.

Internal tides and currents interacting with sub-surface topography such as continental slopes, islands and seamounts give rise to turbulent processes. The subsequent mixing of waters of different densities is important for maintaining the global overturning circulation. This thesis investigates the turbulent processes occurring at two different topographic features; a steep headland in the eastern Pacific and a continental slope canyon in the north-west Atlantic.

Tidal currents and mean-flows interacting with headlands and islands give rice to turbulent wakes. In Chapter 1, surveys in the wake of a steep headland in both tidal and mean flow regimes reveal turbulent wakes driven by shear and vorticity in the flow. Recent theoretical literature has suggested that turbulence at steep topographic features may drive upwelling necessary to close the abyssal overturning circulation. In Chapter 2, a novel near-bottom dye release provides the first direct evidence of upwelling within a steep canyon and suggests that current 1-d models of near-bottom processes are insufficient to understand the importance of bottom boundary upwelling processes. In Chapter 3, this work is extended using a suite of moorings which observe convergence and divergence in the along-canyon direction indicative of exchange between the boundary and interior, a process that is not included in 1-d models.

This thesis examines two phenomena that occur over the continental shelf and generate mixing in stratified coastal systems.

In Chapter 1, we examine the generation, propagation, and dissipation of nonlinear internal waves (NLIW) in sheared background currents on continental shelf offshore of Washington state. At the shelf break, semi-diurnal (M2) energy flux is onshore and the incident M2 internal tide is partially reflected and partially transmitted. NLIW appear at an inshore mooring at the leading edge of the onshore phase of the baroclinic tide, consistent with nonlinear transformation of the shoaling internal tide as their generation mechanism. Of the M2 energy flux observed at the eastern extent of the generation region approximately 30% goes into the NLIW observed inshore. NLIWs are tracked into shallow (30-40 m) water, where a vertically sheared, southward current becomes strong. As train-like waves propagate onshore, wave amplitudes of 25-30 m and energies of 5 MJ decrease to 12 m and 10 kJ, respectively. The observed direction of propagation rotates from 30◦ N of E to ∼ 30◦ S of E in the strongly sheared region. Bore-like waves do not

rotate and do not appear to lose energy within the observed ranges.

Chapters 2-4 are focused on the dynamics of internal tides in the La Jolla Canyon System (LJCS) a steep, shelf-incising submarine canyon off the coast of San Diego, California.

In Chapter 2 we present the results from a short student-led study at the steep head of the canyon system. Baroclinic energy flux is oriented up-canyon and decreases from 182 ± 18 W m−1 at the canyon mouth to 46 ± 5 W m−1 near the head. Variance is dominated by the semi-diurnal (M2) tide which is partially standing. Moving up-canyon, the relative importance of M2 decreases and its higher harmonics are needed to account for a majority of the observed variance, indicating steepening. Steep internal tides cause large isopycnal displacements (∼50 m in 100m water depth) and high strain events. These events coincide with enhanced dissipation of turbulent kinetic energy at mid-depths.

In Chapter 3, we present results from spatial surveys of the LJCS made during a more extensive follow-up experiment. M2 energy flux was oriented up-canyon and contained mostly in mode 1. High values of dissipation occurred near the canyon head at mid-depths associated primarily with high strain. Modal analyses suggest that LJCS was reflective to the mode 1 M2 tide. Higher modes were found to be progressive, and smaller energy fluxes associated with them were oriented down-canyon, suggesting that incident low-mode waves were both back-reflected and scattered. Flux integrated over a canyon cross-section was always onshore, but generally decreased moving shoreward (240 kW to 5 kW), with a jump in flux occurring on a section just inshore of the canyon’s major bend due to reflection of incident waves from the steep sidewalls of the meander. Flux convergence from canyon mouth to head was balanced by the volume integrated dissipation observed. By comparing simple energy budgets from all canyons with sufficient observations, a similar balance is found for most canyons suggesting that much of the elevated turbulence in these canyon systems is driven by internal tide dynamics.

In Chapter 4, we use long, coincident time series from the same experiment to examine temporal variability of the internal wave field, internal tide dynamics, and associated mixing in the LJCS. Results from Chapter 3 are confirmed: the M2 internal tide dominates the signal in both velocity and isopycnal displacement and it is partially standing throughout the measurement period; near the canyon mouth, dissipation is bottom-enhanced and occurs during up-canyon flow periods, while near the canyon head dissipation is elevated during high strain events at mid-depths and time-averaged dissipation is elevated throughout the water column.

At the canyon head spring-neap cycles in depth-integrated energy, energy flux, stratifica- tion, and dissipation occur at times throughout the year. Depth-integrated chlorophyll fluorescence also demonstrates spring-neap variability; maxima in [Chl] lag behind maxima in M2 internal tide energy by 2-3 days. Phase offsets between the surface tide and harmonic fits to isopycnal displacement (η) are relatively constant, suggesting that the internal tide incident at the canyon mouth is generated nearby. At the head of the canyon the skill of a harmonic fit to η over a 90-day fit window is 49.7% – very high when compared to most other coastal mooring records that have been similarly assessed and indicating surprisingly high stationarity of the internal tide in the canyon. Long time series within canyon systems are rare, and this result could motivate an investigation into whether the skill of internal tide predications may be useful near canyon systems or steep topographic features more generally, and whether such regular tidal motions foster significant enhanced productivity nearby.

The Arctic Ocean is characterized by a halocline rather than a thermocline, so that fresh water is found at the surface, with saltier (and often warmer) water beneath. In the western Arctic, the Pacific and Atlantic Oceans provide sources of oceanic heat. The year-round Pacific Summer Water (PSW) temperature maximum is found 40-80 m beneath the sea surface, while Atlantic Water (AW) is generally found deeper than 200 m. Both the Atlantic and Pacific source waters are warming, and the heat content of PSW has nearly doubled over the last three decades.

The broad goals of this disseration are to identify the processes that regulate the release of heat out of these two water masses, to quantify the heat transported upwards where it may affect sea ice, and to understand how these processes may evolve in a changing Arctic Ocean. Microstructure data collected during the 2015 ArcticMix and 2018 Stratified Ocean Dynamics of the Arctic (SODA) process cruises provide a unique opportunity to observe the small-scale mixing associated with subsurface heat in the Canada Basin.

Multiple processes play a role in setting turbulence rates in the western Arctic. Widespread double diffusive convection results in small upwards heat fluxes out of the AW. We investigate the possible influence of observed wind-generated near-inertial internal waves on AW turbulent heat fluxes and find that the impact is minimal, with AW heat fluxes consistently low due to a number of factors inhibiting the creation and vertical propagation of internal waves. We additionally investigate the relative importance of mixing processes associated with anomalously warm PSW intrusions. We find double diffusive convection results in high upwards heat fluxes out of two separate warm intrusions, one warm-core eddy observed in the 2015, and one slope current intrusion observed in 2018. We document a relationship between the vertical complexity of warm intrusions and the total heat flux out of them, with more complex vertical structure corresponding to increased overall mixing. This relationship is supported by 12 microstructure surveys that sampled PSW conducted across both 2015 and 2018.

Cycles of heat transfer between the tropical ocean and atmosphere help drive some of the Planet’s most important modes of climate variability. Subsurface turbulent heat fluxes (Jq) that redistribute ocean heat are known to mediate ocean-atmosphere dynamics at multiple timescales. However, a relative scarcity of observations has slowed down the inclusion of Jq in dynamical frameworks and theories of coupled tropical climate. This dissertation includes three chapters that use in-situ ocean measurements to explore the relations between atmospheric forcing, Jq, and other oceanic variables. Two case studies give special attention to the immediate and long-term ocean response to tropical cyclones (TCs), while a third chapter uses year-long records of Jq to explain their relation to the Asian Monsoon, and interpret their impacts on the seasonality of ocean temperatures in the Western Pacific Warm Pool (WPWP). Overall, the results highlight the importance of atmospheric extremes in regulating Jq and ocean heat uptake by the WPWP and by cold water masses that lay beneath it. Most attention is given to TCs, the near-inertial internal waves (NIWs) they generate, and how NIWs allow TCs to transform ocean currents and thermodynamics for weeks after the storms have dissipated. Besides TCs, it is also emphasized that large scale dynamics can also give rise to atmospheric extremes in which background conditions cause dramatic seasonal shifts in Jq and ocean heat uptake.