Submarine canyons incising the continental shelf and slope are hot spots for topography-internal wave interactions, with elevated dissipation and mixing contributing to regional transport and biological productivity. At two Barkley Canyon sites (the continental slope below the shelf-break, and deep within the canyon), four overlapping years of horizontal velocity time-series data are used to examine the effects of irregular topography on the internal wave field. Mean currents are topographically guided at both sites, and in the canyon there is an inter-annually consistent, periodic (about a week) up-canyon flow (-700 to -900 m) above a near-bottom down-canyon layer. There is elevation of internal wave energy near topography, up to a factor of 10, 130 m above the slope, and up to a factor of 100, 230 m above the canyon bottom. All bands display weak inter-annual variability, but significant seasonality. Sub-diurnal and diurnal flows are presumably sub-inertially trapped along topography, and the diurnal band appears to be forced locally (barotropically). Both sites have high near-inertial energy. At the slope site, near-inertial energy is attenuated with depth, while in the canyon it is amplified near the bottom. Both sites show intermittent near-inertial forcing associated with wind events, downward propagation of high-mode internal waves, and the seasonal mixed-layer depth, though fewer events are observed in the canyon. Free semidiurnal internal tides are focused and reflected near critical shelf-break and canyon floor topography, and appear to experience both local and remote (baroclinic) forcing. The high-frequency internal wave continuum has enhanced energy near bottom at both sites (up to 7 times the open-ocean Garrett-Munk spectrum), and inferred dissipation rates increasing from a background of less than 10^-9 W kg^-1 and reaching 10^-7 W kg^-1 near topography. Dissipation is most strongly correlated with the semidiurnal (M2) constituent at both sites, with secondary contributions from the sub-diurnal (Sub_K1) band on the slope, and the near-inertial (NI) band in the canyon. Power laws for these dependencies are dissipation ~ M2^0.83 + Sub_K1^0.59 at the slope, and dissipation ~ M2^1.47 + NI^0.24 in the canyon. There is evidence in spectra of a near-buoyancy frequency build-up of energy correlated with high-frequency continuum variability, with a power law fit of 'shoulder' power ~ dissipation^0.34 that is independent of site topography. Though some general results are expected from observations at other slope and canyon sites, the greater temporal extent of these data provide a uniquely long-term evaluation of such processes. / Graduate

An alternative formulation of the governing equations of a dynamical system, called the multi-scale localized perturbation method, is introduced and derived for the purpose of solving complex geophysical flow problems. Simulation variables are decomposed into background and perturbation components, then assumptions are made about the evolution of these components within the context of an environmental flow in order to close the system. Once closed, the original governing equations become a set of one-way coupled governing equations called the "delta form" of the governing equations for short, with one equation describing the evolution of the background component and the other describing the evolution of the perturbation component. One-way interaction which arises due to non-linearity in the original differential equations appears in this second equation, allowing the background fields to influence the evolution of a perturbation. Several solution methods for this system of equations are then proposed. Advantages of the delta form include the ability to specify a complex, temporally- and spatially-varying background field separate from a perturbation introduced into the system, including those created by natural or man-made sources, which enhances visualization of the perturbation as it evolves in time and space. The delta form is also shown to be a tool which can be used to simplify simulation setup. Implementation of the delta form of the incompressible URANS equations with turbulence model and scalar transport within OpenFOAM is then documented, followed by verification cases. A stratified wake collapse case in a domain containing a background shear layer is then presented, showing how complex internal gravity wave-shear layer interactions are retained and easily observed in spite of the variable decomposition. The multi-scale localized perturbation method shows promise for geophysical flow problems, particularly multi-scale simulation involving the interaction of large-scale natural flows with small-scale flows generated by man-made structures. / Master of Science / Natural flows, such as those in our oceans and atmosphere, are seen everywhere and affect human life and structures to an amazing degree. Study of these complex flows requires special care be taken to ensure that mathematical equations correctly approximate them and that computers are programmed to correctly solve these equations. This is no different for researchers and engineers interested in studying how man-made flows, such as one generated by the wake of a plane, wind turbine, cruise ship, or sewage outflow pipe, interact with natural flows found around the world. These interactions may yield complex phenomena that may not otherwise be observed in the natural flows alone. The natural and artificial flows may also mix together, rendering it difficult to study just one of them. The multi-scale localized perturbation method is devised to aid in the simulation and study of the interactions between these natural and man-made flows. Well-known equations of fluid dynamics are modified so that the natural and man-made flows are separated and tracked independently, which gives researchers a clear view of the current state of a region of air or water all while retaining most, if not all, of the complex physics which may be of interest. Once the multi-scale localized perturbation method is derived, its mathematical equations are then translated into code for OpenFOAM, an open-source software toolkit designed to simulate fluid flows. This code is then tested by running simulations to provide a sanity check and verify that the new form of the equations of fluid dynamics have been programmed correctly, then another, more complicated simulation is run to showcase the benefits of the multi-scale localized perturbation method. This simulation shows some of the complex fluid phenomena that may be seen in nature, yet through the multi-scale localized perturbation method, it is easy to view where the man-made flows end and where the natural flows begin. The complex interactions between the natural flow and the artificial flow are retained in spite of separating the flow into two parts, and setting up the simulation is simplified by this separation. Potential uses of the multi-scale localized perturbation method include multi-scale simulations, where researchers simulate natural flow over a large area of land or ocean, then use this simulation data for a second, small-scale simulation which covers an area within the large-scale simulation. An example of this would be simulating wind currents across a continent to find a potential location for a wind turbine farm, then zooming in on that location and finding the optimal spacing for wind turbines at this location while using the large-scale simulation data to provide realistic wind conditions at many different heights above the ground. Overall, the multi-scale localized perturbation method has the potential to be a powerful tool for researchers whose interest is flows in the ocean and atmosphere, and how these natural flows interact with flows created by artificial means.

Thesis (S.M. in Oceanographic Engineering)--Joint Program in Ocean Engineering/Applied Ocean Physics and Engineering (Massachusetts Institute of Technology, Dept. of Mechanical Engineering; and the Woods Hole Oceanographic Institution), 2006. / Includes bibliographical references (p. 74-75). / Mean maximum start-up accelerations and velocities achieved by the fast-start specialist, northern pike, are reported at 120 ms-2 and 4 ms-1, respectively (Harper and Blake, 1990). In this thesis, a simple mechanical system was created to closely mimic the startle response that produces these extreme acceleration events. The system consisted of a thin metal beam covered by a urethane rubber fish body. The mechanical fish was held in curvature by a restraining line and released by a pneumatic cutting mechanism. The potential energy in the beam was transferred into the fluid, thereby accelerating the fish. The fish motion was recorded and the kinematics analyzed while using a number of different tail shapes and materials. Performance of the mechanical fish was determined by maximum acceleration, peak and averaged maximum velocity, and hydrodynamic efficiency. Maximum start-up acceleration was calculated at 48 ms-2. Peak and averaged maximum velocity was calculated at 0.96 ms-1 and 0.8 ms-1, respectively. The hydrodynamic efficiency of the fish, calculated by the transfer of energy, was 11%. Flow visualization of the mechanical fast-start wake was also analyzed. The visualization uncovered two specific vortex-shedding patterns; a single and a double-vortex pattern are described. / by Matthew Nicholas Watts. / S.M.

Thesis (S.M.)--Joint Program in Applied Ocean Physics and Engineering (Massachusetts Institute of Technology, Dept. of Mechanical Engineering; and the Woods Hole Oceanographic Institution), 2006. / Includes bibliographical references (leaves 94-96). / Recent studies have shown that the presence of sand ripples on the seabed improves sonar detection of buried mines at sub-critical angles. Sidescan sonar data of ripples off on the west Florida shelf were collected as part of ONR's Ripples Departmental Research Initiative (DRI) September 26-29th and November 7-9th, 2004. Hurricane Ivan, the strongest storm of the 2004 hurricane season, passed over the experiment site a week before the first data collection. This study focuses on the ripples created by Ivan. Average relict ripple wavelengths left after the storm were found to increase with water depth (50 cm, 62 cm, and 83 cm in 20, 30, and 50 meter water depths) despite the fact that orbital diameter decreases with water depth. Ripple prediction requires information about surface gravity waves and sediment grain size. The most reliable offshore wave field available was created with Wavewatch III by Naval Postgraduate School scientists. These waves were inputted into Delft3D WAVE, incorporating the nearshore wave model SWAN to predict waves at the locations where ripples were measured. Orbital motions at the seabed and grain size were inputted into a time-dependent ripple model with varying dissipation parameters to estimate sand ripples created by Hurricane Ivan. Ripple wavelength was found to be more strongly dependent on grain size than wave dissipation. / by Colleen Marie Bowers. / S.M.

Thesis (Ph. D.)--Joint Program in Applied Ocean Physics and Engineering (Massachusetts Institute of Technology, Dept. of Mechanical Engineering; and the Woods Hole Oceanographic Institution), 2007. / Includes bibliographical references (p. 124-126). / This thesis introduces an algorithm for inverting for the geoacoustic properties of the seafloor in shallow water. The input data required by the algorithm are estimates of the amplitudes of the normal modes excited by a low-frequency pure-tone sound source, and estimates of the water column sound speed profiles at the source and receiver positions. The algorithm makes use of perturbation results, and computes the small correction to an estimated background profile that is necessary to reproduce the measured mode amplitudes. Range-dependent waveguide properties can be inverted for so long as they vary slowly enough in range that the adiabatic approximation is valid. The thesis also presents an estimator which can be used to obtain the input data for the inversion algorithm from pressure measurements made on a vertical line array (VLA). The estimator is an Extended Kalman Filter (EKF), which treats the mode amplitudes and eigenvalues as state variables. Numerous synthetic and real-data examples of both the inversion algorithm and the EKF estimator are provided. The inversion algorithm is similar to eigenvalue perturbation methods, and the thesis also presents a combination mode amplitude/eigenvalue inversion algorithm, which combines the advantages of the two techniques. / by Travis L. Poole. / Ph.D.

Thesis (Elec.E. and S.M. in Electrical Engineering)--Joint Program in Applied Ocean Physics and Engineering (Massachusetts Institute of Technology, Dept. of Electrical Engineering and Computer Science; and the Woods Hole Oceanographic Institution), 2003. / "February 2003." / Includes bibliographical references (leaves 139-141). / by Andrea L. Kraay. / Elec.E.and S.M.in Electrical Engineering

Thesis (Ph. D.)--Joint Program in Applied Ocean Physics and Engineering (Massachusetts Institute of Technology, Dept. of Ocean Engineering and the Woods Hole Oceanographic Institution), 2001. / Includes bibliographical references (leaves 149-155). / This thesis takes a look at the near boundary flow about fish-like swimming bodies. Experiments were performed up to Reynolds number 106 using laser Doppler velocimetry and particle imaging techniques. The turbulence in the boundary layer of a waving mat and swimming robotic fish were investigated. How the undulating motion of the boundary controls both the turbulence production and the boundary layer development is of great interest. Unsteady motions have been shown effective in controlling flow. Tokumaru and Dimotakis (1991) demonstrated the control of vortex shedding, and thus the drag on a bluff body, through rotary oscillation of the body at certain frequencies. Similar results of flow control have been seen in fish-like swimming motions. Taneda and Tomonari (1974) illustrated that, for phase speeds greater than free stream velocity, traveling wave motion of a boundary tends to retard separation and reduce near-wall turbulence. In order to perform experiments on a two-dimensional waving plate, an apparatus was designed to be used in the MIT Propeller tunnel, a recirculating water tunnel. It is an eight-link piston driven mechanism that is attached to a neoprene mat in order to create a traveling wave motion down the mat. A correlation between this problem and that of a swimming fish is addressed herein, using visualization results obtained from a study of the MIT RoboTuna. The study of the MIT RoboTuna and a two-dimensional representation of the backbone of the robotic swimming fish was performed to further asses the implications of such motion on drag reduction. PIV experiments with the MIT RoboTuna indicate a laminarisation of the near boundary flow for swimming cases compared with non-swimming cases along the robot body. Laser Doppler Velocimetry (LDV) and PIV experiments were performed. / (cont.) LDV results show the reduction of turbulence intensity, near the waving boundary, for increasing phase speed up to 1.2 m/s after which the intensities begin to increase again through Cp = 2.0 where numerical simulations by Zhang (2000) showed separation reappearing on the back of the crests. Velocity profiles who an acceleration of the fluid beyond the inflow speed at the crest region increases with increased phase speed and no separation was present in the trough for the moving wall. The experimental techniques used are also discussed as they are applied in these experiments. / by Alexandra Hughes Techet. / Ph.D.

Thesis (M.S.)--Joint Program in Oceanographic Engineering (Massachusetts Institute of Technology, Dept. of Ocean Engineering; and the Woods Hole Oceanographic Institution), 1997. / Includes bibliographical references (p. 65-66). / The Surface Suspended Acoustic Receiver (SSAR) is a free-drifting platform intended for use as a receiver in large scale acoustic tomography experiments. Early prototypes of the SSAR exhibited very poor signal-to-noise ratios in the frequency band of the hydrophones. This thesis details efforts to reduce the hydrophone noise level by combining the analysis of experimental data with the results from numerical models. Experiments were conducted to quantify both the frequency content and magnitude of noise generated on the SSAR. Through a program of sea trials and pond testing, two noise sources were identified. The dominant source of noise in the SSAR is velocity dependent flow noise that results from turbulent pressure fluctuations on the hydrophones. A second noise source results from the acceleration sensitivity of the hydrophones in conjunction with high frequency accelerations present in the hydrophone array cable. These high frequency accelerations also show a velocity dependence. The presence of the acceleration-induced noise leads to correlations between the signals from adjacent hydrophones, thus distorting the typical picture that flow noise should be uncorrelated along an array. The primary methods of eliminating the noise are encapsulating the hydrophone in a flow shield, eliminating the array cable, and slowing the system down by replacing the wave following surface buoy with a spar buoy. Using the experimental results, empirical relationships between hydrophone velocity and expected noise level are formed for both shielded and unshielded hydrophones. The numerical models developed as a part of this effort are then used to predict the velocities for a wide range of possible SSAR configurations. The models can also provide information, such as system tensions, that is useful in evaluating the longevity and survivability of SSARs. Modeled design fixes include subsurface component changes as well as comparing a wave following surface buoy to a spar buoy. / by Jason I. Gobat. / M.S.

Thesis (Ph. D. in Applied Ocean Sciences)--Joint Program in Applied Ocean Physics and Engineering (Massachusetts Institute of Technology, Dept. of Ocean Engineering; and the Woods Hole Oceanographic Institution), February 2002. / Includes bibliographical references (leaves 161-170). / Sound propagation in shallow water is highly dependent on the interaction of the sound field with the bottom. In order to fully understand this problem, it is necessary to obtain reliable estimates of bottom geoacoustic properties that can be used in acoustic propagation codes. In this thesis, perturbative inversion methods and exact inverse methods are discussed as a means for inferring geoacoustic properties of the bottom. For each of these methods, the input data to the inversion is the horizontal wavenumber spectrum of a point-source acoustic field. The main thrust of the thesis work concerns extracting horizontal wavenumber content for fully three-dimensionally varying waveguide environments. In this context, a high-resolution autoregressive (AR) spectral estimator was applied to determine wavenumber content for short aperture data. As part of this work, the AR estimator was examined for its ability to detect discrete wavenumbers in the presence of noise and also to resolve closely spaced wavenumbers for short aperture data. As part of a geoacoustic inversion workshop, the estimator was applied to extract horizontal wavenumber content for synthetic pressure field data with range-varying geoacoustic properties in the sediment. The resulting wavenumber content was used as input data to a perturbative inverse algorithm to determine the sound speed profile in the sediment. It was shown using the high-resolution wavenumber estimator that both the shape and location of the range-variability in the sediment could be determined. / (cont.) The estimator was also applied to determine wavenumbers for synthetic data where the water column sound speed contained temporal variations due to the presence of internal waves. It was shown that reliable estimates of horizontal wavenumbers could be obtained that are consistent with the boundary conditions of the waveguide. The Modal Mapping Experiment (MOMAX), an experimental method for measuring the full spatial variability of a propagating sound field and its corresponding modal content in two-dimensions, is also discussed. The AR estimator is applied to extract modal content from the real data and interpreted with respect to source/receiver motion and geometry. For a moving source, it is shown that the wavenumber content is Doppler shifted. A method is then described that allows the direct measure of modal group velocities from Doppler shifted wavenumber spectra. Finally, numerical studies are presented addressing the practical issues associated with using MOMAX type data in the exact inversion method of Gelfand-Levitan. / by Kyle M. Becker. / Ph.D.

Thesis: S.M., Joint Program in Applied Ocean Physics and Engineering (Massachusetts Institute of Technology, Department of Electrical Engineering and Computer Science; and the Woods Hole Oceanographic Institution), 2018. / This electronic version was submitted by the student author. The certified thesis is available in the Institute Archives and Special Collections. / Cataloged from student-submitted PDF version of thesis. / Includes bibliographical references (pages 97-107). / Mobile robots and autonomous sensors have seen increasing use in scientific applications, from planetary rovers surveying for signs of life on Mars, to environmental buoys measuring and logging oceanographic conditions in coastal regions. This thesis makes contributions in both planning algorithms and model design for autonomous scientific information gathering, demonstrating how theory from machine learning, decision theory, theory of optimal experimental design, and statistical inference can be used to develop online algorithms for robotic information gathering that are robust to modeling errors, account for spatiotemporal structure in scientific data, and have probabilistic performance guarantees. This thesis first introduces a novel sample selection algorithm for online, irrevocable sampling in data streams that have spatiotemporal structure, such as those that commonly arise in robotics and environmental monitoring. Given a limited sampling capacity, the proposed periodic secretary algorithm uses an information-theoretic reward function to select samples in real-time that maximally reduce posterior uncertainty in a given scientific model. Additionally, we provide a lower bound on the quality of samples selected by the periodic secretary algorithm by leveraging the submodularity of the information-theoretic reward function. Finally, we demonstrate the robustness of the proposed approach by employing the periodic secretary algorithm to select samples irrevocably from a seven-year oceanographic data stream collected at the Martha's Vineyard Coastal Observatory off the coast of Cape Cod, USA. Secondly, we consider how scientific models can be specified in environments - such as the deep sea or deep space - where domain scientists may not have enough a priori knowledge to formulate a formal scientific model and hypothesis. These domains require scientific models that start with very little prior information and construct a model of the environment online as observations are gathered. We propose unsupervised machine learning as a technique for science model-learning in these environments. To this end, we introduce a hybrid Bayesian-deep learning model that learns a nonparametric topic model of a visual environment. We use this semantic visual model to identify observations that are poorly explained in the current model, and show experimentally that these highly perplexing observations often correspond to scientifically interesting phenomena. On a marine dataset collected by the SeaBED AUV on the Hannibal Sea Mount, images of high perplexity in the learned model corresponded, for example, to a scientifically novel crab congregation in the deep sea. The approaches presented in this thesis capture the depth and breadth of the problems facing the field of autonomous science. Developing robust autonomous systems that enhance our ability to perform exploratory science in environments such as the oceans, deep space, agricultural and disaster-relief zones will require insight and techniques from classical areas of robotics, such as motion and path planning, mapping, and localization, and from other domains, including machine learning, spatial statistics, optimization, and theory of experimental design. This thesis demonstrates how theory and practice from these diverse disciplines can be unified to address problems in autonomous scientific information gathering. / by Genevieve Elaine Flaspohler. / S.M.