ALBERT

Description: Submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy at the Massachusetts Institute of Technology and the Woods Hole Oceanographic Institution September, 1975

Description: Erosion processes involving fine-grained marine sediments were studied by using an in situ flume to erode undisturbed bottom sediments on the sea floor in Buzzards Bay, a shallow marine embayment off the Massachusetts coast. Tte muddy sea floor in that area is characterized by a deposit-feeding infauna that reworks the sediments. Observations made with the in situ flume suggest that erosion resistance of compacted bottom sediments is up to twice as great as the erosion resistance of biogenically reworked sediments. Estimates of erosional bed shear stress from the in situ flume experiments are similar to estimates made during this study of bed shear stress developed in near-bottom tidal currents. It is inferred that erosion by the in situ flume produces reasonable estimates of bed shear stress necessary to erode undisturbed bottom sediments on the sea floor. Buzzards Bay muds were redeposited in a laboratory flume and eroded after various periods of reworking by the deposit-feeding organisms contained in them. Other Buzzards Bay mud samples were treated to remove organic matter, and the erosion resistance of flat beds of these sediments was also investigated in a laboratory flume. The surface of a biogenically reworked bed after two months was covered with mounds, burrows, trails, and aggregates composed of sediments and organic material. This bed was similar in appearance to many of the beds eroded by the in situ flume. The two month bed eroded at an erosional shear stress similar to the erosional shear stress necessary to erode the in situ Buzzards Bay muds (0.8 dynes/cm2 ) . Beds biogenically reworked for shorter periods had high values of erosional shear stress, up to twice that of the two month bed. The bed shear stress necessary to erode flat beds of Buzzards Bay sediments increased as the concentration of organic matter in the sediments increased. Deposit-feeders were absent in these beds, and the mode of deposition was kept uniform, so the increase of erosion resistance with increase in organic content is considered a reliable indication of sediment behavior, and not an artifact of experimental conditions. During the in situ experiments, lee drifts were created behind resistant roughness elements on the sea floor. A brief study of lee drift formation in the laboratory suggests that the formation of lee drifts from fine-grained sediments can be predicted to take place when the body Reynolds number of the resistant roughness elements is below a critical value.

Description: The Office of Naval Research supported this research and provided salary support through grants to the Woods Hole Oceanographic Institution and the Massachusetts Institute of Technology.

Description: Submitted in partial fulfillment of the requirements for the degree of Doctor of Science at the Massachusetts Institute of Technology and the Woods Hole Oceanographic Institution March, 1980

Description: The Southern Ocean as defined here is the body of water between the Antarctic Continent and the Antarctic Polar Front, (APF). This ocean is considered important in the global thermodynamic balance of the ocean-atmosphere system because large planetary heat losses are believed to occur at high latitudes. The ocean and atmosphere must transport heat poleward to balance these losses. In the Southern Hemisphere, the oceanic contribution to this flux involves a southward transport of heat across the APF into the Southern Ocean where it is given up to the atmosphere through air-sea interactions. In Part I, the air-sea interactions and structure of the near surface waters of the Southern Ocean are investigated with a three dimensional time dependent numerical model. The surface waters in this region in summer are characterized by a relatively warm surface mixed layer with low salinity. Below this layer, a cold temperature extremum is usually observed in vertical profiles which is believed to be the remnant of a deep surface mixed layer produced in winter. The characteristics of this layer, the surface mixed layer and the observed distribution of wintertime sea ice are reproduced well by this model. Unlike some other sea-ice models the air-sea heat exchange is a free variable. Model estimates of the annual heat loss by the Southern Ocean exhibit the observed meridional variation of heat gained by the ocean along the APF with heat lost further south. The model's area average heat loss is much smaller than that estimated with direct observations. While several model parameterizations were made which could be in error, the model results suggest that the Southern Ocean does give up vast amounts of heat to the atmosphere away from the continental margins. The model results and direct calculations of air-sea exchanges suggest a southward heat flux must occur across the APF. The lateral water mass transition across the front is not discontinuous but occurs over a finite sized zone of fluid which is dominated by intrusive finestructure. The characteristics and dynamics of these features are investigated in Part II to try and assess their importance in the meridional heat budget. Observations made on two cruises to the APF are presented and the space-time scales of the features and thermohaline characteristics are discussed. It is suggested that double diffusive processes dominated by salt fingering are active within the intrusions. An extension of Stern's (1967) model of the stability of a thermohaline front to intrusive finestructure driven by saltfingering where small scale viscous processes are included, is presented to explain why intrusions are observed in frontal zones. The model successfully predicts vertical scales of intrusions observed in the ocean and the observed dependence of the intrusions' slopes across density surfaces on the vertical scale. Since the fastest growing intrusion is not strongly determined by the model, though, it is likely that finite amplitude effects determine the dominant scale of interleaving in the ocean. The analysis predicts that intrusions transport heat, salt and density down the mean gradients of the front. For the APF, this heat flux is poleward which is the direction required by the global heat budget. This model does not describe intrusions at finite amplitude or in steady state and so cannot be used to estimate the magnitude of the poleward heat flux due to intrusions in the APF.

Description: The research reported on here, and my support as a graduate student was provided by the National Science Foundation through grants OCE 75 14056. OCE 76 82036 and OCE 77 28355.

Description: Submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy at the Massachusetts Institute of Technology and the Woods Hole Oceanographic Institution June 2007

Description: This thesis presents three studies which apply geophysical tools to the task of better understanding mantle melting phenomena at the upper and lower boundaries of the mantle. The first study uses seafloor bathymetry and small variations in the gravitational acceleration over the Hawaii-Emperor seamount chain to constrain the changes in the igneous production of the hot spot melting in the mantle which has created these structures over the past 80 My. The second study uses multichannel seismic reflection data to constrain the location and depth of axial magma chambers at the Endeavour Segment of the Juan de Fuca spreading ridge, and then correlates these magma chamber locations with features of the hydrothermal heat extraction system in the upper crust such as microseismicity caused by thermal cracking and high temperature hydrothermal vent systems observed on the seafloor. The third study uses two-dimensional global pseudospectral seismic wave propagation modeling to characterize the sensitivity of the SPdKS seismic phase to two-dimensional, finite-width ultra-low velocity zones (ULVZs) at the core-mantle boundary. Together these three studies highlight the dynamic complexities of melting in the mantle while offering new tools to understand that complexity.

Description: This thesis was funded by a National Science Foundation Graduate Research Fellowship, NSF grant OCE-0002551 to theWoods Hole Oceanographic Institution (WHOI), the WHOI Academic Programs Office, the Earth, Atmospheric, and Planetary Science Department at MIT, and by the WHOI Deep Ocean Exploration Institute.

Description: Submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy and the Woods Hole Oceanographic Institution February 1982

Description: In this thesis, seismic waves generated by sources ranging from 2.7 kg shots of TNT to magnitude 5 earthquakes are studied in order to determine the seismic activity and crustal structure of the Orozco transform fault. Most of the data were collected by a network of 29 ocean bottom seismometers (OBS) and hydrophones (OBH) which were deployed as part of project ROSE (Rivera Ocean Seismic Experiment). Additional information is provided by magnetic anomaly and bathymetric data collected during and prior to ROSE and by teleseismic earthquakes recorded by the WWSSN (Worldwide Seismic Station Network). In Chapter II, the tectonic setting, bathymetry and teleseismic history of the Orozco Fracture Zone are summarized. Covering an area of 90 x 90 km which includes ridges and troughs trending both parallel and perpendicular to the present spreading direction (approximately east-west), the bathymetry of the transform portion of the fracture zone does not resemble that of other transform faults which have been studied in detail. A detailed study of one of the largest teleseismic earthquakes (mb=5.1) indicates right lateral strike-slip faulting with a strike parallel to the present spreading direction and a focal depth of less than 5 km. The moment sum from teleseismic earthquakes suggests an average fault width of at most a few kilometers. Because the teleseismic earthquake locations are too imprecise to define the present plate boundary and the magnetic anomaly data are too sparse to resolve the recent tectonic history, more questions are raised than are answered by the results in this chapter. These questions provide the focus for the study of the ROSE data. Chapter III contains an examination of the transfer function between seafloor motion and data recorded by the MIT OBS. The response of the recording system is determined and the coupling of the OBS to the seafloor during tests at two nearshore sites is analysed. Applying these results to the ROSE data, we conclude that the ground motion in the absence of the instrument can be adequately determined for at least one of the MIT OBS deployed during ROSE. Hypocentral parameters for 70 earthquakes, calculated for an assumed laterally homogeneous velocity structure which was adapted from the results of several refraction surveys in the area, are presented in Chapter IV. Because of the large number of stations in the ROSE network, the epicentral locations, focal depths and source mechanisms are determined with a precision unprecedented in marine microseismic work. Relative to the assumed model, most horizontal errors are less than ±1 km; vertical errors are somewhat larger. All epicenters are within the transform region of the Orozco Fracture Zone. About half of the epicenters define a narrow line of activity parallel to the spreading direction and situated along a deep topographic trough which forms the northern boundary of the transform zone (region 1). Most well determined depths are very shallow (〈4km) and no shallowing of activity is observed as the rise-transform intersection is approached. In fact, the deepest depths (4-10km) are for earthquakes within 10 km of the intersection; these apparent depth differences are supported by the waveforms recorded a t the MIT OBS. First motion polarities for all but two of the earthquakes in region 1 are compatible with right lateral strike-slip faulting along a nearly vertical plane striking parallel to the spreading direct ion. Another zone of activity is observed in the central part of the transform (region 2). The apparent horizontal and vertical distribution of activity is more scattered than for the first group and the first motion radiation patterns of these events do not appear to be compatible with any known fault mechanism. No difference can be resolved between the stress drops or b values in the two regions. In Chapter V, lateral variations in the crustal structure within the transform region are determined and the effect of these structures on the results of the previous chapter is evaluated. Several data sources provide information on different aspects of the crustal structure. Incident angles and azimuths of body waves from shots and earthquakes measured at one of the MIT OSS show systematic deflections from the angles expected for a laterally homogeneous structure. The effect of various factors on the observed angles and azimuths is discussed and it is concluded that at least some of the deflection reflects regional lateral velocity heterogeneity. Structures which can explain the observations are found by tracing rays through three dimensional velocity grids. High velocities are inferred at upper mantle depths beneath a shallow, north-south trending ridge to the west of the OBS, suggesting that the crust under the ridge is no thicker, and perhaps thinner, than the surrounding crust. Observations from sources in region 2 suggest the presence of a low velocity zone in the central transform between the sources and the receiver. That the presence of such a body provides answers to several of the questions raised in Chapter IV about the hypocenters and mechanisms of earthquakes in region 2 is circumstantial evidence supporting this model. These proposed structures do not significantly affect the hypocenters and fault plane solutions for sources in region 1. The crustal velocity structure beneath the north-south trending ridges in the central transform and outside of the transform zone is determined by travel time and amplitude modeling of the data from several lines of small shots recorded at WHOI OBH. Outside of the transform zone, a velocity-depth structure typical of oceanic crust throughout the world oceans is found from three unreversed profiles: a 1 to 2 km thick layer in which the velocity increases from about 3 to 6.7 km/sec overlies a 4 to 4.5 km thick layer with a nearly constant velocity of 6.8 km/sec. A reversed profile over one of the north-south trending ridges, on the other hand, indicates an anomalous velocity structure with a gradient of 0.5 sec-1 throughout most of the crust ( from 5.25 km/sec to 7.15 km/sec over 3.5 km). A decrease in the gradient at the base of the crust to about 0.1 sec-1 and a thin, higher gradient layer in the upper few hundred meters are also required to fit the travel time and amplitude data. A total crustal thickness of about 5.4 km is obtained. An upper mantle velocity of 8.0 to 8.13 km/sec throughout much of the transform zone is determined from travel times of large shots of TNT recorded at MIT and WHOI instruments. "Relocations" of the large shots relative to the velocity model assumed in Chapter IV support the conclusion from the ray tracing that results from region 2 may be systematically biased because of lateral velocity heterogeneity whereas results from region 1 are not affected. In the last chapter, the results on crustal structure and seismicity are combined in order to define the present plate boundary and to speculate on the history of the present configuration.

Description: This research was supported by the Office of Naval Research, under contracts N00014-75-C-0291 and N00014-80-C-0273

Description: Submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy at the Massachusetts Institute of Technology and the Woods Hole Oceanographic Institution February 2003

Description: Oceanic observations indicate that abyssal mixing is localized in regions of rough topography. How locally mixed fluid interacts with the ambient fluid is an open question. Laboratory experiments explore the interaction of mechanically induced boundary mixing and an interior body of linearly stratified rotating fluid. A single oscillating bar produces a small region of turbulence along the wall at middepth. Mixed fluid quickly reaches a steady state height set by a turbulent-buoyant balance, independent of rotation. Initially, the bar is exposed on three sides. Mixed fluid intrudes directly into the interior rather than forming a boundary current. The circulation patterns suggest a model of unmixed fluid being laterally entrained into the turbulent zone. In accord with the model, observed outflux is constant, independent of stratification and restricted by rotation. Later the bar is laterally confines between two walls, which form a channel opening into the basin. A small percentage of mixed fluid enters a boundary current, which exits the channel. The bulk forms a cyclonic circulation in front of the bar, which blocks the channel and restricts horizontal entrainment. In the confined case, the volume flux of mixed fluid decays with time.

Description: This work was supported by the Ocean Ventures Fund, the Westcott Fund and the WHOI Education Office. Financial support was also provided by the National Science Foundation through grant OCE-9616949.

Description: Submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy at the Massachusetts Institute of Technology and the Woods Hole Oceanographic Institution February 2009

Description: Observations and inverse models suggest that small-scale turbulent mixing is enhanced in the Southern Ocean in regions above rough topography. The enhancement extends 1 km above the topography suggesting that mixing is supported by breaking of gravity waves radiated from the ocean bottom. In other regions, gravity wave radiation by bottom topography has been primarily associated with the barotropic tide. In this study, we explore the alternative hypothesis that the enhanced mixing in the Southern Ocean is sustained by internal waves generated by geostrophic motions flowing over bottom topography. Weakly-nonlinear theory is used to describe the internal wave generation and the feedback of the waves on the zonally averaged flow. A major finding is that the waves generated at the ocean bottom at finite inverse Froude numbers drive vigorous inertial oscillations. The wave radiation and dissipation at equilibrium is therefore the result of both geostrophic flow and inertial oscillations and differs substantially from the classical lee wave problem. The theoretical predictions are tested versus two-dimensional and three-dimensional high resolution numerical simulations with parameters representative of the Drake Passage region. Theory and fully nonlinear numerical simulations are used to estimate internal wave radiation from LADCP, CTD and topography data from two regions in the Southern Ocean: Drake Passage and the Southeast Pacific. The results show that radiation and dissipation of internal waves generated by geostrophic motions reproduce the magnitude and distribution of dissipation measured in the region.

Description: Submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy at the Massachusetts Institute of Technology and the Woods Hole Oceanographic Institution June 1998

Description: A freshwater plume often forms when a river or an estuary discharges water onto the continental shelf. Freshwater plumes are ubiquitous features of the coastal ocean and usually leave a striking signature in the coastal hydrography. The present study combines both hydrographic data and idealized numerical simulations to examine how ambient currents and winds influence the transport and mixing of plume waters. The first portion of the thesis considers the alongshore transport of freshwater using idealized numerical simulations. In the absence of any ambient current, the downstream coastal current only carries a fraction of the discharged fresh water; the remaining fraction recirculates in a continually growing "bulge" of fresh water in the vicinity of the river mouth. The fraction of fresh water transported in the coastal current is dependent on the source conditions at the river mouth. The presence of an ambient current augments the transport in the plume so that its freshwater transport matches the freshwater source. For any ambient current in the same direction as the geostrophic coastal current, the plume will evolve to a steady-state width. A key result is that an external forcing agent is required in order for the entire freshwater volume discharged by a river to be transported as a coastal current. The next section of the thesis addresses the wind-induced advection of a river plume, using hydrographic data collected in the western Gulf of Maine. The observations suggest that the plume's cross-shore structure varies markedly as a function of fluctuations in alongshore wind forcing. Consistent with Ekman dynamics, upwelling favorable winds spread the plume offshore, at times widening it to over 50 km in offshore extent, while downwelling favorable winds narrow the plume width to a few Rossby radii. Near-surface current meters show significant correlations between cross-shore currents and alongshore wind stress, consistent with Ekman theory. Estimates of the terms in the alongshore momentum equation calculated from moored current meter arrays also indicate an approximate Ekman balance within the plume. A significant correlation between alongshore currents and alongshore wind stress suggests that interfacial drag may be important. The final section of the thesis is an investigation of the advection and mixing of a surface-trapped river plume in the presence of an upwelling favorable wind stress, using a three-dimensional model in a simple, rectangular domain. Model simulations demonstrate that the plume thins and is advected offshore by the crossshore Ekman transport. The thinned plume is susceptible to significant mixing due to the vertically sheared horizontal currents. The first order plume response is explained by Ekman dynamics and a Richardson number mixing criterion. Under a sustained wind event, the plume evolves to a quasi-steady, uniform thickness. The rate of mixing slowly decreases for longer times as the stratification in the plume weakens, but mixing persists under a sustained upwelling wind until the plume is destroyed. Mixing is most intense at the seaward plume front due to an Ekman straining mechanism in which the advection of cross-shore salinity gradients balances vertical mixing. The mean mixing rate observed in the plume is consistent with the mixing power law suggested by previous studies of I-D mixing, in spite of the two-dimensional dynamics driving the mixing in the plume.

Description: This research was funded by a National Science Foundation graduate fellowship, and Gulf of Maine Regional Marine Research Program grants UM-S227 and UM-S276.

Description: Submitted in partial fulfillment of the requirements for the degree of Master of Science at the Massachusetts Institute of Technology and the Woods Hole Oceanographic Institution January 1999

Description: Today, deep waters produced in the North Atlantic are exported through the western South Atlantic. Antarctic intermediate water AAJW also enters the Atlantic in this region. Circumpolar deep water (CDW) fills the depths below AAIW and above and below northern source waters. A depth transect of cores from 1567-3909 m water depth in the western South Atlantic are ideally located to monitor inter-ocean exchange of deep water, and variations in the relative strength of northern versus southern source water production. Last glacial maximum (LGM) Cd/Ca and δ13C data indicate a nutrient-depleted intermediate-depth water mass. In the mid-depth western South Atlantic, a simple conversion of LGM δ13C data suggests significantly less nutrient enrichment than LGM Cd/Ca ratios, but Cd/Ca and δ13C data can be reconciled when plotted in CdW/δ13C space. Paired LGM Cd/Ca and δ13C data from mid-depth cores suggest increasingly nutrient rich waters below 2000 m, but do not require an increase in Southern Ocean water contribution relative to today. Cd/Ca data suggest no glacial-interglacial change in the hydrography of the deepest waters ofthe region. To maintain relatively low Cd/Ca ratios low nutrients in the deepest western South Atlantic waters, and in CDW in general, during the LGM requires an increased supply ofnutrient-depleted glacial North Atlantic intermediate water (GNA1W) and/or nutrient-depleted glacial Subantarctic surface waters to CDW to balance reduced NADW contribution to CDW. LGM Cd/Ca and δ13C data suggest strong GNA1W influence in the western South Atlantic which in turn implies export of GNAIW from the Atlantic, and entrainment of GNA1W into the Antarctic Circumpolar current.

Description: Submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy at the Massachusetts Institute of Technology and the Woods Hole Oceanographic Institution February 2018

Description: This thesis examines abiotic processes controlling the transformation and distribution of carbon compounds in seafloor hydrothermal systems hosted in ultramafic rock. These processes have a direct impact on carbon budgets in the oceanic lithosphere and on the sustenance of microorganisms inhabiting hydrothermal vent ecosystems. Where mantle peridotite interacts with carbon-bearing aqueous fluids in the subseafloor, dissolved inorganic carbon can precipitate as carbonate minerals or undergo reduction by H2(aq) to form reduced carbon species. In Chapters 2 and 3, I conduct laboratory experiments to assess the relative extents of carbonate formation and CO2 reduction during alteration of peridotite by CO2(aq)-rich fluids. Results from these experiments reveal that formation of carbonate minerals is favorable on laboratory timescales, even at high H2(aq) concentrations generated by serpentinization reactions. Although CO2(aq) attains rapid metastable equilibrium with formate, formation of thermodynamically stable CH4(aq) is kinetically limited on timescales relevant for active fluid circulation in the subseafloor. It has been proposed that CH4 and potentially longer-chain hydrocarbons may be sourced, instead, from fluid inclusions hosted in plutonic and mantle rocks. Chapter 4 analyzes CH4-rich fluid inclusions in olivine-rich basement rocks from the Von Damm hydrothermal field and the Zambales ophiolite to better understand the origin of abiotic hydrocarbons in ultramaficinfluenced hydrothermal systems. Comparisons of hydrocarbon abundances and stable isotopic compositions in fluid inclusions and associated vent fluids suggest that fluid inclusions may provide a significant contribution of abiotic hydrocarbons to both submarine and continental serpentinization systems.

Description: This thesis research was funded by the National Science foundation through grants OCE- 1427274 and OCE-1634032. Louise Von Damm generously contributed financial support for research conducted in Chapter 4.

Description: Submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy at the Massachusetts Institute of Technology and the Woods Hole Oceanographic Institution September 2018

Description: Many chemical constituents are removed from the ocean by attachment to settling particles, a process referred to as “scavenging.” Radioisotopes of thorium, a highly particle-reactive element, have been used extensively to study scavenging in the ocean. However, this process is complicated by the highly variable chemical composition and concentration of particles in oceanic waters. This thesis focuses on understanding the cycling of thorium as affected by particle concentration and particle composition in the North Atlantic. This objective is addressed using (i) the distributions 228,230,234Th, their radioactive parents, particle composition, and bulk particle concentration, as measured or estimated along the GEOTRACES North Atlantic Transect (GA03) and (ii) a model for the reversible exchange of thorium with particles. Model parameters are either estimated by inversion (chapter 2-4), or prescribed in order to simulate 230Th in a circulation model (chapter 5). The major findings of this thesis follow. In chapters 2 and 3, I find that the rate parameters of the reversible exchange model show systematic variations along GA03. In particular, 𝑘1, the apparent first-order rate "constant" of Th adsorption onto particles, generally presents maxima in the mesopelagic zone and minima below. A positive correlation between 𝑘1 and bulk particle concentration is found, consistent with the notion that the specific rate at which a metal in solution attaches to particles increases with the number of surface sites available for adsorption. In chapter 4, I show that Mn (oxyhydr)oxides and biogenic particles most strongly influence 𝑘1 west of the Mauritanian upwelling, but that biogenic particles dominate 𝑘1 in this region. In chapter 5, I find that dissolved 230Th data are best represented by a model that assumes enhanced values of 𝑘1 near the seafloor. Collectively, my findings suggest that spatial variations in Th radioisotope activities observed in the North Atlantic reflect at least partly variations in the rate at which Th is removed from the water column.

Description: This work was supported by the US National Science Foundation. Two US NSF grants have supported the research in this thesis (OCE-1232578 and OCE-155644).