Marine monopiles can suffer from removal of sediment around their foundations by waves and currents, a process termed scour, which can negatively affect structure stability and integrity of associated infrastructure. Scour is a function of the interaction of local hydrodynamics with the geotechnical properties of the seabed, the feedbacks of which are not well understood. Using the largest prototype scour data base available to date, assembled from field data routinely collected during the consents and design phase of wind farms, this study aims to offer a detailed characterisation of marine monopile scour and conduct critical testing of the current, experimentally-derived, state-of-the-art knowledge and practices. Scour research has been hampered by a dearth of prototype scour observations and much of the existing knowledge is derived from physical and numerical work which has had very little validation with field data. This study addresses the dearth of prototype scour analysis and by adding observations from 281 monopiles more than doubles the size of the currently existing knowledge base on marine monopile scour. Furthermore, the scope, variety and quality of data available in this study have enabled a wider-ranging and more in-depth and problem-focussed analysis of scour to be conducted. The data used in this study comes from a "natural offshore laboratory", consisting of three offshore wind farms in the Outer Thames Estuary, which were strategically chosen to minimise the flow variability within the data set in order to focus on identifying and quantifying the controls on scour exerted by the sea bed substrate. The effect of geotechnical conditions on scour is, so far, little understood as most scour research has focused on unconsolidated sandy sediments. Nevertheless, scour experiments in cohesive substrates have revealed the great complexity of the scour response in such materials. For this reason, quantitatively scrutinizing prototype scour in various substrate types and attempting to establish causal links between geotechnical properties and scour development from real data is important. In order to fulfill the remit of this study, the research is guided by a set of questions, derived from a review of the current scour framework, which pose testable hypotheses and identify knowledge gaps which will be evaluated throughout the course of the analysis. The outcomes of the study include an extensive quantitative description and contextualisation of observed scour with existing prototype observations, a critical validation of current knowledge and methods and an investigation of hydrodynamic and geotechnical controls on scour. Some key findings include improved predictive models for scour depth based on mean water depth as well as secondary relationships for lateral extent and scoured volume. For scour in consolidated and cohesive materials, equations for the estimation of scour-limiting material strength as a function of erosion depth are also proposed. The study concludes with a discussion of temporal, survey resolution and geotechnical issues and recommendations for optimised field data collection and survey strategy, alongside suggestions for additional research to fully resolve some of the findings of this research.

Hydrothermal circulation plays a fundamental role in the chemical transfer from deep in the Earth’s interior to the ocean crust, the oceans and the atmosphere. It is also on of the principal mechanisms for heat transfer from the mantle to the oceans, atmosphere and ultimately, outer space. This process fundamentally influences the composition of the ocean crust during formation and aging as it spreads away from the ridge axis. However, despite much research into hydrothermal alteration of oceanic crust questions still remain including: the thermal and chemical evolution of hydrothermal fluids, the geometry of hydrothermal fluid flow, and the factors that control the nature and extent of hydrothermal alteration of oceanic crust. In this study, whole rock and secondary mineral characteristics of drilled-in situ ocean crust are used to (i) Characterise hydrothermal alteration for a range of drilled, in-situ fast spread ocean crust sites (ii) assess the factors that control hydrothermal alteration within fast spread ocean crust and (iii) assess the evolution and architecture of hydrothermal fluid. Deep Sea Drilling Project, Ocean Drilling Program, and Integrated Ocean Drilling Program Sites 504, 896, 843, 1179, 1149, 1224, 1243 and 1256 represent some the most significant penetrations into the upper portion of intermediate and fast spread crust to date. Analyses of whole rock chemical changes, Sr, O. C and S isotope systematics, petrographic observations and analysis of secondary minerals indicate that all sites underwent variable degrees of cold seawater dominated hydrothermal alteration. All these sites represent variations in the composition of the upper crust, basement topography, sedimentation rates, spreading rates, capping rocks, and age. Comparisons between these factors and style and intensity of alteration for each site indicate that spreading rate and age exherts the strongest influence on hydrothermal activity. Sites 1256 and 504 are the only sites in which both low temperature and high temperature alteration are recovered, both sites now have complete chemical and isotopic records which trace the evolution of hydrothermal fluid through the crust. Chemical and isotopic analyses of anhydrite within the ocean crust and consideration of the sulfur budget at these sites imply that the majority of hydrothermal fluid is heated to moderate temperatures (~250oC) and returns to the oceans as warm diffuse fluids at unaccounted for venting sites.

The eastern equatorial and tropical Pacific regions are areas of significant carbon fluxes from the atmosphere to the ocean interior. Changes in the function of marine biogeochemical cycles in this region potentially exert an important control on global climate. Understanding controls on and changes to ocean chemistry and circulation in this region is therefore of great importance. Redox sensitive metal distributions in hydrothermal sediments have yet to be exploited effectively as proxies of past ocean conditions. This work presents a 740 ka sediment record from an archived core collected at 14o47’S overlying 1.1 Ma crust on the western flank of the EPR. The metalliferous sedimentation is overprinted by diagenetic mobilisation arising from variations in the sediment redox status of the sediments. Amorphous ferrihydrite phases delivered to the sediment have undergone significant alteration to more stable crystalline forms. Under glacial conditions, the transformation of ferrihydrite appears to be impeded, which is inferred to be a function of a distinct change in the redox status of the sediments under glacial conditions. Oxyanions coprecipitated with Fe and Mn (hydr)oxides from the hydrothermal plume (P, V, U) are partitioned during Fe-oxide alteration. V is preferentially incorporated into goethite and residual phases, locking the plume derived V within the sediment. Sediment P/Fe ratios are lower than overlying plume values, and vary systematically with variations in ferrihydrite transformation to goethite on glacial-interglacial timescales. This transformation is inferred to lead to P loss from the particulate/sediment phase. Uranium is highly enriched in sulphide rich EPR plume particles and the sediments at 14oS. U/Fe ratios indicate there has been enhanced release of U under interglacial conditions, and preservation of plume U/Fe ratios under glacial conditions. Mo/Mn ratios are used to confirm the changes in redox status on glacial-interglacial timescales at this site. There is a general trend over Marine Isotope Stages 1-14 of a deepening of the sediment redox front through interglacial stages with a shallowing at the onset of glaciation. Enhanced sub-oxic conditions associated with glacial conditions (in particular MIS 12) are attributed to enhanced productivity (and carbon export to the seafloor) and decreased bottom water O2 (and therefore reduced ventilation of the deep water). This is consistent with paleoproductivity data from other parts of the Eastern Pacific and adds new information of past conditions in a region which has not been studied.

Submarine sediment density flows are a major process for transporting sediment from the continental shelf to the deep-ocean. Understanding submarine flow dynamics relies upon analysis of their deposits (beds) because monitoring them directly is difficult. However, it is rare to be able to correlate individual beds for long distances. This limits our understanding to ‘idealized’ models based on field data with limited lateral extent. Validation of these models requires individual beds to be mapped out. Using > 100 shallow sediment cores this thesis correlates individual beds across their depositional extent (over 2000 km), within the Late Quaternary Moroccan Turbidite System, offshore NW Africa. The vertical and spatial distributions of facies and grain size are examined in each bed to understand the dynamics of the parent flows. The height to which deposits drape up topography is used to infer flow thicknesses. Proximally, synchronous flows passed into the system from multiple disparate entry points. Earthquakes could have triggered these flows. However, it is not possible to determine if these beds were related to earthquakes, highlighting the difficulties faced extending turbidite palaeoseismology beyond the historical earthquake record. Across the central parts of the system flows are interpreted to have been relatively thin and slow moving, yet able to run out for hundreds of kilometers on slopes of < 0.02º. Current, models cannot explain how this is possible. Distally, channels develop and connect two basins. Examination of these channels reveals they are purely constructional features. Flows were able to bypass > 100 km3 of sediment through the channel axes without eroding. Channel relief was built and maintained by deposition along the channel margins and no erosion. The distribution of grain-size breaks is examined within individual beds across the entire system. Grain-size breaks between sand and mud occur almost everywhere. This is attributed to fluid mud layers bypassing intermediate grain sizes down slope. Such a process should (almost) always occur; hence this type of grain-size break should be recognized as a typical feature rather than an exception. The ability to map out individual beds over such distances provides a rare and valuable opportunity to validate models; developed from laterally restricted outcrops, laboratory experiments and theory. Results from this thesis demonstrate current models are limited and that we still have much to learn about the dynamics of submarine flows and how they transport sediment across vast swathes of the seafloor.

Submarine flows can transport huge volumes of sediment across the large submarine fans that dominate many parts of the deep ocean floor. Active flow events are notoriously difficult to monitor directly, and therefore our understanding of such flows still strongly relies on the analysis of the deposits they leave behind. This thesis aims to investigate the transport and depositional processes, the stacking patterns and the time frequency distribution of turbidites and debrites deposited in the Miocene Marnoso Arenacea Formation (Italian Apennines). This location is unique because deposits from individual flow events (beds) can be traced for long distances, allowing the lateral and down flow evolution of single flow events to be analyzed in detail. Lateral changes in individual flow deposits are documented through extensive correlation of beds deposited in a stratigraphic interval below the most prominent Contessa Marker bed. The observed transitions in facies, and the external shape of different types of deposit, are used as an independent test of models that capture our understanding of submarine flow processes. This work highlights how deposits of submarine density flows can be complex, even in relatively simple basin plain settings. A single event can comprise different flow types, and transformations can occur between these flow types. The initial volume, sediment concentration and grain size (including the proportion of fine cohesive mud) control the external shape of the deposits. Low density turbidity currents deposit clean sandstone beds with an exponentially tapering shape, while coarser grained high density turbidity currents produce massive or parallel laminated layers that maintain their thickness for longer (10’s of kilometers) distances. Cohesive debris flows form istinctive ungraded mud-rich sandstone that can either pinch-out abruptly or gradually taper. Liquefied debris flows with elevated pore pressures can deposit clean (mud-poor) sand over large areas (up to 30 km) of the Marnoso Arenacea basin plain. This is suggested by the distinctive swirly, patchy fabric of a particular type of clean sandstone, that records pervasive liquefaction during the late stages of the flow, and confirmed by the rapid pinch-out geometry of flow deposits at their margins. Such debris flows most likely form through transformation from an initial high density turbidity current. A similar flow process may characterize the distal, rapid pinch out of sandstone lobes in Fan 4 of the Skoorsteenberg Formation (Karoo basin, South Africa). The observed stacking pattern of turbidite beds in a 530 meters thick stratigraphic section indicates a long-term clustering. Debrite intervals however occur randomly, and bed correlation suggest that almost every large volume flow deposited clean or muddy debrite (or both) intervals in different positions of the basin. Hemipelagic marl thickness is used as a proxy for time between flow events. The distribution of time between events is exponential, therefore related to a Poisson Process. This indicates that flow events (most likely triggered by submarine slope failures) occur independently one from the other through time.

Submarine landslides are one of the most important processes for moving sediment across our planet. Landslides that are fast enough to disintegrate can generate potentially hazardous tsunamis, and produce long run-out turbidity currents that break strategically important cable networks. It is therefore important to understand their frequency and triggers. This thesis aims to do so using extensive datasets (N>100) suitable for statistical analysis. The influence of temporally non-random variables on landslide and turbidity current frequency is assessed statistically. In light of predicted future global and sea level rises this is a timely study. Analysis of large volume turbidites (>0.1 km3) reveals two distinct frequency distribution forms for submarine landslide recurrence. A Poisson (time-independent) form is observed in three basins which may indicate similar controls on landslide frequency and triggers occur in disparate areas. A log-normal (time-dependent) distribution is seen in the Iberia Abyssal Plain over much longer timescales (20 Myr). Physiographic and palaeoclimatic effects are thought to explain the two different distribution forms. The influence of sea level is either shown to be statistically insignificant (Poisson form) or has a significantly delayed (~1.2 Myr) influence (log-normal form) on landslide and turbidite recurrence. Two sequences that cross the Initial Eocene Thermal Maximum (IETM) hyperthermal, show a reduction in turbidity current and landslide activity, rather than the increase hypothesised by many studies. Therefore, predicted future sea level rise and global warming may not necessarily result in significantly increased submarine landslide or turbidity current frequency on human timescales. Finally, a unique direct monitoring dataset from the Squamish Prodelta, British Columbia provides new insights into the links between rivers and offshore deltas on very short (<months) timescales. It is shown that elevated river discharge is the primary control for the ‘switch on’ of turbidity current activity, but the trigger for flows is related to a combination of tide and river-controlled elevated shear stresses on the submarine delta top. Extreme flood peaks do not correspond to hyperpycnal flows, but instead cause rapid accumulation of sediment leading to delta lip failures.

Submarine mud volcanism is an important pathway for transfer of deep-sourced fluids enriched in hydrocarbons and other elements into the ocean. Mud volcanoes (MVs) occur in abundance on all oceanic plate margins, so fluxes of methane (CH4) and other chemical constituents from mud volcanism are likely to be significant for the oceanic budgets of some elements. Here, I present a detailed study of the spatial and temporal variation in fluid and chemical fluxes and mud flow activity at the Carlos Ribeiro and the Darwin MVs in the Gulf of Cadiz. Analyses of the chemical composition of pore fluids, sediments and authigenic carbonates are combined with a 1-D transport-reaction pore fluid model. Pore fluids from both MVs contain high concentrations of hydrocarbons (up to 16 mM), while pore fluids from Carlos Ribeiro MV (CRMV) are also enriched in lithium (Li+) and boron (B) but depleted in chloride relative to seawater. Oxygen, hydrogen and strontium isotope data suggest that the pore fluids are derived from depth and are affected by the transformation of smectite to illite. This process also produces pore fluids that are depleted in chloride and potassium, while B and Li+ appear to be leached from the sediments during this transformation process. The CRMV is the most active of the two MVs: fluid flow velocities are as high as 4 cm yr−1 at the eye of the MV but rapidly decrease to 0.4 cm yr−1 at the periphery. The associated fluxes of B, Li+ and CH4 are 7-301, 0.5-6 and 0-806 mmol m−2 yr−1, respectively. Fluid flow velocities at the Darwin MV are lower, ῀0.09 cm yr−1, and show little spatial variation: seepage activity appears to be controlled by the distribution of slabs of authigenic carbonate that are found on the seafloor at the summit of the MV. Results of radiocarbon dating of the hemipelagic sediments covering the CRMV suggest that there has been recent mudflow activity at the eye of the MV, and frequent mud expulsions over the past ῀1000 Cal yrs BP. The distribution of barite fronts at the margin of the MV and within the mudflow to the SE of the summit suggests that fluid advection has ceased over the past 340 Cal yrs, but degassing of these mudflows is ongoing and is potentially an important source of CH4. Geochemical and petrographic analyses of carbonates from the Darwin MV suggest that the MV formed in stages, with periods of intense fluid flow alternating with phases of mud extrusion and tectonic uplift. The results of this study demonstrate that fluid (and chemical) fluxes from MVs vary over relatively small time and space scales and that mud volcanism may contribute significantly to the oceanic inventories of Li+ and B. Moreover, anaerobic oxidation of methane appears to be an important control on methane emissions into the overlying water column, and a better understanding of this process is urgently required to properly quantify the impact of mud volcanism on the global oceanic methane budget.

The Central Indian Ridge (CIR) between 18° and 21°S shows significant morphological variation at a relatively constant spreading rate (between 47 and 51mmyr-1). High resolution TOBI sidescan sonar data (cruise CD127) and multibeam bathymetry data (Magofond 2 cruise), complemented by regional geophysical and geochemical datasets, provide an important opportunity to examine the processes controlling morphological and volcano-tectonic variations along the CIR. The CIR is situated on an elevated plateau formed from a temporally persistent and robust melt supply to the ridge axis. Analysis of the data shows that the fracture zones have a significant structural control on along-axis morphology and melt supply, partitioning the ridge into three morphologically defined regions. The central region (Region B) bounded by two of the fracture zones, contains segments which show rifted axial morphologies characteristic of slow-spreading ridges. Regions A and C show patterns of variable volcanic robustness along-axis associated with the supply and availability of melt to the ridge. Tectonic analysis correlates with inter- and intra-segment trends in morphology and volcanic structure, further highlighting regions of robust or deficient melt supply. Strong relationships are also revealed between the tectonic parameters of fault length, spacing and density along-axis. Finite difference analysis is used to model the Non-Transform Discontinuities along the CIR and Mid-Atlantic Ridge to understand patterns of strain within their interiors. The results indicate the presence of a damage zone ahead of a propagating segment tip, providing increased crustal permeability and a greater potential for hosting hydrothermal circulation. Analysis of geochemical data along the CIR identifies five hydrothermal plume signatures, three of which are coincident with the locations of NTDs in the study area, correlating with the findings of the numerical analysis. A detailed model of the CIR demonstrates that thermal variation influenced by ridge structure can produce significant variations in morphology and volcano-tectonic distribution at a constant spreading rate.

The Middle Eocene (37 to 49 million years ago, Ma) was characterised by warmer global temperatures and higher atmospheric carbon dioxide (pCO2) levels than today with only small/non-existant icesheets. Because predicted pCO2 levels for the coming century have not been seen on Earth since at least ~40 Ma thus, the primary objective of this thesis is to improve our understanding of the nature, timing and consequences of carbon perturbations in the Middle Eocene between 39 and 43 Ma. In Chapter 3, a new (instrument specific) methodology for measuring Mg/Ca in foraminiferal calcite is developed to enable the simultaneous measurement of additional trace elements indicative of foraminiferal test contamination. This new methodology enables Mg/Ca data to be screened more efficiently for contamination and increases confidence in palaeoceanographic reconstructions based on the Mg/Ca palaeotemperature proxy. In Chapter 4, new foraminiferal stable isotope records (~5 kyr resolution) from Demerara Rise in the equatorial Atlantic are generated to test the hypothesis that the onset of continental ice sheets in the Northern Hemisphere occurred at ~41.6 Ma in the Middle Eocene, 30 million years earlier than previously thought. The new data herein, indicate that if continental ice sheets were present, they were small and easily accommodated on Antarctica with no need to invoke storage of ice in the Northern Hemisphere. The dearth of appropriate Middle Eocene sedimentary sections on which to work means that a well calibrated timescale for this interval remains to be produced. In Chapters 5 and 6, this problem is addressed using Middle Eocene sediments recovered from the Blake Nose plateau in the western North Atlantic. A new high resolution magnetic stratigraphy and new quantitative foraminiferal biostratigraphic counts were developed between 39.5 and 42.0 Ma, which allows re-assessment and refinement of previous magnetostratigraphic and biostratigraphic interpretations. This provides excellent age control for these sediments and new calibrations to the Geomagnetic Polarity Time Scale (GPTS). In Chapter 7, the first high-resolution quantitative planktic foraminiferal assemblage counts were developed for the global warming event the Middle Eocene Climatic Optimum (MECO) between ~39.5 and 41.5 Ma. New biotic records show that the MECO was accompanied by significant biotic shifts that suggest a shift from warm, oligotrophic surface waters to warmer, more productive surface waters during the MECO.

The magnitudes and pathways of sand transport in Chioggia Inlet, southern Venice Lagoon, and the resulting morphological evolution have been investigated. Bathymetric analysis has established that there is a net export of sediment from the southern Lagoon Basin, and that significant morphodynamic changes in the barrier-inlet system have taken place. The total loss of sediment from the lagoon was evaluated as 125,770 m3/a, 10% of which is estimated to be sand. In addition, the presence of an ebb-tidal delta, extending from the mouth of Chioggia Inlet, as well as two subaqueous spits, were identified. The ebb-tidal delta annual growth was estimated as 50,000 m3/a, suggesting the existence of an additional source of sand for the delta, in addition to the Lagoon. The pathways and provenance of sand in transport were determined through seabed sampling and mineralogical analysis. Three predominant sources were identified: sand eroded from the tidal flats in the Southern Basin; a northelrly littoral drift of quartz-rich sand supplied by the Adige and Brenta rivers, to the south of the Chioggia Inlet; and a southerly littoral drift of carbonate-rich renourishment sand from Pellestrina, to the north of the Inlet. The nature and magnitude of transport through the inlet was investigated through direct measurements and modelling. It was found that the export of sand through the inlet is bed grain size dependant, ebb dominant and mostly in the bottom 1 m of the water column. The shape of the profile of concentration is well reproduced by the Rouse parameter, and thresholds for motion are between those derived by Van Rijn (1993) and by Amos et al. (2010b) for the Venetian inlets. Modelling results suggest that present export rates are 40% lower than those evaluated from longterm bathymetry changes. Volumetric and modelling estimates of drift along the shores adjacent to the Inlet indicate that the main contribution to the inlet features is littoral drift, with a longshore transport rate of 118,000 m3/a, from the south, and 91,000 m3/a, from the north. A conceptual sand budget for Chioggia Inlet was proposed. Littoral drift was found to be dominant over transport through the Inlet; thus, it is suggested that although the overall budget of sand in Chioggia Inlet is still negative, a small import of sand is starting to take place. Furthermore, the contribution of Chioggia Inlet to the net export of sand from the lagoon, based on present net budget estimates was calculated as 15%.