The Strait of Gibraltar limits the exchange between the Atlantic and the Mediterranean Sea and therefore plays an important role in determining the water properties of the evaporation-dominated Mediterranean Sea. On the other hand, the strait dynamics depends on the boundary conditions set by the basin. To investigate this fundamental feedback between strait and basin processes, a 3-box model of the Mediterranean with a hydraulically controlled strait was programmed. It accommodates both maximal and submaximal strait exchange and does not impose steady state budget constraints, making it particularly useful for investigating transitional and non-equilibrium situations. The model is used in an explorative study to find new dynamical aspects of the system. First, the response of the system to changing air-sea-fluxes is modelled. Increasing evaporation shifts the steady state of the system to a more saline and - as a secondary effect - slightly warmer state. Increased heat loss leads to a colder and slightly less saline basin. Besides shifting the steady state, changes in heat flux and net evaporation can also lead to a nonlinear response in which the pycnocline deepens considerably for a transitional period of decades or centuries before returning to its steady state depth. Second, the effect of rising sea levels since the Last Glacial Maximum (18kyr BP) was modelled. At times of rapid sea level rise, the long residence time leads to stronger stratification and reduced circulation in the basin, providing a possible mechanism for the formation of sapropel S1. Finally the effect of mixing in the hydraulic jump between the Mediterranean and the Strait of Gibraltar is included in the model. The entrainment of inflowing water into the outflow reduces the effective exchange between Atlantic and Mediterranean, and the system develops multiple equilibria. After a comparatively short perturbation (e.g. a 20% larger evaporation for 10 years), the system can move from the stable, well-ventilated state to an almost stagnant meta-stable state which persists for centuries before the well-ventilated state is reestablished.

An experimental study on kinematics and turbulent structures in surf-zone breaking waves has been carried out in laboratory wave flumes. Particle Image Velocimetry (PIN) was used as a measurement technique to obtain the full field, instantaneous velocity throughout the breaking process. The experiments were performed on a single plunging breaker, a train of weak plunging breakers and a train of spilling breakers; all of which were generated on a 1/13 plane sloping beach. Measurements were carried out in a way which would allow mean flow and turbulent characteristics to be determined. The PIV velocity data were post-processed by means of an ensemble average (for the single plunging breaker) and a phase average (for the trains of breakers). The spatial distributions of mean velocity, mean vortic­ity and turbulent intensity, for each type of breaking wave, are presented. The results ob­tained from the single plunging breaker experiments are also compared to those predicted by a numerical model based on a Navier-Stokes solver. Furthermore, turbulent kinetic energy transportation is examined, based on the<i> k</i>-equation. The turbulent advection and production are calculated directly from the PIV data while the turbulent diffusion and dissipation are es­timated. By this means the turbulent transport process for different types of the surf-zone breaker is investigated. The influence of wave and turbulence characteristics on successive waves is also determined. Comparisons are made with the results of other researchers; some of which were carried out with different optical measuring techniques.

This study assessed a suite of geochemical proxies in the coastal Antarctic sea ice environment over two growing seasons (2004-2006). Time series measurements or productivity and export production were carried out over two contrasting growing seasons in Ryder Bay, a coastal embayment on the Western Antarctic Peninsular with a heavy sea ice influence. Concurrent measurements of elemental (C, N and Ba) and isotopic (¹³C and ¹⁵N) tracers were carried out on nutrients, suspended particulate matter and sinking particulate matter. The results of these measurements identified the following mechanisms that affect the use of these proxies as palaeoceanographic tracers. Variability of δ¹³C_PCC at this site is not caused by changes in [CO_2(aq)] or δ¹³C-CO₂, but appears to be affected by changes in diatom assemblages with different surface area to volume ratios and biochemical pathways for C fixation. Although the δ¹³C_POC signal in surface waters is transferred to depth, these findings preclude the use of δ¹³C as a tracer of paleo-<i>p</i>CO₂ concentration. δ¹⁵N, the proxy for nitrate utilisation in the Southern Ocean shows a predicted response to nutrient utilisation and the physical properties of the water column in each season, with each season displaying a different evolution of δ¹⁵N over the course of spring and summer. However, this proxy for utilisation of nitrate does not reflect the absolute changes in C flux. It is therefore necessary to consider this scenario when making connections between nitrate utilisation and changes in atmospheric CO₂ in sediment cores. This study also better constrained the biogeochemical cycling of Ba in the coastal Antarctic sea ice environment.

Observations of subsurface temperatures in the North Atlantic between 1950 and 1994 have been analysed to investigate interdecadal variability. Previous studies have concentrated on the more plentiful sea surface temperature data or on limited hydrographic data. Here, all subsurface temperature data, principally from bathythermographs, are used to investigate the development and evolution of temperature anomalies. During most of the period, a dipole feature was seen. The western subtropical gyre and the ocean north of 50°N were occupied by anomalies of opposite sign. These regions appear to show a cycle of warming and cooling with a period of approximately 30 years but the data record is too short to determine whether this is a regular oscillation. There is evidence of propagation of temperature anomalies along the paths of the major upper ocean currents, namely along the Gulf Stream and North Atlantic Current and also along the return currents on the southern edge of the subtropical gyre. An intense warm anomaly on the northern edge of the subtropical gyre developed between 1988 and the end of the data in 1994. The Gulf Stream front became steeper and isotherms were depressed by up to 150m. This is consistent with a stronger Gulf Stream and an intensification of the circulation of the subtropical gyre. This is compared with a cold anomaly in 1966-1972 which is opposite in nature, indicating a weaker subtropical gyre circulation. A scheme developed by Cooper and Haines (1996) has been employed to assimilate sea level anomaly data from the TOPEX/POSEIDON altimeter with a climatological hydrography of the North Atlantic.

The technique of Planer Laser Induced Fluorescence (PLIF) was applied to investigate the dispersion of surface films by breaking waves. The mixing caused by both isolated single breaking waves and pairs of identical breaking waves was examined. The results have a direct application to the study of dispersion of surface pollutants in the sea. PLIF was used to obtain images of the concentration distribution of a dispersing methanol-rhodamine solution used to mimic a surface film. Experiments were carried out on seven different non-dimensional amplitudes of single breaking wave and three different non-dimensional amplitude of pairs of breaking waves, ranging from mild spilling to large plunging breakers. Spatial and temporal information was extracted form the recorded images in order to quantify the dispersion in terms of maximum depth reached, area covered, centre of mass motion, dispersion coefficients and fractal dimension of the water-dye boundary. The result expanded significantly upon, and compared favourably with, previous work carried out on isolated single breaking waves.

With strong diurnal warming, the temperature at the air-sea interface could be several degrees warmer than the temperature at one or two metres depth. To correctly interpret SST measurements during conditions of diurnal-warming of SST, the diurnal response to environmental conditions must be understood. This thesis is a study of the response of diurnal-warming of SST to the primary environmental conditions that cause it. A one-dimensional ocean turbulence model is used to simulate the diurnal-cycle of warming of SST. The model is developed and enhanced to enable accurate predictions of amplitudes of the night to day difference in SST and the stratification associated with strong warming events. The enhanced model is validated with data from in-situ instrumented moorings. The model is used to investigate the shape and timing of the warming response to environmental causes, including the timing of those causes. The one-dimensional turbulence model must be ‘forced’ with air-sea fluxes. Available data sets for these fluxes have various temporal resolutions, from just a few minutes (high resolution) to daily averages. The performance of the model is tested against temporal resolution of the air-sea fluxes. This allows for a realistic interpretation of the modelled SST for applications where data is only available at low temporal resolution. SSTs from the Meteosat Second Generation (MSG) satellite have recently become available. SSTs (at the air-sea interface) from the new model are compared with the satellite SSTs at buoy locations in the Atlantic and show useful agreement with the shape and amplitude of the diurnal cycle for several events, (within the limits imposed by the low-resolution forcing data presently available for the satellite/buoy match-ups).

This thesis presents a new method for sea surface multiple removal which includes in its solution the information at the source from the oscillating bubbles which generate the propagating wavefield. Arrays of interacting oscillating bubbles excite the earth with sound energy which is recorded as marine seismic reflection data. In the manipulation of the data to produce a detailed image of the earth's subsurface, the first task is the removal of multiple reflections related to the sea surface. The new formulation, in common with other wave-theoretical methods, requires a two-dimensional grid of receivers to record a two-dimensional grid of shots. Using the data themselves, and no information about the subsurface, all multiples are removed in a three-dimensional earth by calculating the plane wave reflection response in the frequency-wavenumber domain. Applying the plane wave concept to waves in three dimensions permits a complicated reflected wave to be decomposed into plane wave components. Each of these scattered upgoing plane wave components comprises primary and multiple reflections. The primaries result from a multiplication of an incident downgoing plane wave component with a plane wave reflection response, unique for the particular downgoing and upgoing plane wave components chosen. The multiples represent a sum of products of scattered downgoing plane wave components with plane wave reflection responses, related to the chosen downgoing and upgoing plane wave components. The result is a set of simultaneous equations whose unknowns form the three-dimensional reflection response. The wavefield from the oscillating bubbles produces the downgoing incident plane wave. There are various methods of measuring this input energy. They depend on the acquisition configuration being used, and the feasibility of placing hydrophones at certain distances from the sources of the oscillating bubbles. Consideration of the fluid dynamics of these oscillating bubbles, in isolation and together, and the wave propagation produced by them, reveals the requisite measurements to describe fully the acoustic output of the source.

A comprehensive experimental study has been carried out to measure the velocity field of waves breaking on two types of beach. The measurement technique of Particle Image Velocimetry (PIV) has been used throughout to measure the full-field, instantaneous velocity distributions. This technique is described in some detail with particular attention paid to the errors inherent when using PIV in this application. The first type of beach studied has a mild slope, typical of sandy beaches. PIV measurements have been carried out on 1:30 and 1:100 plane sloping beaches with monochromatic waves. The measured internal kinematics are compared to those predicted by the Boussinesq and Serre wave models, with some good agreement. However, there are some significant differences in the near-surface region of the wave crest which tend to increase as the waves propagate up the beach. In addition to these theoretical comparisons, the Integral Properties of the waves on the 1:30 slope are calculated from the PIV measurements. The second type of beach examined is typical of a shingle beach. In addition to being steeper than the previous case, these beaches are also profiled, representing the shape of the beach formed under storm conditions. Two profiles have been modelled and three different monochromatic waves have been measured on each. The wave breaking processes are described and velocity measurements from each wave on both beaches are shown. Furthermore, a near-bed velocity comparison between the three waves at various positions along each beach has also been carried out in order to explain any potential on-shore and off-shore sediment motion.

Solitary waves are nonlinear, non-oscillatory disturbances of permanent form. Recent advances in synthetic aperture radar imaging and analysis techniques have confirmed <i>in situ</i> observations and measurements that the passage of oceanic internal waves, and in particular internal solitary waves, is associated with modulations in sea surface roughness. It has not only revealed the ubiquity of this phenomenon but also highlighted the global existence of large amplitude, tidally induced, internal solitary waves. It appears, however, that little laboratory-based research has been carried out in this field. This work, therefore, focusses on the study of surface wavetrain modulations resulting from the passage of a single internal solitary wave. Digital Particle Image Velocimetry (DPIV) and Planar Laser Induced Fluorescence (PLIF) were employed to provide two-dimensional instantaneous velocity and density information respectively. Previous studies in this field have been performed with intrusive probe techniques, disturbing the fluid flow during measurement. Preliminary DPIV and PLIF experiments were performed on single internal solitary waves in a two-layer brine - fresh water stratification. To the author’s knowledge, the application of PLIF to the study of these waves had not been done previously.  Results from the DPIV measurements concurred with previous research and highlighted the constraints of the DPIV system. The results were also compared to a recently developed and validated fully nonlinear numerical method. From the interaction investigations, both wavelength and amplitude modulations of the surface waves as a function of solitary wave phase were observed. In some cases, the shape of the internal wave was also affected. Velocity profiles were compared to the linear superposition of surface wave linear theory and the fully nonlinear numerical method. In addition, the PLIF analysis showed that, for the wave and stratification parameters investigated, there was no evidence for the compression and expansion of the density interface during the interaction.

No description available.