Phytoplankton in the Sub-Antarctic Southern Ocean have a distinct seasonal cycle, which is highly variable in both space and time. The seasonal and spatial distribution of chlorophyll can be attributed to the complex nature of the physical and biogeochemical factors controlling phytoplankton production. Studies show that high-resolution sampling is required to understand variability in phytoplankton distribution and primary production. In this study, high-resolution glider data sampled in the Atlantic Sub-Antarctic Zone are used to characterise the scales of phytoplankton variability. Continuous glider data provide a novel way to assess phytoplankton variability at small time and space scales (meso- to submesoscale), especially in an area that has a lack of continuous measurements, which are necessary for addressing climate related questions. Temporal variability of phytoplankton was investigated using Empirical Mode Decomposition of surface chlorophyll-a concentrations collected from a Seaglider over a period of 5.5 months (25 September 2012 to 15 February 2013). This study found that during spring, chlorophyll-a concentrations were dominated by small scale daily fluctuations as well as by the rising seasonal ramp due to seasonal stratification. The removal of these signals revealed that the chlorophylla variability was dominated by submesoscales. In spring, phytoplankton blooms occurred as a result of features that shoaled the mixed layer depth when the wind stress weakened, elevating light conditions for short periods and allowing increased growth. In summer phytoplankton blooms were found to occur at submesoscales periods as well. This variability was found to be driven by synoptic storms varying the strength of the wind stress and consequently the mixed layer depth (that alters the nutrient and light environment). Additionally, through reconstructing the time series through subsampling at the dominant signals, this study found that in order to accurately resolve and characterise the multiseasonal variability of phytoplankton, chlorophyll needs to be sampled at high frequencies (<10 days). Spatial variability was investigated using daily MODIS ocean colour and sea surface temperature images coincident with the glider track. Spatial variability was characterised by the variance calculated at different length scales. Spatial analysis found that phytoplankton were patchier in both spring and summer when compared to sea surface temperature, at all length scales and that a greater variance was contained at small scales. There was also a greater variance in summer chlorophyll-a compared to spring due to higher maximum biomass. Further spatial analysis compared satellite spatial variance with glider measurements at the same length scale (70 km). This study found that a third of the variability found by the glider was caused by spatial patchiness, while the remainder could be contributed by local growth. These dominant meso- and submesoscale changes in chlorophyll-a at both temporal and spatial scales, highlights the need to resolve for both meso- to submesoscales in order to accurately reflect phytoplankton seasonal variability and ultimately to understand the impact of phytoplankton variability on carbon flux.

Bibliography: p. 142-149. / The dissertation seeked to understand the effectiveness of satellite based data collection of wind fields along the west coast of southern Africa, pertaining to particular synoptic atmospheric systems. A comparison between data from two automatic coastal weather stations and measurements obtained by the NASA scatterometer (NSCAT) for the period 1 December 1996 to 31 May 1997 yielded a correlation coefficient of 70%.

Includes bibliographical references (leaves 76-86) / This study explores the use of the 1/4° and the 1/12° Ocean Circulation and Climate Advanced Modelling (OCCAM) project. The model's representation of the dynamic nature of this region is assessed. On average 2 - 3 intense and well-defined eddies are generated per year within the model; having mean longevities of 4.89 ± 2.20 months with average advection speeds of 5.51 ± 1.57 km day¯¹

Includes bibliographical references. / The methods of analysing, interpreting and computing surf ace geostrophic currents from GEOSAT altimeter data are explained. These methods were implemented in an examination of GEOSAT altimeter data for the period 12 July 1988 to 13 May 1989 with application to the identification and tracking of Agulhas rings in the southeast Atlantic Ocean. A population of as many circulation features of as five anticyclonic (and cyclonic) annular form consistent with that reported for Agulhas rings was identified from the altimeter data. The positive identification of the same feature in subsequent repeat cycles was, however, difficult. The Verna ring was identified as it travelled approximately along the ground track of GEOSAT pass A 158. Its horizontal (approximately 200 km diameter) and vertical (>30 cm sea surface elevation) dimensions confirmed its identity as an Agulhas ring. The translation speed of the Verna ring along pass A 158 was determined as 6.8 cm·s⁻¹ consistent with the speeds reported in the literature. Surface geostrophic speeds calculated from the altimeter-derived sea surface slopes showed maximum anticyclonic speeds of nearly 40 cm·s⁻¹ toward the SW at the northern rim and 25 cm·s-1 toward the NE at the southern edge of the ring.

Bibliography: leaves 111-126. / An anomalous leakage of Agulhas Current water into the south-east Atlantic Ocean, exhibiting a mushroom-like shape, was observed during routine observations of A VHRR satellite imagery in early December 1996. The development of this anomaly was followed on the sea surface temperature (SST) imagery and it was tentatively identified as a consequence of filament interaction between the Agulhas retroflection and an occluding Agulhas ring. This interpretation prompted a cruise onboard the FR.S Africana with the objective of conducting a hydrographic survey of the Agulhas ring and the associated filament near Cape Town. A descriptive analysis, gleaned from A VHRR satellite imagery and in situ data, of the hydrographic characteristics of a vortex dipole, surveyed during this cruise, is presented in this thesis. An analysis of water mass properties and geostrophic flow patterns determined that an Agulhas ring and a cyclonic eddy, containing Benguela Current water in its core, constituted a dipole vortex in the south-eastern Atlantic Ocean. During the period of the hydrographic survey, the secondary vortex exhibited an anticlockwise rotation of 8. 6°/day around the Agulhas ring A warm filament, originating from the western Agulhas Bank, was entrained between the two counter-rotating eddies, thus resulting in the mixing of Agulhas Bank water into the South Atlantic Ocean. Satellite altimetry and A VHRR imagery were used to "backtrack" the vortex dipole to its origin at the Agulhas retroflectiOn. By combining interpretations from the altimetry and A VHRR imagery, it was possible to describe the complex interactions the dipole displayed with the retroflection and the Agulhas Bank as it franslated in a north-westerly direction. The mushroom configuration, identified earlier on SST imagery, betrayed the presence of an adjacent pair of circulatory features of opposing spm. As the dipole translated northward, it interacted with the Agulhas Bank and the cyclone was strained, becoming a filament as it was forced between the Agulhas ring and the Agulhas Bank. West of Cape Town the dipole was re-established when the cyclone redeveloped, changing the orientation of the dipole so that a filament was drawn directly from the Agulhas Bank.

Bibliography: leaves 152-159. / The continued sustainable beneficial use of the sea, and in particular Simon's Bay, as an important industrial Naval Harbour, beach and recreational retreat, tourist destination, penguin haven, trek-net fishery, maritime archaeological area and research zone has been highlighted. If these benefits are to be maintained or in the context of expanded future use, then the quality of seawater in the region is seen as a key maritime health indicator of this interdependent system. Information about near-shore interaction with False Bay and beyond is important for good management. This work is primarily concerned with the marine coastal region defined by the port limits of Simon's Bay. Simon's Bay cannot, however, be studied, without consideration of the external influences of the greater False Bay and, still further afield, from the South Atlantic Ocean and the Agulhas bank The primary area of interest includes the port limits of Simon's Bay as defined by the South African Navy, depicted on the SAN 1017, 1998 Simon's Bay hydrographic chart. This work aims to provide an initial understanding of the spatial and temporal environmental dynamics of Simon's Bay and establish baseline values.

Includes abstract. / Includes bibliographical references (leaves 85-92). / The Mozambique Channel and the region south of Madagascar are dominated by high variability of the oceanic flow due to the ubiquitous presence of mesoscale eddies. The bottom topography of this region has several shallow ridges. The water flowing through the Mozambique Channel propagates southwards predominantly as a train of anti-cyclonic eddies, moving towards the Agulhas Current. South of Madagascar, dipolar vortices regularly propagate in a south-westward direction. Their deep extent favours interaction with the shallow bathymetry of the Davie, Mozambique and Madagascar Ridges. The role of the Madagascar Ridge on the mesoscale circulation is investigated using altimetric observations and model simulations.

Bibliography: leaves 63-73. / This study forms part of one of the Benguela Environment Fisheries Interaction and Training (BENEFIT) program projects, whose main objective is to implement effective and quality ADCP data collection in the Benguela region. The thesis has two main components, firstly it undertook to investigate and assess the data quality, processing methods and software of ADCP data that are used to collect and process the ADCP data available in the Benguela region. An inter-comparison was made between the two different formats for data collection, i.e. raw and RDI proprietary "TRANSECT"processed data (which is more readily available in South Africa), to evaluate their respective accuracy in depicting current flow. The raw data were validated (edited and calibrated) using the CODAS software package before the current vectors could be drawn. TRANSECT -processed data could not be validated therefore vectors were drawn from the un-validated data. Data used here was collected during a monthly monitoring survey on 06-08 August 2001 on board one of the South African research vessel Algoa. The comparison results showed that TRANSECT -processed data includes unedited errors, noise and biases, which are already averaged into the data by the program and that the raw data presented a more realistic current flow. Secondly, the study undertook to apply the ADCP data collected to describe the state of oceanography of the Benguela region during the survey periods. Data from monitoring survey in 06-08 August 2001 was used to describe the southern Benguela while the second data set from a cruise conducted off Namibia in October 2000 was used to describe the oceanography of the central and northern Benguela. Most of the structures observed were in support of the literature and confirming previous studies of the region. In the southern region, the equator ward shelf jet off Cape Peninsula was observed to be forced offshore and entrained in an anticyclonic-like feature, which appeared to be part of a wann Agulhas filament. In northern Nambia, a strong poleward movement of the wann Angolan water was measured.

Bibliography: pages 96-100. / Tide gauge data from ten ports around the coast of Southern Africa are used to study the nature and behaviour of extreme high sea levels with a view towards predicting the likelihood of extremes occurring in the future. A recorded sea level height can be thought of as a combination of an astronomical tide and a weather determined component. In Southern Africa tides are typically 2 to 2.5 metres in range and the non-tidal residual accounts for up to 50 cm. Sea level is governed by local tides and local meteorology and there is great similarity in the magnitudes and timing at all ports. However tides are found to be the dominant contribution to extreme levels, hence the long term character of tidal variations is important in Southern African extremes. The fortnightly, equinoctial and 4.4 year tidal cycles determine the expected sea level extremes. A prediction technique developed here makes use of the tidal dominance by calculating the likelihood of exceedance for a specific month in a particular year. For any month the highest tide is known and an extreme will depend on the necessary surge occurring. Probability is derived from the surge distribution for that month, carried out for each month in a year, and the results presented as an exceedance chart.

The interplay of three important parameters, dissolution, dilution and winnowing seem to be controlling the state in which sediment material is preserved in the study area though most samples examined under the SEM were generally well preserved. The distribution of forty-four coccolithophore species in one hundred deep sea core-tops from the southwest Indian Ocean is described. Three coccolith assemblages have been recognised (Maputo, Agulhas Current and deep water) and are delineated by the relative abundances of four ecologically significant coccolithophore species (Gephyrocapsa oceanica, Emiliania huxleyi, Calcidiscus leptoporus and Umbilicosphaera sibogae). These four species are the most abundant in the study area and the major factors influencing their geographical distribution seem to be temperature, nutrient concentration and dissolution. Coccolith and foraminifera preservations indicate that the carbonate lysocline lies somewhere between 3500 and 4000 meters, resulting in the concentration of dissolution resistant microfossils below this depth. Stable oxygen isotope ratios in a planktonic foraminifera and percent abundances of E. huxleyi reveal that apart from cores taken in the Agulhas Passage, most of the core-top samples are probably less than 85,000 yrs. BP. Lightest isotope ratios of -1.5 to -1.0 per mil PDB (equal to 22.8 to 25.1°C) occur in a narrow band on the sea floor beneath the "A" route of the Agulhas Current. These values are about 0.5 per mil PDB lighter than samples analyzed on either side of this band and can be explained by the Agulhas Current's elevated temperature at the ocean surface of between 2 to 3°C. Thus, an oxygen isotope imprint of the Agulhas Current exists beneath it on the sea floor. The Agulhas Current is probably the major factor influencing sedimentation, sediment distribution patterns and geological features in the study area. At present it is voluminous and fast flowing, possibly eroding sediments 2500 meters below the surface. The oxygen-isotope ratios and nannoplankton counts obtained in this study indicate however, that the majority of samples are most probably recent or at least not older than 85,000 years except for sediments found in the Agulhas Passage. This implies that sediments are accumulating on the ocean floor and that the current does not have a pronounced erosional influence, at least in areas from which cores were retrieved for this study.