A two-dimensional, longitudinal-vertical model for partially-mixed estuaries has been developed. The model provides intratidal predictions of surface level, velocity, and salinity through a semi-implicit finite-difference solution to the continuity and momentum equations and an explicit finite-difference solution to the salinity equation. The model was verified through comparison with analytical solutions, laboratory data, and prototype data. Following verification, the model was used to simulate the destratification-stratification cycle which occurs in the James River Estuary, Virginia, coincident with the spring-neap tidal cycle. In a second application to the James, a simulation of the movement of the salinity intrusion following a storm-generated freshwater flow pulse was conducted. Investigations were conducted into the reaction of a hypothetical estuary to step-like and pulse-like alterations in wind stress, tide range, boundary conditions and flow. It was noted that the reaction time-scale of the estuary was much longer than the time-scale of alterations in the forcing functions. Thus, in prototype estuaries in which forcing functions are periodic and/or randomly superimposed, truly steady-state conditions are never attained.

Includes bibliographical references (pages 71-89). / In this model study of the coastal ocean off Tanzania, the Regional Ocean Modelling System (ROMS) was employed to model the coastal ocean off Tanzania over the domain of 5°N-15°S and 38-55°E. It was integrated for ten years with monthly mean Comprehensive Ocean and Atmosphere Data Sets (COADS) winds and heat fluxes. Initial and lateral boundary conditions were derived from the World Ocean Atlas. The model was used to simulate the annual cycle, and the sea surface temperature (SST) output compared with the Tropical Rainfall Measuring Mission (TRMM) Microwave Imager (TMI) sea surface temperature (SST) measurements for the same region. Although broadly comparable, the model SST was generally warmer than that of TMI data. The high SSTs in the Tanzanian coastal waters (greater than 28°C) occur from December to May while SSTs of less than 28°C occur during the rest of the year. The East African Coastal Current (EACC) experiences its lowest spatial and temporal average speeds (about 0.4ms- 1) in February and its maximum speed (1.7 ms⁻¹) in July. Speeds of greater than 1 ms⁻¹ occur during both transition seasons north of 6°S. The meridional wind stresses appear to be positively correlated with the EACC(r>0.6) in all locations and they are statistically significant (p<0.05). The annual cycle of the model flow in the southern Tanzanian waters seems to be positively correlated with the flow to the north of Madagascar (r=0.57 and p=O.O5). The flow in these regions changes in phase with each other from October to April and June to July with minimum speeds in November. For the other months, the flow in these regions is out of phase with each other. The model currents off southern Tanzania attain their maximum speeds in August when the South West monsoon is fully developed while the flow north of Madagascar attains its maximum speed in September when the South West monsoon fades. However, the flow in the southern Tanzanian waters is more affected by the reversal of winds over the tropical western Indian Ocean (r=0.69, p=0.01) than that north of Madagascar (r=0.51, p=0.09). This difference results in a larger annual speed range in the flow off southern Tanzania (about 0.4 ms⁻¹ ) than that to the north of Madagascar (about 0.3ms⁻¹). The ROMS model realistically simulates the annual cycle of the sea surface temperature and heat flux, the East African Coastal Current and the annual cycle of the flow entering the coastal ocean off the southern part of Tanzania. However, studies which integrate the large scale domain and regional coupled ocean-atmosphere interactions are needed to better understand of the East African climate and ocean variability. Such model results combined with suitable remote sensing and in situ observations will help improve understanding of the circulation and properties of the coastal ocean off Tanzania.

The alternation between summertime nitrate drawdown and wintertime nitrate recharge is central to the role of the Southern Ocean in setting atmospheric CO2. However, active cycling of nitrogen (N) in the seasonally-varying mixed layer – including the production of ammonium and its subsequent removal via phytoplankton uptake and nitrification (i.e., the oxidation of ammonium to nitrite and then nitrate) – remains poorly understood. Following the “new production paradigm”, phytoplankton production fueled by ammonium (“regenerated production”) results in no net drawdown of CO2 to the deep ocean, while growth supported by nitrate (“new production”) can be equated to CO2 removal provided that mixed-layer nitrification is negligible. While non-zero mixed-layer nitrification has been measured in many ocean regions, very few data exist for the Southern Ocean. This thesis presents new N cycle data collected across the Southern Ocean south of Africa in winter and summer that emphasize the integral role of mixed-layer N transformations in Southern Ocean productivity and biological CO2 drawdown. To evaluate the new production paradigm as a framework for quantifying Southern Ocean carbon export potential, rates of net primary production (NPP), N uptake (as ammonium and nitrate) and nitrification (ammonium and nitrite oxidation) were measured across the Atlantic sector in winter and summer. Winter mixed-layer NPP and total N (i.e., ammonium + nitrate) uptake were strongly decoupled, likely due to elevated heterotrophic bacterial consumption of ammonium. In summer, NPP and total N were generally well-coupled, although dissolved organic N apparently supported more than a third of NPP at some stations. Nitrification accounted for >100% of the nitrate consumed by phytoplankton in winter, rendering the new production paradigm ill-suited for quantifying carbon export in this season. By contrast, of the >50% of summertime NPP fueled by nitrate, < 4% on average derived from mixed-layer nitrification. While the near-zero mixed-layer nitrification rates measured in summer could be taken as confirmation that nitrate uptake is a good proxy for Southern Ocean carbon export potential, a portion of the nitrate consumed in the summertime euphotic zone was produced in the winter mixed layer and will thus not support net carbon dioxide removal on an annual basis. Despite the high rates of ammonium uptake and oxidation measured in winter Southern Ocean surface waters, mixed-layer ammonium concentrations remain fairly high, indicating an imbalance between ammonium production and consumption. Kinetics experiments conducted across the Indian sector (37-62ºS) reveal a seasonal switch from a phytoplankton community with a high affinity for ammonium in summer to one with a lower affinity in winter, even though phytoplankton at similar latitudes achieved a comparable maximum specific ammonium uptake rate in summer and winter. Rates of ammonium oxidation showed a Michaelis-Menten response to substrate availability only when the ambient ammonium concentration was ≤90 nM. This, coupled with half-saturation constants (Km values) of 28-137 nM (i.e., indicating a very high affinity for ammonium) suggest a dominant role for ammonia oxidizing archaea in mixed-layer nitrification. The maximum rate of ammonium oxidation was near-constant across the transect (37-62ºS), despite a significant gradient in sea surface temperature, light availability and ammonium concentration, perhaps due to iron limitation of ammonium oxidation, which has been hypothesized from culture experiments but not yet shown in the environment. It is possible that iron depletion in the surface Southern Ocean may limit the role of winter mixed-layer nitrification in offsetting phytoplankton CO2 drawdown annually. To better understand the controls on nitrifier ecology in the surface Southern Ocean, a series of nitrite oxidation kinetics experiments were conducted across the Indian sector (37- 62ºS) in winter. All experiments yielded a Michaelis-Menten relationship with substrate concentration, yet the nitrite oxidation rates only increased significantly at nitrite concentrations >115-245 nM, suggesting that nitrite oxidizers require a minimum (i.e., “threshold”) nitrite concentration to produce nitrate. Low derived Km values (134-403 nM) that increased with increasing ambient nitrite indicate a high affinity of nitrite oxidizers for substrate, in contrast to results from culture experiments. Throughout the Southern Ocean mixed layer, ambient nitrite concentrations are rarely less than 150 nM, regardless of season. Coincident measurements of ammonium and nitrite oxidation in the mixed layer suggest that nitrite oxidation is the rate-limiting step for nitrification in the winter Southern Ocean. This, combined with a possible nitrite concentration threshold for nitrite oxidation, may explain the perennial non-zero mixed-layer nitrite. A possible explanation for the apparent threshold nitrite requirement of nitrite oxidizers is undersaturation of the hemerich nitrite oxidoreductase enzyme, perhaps driven by the limited availability of iron in Southern Ocean surface waters. The findings described in this thesis yield new insights into the active cycling of N within the Southern Ocean's mixed layer, and particularly emphasize the need for seasonally-resolved parallel N- and iron cycle investigations to fully understand the role of nitrification in biological CO2 removal and N supply. Climate change-driven warming and acidification of Southern Ocean surface waters is already driving changes in microbial community composition, nutrient supply, and primary productivity. If we are to better predict the Southern Ocean's future role in CO2 sequestration and global ocean fertility, an improved understanding of the controls on mixed layer N cycling, particularly nitrification, is essential.

The intraseasonal variability of low-level winds over the southeast Atlantic Ocean is investigated using high-resolution satellite derived QuikSCA T wind data spanning a 5- year period from August 1999 to July 2004. Wavelet analysis is applied to 20-70 day bandpass filtered zonal and meridional wind time series over six tropical and subtropical southeast Atlantic Ocean regions identified in previous research as being distinct. The influence of the winds over the tropical southeast Atlantic on southern African rainfall during late austral summer to early autumn (January - April) is studied. Composite analysis of QuikSCA T winds and National Center for Environmental Prediction (NCEP) re-analyses atmospheric circulation variables is performed to investigate the potential mechanisms influencing rainfall. It is found that intraseasonal wind variability over the six regions is characterised by dominant frequencies ranging between 20 and 40 days. A single spectral peak with a period that shifts from 32 days in the tropical regions (1 0-l 8.5°S) to a slightly shorter 26- 28 day period over the subtropical zones (l 9-35°S) is evident in the meridional wind spectra. The zonal wind spectra contain two main frequency peaks at 24-28 and 36-40 days. The 36-40 day frequency peak is somewhat less pronounced between 10° and 23.5°S, becoming more defined towards the south (24-35°S). The intraseasonal wind oscillations in the tropical southeast Atlantic seem to be related to convective activity over the Angola low and the West African monsoon region during late austral summer and winter, respectively. In the subtropical regions, the intraseasonal wind oscillations appear to be associated with eastward propagating midlatitude waves. Significant year-to-year variability in the intraseasonal oscillations is revealed, particularly in the timing and magnitude of dominant oscillations. Enhanced wavelet power is apparent during the 1999-2001 and 2002-2003 La Nina and El Nino events, respectively. This confirms the impact of ENSO signals over the southeast Atlantic Ocean. Westerly winds over the tropical southeast Atlantic seem to have an important influence on southern African rainfall. Generally, significantly enhanced westerly flow over thetropical southeast Atlantic coincides with increased moisture flux that feeds into the Angola low and enhanced rainfall over parts of southern Africa. The occurrence, magnitude and spatial extent of the rainfall activity seem to depend primarily on the amount of moisture inflow from the southwest Indian and tropical southeast Atlantic oceans, atmospheric circulation anomalies over southern Africa as well as the intensity of the Angola low. While the enhanced westerly winds over the tropical southeast Atlantic seem to have an influence on southern African rainfall, they do not appear by themselves to lead to above-average rainfall over the region, nor are all periods of above-average rainfall associated with enhanced westerly winds over the tropical southeast Atlantic.

The story of Southern Hemisphere humpback whales through time presents multiple narratives. This study integrated two of those narratives to better understand our climate over time - the ecological behaviour of this species as well as their exploitation in the last century. The changing of our climate is largely better understood from the introduction of satellites in the late 20th century when data could be collected at higher spatial and temporal levels. Prior to this, data were scarce, especially for remote areas such as the Southern Ocean. The trust on climate models to produce valuable projections rely on how skilled they are in reproducing the historical climate; therefore their results require assessments against as many observations as possible to further increase their reliability. The Southern Ocean being an integral component to climate regulation, it is important to try understand its oceanographic features. The seasonal sea ice cover represents one major feature of this system. This study proposes to use other sources of data for the early 20th century that will help closing the gap prior to satellite observations. Humpback whales migrate poleward during the austral summer to feed on Antarctic krill at a proximity to the ice edge. Humpback whale catch locations in the early 20th century corroborate with this foraging behaviour. Using humpback whale catch location data as a benchmark, the study aims to assess the skill of climate models in simulating sea ice edge location for the early 20th century. Sea ice edge is directly related to sea-ice extent, which is an important variable in the research of sea ice dynamics over time. This is especially true in the face of rapid climate change where accuracy of sea ice changes is very important. The study therefore also aims to assess climate model climatological seasonal cycle of sea ice extent results from climate models against literature and contemporary observations. Comparisons between each model's results are also carried out. The humpback whale catch effort, as per IWC data, mostly covered the Atlantic and the Indian sector of the Southern Ocean. For this reason, the study focused on these two sectors for the analyses. Decade 1930-1939 showed the highest catch numbers consistently throughout the months of the study. The simulated past century and recent climatological seasonal cycle sea ice extent show a wide variety of responses between the models, with the majority of them underestimating the seasonal cycle based on previous literature and contemporary observations. This indicates the need to improve the sea ice physical processes in models to better capture the specific Southern Ocean processes. The ensemble median of ice edge location from the models apparently follow the latitudinal pattern of the whale catch locations, which are assumed in this study to mark the topography of the sea ice edge. However, they simulate a sea ice edge equatorward of the edge derived from an ensemble analysis of humpback whale catch locations. The variance explained by the coefficient of determination between the models and the whale catch distribution is low, with the highest value of one month being low as well. This indicates that only a portion of the simulated edge follows the reconstructed sea ice features.

South Africa's estuaries and their surrounding communities are becoming increasingly vulnerable to storm surges and accompanied estuary flooding. These events are largely due to increasing severity of storm surges combined with growing housing and commercial developments. A particularly severe weather event in 2007/2008 highlighted the pressing need to understand the processes involved and the urgency to develop proactive response and management actions to mitigate the effects of future storm events on these coastal areas. Scientific research on estuarine flooding is limited not only for South Africa but within the international community as well and only recently has received committed attention from policy makers. It is clear that our current knowledge of South African estuary flooding events remains rudimentary; while necessary action to mitigate such events are poorly understood and planned. The aim of this PhD thesis is to devise and implement an Estuary Early Warning – Emergency Preparedness and Response Guide for stakeholders and government policymakers. This guide will target South Africa's coastal region by analysing past information on storm surges and estuary flooding, particularly in the low-lying southern coast region of the Western Cape, South Africa. The key objective of this thesis is to assess the best processesfor the issuing of estuary alerts and to better standardise them so that the response remains in line with multi-hazard early warning standard procedures and practices within South Africa. A further aim is to provide a comprehensive national guideline on how best to effectively disseminate and communicate such information and to establish an Estuary Early Warning (EEW) – Emergency Preparedness and Response Guide (EPRG), which forms part of the South African Multi-Hazard Early Warning System (MHEWS). It is critical that this EEW meets general principles accepted internationally for an effective Early Warning System. This thesis addresses the following key elements namely: (1) Risk identification, (2) Key drivers and contributions to estuary flooding, (3) Monitoring and alert early warning system, (4) Alert dissemination and (5) Response actions. Such pioneering work is an essential tool to translate science into policy, a crossover field, which remains poorly implemented.

Bibliography: leaves 107-115. / The greater Agulhas Current system is believed to form a key link in the global ocean circulation since it is the inter-ocean conduit for warm Indian Ocean water to the Atlantic Ocean. This system has been thought to derive its water from the South Equatorial Current via two sources: the Mozambique Current, and secondly, the East Madagascar Current. In spite of their global significance surprisingly few observations have been made in these source regions. In March 2000 a multidisciplinary cruise, the first one in 25 years, the Agulhas Current Sources Experiment (ACSEX-1) was carried out in the Mozambique Channel. The main aim of the ACSEX-1 cruise was to establish the existence, trajectory and hydrographic structure of the Mozambique Current. The use of satellite altimetry and numerical modeling revealed high mososcale activity in the Mozambique Channel. Thus guided by real-time altimetric data, the cruise sections intersected the main regions of high mesoscale activity in the centre of the Channel. From this dataset we are now able to determine whether the Mozambique Current is a continuous current or whether it exists merely of a train of eddies, as the altimetric data suggest.

Ocean gravity waves have been recorded near Melkbosstrand with a Wemelsfelder float type wave recorder on a sea tower situated in water 13m deep and one Kilometre offshore. The half hour records, taken twice daily between July 1972 - June 1973, have been analysed in a format uniform with that of Draper (1966). Measurements have been made from the wave records to check the value of the statistical ratios of Longuet- Higgins (1952) and Draper (1963). Ratios of maximum wave height to root mean square amplitude are lower than the theoretical values and reasons for this are discussed in the light of the records being taken in shallow water. Ratios of maximum wave height to the average upper one tenth wave height and average upper one third wave height agree closely with the theoretical values. Spectral analysis using the methods of Blackman and Tukey (1958), with 36 degrees of freedom and 100 lags have been applied to half hour records digitized once per second. Frequency, time diagrams contoured at equal energy values have been constructed for the period May and part of June 1973 on the basis of twice daily spectral measurements at 5 mHz intervals. The wave events, which include generation of swell from near South Georgia, (54°S, 37°W), have been compared with the South African Weather Bureau synoptic charts in order to identify generation centres. Repeated energy values near the 50 mHz band seem to be generated by winds of a lower velocity than required by Moskowitz (1964). Predominant siting of fetches at distances of the order of 1000 nautical miles accounts for the bulk of wave generation between latitudes 40° - 50°S. One storm appears to have originated near the Drake passage, but no events have been detected with waves passing through the passage. Analysis of wave spectra from several storms generating maximum wave heights over 5 metres, show that fully developed seas are seldom, if ever, present in Cape waters.

Bibliography: pages 103-114. / This study will attempt to begin the bio-optical characterisation of gelbstoff in southern African waters. Gelbstoff is a collective term, in itself perhaps an indication of a poorly understood phenomenon, given to a complex group of macromolecular organic compounds. It is the common bio-optical properties of these compounds that cause such an association, specifically the exponential decrease of absorption with increase in wavelength, resulting in typical absorption spectra decaying exponentially from a maximum in the ultra-violet. It is the accurate measurement or inferral of these spectra that is the primary aim of any bio-optical investigation of gelbstoff.

Sea level rise is highly interdisciplinary and its study entails not only oceanography, but other fields such as geomatics, climatology and geology. In this study we relied on the tools from geomatics to produce sea level rise maps in order to assess the vulnerability of the coastline of Cape Town, South Africa. After generating a DEM of a spatial resolution of 2 m from LiDAR point cloud data, we made use of GIS to design 4 sea level rise scenarios based on the RCP 2.6, RCP 4.5, RCP 6.0 and RCP 8.5 scenarios from the IPCC. Among the findings, it was found that 2.16 – 3.09 km² of land would be potentially inundated by 2100. The main receptors which were identified were sandy beaches, rocky shores and built-up land. Permanent inundation would possibly change the appeal and the nature of the beaches and affect the tourism industry. Hence the coastline requires immediate attention as it is one of the most valuable assets in the tourism industry. Tidal effect and storm surge effect were also identified as additional factors which brought temporary changes to the sea level in Cape Town. These impacts were further investigated in 8 coastal suburbs (Tableview, Woodbridge Island, Paarden Eiland, Foreshore, Sea Point, Glencairn, Fish Hoek and Strand.) Suitable adaptation strategies including hard protection measures (e.g groynes, sea walls, barriers) and soft protection measures (e.g beach nourishment) were also proposed for these 8 suburbs.