No description available.

Slow-spreading mid-ocean ridges such as the Mid-Atlantic Ridge are characterized by deep axial rift valleys which are isolated from the water on the ridge flanks. Topographic effects therefore have a significant impact on the rift-valley hydrography and dynamics but little is known about the details. Known processes of global importance acting near the axes of midocean ridges include high rates of diapycnal mixing associated with the rough topography and high-temperature hydrothermal circulation, a major source for a number of chemical constituents of the ocean. Physical data sets from the rift valley of two connected segments of the Mid-Atlantic Ridge, which include the largest known hydrothermal vent field of the Atlantic, were analyzed to investigate the segment-scale hydrography, dynamics and geothermal fluxes. The data include two quasi-synoptic hydrographic and particle plume surveys (one year apart) and one-year-long records from an array of moored current meters. The hydrographic properties of the rift-valley water were similar during the two surveys, suggesting a stable state characterized by inflow from the eastern ridge flank, unidirectional along-segment flow (directly observed during an entire year), and monotonic along-valley hydrographic gradients consistent with high rates of diapycnal mixing. Geothermal processes do not appear to contribute significantly to these patterns. The data contain signatures of a range of dynamical processes consistent with high rates of diapycnal mixing, including hydraulically controlled sill flows, topographic lee waves and high-energy tidal flows. The spatial distribution of the light-scattering anomalies associated with the dispersing hydrothermal particle plume are consistent with the dynamical observations. Close to the vent field the particle distribution is highly inhomogeneous but density-averaged profiles in¬ dicate that the mean plume is Gaussian in depth. To quantify the fluxes associated with the hydrothermal plume the corresponding hydrographic anomalies were determined. The ABSTRACT ii complexity of the hydrography within the rift valley precludes the application of "standard" methods ho that a new method had to be developed resulting in the first quantitative hydrographic anomaly measurements of an Atlantic hydrothermal plume. The hydrographic and particle anomalies of this plume are linearly correlated, indicating that the particles behave conservatively in the near field. Estimates for the heat flux associated with the hydrothermal plume were derived using two established methods, one based on plume-rise modeling and the other on the advection of heat anomalies away from the vent field. Height-of-rise modeling yields values which are an order of magnitude too low because the plume model relies on a point-source assumption which is violated by the geometry of the vent field. The uncertainties associated with the advection method are particularly small at the site studied because of the uni-directionality of the flow field, the small uncertainties of the hydrographic anomaly measurements, and the Gaussian shape of the averaged near-field plume. The resulting estimate for the heat flux associated with the particle plume is 2.5 GW. Mass and heat budgets of the rift valley indicate that high diapycnal diffusivities are required to account for the hydrographic observations and suggest that a portion of the water flowing along the rift valley may be lost to the overlying water column. Inspired by the observations a simple analytical and numerical model for the flow within the rift valley was developed. The results indicate that the rift valley acts as an efficient low-pass filter with characteristic time scales of weeks to months, providing a plausible explanation for the persistence of the along-segment flow.

This work addresses two problems: 1) the dispersion induced by a cloud of vortices near a straight coast-line when the bottom is flat and the coastal boundary is a vertical plane or when the bottom is a planar slope; 2) the dynamics of vortices moving over a planar sloping bottom. Vortices near a vertical boundary are studied by the well-known method of images. For a plane sloping bottom we describe and develop a model, first introduced by Peregrine (1996) that uses a sector of a vortex ring to model a vortex in a wedge of fluid, where the wedge is formed by the water surface and by the planar sloping bottom. Numerical simulations using these free-slip analytical models are used to investigate the dispersion of vorticity and of a passive tracer induced by clouds of vortices. The results of the two models are compared. The dispersion of vortices and particles is mainly affected by the formation of vortex dipoles. The shoreline sets a preferential direction for the dispersion process and the dispersion normal to the shoreline is generally smaller, or bounded when the vortices forming the dipole have different absolute circulation. The dispersion of particles is generally smaller than the dispersion of vortices. In the second part of this work the analytical model of Peregrine (1996) for vortices moving over a planar slope at an angle a with the horizontal is tested against a set of laboratory experiments. Experiments were made by studying the dynamics of a vortex dipole moving towards a planar sloping beach. We measured the minimum distance from the shoreline reached by the vortices and their along-shore speed. The parameter ranges examined were 3<a<45, and 1-103<Re<6-103 (where Re is the Reynold's number of the vortices). We find a good agreement between the predictions and the observations when Re >~ 1500.

Phytoplankton make up approximately half of the global biosphere production. Climate change is predicted to affect phytoplankton productivity. Detecting the climate change signal in satellite records of productivity would imply that ocean primary production has been affected by anthropogenic influences. Long-term trends in chlorophyll (chl) concentration in the ocean have been observed by several studies. However, the effect of internal variability in chl was not taken into account in these observed trends. This thesis aims to perform a formal detection and attribution analysis on observed chl concentration using the optimal fingerprint (OF) method. The methodology has been applied to detect and attribute greenhouse gas induced climate change in sea-surface temperature records, ocean heat content, atmospheric air temperature etc., but this is the first attempt to apply it to ocean productivity records. The OF method was applied to monthly observations of chl data (1999-2005) from NASA’s Ocean Biogeochemical Model (NOBM) which assimilates satellite-derived chl. Control run and forced simulations from four Earth System Models were used to derive the internal variability of chl and response of chl to climate forcings (anthropogenic and natural), respectively. Three metrics were defined to describe the climate change signal in chl - spatial linear trend of chl; linear trend of zonal average; and time series of the size of the oligotrophic gyres. The OF technique of detection and attribution was implemented on the observational datasets for each of the three metrics. The amplitude of the responses provide an indication of whether a climate forcing signal is present in the observations. Out of the three metrics, the study demonstrated that the second metric (linear trend of zonal average in chl) is the best, and the third metric (size of the oligotrophic gyres) is the worst, 'direction' to look for a climate change signal in chl. Thus, metrics should be defined such that they capture the relevant change in chl and at the same time do not contain too much small scale variability which leads to noise. It was also illustrated that climate models do not necessarily simulate the internal variability of chl well, or the response of chl to climate forcings, indicating the need to improve the performance of climate models. A greenhouse gas signal was detected in observations in some regions of the ocean indicating that chl concentration is likely being affected by climate change. The canonical model of chl response to global warming, i.e. decrease in chl in lower latitudes and increase in chl in higher latitudes, was not consistently observed in all the regions of the ocean. This signifies that changing climate is affecting chl in a way which is not yet completely understood and in future the effects of climate change on chl may be surprisingly different from our current conceptual model.

The Nordic Seas provide the main oceanic connection between the Arctic and the deep global oceans via dense overflows between Greenland and Scotland, into the North Atlantic. An understanding of the circulation and exchanges of this region is vital for any consideration of the implications of high latitude climate change to variability in the Atlantic thermohaline circulation and consequences for regional (European) climate. This thesis makes use of a unique data set of near synoptic hydrographic and LADCP (lowered acoustic Doppler current profiler) measurements across the entire region during summer 1999. The box inverse method is applied to this hydrographic data, using computed geostrophic velocities referenced to detided LADCP measurements. The full summer Nordic Sea flux field (volume, heat and freshwater) is quantified, studying both the exchanges across the openings to the Nordic Seas, and the interior circulation. The total volume transports imply an inflow of 1.3 ± 0.5 Sv to the Nordic Seas from the Arctic Ocean, and a net export of 1.2 ± 0.5 Sv across the Greenland-Scotland Ridge into the North Atlantic. Within the Nordic Seas 4.0 ± 1.3 Sv of the warm saline inflow (s0 < 27.8) are converted to more dense waters, with the majority of the transformation (and ocean-atmosphere heat loss) occurring over the southern part of the Nordic Seas. The total heat convergence within the Nordic Seas is 137 ± 44 TW, giving an average flux of 51 ± 16 W m–2, and the net input of freshwater to the Nordic Seas is 0.059 ± 0.019 Sv. The sensitivity of the summer circulation and fluxes is investigated; considering the formal error estimates from the inverse model, together with the errors implied from inverse and oceanographic sensitivity tests. Supplementary winter data is used to construct a winter circulation providing an indication of significant seasonal variability. This infers that an estimate of the annual mean fluxes based on summer data alone cannot be justified.

The tropical ocean and atmosphere are a highly active and very important region of the globe. Climate phenomena such as El Ni˜no (Philander, 1990), the Tropical Atlantic Dipole, and the Indian Ocean Dipole, play an important role in global climate variability. The tropical atmospheric boundary layer is very sensitive to even small changes in the sea surface temperature (SST). Small SST anomalies in the tropics can lead to shifts in the large scale convection cells and result in atmospheric heating. There is potential for positive feedback between the tropical ocean and atmosphere. Ocean waves are capable of propagating long distances very fast. Barotropic waves (adjustments in free surface height) can propagate round the globe within days. Baroclinic waves, propagating along the thermocline are able to cross the equatorial Atlantic in 2 – 3 months. This work shows the potential for ocean wave propagation to influence global climate, by linking high latitude anomalies to tropical climate phenomena. The first part of this thesis is a detailed examination of the “Tropical Atlantic Dipole” (TAD). Analysis of model data shows a dipole pattern in the SST, with strong cross-equatorial asymmetry in the surface mixed layer. Below the mixed layer the pattern becomes symmetric, and Kelvin and Rossby wave like adjustment can be seen to occur. However, the timeseries is not sufficiently long to provide confidence in resolving the power spectrum, and as such the results are inconclusive. The complexity of the model makes it difficult to identify the mechanism(s) which are responsible for driving the dipole. An idealised basin model is used to examine high latitude anomalies which create equatorward propagating coastal Kelvin waves as a possible driving mechanism for the TAD. The results show that coastal Kelvin wave propagation can quickly transmit a signal from the high latitude anomaly to the equator, and equatorial Kelvin and Rossby wave propagation can quickly influence the entire tropical ocean. This suggests that forcing of the TAD may come from higher latitudes, although it is still not fully understood how a symmetric sub-surface signal can become asymmetric at the surface. Restoring surface boundary conditions limit the response of the model, restricting the formation of a TAD. A similar experiment, using an idealised coupled model configuration is suggested, but not possible in the time available. The second part of this thesis looks in detail at the role of the ocean in rapidly transmitting a high latitude response to the equator, using an existing coupled climate model configured with realistic land geometry and bottom topography. Simulations of a salinity anomaly in the Southern Ocean show that it is possible to create an equatorial response in SST within a month, with SST anomalies of 2.5± after 6 months. Barotropic Kelvin and Rossby wave propagation is shown to be important in creating such a rapid equatorial response. Two points that are identified from this experiment are examined in further detail using an idealised basin model. Firstly, a mechanism for energy exchange within the equatorial waveguide is tested. Results suggest that it is not the mechanism responsible for the signals seen in the coupled climate model. Secondly, idealised model integrations confirm that transmission of signals along topographic ridges is possible. Signals strong enough to excite equatorward coastal Kelvin wave propagation are able to use topography to cross the Southern Ocean and reach the coast of Australia.

The aim of this research was to develop a 'transportable' water quality model for the Solent and Southampton Water estuarine system, as a part of an effort to examine the effects of human activity and natural processes on estuarine water quality. Dissolved oxygen (DO), as a main indicator of water quality, is influenced by physical, chemical and biological processes, and has been chosen as the core parameter to link the different processes to be modelled. Monthly surveys of DO, planktonic community respiration rates and other water quality parameters (temperature, salinity, chlorophyll, suspended particulate matter, inorganic nutrients etc.) in the Itchen Estuary and Southampton Water were conducted from January 1998 to April 1999. DO data shows that Southampton Water is a relatively healthy estuary, despite receiving considerable loads of oxygen demanding organic sewage effluent discharged from a number of points. A persistent moderate DO sag (DO saturation > 80%) was observed in the upper Itchen Estuary throughout the year. In the lower Itchen Estuary and Southampton Water, the waters were DO saturated during the non-phytoplankton growth season. Surface DO supersaturation was observed during the phytoplankton growth season especially during algal blooms, but no severe DO depletion was detected following the bloom collapse. Community respiration rates maintained a substantial level in the upper Itchen estuary, while in the lower estuary respiration rates were low during the non-phytoplankton season and increased during the phytoplankton growth season. It is suggested that the high winter respiration rate in the upper Itchen Estuary are sustained by inputs from external sources (rivers, sewage and industrial effluents) and that the summer increase in the lower estuary is a consequence of phytoplankton photosynthesis. Nutrients in the Itchen Estuary and Southampton Water show mainly conservative behaviour in a plot of nutrient concentration against salinity. The removal of the nutrients by phytoplankton activity occurred at high salinities during the spring to summer period. A 3-D finite element baroclinic hydrodynamic model with two-equation q2-q2l turbulence closure has been developed including a mass conservation scheme. The model successfully simulated the tides, tidal currents, and estuarine circulation in the Southampton Water and Solent estuarine system. The modelled tidal induced residual currents and water mass transportation in Southampton Water and the Solent have been examined. Model results show the existence of a predominant westward tidal induced residual current in the Solent. The tidal induced residual water mass transport is extremely limited in Southampton Water, except near the entrance to Southampton Water, where it joins the Solent. The estuarine circulation with surface, seaward flowing fresher water and bottom, landward flowing saltier water provides the main mechanism for water mass transport in the model. The short residence time of waters in the estuary estimated from the survey salinity data confirmed how effective the estuarine circulation is for sea water from the Solent to replace the water within Southampton Water. The trapping effect of estuarine circulation is also crucial for the water quality in the estuary. A water quality model has been developed and coupled with the 3-D hydrodynamic model. The water quality model consists of an external (dissolved oxygen-biochemical oxygen demand) model, which models the direct impact of external inputs (riverine discharge, domestic and industrial effluents) to the water quality, and an internal model, which simulates the impact of local estuarine phytoplankton growth on the water quality. DO and dissolved inorganic nutrients are the 'link substances' between the external model and internal model. The integrated water quality model output has been compared against the survey data for 1998, and has been shown to reproduce the spatial and temporal change in oxygen, nutrients, chlorophyll and planktonic respiration in Southampton Water.

This study examines the response of the thermohaline circulation in the north Atlantic to steady and variable buoyancy forcing. The model used is a version of the MOMA model (Webb, 1996), updated to include a free surface and Gent & McWilliams mixing. The model’s resolution is coarse, 4 x4 degrees with 15 levels in the vertical. In a first instance, the model’s response to 14 different fixed thermal profiles is investigated, by systematically keeping the equator temperature fixed and then the northernmost temperature fixed. The results show that the models response differs for these two sets of experiments as one setup favours stratification while the other favours convection. In a second instance, the restoring field is made to oscillate over 17 different periods, ranging from 6 months to 32,000 years. The model's meridional overturning circulation (MOC) exhibits a very strong response on all timescales greater than 15 years, up to and including the longest forcing timescales examined. The peak-to-peak values of the MOC oscillations reach up to 125% of the steady-state maximum MOC and exhibit resonance-like behaviour, with a maximum at centennial to millennial forcing periods (depending on the vertical diffusivity). This resonance-like behaviour stems from the existence of two adjustment time scales, one of which is set by the vertical diffusion and another, which is set by the basin width. Finally, the study is extended to a double hemisphere basin. Again, the model's MOC exhibits a very strong response on all timescales in both hemispheres, up to and including the longest forcing timescales examined for either set of experiments with the amplitude of the oscillations reaching up to 140% of the steady-state maximum MOC and exhibiting resonance-like behaviour, with a maximum at centennial to millennial forcing periods. This resonance like behaviour is identical to what has been observed in a single hemisphere and occurs for the same reasons. What is novel is that when the forcing in the southern subordinate hemisphere lags that of the northern by half a period, the amplitude of the response is substantially greater for large forcing periods (millennial and above), particularly in the subordinate (southern) hemisphere. This happens because the basin has in effect two sources of deep water. This leads to colder bottom waters and thus greater stratification, which in turn stabilises the water column and thus reduces the value of the minimum overturning. The considerable deviation from the quasi-equilibrium response at all timescales above 15 years for both the single hemisphere and the double hemisphere experiments is surprising and suggests a potentially important role of the ocean circulation for climate even at Milankovich timescales.

Adjoint models calculate the exact sensitivity of an output function of a model to innitesimal perturbations in the forcing or initial conditions. In eddy resolving ocean models the presence of chaotic eddies is expected to lead to sensitivities to infinitesimal perturbations that are very dierent from the sensitivity to large perturbations and that no longer contain useful information. Previous studies disagree as to whether adjoint models can be used with eddy resolving ocean models on timescales longer than a few months. Here the MIT ocean general circulation model and its adjoint are used to look at the sensitivity of the time mean heat content, kinetic energy, available potential energy and thermocline depth to the sea surface temperature, zonal wind stress, and vertical diffusivity in an eddy resolving model of a zonally reentrant channel. Using the tangent linear model the non linear timescale of the eddy resolving model is estimated at around 200 days. The adjoint model is integrated over 278 days and 690 days to see whether useful information remains in the sensitivities calculated by the adjoint model for longer than the non linear timescale of the system. The usefulness of the information in the sensitivities calculated by the adjoint model is assessed by comparison with integrations of the full non linear forward model with large spatial scale perturbations to the forcing, finite difference gradient checks, and sensitivities calculated by an adjoint model in a non eddy resolving channel where the adjoint method is known to provide useful information. Finite difference gradients are found to be unsuitable for calculating sensitivities of time averaged climate quantities in an eddy resolving ocean model as they are also affected by chaos. Comparison of the sensitivities calculated by the adjoint model in the eddy and non eddy resolving models shows that information remains in the spatial structure of the adjoint model results in the eddy resolving model on a time scale of 278 days. In the non eddy resolving case the adjoint model results agree well with the perturbed forward model experiments, and are clearly climatically relevant on a timescale of 690 days. Use of a parameterisation scheme that reduces the eddy kinetic energy gives adjoint sensitivities that agree with well the perturbed forward model experiments after 690 days, although there are areas of extremely high adjoint sensitivity that may not be physically realistic. Without this parameterisation scheme, adjoint sensitivities involving dynamic variables grow exponentially with time as expected in a chaotic system, but at the end of the integration time of 690 days there is some agreement between the adjoint and forward model results for sensitivities involving thermodynamic variables only. These results show that even in the presence of chaotic eddies some useful information is retained in the adjoint model solution beyond the nonlinear timescale of the system.

The value of ocean diapycnal diffusivity (v) sets the rate at which dense bottom water can be mixed up through the stratified water column and thus plays an important role in the meridional overturning circulation (MOC). Previous idealised experiments and simplified theory suggest that the strength of the MOC and the ocean heat transport scale with the v. This study investigates the dependence of the MOC and other parameters on v using atmosphere-ocean general circulation models (AOGCM). Firstly, the dependence of the MOC strength on v is studied using a low resolution AOGCM with realistic geometry, FORTE, with spatially constant v values ranging from 0.1 cm2/s to an unrealistic high value of 5 cm2/s. At the cyclostationary state, global MOC strength is found to scale with v (in agreement with previous studies) according to a power law of 0.5. No power law is found for the MOC in the individual basins. The increase in MOC strength in the Atlantic and Pacific Oceans is associated with an increase in the ocean heat transport. The atmosphere responds to the change in the ocean state by a decrease of its energy transport and surface winds. Only a partial compensation is found between the ocean and atmosphere energy transport. The strength of v is found to have a strong impact on coupled phenomena, such as a cessation of El Niño at high v. Secondly, similar experiments are conducted with a state-of-the-art AOGCM, ECHAM5/ MPIOM. In this model, v is derived from a constant background diapycnal diffusion (b), wind induced mixing, the Richardson number and the convective adjustment. A set of 3 coupled experiments is conducted, with b = 0.1, 0.25 and 1 cm2/s. The scaling law from simple theory and the previous experiments with FORTE is not observed with this coupled model. At the cyclostationary state, the MOC strength weakens by 16% as b increases from 0.1 to 1 cm2/s. This behavior is not found when the experiments are repeated with an ocean-only model. The reduction in MOC in the coupled model is linked to a strong reduction in the convective mixing at high latitudes. The convective mixing is reduced by a continuous strong freshening in the Arctic region due to an increase in surface air temperature and melting of the sea-ice in the coupled experiments, which is not observed in the ocean-only experiments. The responses of the two coupled models show many similarities as b increases. Both models show convection in the Pacific for high values of b. The main difference is the response of the MOC in the Atlantic is linked to the different locations of the deep convection and their relative changes in the models. I conclude that the diapycnal mixing and the ocean-atmosphere interactions both control the strength of the MOC, and their influences cannot be considered separately.