Diapycnal nutrient fluxes on the northern boundary of Cape Ghir upwelling region

In this study we estimate diffusive nutrient fluxes in the northern region of Cape Ghir upwelling system (Northwest Africa) during autumn 2010. The contribution of two co-existing vertical mixing processes (turbulence and salt fingers) is estimated through micro- and fine-structure scale observations. The boundary between coastal upwelling and open ocean waters becomes apparent when nitrate is used as a tracer. Below the mixed layer (56.15±15.56 m), the water column is favorable to the occurrence of a salt finger regime. Vertical eddy diffusivity for salt (K_s) at the reference layer (57.86±8.51 m, CI 95%) was 3×10⁻⁵ (±1.89×10⁻⁹, CI 95%) m² s⁻¹. Average diapycnal fluxes indicate that there was a deficit in phosphate supply to the surface layer (6.61×10⁻⁴ mmol m⁻² d⁻¹), while these fluxes were 0.09 and 0.03 mmol m⁻² d⁻¹ for nitrate and silicate, respectively. There is a need to conduct more studies to obtain accurate estimations of vertical eddy diffusivity and nutrient supply in complex transitional zones, like Cape Ghir. This will provide us with information about salt and nutrients exchange in onshore–offshore zones.     

Computations of the energy flux to mixing processes via baroclinic wave drag on barotropic tides

The geographical distribution of barotropic to baroclinic transfer of tidal energy by baroclinic wave drag in the abyssal ocean is estimated. Using tidal velocities from a state-of-the-art numerical tidal model, the total loss of barotropic tidal energy in the deep ocean (between 70°S and 70°N and at depths greater than 1000 m) is estimated to be about 0.7 TW (M₂) corresponding to a mean value of the energy flux (e) of 2.4×10⁻³ W/m². The distribution of e is however highly skewed with a median of about 10⁻⁶ W/m². Only 10% of the area is responsible for more than 97% of the total energy transfer.To assess the possible influence of the relatively coarse bathymetry representation upon the present estimate, complementary calculations using better resolved sea floor topography are carried out over a control area around the Hawaiian Ridge. There are no major differences between the results achieved using the two different bathymetry databases. Fluxes of about 16 GW or 6×10⁻³ W/m² are computed in both cases, and the main contributions to the total fluxes originate in the same range of e-values and cover equally large parts of the total area.It is not clear whether the present model is valid at flat or subcritical bottom slopes. However, for the Hawaiian region, only 2% of the total energy flux as calculated in the present study originates in areas of critical and subcritical slopes.     

Mode-1 internal tides in the western North Atlantic Ocean

Mode-1 internal tides were observed the western North Atlantic using an ocean acoustic tomography array deployed in 1991–1992 centered on 25°N, 66°W. The pentagonal array, 700-km across, acted as an antenna for mode-1 internal-tides. Coherent internal-tide waves with O(1 m) displacements were observed traveling in several directions. Although the internal tides of the region were relatively quiescent, they were essentially phase locked over the 200–300 day data record lengths. Both semidiurnal and diurnal internal waves were detected, with wavenumbers consistent with those calculated from hydrographic data. The M₂ internal-tide energy flux was estimated to be about 70 W m⁻¹, suggesting that mode-1 waves radiate 0.2 GW of energy, with large uncertainty, from the Caribbean island chain at this frequency. A global tidal model (TPXO 5) suggested that 1–2 GW is lost from the M₂ barotropic tide over this region, but the precise value was uncertain because the complicated topography makes the calculation problematic. In any case, significant conversion of barotropic to baroclinic tidal energy does not occur in the western North Atlantic basin. It is apparent, however, that mode-1 internal tides have very weak decay and retain their coherence over great distances, so that ocean basins may be filled up with such waves. Observed diurnal amplitudes were an order of magnitude larger than expected. The amplitude and phase variations of the K₁ and O₁ constituents observed over the tomography array were consistent with the theoretical solutions for standing internal waves near their turning latitude. The energy densities of the resonant diurnal internal waves were roughly twice those of the barotropic tide at those frequencies.     

The importance of tides for sediment dynamics in the deep sea—Evidence from the particulate-matter tracer 234Th in deep-sea environments with different tidal forcing

Key aspects of deep-ocean fluid dynamics such as basin-scale (residual) and tidal flow are believed to have changed over glacial/interglacial cycles, with potential relevance for climatic change. To constrain the mechanistic links, magnitudes and temporal succession of events analyses of sedimentary paleo-records are of great importance. Efforts have been underway for some time to reconstruct residual-flow patterns from sedimentary records. Attempts to reconstruct tidal flow characteristics from deep-sea sediment deposits, however, are at a very early stage and first require a better understanding of the reflection of modern tides in sediment dynamics. In this context internal (baroclinic) tides, which are formed by the surface (barotropic) tide interacting with seafloor obstacles, are believed to play a particularly important role. Here we compare two modern deep-sea environments with respect to the effect of tides on sediment dynamics. Both environments are influenced by kilometre-scale topographic features but with vastly different tidal forcing: (1) two sites in the Northeast Atlantic (NEA) being surrounded by, or located downstream of, fields of short seamounts (maximum barotropic tidal current velocities ～5 cm s^?1); and (2) a site next to the Anaximenes seamount in the Eastern Mediterranean (EMed) (maximum barotropic tidal current velocities ～0.5 cm s^?1). With respect to other key fluid-dynamical parameters both environments are very similar. Signals of sedimentary particle dynamics, as influenced by processes taking place in the bottom boundary layer, were traced by the vertical water-column distribution of radioactive disequilibria (daughter/parent activity ratios≠1) between the naturally occurring, short-lived (half-life: 24.1 d) particulate-matter tracer ²³⁴Th relative to its very long-lived and non-particle-reactive parent nuclide ²³⁸U. Activity ratios of ²³⁴Th/²³⁸U<1 in water samples collected near the seafloor indicate fast ²³⁴Th scavenging onto particles followed by fast settling of these particles from the sampled parcel of water and, therefore, imply active sediment resuspension and dynamics on time scales of up to several weeks. In the Northeast Atlantic study region tides (in particular internal tides) are very likely to locally push total current velocities near the seafloor across the critical current velocity threshold for sediment erosion or resuspension whereas in the Eastern Mediterranean the tides are much too weak for this to happen. This difference in tidal forcing is reflected in a difference of the frequency of the occurrence of radioactive disequilibria <1 between total ²³⁴Th and ²³⁸U: In the near-bottom water column of the Northeast Atlantic region 59% of samples had detectable ²³⁴Th/²³⁸U disequilibria whereas at the Eastern Mediterranean site this fraction was only 7% (including a few disequilibria >1). The results of this study, therefore, add to the evidence suggesting that tides in the deep sea of the open oceans are more important for sediment dynamics than previously thought. It is hypothesised that (a) tide/seamount interactions in the deep open ocean control the local distribution of erosivity proxies (e.g., distributions of sediment grain sizes, heavy minerals and particle-reactive radionuclides) in sedimentary deposits and (b) the aforementioned topographically controlled sedimentary imprints of (internal) tides are useful in the reconstruction of past changes of tidal forcing in the deep sea.     

Tuning a physically-based model of the air–sea gas transfer velocity

Air–sea gas transfer velocities are estimated for one year using a 1-D upper-ocean model (GOTM) and a modified version of the NOAA–COARE transfer velocity parameterization. Tuning parameters are evaluated with the aim of bringing the physically based NOAA–COARE parameterization in line with current estimates, based on simple wind-speed dependent models derived from bomb-radiocarbon inventories and deliberate tracer release experiments. We suggest that A = 1.3 and B = 1.0, for the sub-layer scaling parameter and the bubble mediated exchange, respectively, are consistent with the global average CO₂ transfer velocity k. Using these parameters and a simple 2nd order polynomial approximation, with respect to wind speed, we estimate a global annual average k for CO₂ of 16.4 ± 5.6 cm h^?1 when using global mean winds of 6.89 m s^?1 from the NCEP/NCAR Reanalysis 1 1954–2000. The tuned model can be used to predict the transfer velocity of any gas, with appropriate treatment of the dependence on molecular properties including the strong solubility dependence of bubble-mediated transfer. For example, an initial estimate of the global average transfer velocity of DMS (a relatively soluble gas) is only 11.9 cm h^?1 whilst for less soluble methane the estimate is 18.0 cm h^?1.     

Influence of surface forcing on near-surface and mixing layer turbulence in the tropical Indian Ocean

An autonomous upwardly-moving microstructure profiler was used to collect measurements of the rate of dissipation of turbulent kinetic energy (ε) in the tropical Indian Ocean during a single diurnal cycle, from about 50 m depth to the sea surface. This dataset is one of only a few to resolve upper ocean ε over a diurnal cycle from below the active mixing layer up to the air–sea interface. Wind speed was weak with an average value of ~5 m s⁻¹ and the wave field was swell-dominated. Within the wind and wave affected surface layer (WWSL), ε values were on the order of 10⁻⁷–10⁻⁶ W kg⁻¹ at a depth of 0.75 m and when averaged, were almost a factor of two above classical law of the wall theory, possibly indicative of an additional source of energy from the wave field. Below this depth, ε values were closer to wall layer scaling, suggesting that the work of the Reynolds stress on the wind-induced vertical shear was the major source of turbulence within this layer. No evidence of persistent elevated near-surface ε characteristic of wave-breaking conditions was found. Profiles collected during night-time displayed relatively constant ε values at depths between the WWSL and the base of the mixing layer, characteristic of mixing by convective overturning. Within the remnant layer, depth-averaged values of ε started decaying exponentially with an e-folding time of 47 min, about 30 min after the reversal of the total surface net heat flux from oceanic loss to gain.     

The tracer signature of Antarctic Bottom Water and its spread in the Southwest Indian Ocean: Part I—CFC-derived translation rate and topographic control around the Southwest Indian Ridge and the Conrad Rise

In this paper we use a temperature and salinity based mixing model to assess the dilution of Antarctic Bottom Water (AABW) as it moves away from the Weddell Sea and into the Southwest Indian Ocean. By combining these results with CFC tracer measurements we have been able to make direct estimates of the large-scale translation rates of AABW in this region. We confirm that there is a major northward flow of AABW via a gap in the Southwest Indian Ridge at 30°E, and thence across the Agulhas Basin into the Mozambique Basin, with a translation rate from the Greenwich Meridian of 0.8–1.0 cm s⁻¹ and a volume transport between the two basins of 1.5×10⁶ m³ s⁻¹. A second, smaller flow cuts the Del Cano Rise through the Prince Edward Fracture Zone but is indistinguishable from the general bottom waters once on the northern side of the rise. The third flow moves eastward along the southern flank of the Del Cano Rise to pass north of the Conrad Rise. This has bottom velocities of 0.7 cm s⁻¹ and a volume transport of 1.6×10⁶ m³ s⁻¹. This water is probably the source of the AABW-rich Circumpolar Deep Water that flows through the gap to the west of Crozet Island, and which is traceable again at stations on the northern flanks of the ridge. Flow between the Conrad Rise and the Del Cano Rise is complicated by the influence of a fourth flow, the AABW that passes south of the former and thence into the Crozet Basin via the Crozet-Kerguelen Gap. We suggest that a portion of this flow loops into the channel between the Del Cano Rise and the Conrad Rise, modifying the bottom waters at the easternmost stations within this channel. We will go on in Part 2 of this paper to use these results to estimate the dissolution rates of silica in the SWINDEX area.     

Fluxes and exchange of suspended sediment in tidal inlets draining a degraded mangrove forest in Kenya

This study focuses on sediment exchange in the degraded Mwache mangrove forest wetland located in southern Kenya. It involved measurement of total and particulate organic suspended sediment concentrations (TSSC and POSC), tidal water elevation and current velocities. Results showed that in the heavily degraded backwater zone mangrove forest, the ebb and flood tide total sediment fluxes were of same order of magnitude, however, flood tide sediment fluxes were slightly higher than the ebb ones. In the moderately degraded frontwater zone mangrove forest, the flood tide sediment fluxes were more than 50% higher than the ebb tide fluxes. The peak net sedimentation in the highly degraded backwater zone was 4 g m⁻² tide⁻¹ but that in the moderately degraded frontwater zone was 63 g m⁻² tide⁻¹. In the frontwater zone of the mangrove forest, the peak instantaneous ebb tide sediment flux was 3206 kg tide⁻¹ equivalent to 35.6 g m⁻² tide⁻¹ and the flood one 8574 kg tide⁻¹ (95 g m⁻² tide⁻¹). The peak instantaneous flood and ebb tide particulate organic sediment (POS) fluxes in the frontwater zone mangrove forest were 1316 kg tide⁻¹ (15 g m⁻² tide⁻¹) and 587 kg tide⁻¹ (6.5 g m⁻² tide⁻¹), respectively. The peak ebb and flood tide sediment fluxes in the backwater mangrove forest were 3206 kg tide⁻¹ (36 g m⁻² tide⁻¹) and 3305 kg tide⁻¹ (36.7 g m⁻² tide⁻¹), respectively. In case of POS fluxes in the backwater zone mangrove forest, the peak flood period POS flux was 969 kg tide⁻¹ (10.7 g m⁻² tide⁻¹) while the ebb period one was 484 kg tide⁻¹ (5.4 g m⁻² tide⁻¹). In both highly degraded backwater and moderately degraded frontwater zone of the mangrove forest, there is net import of sediments. However, the net import is relatively lower in the backwater zone forest where the trapping efficiency is 27%. In the moderately degraded frontwater zone of the mangrove forest, the sediment trapping efficiency is 65%. The net sediment import occurs mainly in periods of high river discharge in both neap and spring tides, but occurs only in spring tides during dry season. The net accretion rates in the backwater and frontwater zone mangrove forests are 0.25 and 3.5 cm year⁻¹, respectively.     

The swarm dynamics of northern krill (Meganyctiphanes norvegica) and pteropods (Cavolinia inflexa) during vertical migration in the Ligurian Sea observed by an acoustic Doppler current profiler

A ship-mounted 153 kHz narrow-band ADCP and 1 m² MOCNESS were deployed between 16 and 24 Sept. 1997 in the Ligurian central zone (∼43°20′N 7°48′E). Results from both instruments showed that the zooplankton community performed vertical migrations that conformed to the classical pattern of ascent at dusk (∼18:30 h) and descent at dawn (∼06:30 h). Depth-discrete net samples between 0 and 500 m showed that the community was dominated by two species, the euphausiid Meganyctiphanes norvegica (Northern krill) and the pteropod Cavolinia inflexa, which migrated in separate discrete bands that were detectable by the ADCP. Information from the ADCP was used to estimate vertical migration speed in two ways: (i) from the trajectory of the back-scattering bands over time and (ii) from the Doppler-shift vertical velocity measured within depth zones at the corresponding time and depth of these bands. Estimates of the migration speed of C. inflexa were between 2 and 7 cm s⁻¹ upwards and between 4 and 7 cm s⁻¹ downwards. M. norvegica was estimated to migrate at speeds between 7 and 8 cm s⁻¹ upwards and over 11 cm s⁻¹ downwards. The consistently lower migration speeds estimated from Doppler measurements as compared with estimates obtained from measuring trajectories of back-scattering bands over time was believed to result from a methodological artefact. The Doppler measurements were nevertheless useful in a relative sense in revealing the relative speed of individuals within swarms. It was shown that individuals at the front of the upwardly migrating band of M. norvegica moved more slowly than those at the rear. These results illustrate the extra biological information that can be obtained by ADCPs compared with conventional echo-sounders.     

A note on salt intrusion in funnel-shaped estuaries: Application to the Incomati estuary,Mozambique

Salt intrusion in estuaries is important for ecological reasons as well as water extraction purposes. The distance salt intrudes upstream depends on a number of factors, including river discharge, tidal and wind mixing and gravitational circulation. In this paper, an analytical solution is presented for the salt intrusion in a well mixed, funnel-shaped estuary whose cross sectional area decreases exponentially (with decay coefficient β) with distance, x, inland, and in which longitudinal mixing is constant along the length of the estuary. The solution predicts that a graph of the logarithm of salinity against exp (βx) should be a straight line, with slope proportional to the mixing coefficient K_x. The solution is tested against observations from 15 surveys over a four-year period in the Incomati estuary. Good straight line fits, as predicted, are observed on all surveys, with a mean R² = 0.97. The average value of K_x for all surveys is 38 m² s⁻¹. The solution is used to make predictions about the minimum river flow required to prevent salt intruding to an extent where it causes a detrimental effect on water extraction. The minimum recommended river flow required to prevent this is 35 m³ s⁻¹. In recent years, flow has fallen below this level for several months each year.     

Biogeochemical patterns in the Atlantic Inflow through the Strait of Gibraltar

The effects of tidal forcing on the biogeochemical patterns of surface water masses flowing through the Strait of Gibraltar are studied by monitoring the Atlantic Inflow (AI) during both spring and neap tides. Three main phenomena are defined depending on the strength of the outflowing phase predicted over the Camarinal Sill: non-wave events (a very frequent phenomenon during the whole year); type I Internal wave events (a very energetic event, occurring during spring tides); and type II Internal wave events (less intense, occurring during neap tides).During neap tides, a non-wave event comprising oligotrophic open-ocean water from the Gulf of Cádiz is the most frequent and clearly dominant flow through the Strait. In this tidal condition, the inflow of North Atlantic Central Water (NACW) provides the main nutrient input to the surface layer of the Alboran Sea, supplying almost 70% of total annual nitrate transport to the Mediterranean basin. A low percentage of active and large phytoplankton cells and low average concentrations of chlorophyll (0.3–0.4 mg m⁻³) are found in this tidal phase. Around 50% of total annual phytoplankton biomass transport into the Mediterranean Sea through the Strait presents these oligotrophic characteristics.In contrast, during spring tides, patches of water with high chlorophyll levels (0.7–1 mg m⁻³) arrive intermittently, and these are recorded concurrently with the passage of internal waves coming from the Camarinal Sill (type I internal wave events). When large internal waves are arrested over the Camarinal Sill this implies strong interfacial mixing and the probable concurrent injection of coastal waters into the main channel of the Strait. These processes result in a mixed water column in the AI and can account for around 30% of total annual nitrate transport into the Mediterranean basin. Associated with type I internal wave events there is a regular inflow of large and active phytoplankton cells, transported in waters with relatively high nutrient concentrations, which constitutes a significant supply of planktonic resources to the pelagic ecosystem of the Alboran Sea (almost 30% of total annual phytoplankton biomass transport).     

Variations of deep western boundary currents in the Melanesian Basin in the western North Pacific

Five moorings ML1–ML5 were deployed on the slope of the Solomon Rise in the Melanesian Basin in the western North Pacific, northeastward at increasing water depths. We measured the velocities of the western branch current of the deep western boundary current (DWBC) and the upper deep current carrying the Lower and Upper Circumpolar Waters (LCPW, UCPW), respectively. The daily mean velocity data from 1–3 February 1999 to 24–26 February 2000 were analyzed, and variability of the DWBCs was clarified. Although the current meters did not entirely cover the western branch current of the DWBC composed of two or three streams, a stream of the western branch current was observed at a depth of 4700 m at ML4 or 4260 m at ML5 for more than half of the observation period. The stream had a mean velocity of 3.7 cm s⁻¹ and alternated between ML4 and ML5 at 20- to 40-day intervals without occupying both of ML4 and ML5 simultaneously. This shows that the width of the stream is less than 120 km (distance between ML4 and ML5), and the position changes in a similar range. In contrast to the velocity of the eastern branch current of the DWBC, that of the western branch current did not decrease with decreasing depths to 4000 m. This reflects the vertical division into the branch currents by the bifurcation of the DWBC. The western branch current of the DWBC is located at the deep side of the countercurrent which was almost always observed at depths of 3880 and 4080 m at ML3. The countercurrent was thought to be the return flow of the western branch current that is partly reversed in the East Mariana Basin. The previous estimate of geostrophic transport of LCPW at the time of the mooring deployment was corrected to 1.4 Sv (10⁶ m³ s⁻¹) in the western branch current, 1.7 Sv in the countercurrent, and 1.1 Sv in the inflow to the East Caroline Basin. The upper deep current was located over the slope of the Solomon Rise with water depth less than 4500 m including ML1–ML3. It flowed at depths of approximately 2000–3500 m with the highest velocity in the middle of this layer and seldom reached the near-bottom where eddy-like disturbances existed. Its volume transport at the mooring deployment was 10.4 Sv. The upper deep current during the first half of the observation period had double cores divided by the countercurrent at ML1, whereas that during the second half had a single core, as the countercurrent at ML1 disappeared in early September 1999. The vector mean velocities of the upper deep current were 5.0 (2650 m, ML2) and 3.6 cm s⁻¹ (1880 m, ML3) during the first half of the observation period and 7.0 cm s⁻¹ (2670 m, ML1) during the second half; they ranged from 3 to 7 cm s⁻¹. Similarly, those of the countercurrent at ML1 during the first half were 6.4, 3.8, 4.6 cm s⁻¹ (2170, 2670, 3570 m).     

Oxygen dynamics in the North Atlantic subtropical gyre

Dissolved oxygen (DO) in the ocean is a tracer for most ocean biogeochemical processes including net community production and remineralization of organic matter which in turn constrains the biological carbon pump. Knowledge of oxygen dynamics in the North Atlantic Ocean is mainly derived from observations at the Bermuda Atlantic Time-series Study (BATS) site located in the western subtropical gyre which may skew our view of the biogeochemistry of the subtropical North Atlantic. This study presents and compares a 15 yr record of DO observations from ESTOC (European Station for Time-Series in the Ocean, Canary Islands) in the eastern subtropical North Atlantic with the 20 yr record at BATS. Our estimate for net community production of oxygen was 2.3±0.4 mol O₂ m⁻² yr⁻¹ and of oxygen consumption was −2.3±0.5 mol O₂ m⁻² yr⁻¹ at ESTOC, and 4 mol O₂ m⁻² yr⁻¹ and −4.4±1 mol m⁻² yr⁻¹ at BATS, respectively. These values were determined by analyzing the time-series using the Discrete Wavelet Transform (DWT) method. These flux values agree with similar estimates from in-situ observational studies but are higher than those from modeling studies. The difference in net oxygen production rates supports previous observations of a lower carbon export in the eastern compared to the western subtropical Atlantic. The inter-annual analysis showed clear annual cycles at BATS whereas longer cycles of nearly 4 years were apparent at ESTOC. The DWT analysis showed trends in DO anomalies dominated by long-term perturbations at a basin scale for the consumption zones at both sites, whereas yearly cycles dominated the production zone at BATS. The long-term perturbations found are likely associated with ventilation of the main thermocline, affecting the consumption and production zones at ESTOC.     

The natural diet of a hexactinellid sponge: Benthic–pelagic coupling in a deep-sea microbial food web

Dense communities of shallow-water suspension feeders are known to sidestep the microbial loop by grazing on ultraplankton at its base. We quantified the diet, rates of water processing, and abundance of the deep-sea hexactinellid sponge Sericolophus hawaiicus, and found that, like their demosponge relatives in shallow water, hexactinellids are a significant sink for ultraplankton. S. hawaiicus forms a dense bed of sponges on the Big Island of Hawaii between 360 and 460 m depth, with a mean density of 4.7 sponges m⁻². Grazing of S. hawaiicus on ultraplankton was quantified from in situ samples using flow cytometry, and was found to be unselective. Rates of water processing were determined with dye visualization and ranged from 1.62 to 3.57 cm s⁻¹, resulting in a processing rate of 7.9±2.4 ml sponge⁻¹ s⁻¹. The large amount of water processed by these benthic suspension feeders results in the transfer of approximately 55 mg carbon and 7.3 mg N d⁻¹ m⁻² from the water column to the benthos. The magnitude of this flux places S. hawaiicus squarely within the functional group of organisms that link the pelagic microbial food web to the benthos.     

Hydrodynamic controls on cold-water coral growth and carbonate-mound development at the SW and SE Rockall Trough Margin,NE Atlantic Ocean

Long-term (⩽1-year) records obtained by seabed observatories (BOBO) and repeated (24-h) CTD casts show the presence of a highly energetic environment in and around two cold-water carbonate-mound provinces, on the Southwest and Southeast Rockall Trough (SW and SE RT) margin. Carbonate mounds, covered with a thriving coral cover, are embedded mainly in the Eastern North Atlantic Water (ENAW) and are observed in a confined bathymetric zone between 600 and 1000 m water depth. Cold-water corals seem to be restricted in their growth by temperature and food availability. The presence of living corals on top of the carbonate mounds appears linked to the presence of internal waves and tidal currents in the water column, and consequently carbonate mound structures are shaped by the local hydrodynamic regime. Mound clusters have an elongated shape perpendicular to the regional contours and corresponding to the direction of the highest current speeds. On the SW RT margin temperature, salinity and current speed reflect a diurnal tidal pattern, causing maximum temperature variations at 900 m depth of more than 3 °C. Current speeds up to 45 cm s⁻¹ occur, and a residual current of 10 cm s⁻¹ is directed along the slope to the southwest. At the SE RT margin the temperature of the bottom water fluctuates more than 1 °C with a semi-diurnal tidal cyclicity. Amplitudes of average and peak current speeds here are comparable with those measured on the southwest margin, but the residual current in this area is directed to the northeast. Tidal currents and internal waves at both margins force the formation of intermediate and bottom nepheloid layers and bring fresh food particles with increased velocity to the mounds. The distribution of corals in both mound areas is considered directly related to the presence of enhanced turbidity. An increase in temperature can be directly related to an increase in the amount of particles in the water column. Current velocity increases when a transition occurs from cold to warm waters. High current velocities prevent local sedimentation but provide sufficient food particles to the corals, so that the corals thrive at the mound summits.     

Total organic and inorganic carbon exchange through the Strait of Gibraltar in September 1997

The total organic carbon (TOC) and total inorganic carbon (C_T) exchange between the Atlantic Ocean and the Mediterranean Sea was studied in the Strait of Gibraltar in September 1997. Samples were taken at eight stations from western and eastern entrances of the Strait and at the middle of the Strait (Tarifa Narrows). TOC was analyzed by a high-temperature catalytic oxidation method, and C_T was calculated from alkalinity–pH_T pairs and appropriate thermodynamic relationships. The results are used in a two-layer model of water mass exchange through the Strait, which includes the Atlantic inflow, the Mediterranean outflow and the interface layer in between. Our observations show a decrease of TOC and an increase of C_T concentrations from the surface to the bottom: 71–132 μM C and 2068–2150 μmol kg⁻¹ in the Surface Atlantic Water, 74–95 μM C and 2119–2148 μmol kg⁻¹ in the North Atlantic Central Water, 63–116 μM C and 2123–2312 μmol kg⁻¹ in the interface layer, and 61–78 μM C and 2307–2325 μmol kg⁻¹ in the Mediterranean waters. However, within the Mediterranean outflow, we found that the concentrations of carbon were higher at the western side of the Strait (75–78 μM C, 2068–2318 μmol kg⁻¹) than at the eastern side (61–69 μM C, 2082–2324 μmol kg⁻¹). This difference is due to the mixing between the Atlantic inflow and the Mediterranean outflow on the west of the Strait, which results in a flux of organic carbon from the inflow to the outflow and an opposite flux of inorganic carbon. We estimate that the TOC input from the Atlantic Ocean to the Mediterranean Sea through the Strait of Gibraltar varies from (0.97±0.8)10⁴ to (1.81±0.90)10⁴ mol C s⁻¹ (0.3×10¹² to 0.56×10¹² mol C yr⁻¹), while outflow of inorganic carbon ranges from (12.5±0.4)10⁴ to (15.6±0.4)10⁴ mol C s⁻¹ (3.99–4.90×10¹² mol C yr⁻¹). The high variability of carbon exchange within the Strait is due to the variability of vertical mixing between inflow and outflow along the Strait. The prevalence of organic carbon inflow and inorganic carbon outflow shows the Mediterranean Sea to be a basin of active remineralization of organic material.     

Diapycnal mixing by meso-scale eddies

The mean available potential energy released by baroclinic instability into the meso-scale eddy field has to be dissipated in some way and Tandon and Garrett [Tandon, A., Garrett, C., 1996. On a recent parameterization of mesoscale eddies. J. Phys. Oceanogr. 26 (3), 406–416] suggested that this dissipation could ultimately involve irreversible mixing of buoyancy by molecular processes at the small-scale end of the turbulence cascade. We revisit this idea and argue that the presence of dissipation within the thermocline automatically requires that a component of the eddy flux associated with meso-scale eddies must be associated with irreversible mixing of buoyancy within the thermocline. We offer a parameterisation of the implied diapycnal diffusivity based on (i) the dissipation rate for eddy kinetic energy given by the meso-scale eddy closure of Eden and Greatbatch [Eden, C., Greatbatch, R.J., 2008. Towards a meso-scale eddy closure. Ocean Modell. 20, 223–239.] and (ii) a fixed mixing efficiency. The implied eddy-induced diapycnal diffusivity (κ) is implemented in a coarse resolution model of the North Atlantic. In contrast to the vertical diffusivity given by a standard vertical mixing scheme, large lateral inhomogeneities can be found for κ in the interior of the ocean. In general, κ is large, i.e. up to o(10) cm²/s, near the western boundaries and almost vanishing in the interior of the ocean.     

The thermohaline structure and evolution of the deep waters in the Canada Basin,Arctic Ocean

Below the sill depth (at about 2400 m) of the Alpha-Mendeleyev ridge complex, the waters of the Canada Basin (CB) of the Arctic Ocean are isolated, with a ¹⁴C isolation age of about 500 yr. The potential temperature θ decreases with depth to a minimum θ_m≈−0.524°C near 2400 m, increases with depth through an approximately 300 m thick transition layer to θ_h≈−0.514°C, and then remains uniform from about 2700 m to the bottom at 3200–4000 m. The salinity increases monotonically with depth through the deep θ_m and transition layer from about 34.952 to about 34.956 and then remains uniform in the bottom layer. A striking staircase structure, suggestive of double-diffusive convection, is observed within the transition layer. The staircase structure is observed for about 1000 km across the basin and has been persistent for more than a decade. It is characterized by 2–3 mixed layers (10–60 m thick) separated by 2–16 m thick interfaces. Standard formulae, based on temperature and salinity jumps, suggest a double-diffusive heat flux through the staircase of about 40 mW m⁻², consistent with the measured geothermal heat flux of 40–60 mW m⁻². This is to be expected for a scenario with no deep-water renewal at present as we also show that changes in the bottom layer are too small to account for more than a small fraction of the geothermal heat flux. On the other hand, the observed interfaces between mixed layers in the staircase are too thick to support the required double-diffusive heat flux, either by molecular conduction or by turbulent mixing, as there is no evidence of sufficiently vigorous overturns within the interfaces. It therefore seems, that while the staircase structure may be maintained by a very weak heat flux, most of the geothermal heat flux is escaping through regions of the basin near lateral boundaries, where the staircase structure is not observed. The vertical eddy diffusivity required in these near-boundary regions is O(10⁻³) m² s⁻¹. This implies Thorpe scales of order 10 m. We observe what may be Thorpe scales of this magnitude in boundary-region potential temperature profiles, but cannot tell if they are compensated by salinity. The weak stratification of the transition layer means that the large vertical mixing rate implies a local dissipation rate of only O(10⁻¹⁰) W kg⁻¹, which is not ruled out by plausible energy budgets. In addition, we discuss an alternative scenario of slow, continuous renewal of the CB deep water. In this scenario, we find that some of the geothermal heat flux is required to heat the new water and vertical fluxes through the transition layer are reduced.     

Interhemispheric exchanges of mass and heat in the Atlantic Ocean in January–March 1993

Hydrographic, geochemical, and direct velocity measurements along two zonal (7.5°N and 4.5°S) and two meridional (35°W and 4°W) lines occupied in January–March, 1993 in the Atlantic are combined in an inverse model to estimate the circulation. At 4.5°S, the Warm Water (potential temperature θ>4.5°C) originating from the South Atlantic enters the equatorial Atlantic, principally at the western boundary, in the thermocline-intensified North Brazil Undercurrent (33±2.7×10⁶ m³ s⁻¹ northward) and in the surface-intensified South Equatorial Current (8×10⁶ m³ s⁻¹ northward) located to the east of the North Brazil Undercurrent. The Ekman transport at 4.5°S is southward (10.7±1.5×10⁶ m³ s⁻¹). At 7.5°N, the Western Boundary Current (WBC) (17.9±2×10⁶ m³ s⁻¹) is weaker than at 4.5°S, and the northward flow of Warm Water in the WBC is complemented by the basin-wide Ekman flow (12.3±1.0×10⁶ m³ s⁻¹), the net contribution of the geostrophic interior flow of Warm Water being southward. The equatorial Ekman divergence drives a conversion of Thermocline Water (24.58⩽σ₀<26.75) into Surface Water (σ₀<24.58) of 7.5±0.5×10⁶ m³ s⁻¹, mostly occurring west of 35°W. The Deep Water of northern origin flows southward at 7.5°N in an energetic (48±3×10⁶ m³ s⁻¹) Deep Western Boundary Current (DWBC), whose transport is in part compensated by a northward recirculation (21±4.5×10⁶ m³ s⁻¹) in the Guiana Basin. At 4.5°S, the DWBC is much less energetic (27±7×10⁶ m³ s⁻¹ southward) than at 7.5°N. It is in part balanced by a deep northward recirculation east of which alternate circulation patterns suggest the existence of an anticyclonic gyre in the central Brazil Basin and a cyclonic gyre further east. The deep equatorial Atlantic is characterized by a convergence of Lower Deep Water (45.90⩽σ₄<45.83), which creates an upward diapycnal transport of 11.0×10⁶ m³ s⁻¹ across σ₄=45.83. The amplitude of this diapycnal transport is quite sensitive to the a priori hypotheses made in the inverse model. The amplitude of the meridional overturning cell is estimated to be 22×10⁶ m³ s⁻¹ at 7.5°N and 24×10⁶ m³ s⁻¹ at 4.5°S. Northward heat transports are in the range 1.26–1.50 PW at 7.5°N and 0.97–1.29 PW at 4.5°S with best estimates of 1.35 and 1.09 PW.