Description: Three data types are compared in the low-current-velocity regime in the southeastern North Atlantic, between 12-degrees-N and 30-degrees-N, 29-degrees-W and 18-degrees-W: Geosat altimetric sea level and derived surface geostrophic velocities, shallow current meter velocities, and dynamic heights derived from hydrographic data from cruises 4, 6, and 9 of the research vessel Meteor. The four current meter daily time series, at depths around 200 m, were smoothed over 1 month; the altimetric geostrophic velocities were computed from sea surface slopes over 142 km every 17 days. The correlation coefficients between the current meter and altimetric geostrophic velocities range between 0.64 and 0.90 for the moorings near 29-degrees-N but between 0.32 and 0.71 for the two around 21-degrees-N; the associated rms discrepancies between the two measurement types range between 1.5 and 4.4 cm/s, which is 49% to 127% of the rms of the respective current meter time series. Dynamic heights relative to 1950 dbar for the months of November 1986 (d(M4)), November 1987 (d(M6)), and February 1989 (d(M9)) were computed from Meteor cruises 4, 6, and 9. Both dynamic heights and altimetric heights (h(M4), h(M6), h(M9)) were averaged over 1-degrees boxes for the duration of each cruise. Differences d(M4) - d(M6) and d(M9) - d(M6) were computed only at bins where at least one station from both cruises existed, Assuming that dynamic heights d in dynamic centimeters are equivalent to sea level h in centimeters, the standard deviation sigma of the differences ((h(M4) - h(M6)) - (d(M4) - d(M6))) and corresponding M9 - M6 values was 2.1 cm. This value (squared) is only 13% of the (5.8 cm)2 variance of the dynamic height differences and is indistinguishable from the 2.7- to 5.6-cm natural variability of sea level in the area expected between the times when the ship and the satellite sampled the ocean. The areally averaged discrepancy for M9 - M6 was only 0.7 cm, but the corresponding value for M4 - M6 was 5.2 cm. A systematic difference between the water vapor corrections used before and after July 1987 is responsible for the M4 - M6 difference. The average M4 - M6 discrepancy is only 0.1 cm using the Fleet Numerical Oceanography Center correction, with a standard deviation of 3.1 cm. In spite of the underlying differences in sampling and physics, including unknown barotropic components not included in our hydrographic dynamic heights, and in data errors, including water vapor, ionospheric, and orbital effects on the altimetry, consistent interannual changes of the mean sea level from the independently obtained altimetric and hydrographic data sets are obtained, and correlated seasonal changes in surface currents are observed with both altimetry and current meters.

Description: The eastern part of the North Atlantic subtropical gyre is found in the region between the Azores and the Cape Verde Islands. A study of the gyre structure in the area east of 35°W between 8°N and 41°N is presented. The geostrophic flow field determined from historical temperature-salinity data sets by objective analysis indicates seasonal variations in shape but no significant changes in the magnitude of volume transports. The eastern part of the gyre has a larger east-west and smaller north-south extension in summer compared with the winter season. The center shifts by about 2° latitude to the south from winter to summer. Long-term temperature time series (6.5 years) from a mooring near the Azores are consistent with these results, showing always a consistent temperature increase at the beginning of the year which is apparently due to the displacement of the northeastern part of the gyre. A comparison between the mean flow fields and fields obtained from individual zonal sections indicates large deviations north and south of the gyre but small deviations within the gyre.

Description: The Guinea Dome is a permanent, quasi-stationary feature on the eastern side of the thermal ridge extending zonally across the tropical North Atlantic. The dome is a part of the large-scale near-surface flow fields associated with the North Equatorial Current, the North Equatorial Countercurrent and the North Equatorial Undercurrent. In the present study, historical and recently obtained hydrographic data are combined to investigate the thermohaline structure and geostrophic flow field in the vicinity of the dome. It is shown that the Guinea Dome exists throughout the year both in subthermocline and thermocline layers, that it has a corresponding cyclonic geostrophic flow, and that seasonal changes occur with respect to its vertical structure, horizontal extent, and position. The observational results are then compared with simulations from a general circulation model of the tropical Atlantic. A seven-year simulation forced by observed monthly winds is run to compute a monthly climatology. The model adequately simulates the Guinea Dome with respect to its structure, flow field, and seasonal variability.

Description: An analysis is presented of geostrophic volume transport across a zonal line along 28-degrees-N in the eastern Atlantic. The data are from an array of five moorings with 200-km spacing carrying temperature sensors and one current meter each for 1 or 2 years. Transport changes in the main thermocline relative to a fixed depth level are obtained by the use of temperature-salinity relationships. The transport variability is simulated by two propagating waves with first-order baroclinic mode structure. Solutions exist with annual and semi-annual periods and zonal wavelengths of 100-200 km and 300 km, respectively. Assuming quasi-geostrophic dynamics and using results on the Reynolds stress, the dominating waves of annual and semi-annual period are found to propagate to the southwest, with 45-degrees-60-degrees and 25-degrees to the south off the westward direction, respectively. Wave solutions with a 90-day period and a zonal wavelength of about 300 km are interpreted as an effect of barotropic waves arising due to horizontal temperature inhomogeneity. The propagation is about +/-25-degrees off the westward direction. In general, good approximations are obtained with the propagating wave simulations in the western and central part of the array, while large differences occur between observation and simulation close to the Canary archipelago. Possible causes for these differences are discussed.

Description: The Azores Current, south of the Azores Archipelago, is part of the subtropical North Atlantic gyre. Using an international hydrographic data set, we analyze mean and seasonal geostrophic transport fields in the upper 800 m of the ocean in order to determine the origin of the Azores Current in the western basin and seasonal changes in the related flow. Geostrophic currents are obtained by using the method applied by Stramma (1984) in the eastern basin. The Azores Current is found to originate in the area of the Southwest Newfoundland Rise (Figure 10). In winter an almost uniform current connects this region of origin with the Azores Current, while a branching into two current bands is observed in summer, with the southern band forming a marked cyclonic loop. Within the upper 800 m, all of the transport in the northern band and about 70% of the transport in the southern band recirculates in the eastern basin. Additionally, expendable bathythermograph data from the Azores Current region indicate an increase of eddy potential energy from winter to summer.

Description: Data from a large-scale moored array in the Iberian and Canary basins are used to determine the energies of barotropic and baroclinic M2 and S2 tides. An analysis of time-varying dynamical modes is performed. The results for barotropic modes confirm the global surface tide model results of Schwiderski (1980) for this region. The barotropic modes dominate in the deep basins, but increased baroclinic contributions are usually found over rough topography. At three locations near the continental slope in the southern Canary Basin the baroclinic modes dominate the barotropic mode. Results from an array of three moorings at the northern part of the Cape Verde Rise show an inverse behavior of barotropic and baroclinic energies, such that the baroclinic energy is steadily enhanced while the barotropic energy is reduced towards the continental margin. The increase in baroclinic energy is consistent with a generation of internal tides close to the shelf by surface tidal forcing over topography. Further evidence for this process is provided by the 2-week periodicity of the first-order baroclinic mode at the slope, corresponding to the spring-neap cycle of the barotropic tide.

Description: Mechanical energy terms are calculated from moored current meter data in the Cape Verde Frontal Zone (about 20°N, 25°W) and compared with those derived from a mesoscale model of this frontal region. The model is of the Bleck and Boudra [1981] type with isopycnal coordinates. An initially zonal jet, representing the Canary Current, is allowed to develop under the influence of baroclinic and barotropic instability processes. We find reasonable agreement in magnitudes, somewhat smaller in the model, and similar distributions in the vertical. This leads to the conclusion that the energy transfer terms from the model can be expected to be sufficiently close to reality. Determination of the transfer terms confirms that the energy transfer in the zone is dominated by baroclinic instability processes while barotropic instability is of minor importance. Average baroclinic instability energy transfer terms reach values of 2–3 μW m−3 in the pycnocline. Peak layer mean values are of the order 10 μW m−3. It is shown that the spatial distribution of active transfer regions is closely related to the structure of the transient eddy field in the frontal zone and that strong instability processes are restricted to the pycnocline.

Description: Recent hydrographic sections and high-quality historical data sets are used to determine geostrophic currents at subtropical latitudes in the western basin of the South Atlantic. Levels of no motion are determined from water mass information and a mass balance constraint to obtain the transport field of North Atlantic Deep Water (NADW) in this region. The incoming NADW transport of about 20 Sv from the north at 19 degrees S appears to be balanced by only one third of this transport leaving in the south and two thirds leaving to the east or northeast at the Mid-Atlantic Ridge. A simple model is proposed to determine the cause of the NADW branching. It is shown that potential vorticity preservation in the presence of topographic changes leads to a similar flow pattern as observed, with branching near the Vitoria-Trindade-Ridge and also an eastward turning of the southward western boundary current at about 28 degrees S, the latitude where a balance of planetary vorticity change and stretching can be expected.

Description: A 4-year expendable bathythermograph data set (1984–1987) from the area between southern Brazil and the Antarctic Peninsula provides information on the interannual variability of front locations. Two boundaries of subtropical water at different depths are identified north and south of the Brazil Current-Falkland (Malvinas) Current confluence zone. The northern Subtropical Front is displaced over a large part of the Argentine Basin from one observational period to the other. The shallow southern Subtropical Front appears fixed to the Falkland Escarpment. The Polar Front and Subantarctic Front locations do not vary much, except for one case where a cold core eddy in the Polar Frontal Zone causes a large northward displacement of the Subantarctic Front.

Description: The meridional oceanic transports of dissolved inorganic carbon and oxygen were calculated using six transoceanic sections occupied in the South Atlantic between 11 degrees S and 30 degrees S. The total dissolved inorganic carbon (TCO2) data were interpolated onto conductivity-temperature-depth data to obtain a high-resolution data set, and Ekman, depth-dependent and depth-independent components of the transport were estimated. Uncertainties in the depth-independent velocity distribution were reduced using an inverse model. The inorganic carbon transport between 11 degrees S and 30 degrees S was southward, decreased slightly toward the south, and was -2150 +/- 200 kmol s(-1) (-0.81 +/- 0.08 Gt C yr(-1)) at 20 degrees S. This estimate includes the contribution of net mass transport required to balance the salt transport through Bering Strait. Anthropogenic CO2 concentrations were estimated for the sections. The meridional transport of anthropogenic CO2 was northward, increased toward the north, and was 430 kmol s(-1) (0.16 Gt C yr(-1)) at 20 degrees S. The calculations imply net southward inorganic carbon transport of 2580 kmol s(-1) (1 Gt C yr(-1)) during preindustrial times. The slight contemporary convergence of inorganic carbon between 10 degrees S and 30 degrees S is balanced by storage of anthropogenic CO2 and a sea-to-air flux implying little local divergence of the organic carbon transport. During the preindustrial era, there was significant regional convergence of both inorganic carbon and oxygen, consistent with a sea-to-air gas flux driven by warming. The northward transport of anthropogenic CO2 carried by the meridional overturning circulation represents an important source for anthropogenic CO2 currently being stored within the North Atlantic Ocean.