Fronts, meanders and eddies in Drake Passage during the ANT-XXIII/3 cruise in January–February 2006: A satellite perspective

https://doi.org/10.1016/j.dsr2.2011.01.003Get rights and content

Abstract

We used satellite altimetric data to provide a context for the results of the ANT-XXIII/3 cruise in January–February 2006 both in time (16 years) and space (the whole of Drake Passage). The repeat of the hydrographical section within 3 weeks permitted different comparisons between the in-situ datasets and the satellite data products. Comparisons suggested that the multi-satellite product improved the temporal resolution on a Jason-1 track.

A detailed analysis of the four absolute dynamic topography maps contemporaneous with the ANT-XXIII/3 cruise permitted identification of the location of the frontal branches of the Antarctic Circumpolar Current, of the major meanders and eddies. This spatial context proved particularly valuable for the interpretation of the in-situ data (see companion papers of Provost et al., 2011; Renault et al., 2011; Sudre et al., 2011).

The altimetric time-series documented the long-term trends in sea-surface height, the recurrence of major frontal meanders and eddies and the statistical links between them. Negative trends in the Yaghan Basin indicated that both the Subantarctic Front and the Polar Front have shifted to the north of their climatological location. This northward shift in the Yaghan Basin contrasts with the large-scale southward shift in the Polar Front current core described in the literature, and is probably related to the local bottom topography in Drake Passage.

Sea-level anomaly patterns observed during the cruise were related to statistical modes of the corresponding variations in Drake Passage. For example, the southward meander of the Subantarctic Front at the entrance to Drake Passage was part of a dipole comprising an adjacent Polar Front meander and occurred with a close to annual periodicity.

A census of eddies in the Ona Basin revealed that the spatial distribution of anticyclonic eddies was consistent with generation from a meander of the Polar and Southern ACC Fronts over the Ona Seafloor Depression, while cyclonic eddies mostly originated from meanders of southern fronts associated with two rises on the continental slope: the Ona Rise and the Terror Rise.

Introduction

The Southern Ocean, the only ocean that circles the globe without interruption by a land mass, contains the largest of the world's ocean currents: the Antarctic Circumpolar Current (ACC). Drake Passage (DP) is the narrowest strait through which the ACC must pass. The ACC flow is concentrated in three deep-reaching oceanic frontal systems (Fig. 1a), from north to south: the Subantarctic Front (SAF), the Polar Front (PF) and the Southern ACC Front (SACCF). The southern limit of the ACC water masses is called the southern boundary (SB) of the ACC.

Satellite observations are particularly important for Southern Ocean studies, as in-situ observations and long time-series are scarce. In particular, sea-surface temperature (SST) obtained by satellite radiometry and sea-surface height by satellite altimetry have led to major advances in our description and understanding of the ACC.

Satellite sea-surface temperature data have been particularly helpful in describing the PF, which is the only ACC front with a strong signature in the SST. Moore et al. (1999) used the strong SST gradient across the PF to determine its location from weekly composites of the daily images obtained from satellite infrared measurements. However, because of cloud cover, about 45% of the data were unavailable, possibly introducing seasonal biases. Since June 2002, the Advanced Microwave Scanning Radiometer for the Earth Observing System (AMSR-E) provides all-weather SST measurements. From these measurements, Dong et al. (2006b) examined the position of the PF, the SST values and gradient at the PF, as well as their seasonal variability, during 3 years.

Satellite altimeters have provided continuous and high-accuracy data over the whole ocean since 1982. Altimetry has shown that there is high eddy kinetic energy (EKE) along the path of the ACC, and that the geographical distribution of the mean currents and their variability was strongly influenced by the bottom topography (Chelton et al., 1998, Chelton et al., 1990). Modeling and altimetric studies have examined the role of eddies with respect to the mean current. Morrow et al. (1994) found that eddies tend to add momentum to the mean current and on average to accelerate frontal jets, whereas Hughes and Ash (2001) showed that the strongest jets were clearly not accelerated by eddies. Hughes (1995) and Hughes et al. (1999) showed that part of the altimetric mesoscale variability was associated with propagating Rossby waves with wavelengths of about 300 km and periods of 4–12 months. In the mean axis of the ACC, these waves propagated eastward, and westward elsewhere. Sprintall (2003) examined the frequency and propagation paths of eddies between the SAF and the PF in DP from 8 years of TOPEX/Poseidon satellite measurements. The analysis suggested some seasonal variability in the number of eddies and propagation speeds of 20–30 cm s−1. Altimetric data indicated the presence of a significant peak in eddy kinetic energy over the Southern Ocean during 2000–2002. Meredith and Hogg (2006) showed that this EKE peak lagged, by approximately 2 years, a peak in the strength of the circumpolar westerly winds characterized by the Southern Annular Mode (SAM). Extending the method developed by Sokolov and Rintoul (2002), Sallée et al. (2008) used absolute sea-surface height to monitor the position of the SAF and PF during the period 1993–2005. They examined how shifts in the front positions were related to the main climate modes in the southern hemisphere, the SAM and El Niño Southern Oscillation (ENSO). They found strong regional differences between different sectors of the Southern Ocean. Recently, Sokolov and Rintoul, 2009a, Sokolov and Rintoul, 2009b found that each of three primary circumpolar ACC fronts consisted of multiple branches that were associated with local maxima in the gradient of absolute sea-surface height. In their view, nine sea-surface height contours were associated with the ACC jets, with three different contours (thus branches) for the SAF, three for the PF, two for the SACCF and one for the SB. They suggested that, in past studies, some of these branches might have been dismissed as transient or local features of limited interest.

Monitoring the ACC by satellite altimetry is a challenging task. Because the fronts are narrow and variable, they are also difficult to sample by altimetry, and mapped altimetric data tend to smooth the representation of the fronts and reduce the inferred speed of associated currents. The ACC fronts also have a large barotropic component which is measured by altimetry, but is difficult to estimate from hydrographic data. Despite these difficulties, several altimetric studies have attempted to estimate the mean circulation and transport of the ACC. Gille, 2008, Gille, 1994 applied a model of a meandering Gaussian jet to along-track altimetric data to estimate the mean sea-surface height across the Subantarctic Front and the Polar Front. The model estimated that the jets have an average Gaussian width of about 44 km in the meridional direction and meander about 75 km to either side of their mean locations. Chouaib et al. (2006) applied the same method to two satellite tracks in DP and found that the surface transports associated with the SAF and the PF were strongly anticorrelated so that the variance of total surface transport was much smaller than the variance of the surface transport of each individual front.

As a choke point, DP has been the site of numerous in-situ measurements, and there have been several attempts to combine in-situ measurements and altimetric data in the examination of the currents and fronts. Challenor et al. (1996) have estimated the surface geostrophic currents in DP from a combination of altimetric, hydrographic and current data. However, their technique had severe limitations, since it did not take into account the time difference between the ERS-1 pass and the shipboard measurement, leading to serious aliasing. Meredith and Hughes (2005) addressed the sampling frequency required for reliable monitoring of the ACC at interannual periods. They found that a sampling periodicity shorter than 10 days was required. Cunningham and Pavic (2007) used weekly maps of sea-level anomalies obtained by pooling data from several altimetric satellites. They computed time-series of surface geostrophic currents by combining these altimetric data with in-situ current–velocity measurements with a net error of around 15 cm s−1. Lenn et al. (2008) estimated the mean surface currents to the west of 60°W in DP using altimetric data and repeated upper-layer hydrographic measurements. In contrast to the smooth classical dynamic-height climatologies, their mean ACC fronts were associated with steep sea-surface height gradients and swift currents.

Drake Passage has a complex bottom topography with a number of fracture zones and ridges that delimit small basins: the Former Phoenix Plate Basin (FPPB), the Yaghan Basin (YB), the Ona Basin (OB) and the Northeast Scotia Sea Basin (NESSB) (Fig. 1A and B). Eddy kinetic energy is particularly strong at the entrance to and the exit from the Yaghan Basin, and local maxima in the standard deviation of the sea-surface height (>20 cm) are found over deep regions between the PF and SAF (Fig. 2).

In January–February 2006, R.V. Polarstern undertook a cruise across DP (ANT-XXIII/3) (Provost et al., 2011). Within 3 weeks it carried out two high-resolution full-depth hydrographical sections, using a conductivity–temperature–depth (CTD) sounder and a lowered acoustic Doppler current profiler (LADCP), along the Jason-1 ground track 104 (Fig. 1B). The track crossed a relative minimum in the standard deviation of sea-surface height in the Yaghan Basin (Fig. 2). To help with the interpretation of the in-situ along-track data, we examined satellite sea-surface temperature and sea-surface height data, taking advantage of their spatial coverage and the long time period, with two objectives: (i) to describe the mesoscale situation (fronts, eddies) in DP during the ANT-XXIII/3 cruise and (ii) to place the results obtained in the long-term oceanographic context. Furthermore, the repetition of the hydrographic section within a period of three weeks offered the opportunity to “validate” the satellite products.

The repetition of the hydrographic section within a period of 3 weeks offered a unique opportunity to compare different altimetry products with the in-situ data and showed that we could rely on the satellite data. We were therefore able to provide a detailed description of the frontal branches, meanders and eddies in DP during the ANT-XXIII/3 cruise as documented by satellite sea-surface temperature and sea-surface height data. The location of the front branches, in particular the merging or separation of the different branches, was closely related to specific topographic features in DP. The precise description of the meso-scale situation in the DP during the period of the cruise proved very useful for the interpretation of the in-situ along-track data by Provost et al., Renault et al., Sudre et al., 2011.

To put the results obtained during the cruise period in a longer-term context, the 16-year-long satellite altimetric record was used to investigate the following: (i) long-term trends in the location of frontal branches and in the eddy population; (ii) the recurrence of some of the large frontal meanders and the possible statistical relations among them; (iii) the eddy population in two particular regions: the entrance to the Yaghan Basin, following Sprintall's (2003) work, and the Ona Basin, an active region which has received little attention.

Following a brief description of the data in Section 2, we compare satellite data with in-situ data in Section 3. In Section 4, we describe the fronts and eddies during the cruise survey and examine the origin and fate of the two eddies crossed by the research vessel. In Section 5, after estimating the linear trends in sea level in DP over 16 years, we examine the recurrence of major frontal meanders and discover statistical relations among them. Then in Section 6 we focus on eddies at the entrance to the Yaghan Basin and in the Ona Basin. Finally, in Section 7, we summarize and discuss the results.

Section snippets

CTD and LADCP/SADCP datasets

The hydrographical data and LADCP/SADCP data used in the present paper are described in detail by Provost et al. (2011) and Renault et al. (2011), respectively. The data comprised two full-depth hydrographical sections and LADCP observations along the Jason-1 ground-track 104 (Fig. 1b). The first leg (southward) was performed from 16 to 26 January 2006, and the second, return leg (northward), from 31 January to 6 February 2006. The location of the main front branches changed somewhat between

Evaluation of the CNES-CLS09 mean dynamic topography along Jason track 104

Along-track MDT was first compared to the corresponding dynamic heights computed from the CTD data using various reference levels. The first level was the constant pressure surface at 2500 dbar, which is a commonly used as a reference level to estimate ACC transport (Whitworth and Peterson, 1985, Rintoul et al., 2002) (Fig. 6). The CTD-derived sea-surface dynamic heights for legs 1 and 2, the CNES-CLS09 MDT and the ADTs corresponding to the cruise dates (Fig. 6) were set to 0 m at 60°S.

Main SLA features

The strongest SLAs observed in DP during the cruise period are labeled on the SLA maps (Fig. 9), and the temporal evolution of their characteristics (position, size, amplitude) was carefully examined. Positive SLA (Pi, i=1,…,10) corresponded either to southward frontal meanders, or anticyclonic eddies, and negative SLA (Ni, i=1,…,7), either to northward frontal meanders, or cyclonic eddies.

Large positive anomalies (>20 cm) were observed on the western and eastern boundaries of the Yaghan Basin.

Recurrence of the meanders observed during the ANT-XXIII/3 cruise

We used the 16-year altimetric time-series to examine whether the major SLA features observed in DP are recurrent and bear any link to each other. These major features at the time of the cruise were the SAF southward meander at the entrance to DP (P1), the strong northward meander of the two northern branches of the PF around eddy N6 at the exit from DP and the large meander of the two northern branches of the PF to the north of the WSR (N3 and N4) (Fig. 9, Fig. 11). First of all, we examined

Generation and fate of the cyclonic eddy N2

During the cruise survey, a deep-reaching cyclonic eddy was observed at 63.7°W, below the Jason-1 track, and at 56.9°S, south of SAF-M (Renault et al., 2011). It corresponded to SLA N2 (−20 cm and 100 km diameter) (Fig. 9) and was associated with the presence of a cold water mass (T=0 °C at 250 m and T=6 °C at the surface) between stations 7 and 10 (leg 1) (Fig. 3A). It was also observed in SST (∼6 to 6.5 °C) (Fig. 5A). The eddy propagated northeastward and its edge was crossed during the second leg (

Summary and discussion

We used satellite altimetric data to provide a context for the ANT-XXIII/3 cruise in January–February 2006, both in time (16 years) and space (the whole of DP). The spatial context revealed by our detailed analysis of the four ADT maps contemporaneous with the ANT-XXIII/3 cruise proved particularly valuable for the interpretation of the in-situ data (see companion papers of Provost et al., 2011; Renault et al., 2011; Sudre et al., 2011). The temporal context made it possible to expose the

Acknowledgments

The authors are deeply grateful to the CNES (Centre National d'Etudes Spatiales) for the strong and constant support. We also wish to thank M-H Rio (CLS—Collecte Localisation Satellites) for sharing an early release of the mean dynamic topography. Our gratitude goes to Ray C. Griffiths and to the anonymous reviewers for their valuable comments on the manuscript.

References (48)

  • H. Bryden

    The Southern Ocean

  • P.G. Challenor et al.

    Measuring surface currents in Drake Passage from altimetry and hydrography

    J. Phys. Oceanogr.

    (1996)
  • D.B. Chelton et al.

    Geosat altimeter observations of the surface circulation of the Southern Ocean

    J. Geophys. Res.

    (1990)
  • D.B. Chelton et al.

    Geographic variability of the first baroclinic Rossby radius of deformation

    J. Phys. Oceanogr.

    (1998)
  • N. Chouaib et al.

    Variability of the Subantarctic and Polar Fronts in the Drake Passage as deduced from altimetry

    J. Mar. Res.

    (2006)
  • D.S. Crosby et al.

    A proposed definition for vector correlation in geophysics: theory and application

    J. Atmos. Oceanic Technol.

    (1993)
  • S.A. Cunningham et al.

    Transport and variability of the Antarctic Circumpolar Current in Drake Passage

    J. Geophys. Res.

    (2003)
  • S. Dong et al.

    Validation of the advanced microwave scanning radiometer for the Earth Observing System (AMSR-E) sea surface temperature in the Southern Ocean

    J. Geophys. Res.

    (2006)
  • S. Dong et al.

    Location of the Antarctic Polar Front from AMSR-E satellite sea surface temperature measurements

    J. Phys. Oceanogr.

    (2006)
  • N. Ducet et al.

    Global high-resolution mapping of ocean circulation from the combination of T/P and ERS-1/2

    J. Geophys. Res.

    (2000)
  • J.C. Fyfe et al.

    The role of poleward intensifying winds on Southern Ocean warming

    J. Climatol.

    (2007)
  • S.T. Gille

    Mean sea surface height of the Antarctic Circumpolar Current from Geosat data: method and application

    J. Geophys. Res.

    (1994)
  • S.T. Gille

    Warming of the southern ocean since the 1950s

    Science

    (2002)
  • S.T. Gille

    Decadal temperature trends in the Southern hemisphere ocean

    J. Climatol.

    (2008)
  • Cited by (40)

    • The Indian Ocean ‘garbage patch’: Empirical evidence from floating macro-litter

      2021, Marine Pollution Bulletin
      Citation Excerpt :

      The survey thus encompassed four water masses: sub-tropical (north of the Sub-Tropical Front), sub-Antarctic (between Sub-Tropical and Sub-Antarctic fronts), polar frontal zone (between Sub-Antarctic and Antarctic Polar fronts) and Antarctic waters (south of the Antarctic Polar Front) (Fig. S1). Absolute dynamic topography (http://marine.copernicus.eu) was used to define the positions of the fronts at the time of the survey (December 2019 – February 2020) using the following values: Sub-Tropical Front +0.23 m, Sub-Antarctic Front ‐−0.10 m, Antarctic Polar Front ‐−0.62 m (Barré et al., 2011; Park et al., 2019). To investigate the influence of oceanographic patterns on litter densities, the surveyed area was divided into four legs of similar length: inbound north of the Sub-Tropical Front (leg 1), south of the Sub-Tropical Front (sub-Antarctic and Antarctic waters, leg 2), outbound north of Sub-tropical Front until meandering of the ship ends (leg 3), and straight line heading west at 30°S (leg 4; Fig. S1).

    View all citing articles on Scopus
    View full text