On the circulation of bottom water in the region of the Vema Channel

https://doi.org/10.1016/S0967-0637(02)00016-XGet rights and content

Abstract

The circulation and transport of Antarctic Bottom Water (σ4<45.87) in the region of the Vema Channel are studied along three WOCE hydrographic lines, the geostrophic velocities referenced to previously published direct current measurements. The primary supply of water to the deep Vema Channel is from the Argentine Basin's deep western boundary current, with no indication of an inflow from the southeast. In the northern Argentine Basin, detachment of lower North Atlantic Deep Water from the continental slope is associated with a deep thermohaline front near 34°S. To the north of this front, the upper part of the AABW bound for the Vema Channel (σ4<46.01) exhibits a significant NADW influence. Further modification of the throughflow water occurs near 30°30′S, where the channel orientation changes by ∼50°. Southward flow of bottom water on the eastern flank of the Vema Channel, amounting to ∼1.5 Sv, represents a significant countercurrent to the deep channel transport. Inclusion of this countercurrent reduces the net flow of AABW through the Vema Channel from 3.2±0.7 to 1.7±1.1 Sv. Water properties imply that the near-zero net flow over the Santos Plateau results from a near-closed cyclonic circulation fed by the deep Vema Channel throughflow. A disruption of the northward boundary current in the upper AABW (lower circumpolar water) is required by this flow pattern. The extension of the cyclonic circulation on the Santos Plateau enters the Brazil Basin as a ∼1 Sv flow distinct from the outflow in the Vema Channel Extension (6.2 Sv). The high magnitude of the latter suggests a southward recirculation of bottom water near the western boundary to the north of the region of study.

Introduction

The Vema Channel, which lies between the Santos Plateau and the Rio Grande Plateau in the South Atlantic (Fig. 1), acts as a major conduit for bottom waters flowing northward from the Argentine Basin to the Brazil Basin. The Hunter Channel, located some 1200 km to the east-southeast, also accommodates bottom water flow between the two basins. Bottom waters formed in the Southern Ocean ventilate the abyssal ocean. As monitoring their flux is easiest at choke points, several measurement programs have been conducted at these channels (Johnson et al., 1976; Hogg et al (1982), Hogg et al (1999); Zenk et al., 1999; Speer et al., 1992; Speer and Zenk, 1993). Recently, the flow in the Vema Channel and on the Santos Plateau was monitored with current meters as part of the World Ocean Circulation Experiment (WOCE) Deep Basin Experiment (DBE). Hogg et al. (1999) combined these data with hydrographic data collected at the beginning and the end of the 2-year long current meter deployment, and calculated a northward flow of water cooler than 2°C of 3.7 and 2.0 Sv, respectively. These and other estimates of the flux are given in Table 1. Although this table also provides estimates of the meridional transport over the Santos Plateau located between the Vema Channel and the continental slope, the abyssal circulation surrounding the channel is generally less well understood than in the deep passage itself. Classically, the Vema Channel is considered to be fed by a deep western boundary current in the Argentine Basin (Stommel and Arons, 1960), but recent studies have hinted that the throughflow may also be fed from the eastern Argentine Basin (Coles et al., 1996). Downstream of the passage, Sandoval and Weatherly (2001) found a splitting of the abyssal boundary current in two branches, but the details of the connection to the passage are not provided by their large-scale analysis.

In this study we use three hydrographic sections, undertaken during WOCE, that cross the Rio Grande Plateau and the Santos Plateau (Cruises A10, A17, A23; Fig. 1) to address several still unsettled questions concerning the abyssal circulation in the region. The configuration of the data used here allows us to complement analysis of the flow through the deep Vema Channel and over the Santos Plateau (the focus of most previous investigations) with a study of the pathways and fluxes of bottom water approaching and leaving the Vema Channel. Our presentation will also occasionally refer to older hydrographic sections for the discussion of specific points (continuous lines in Fig. 1). These are four quasi-zonal transects to the continental slope between 27°S and 36°S, which were occupied during the Thomas Washington Marathon cruise in 1984 and were analyzed by Zemba (1991) in a study of the boundary currents. Also referred to is a meridional line at 41°W from the SAVE5 cruise (1989), used by Coles et al. (1996) for a study of the Antarctic Bottom Water (AABW) in the Argentine Basin.

The origin of the throughflow is examined by comparing the segments of A17 and A23 in the northern Argentine Basin that sampled possible arrivals from the west and east. The flow through the deep Vema Channel and modifications of the water properties along the passage are examined at all intersections with the WOCE lines. A mean southward flow of AABW has been consistently observed in direct velocity measurements made on the eastern flank of the Vema Channel (Hogg et al (1982), Hogg et al (1999)). This portion of the flow, partially integrated in some of the net flow estimates (Speer and Zenk, 1993; Zenk and Hogg, 1996), is missing in others (Hogg et al., 1999). We will estimate its magnitude from the segments of A23 (115–119) and A10 (12–15) by which it was sampled. Across the Santos Plateau, Hogg et al. (1999) suggested a near-closed cyclonic circulation of AABW that has already passed through the deep Vema Channel. Such a scheme will be investigated with new information provided by A10 and A23. Finally, the transports of bottom water entering the Brazil Basin from the Vema Channel or over the Santos Plateau will be estimated from the northern parts of A17 and A23. The flow field derived from the sections is constructed to be consistent with direct velocity measurements of bottom, deep, and intermediate waters (Hogg et al., 1999; Hogg and Owens, 1999; Boebel et al (1997), Boebel et al (1999)).

Section snippets

Data used

The WOCE CTD-O2 data used in this study are subsets of sections A10, A17 and A23 that took place in the early 1990s (Fig. 1, Fig. 2, Fig. 3, Fig. 4). Each section was split into layers defined by potential density horizons that correspond to the following six water masses: Central Water (CW), Antarctic Intermediate Water (AAIW), Upper Circumpolar Water (UCPW), North Atlantic Deep Water (NADW), Lower Circumpolar Water (LCPW) and Weddell Sea Deep Water (WSDW). We will concentrate on the results

The deep front

One feature of note on the A17 and A23 distributions of salinity and oxygen is the deep front present at around 3000 dbar between stations 63 and 64 on A17 (Fig. 3) and stations 100 and 101 on A23 (Fig. 2). The potential temperature–salinity (θ–S) profiles at these stations (Fig. 5) show that the front is evident at temperatures down to 0.36°C on A23. At this potential temperature the salinity to the south of the front increases by 0.005. In the A17 data the base of the front is less distinct in

Referencing the geostrophic shear for flux calculations

Most previous calculations of the fluxes through the Vema Channel have assumed a single surface of no motion (Table 1). This surface of no motion generally lies between NADW and AABW and has been approximated by θ=2°C and σ4 surfaces at or very close to 45.87. Its choice was based on the large-scale circulation and the assumption that between southward flowing NADW and northward flowing AABW there lies a surface where the velocity is zero. However, what is true on a large scale is not always

Results

The cumulative transport of AABW along each section is shown in Fig. 8. In Fig. 9 AABW transports at key points on each section have been grouped together to address specific questions (Section 6). The uncertainties in the transports have been estimated assuming an uncertainty of ±1.0 cm s−1 in our choice of reference surface velocity. The uncertainties calculated in this way were always greater than the variability in AABW transport produced by changing the bottom triangle formulation, so we do

The origin of the throughflow in the deep Vema Channel

The dominant input to the Vema Channel is from the Argentine Basin's western boundary. A transport volume of 8.5±3.4 Sv crosses A17 and enters the region north of the Deep Front and south of the Vema Channel (Fig. 9). This agrees well with an AABW northward transport of about 9 Sv found by Wienders et al. (2000) inshore of A17 at 36°S and is consistent with estimates of 5–6 Sv across the Marathon lines 34S and 36S by Zemba (1991). Anticipating the forthcoming conclusion that no LCPW flows directly

Summary and conclusion

The transports across, and properties along, the three WOCE lines A10, A17, and A23 were analyzed in the light of previous knowledge for an ensemble view of the flow and modification of AABW in the region of the Vema Channel. The schematic abyssal circulation pattern thus obtained (Fig. 9) provides complementary (often corroborating) information on issues previously addressed, and suggests further investigation on other points. It is not definitive and uncertainties would be reduced by

Acknowledgments

WOCE section A23 was supported by the Natural Environment Research Council UK WOCE Special Topic Grant GST/02/575. E.L.M.'s contribution to this work was supported by the A23 grant and the James Rennell Division, Southampton Oceanography Centre. The contribution of M. Arhan to this study was supported by IFREMER (Grant 210161), INSU (Institut National des Sciences de l’Univers) and the CNRS (Centre National de la Recherche Scientifique), in the framework of the Programme National d’Etude de la

References (22)

  • N. Hogg et al.

    On the transport and modification of Antarctic Bottom Water in the Vema Channel

    Journal of Marine Research

    (1982)
  • Cited by (29)

    • Contourite systems around the northern exit from the Vema Channel

      2022, Marine Geology
      Citation Excerpt :

      Sedimentation rates on the depositional wave flanks exceed 30 cm/kyr (Flood et al., 1993; Ledbetter, 1993). These differences can be explained by the strong variability of current speeds and directions on the SCP (Speer and Zenk, 1993; McDonagh et al., 2002; Frey et al., 2018) resulted in alternated deposition on different wave flanks (Borisov et al., 2020) and formation of short-term hiatuses. The evidence of strong winnowing and reworking under high-energy conditions is provided by the siliciclastic sandy contourites, represented by well-sorted very fine to coarse sand occurring on the upper continental slope of the Mozambican margin, at about 300 m water depth (de Castro et al., 2021).

    • Modeling of bottom currents for estimating their erosional-depositional potential in the Southwest Atlantic

      2022, Journal of Marine Systems
      Citation Excerpt :

      The AABW flow accelerates in narrow deep-water channels; the velocities here can be as high as 30 cm/s on average and occasionally reach 60 cm/s (Hogg et al., 1999; Morozov et al., 2021). The flow over Santa Catarina Plateau is much more variable, the direction of currents frequently changes from northward to southward, and annual mean velocities here do not exceed 10 cm/s (Speer and Zenk, 1993; Hogg et al., 1999; McDonagh et al., 2002). This study includes a comparison of the erosional and depositional potential of bottom currents with high velocities in narrow channels and with low velocities over wider plateaus.

    • Deep Sea Sedimentation

      2022, Treatise on Geomorphology
    • Sediment waves on the Santa Catarina Plateau (western South Atlantic)

      2020, Journal of South American Earth Sciences
    View all citing articles on Scopus
    View full text