Evidence of deep- and bottom-water formation in the western Weddell Sea

https://doi.org/10.1016/j.dsr2.2007.12.015Get rights and content

Abstract

During Ice Station POLarstern (ISPOL; R.V. Polarstern cruise ANT XXII/2, November 2004–January 2005), hydrographic and tracer observations were obtained in the western Weddell Sea while drifting closely in front of the Larsen Ice Shelf. These observations indicate recently formed Weddell Sea Bottom Water, which contains significant contributions of glacial melt water in its upper part, and High-Salinity Shelf Water in its lower layer. The formation of this bottom water cannot be related to the known sources in the south, the Filchner–Ronne Ice Shelf. We show that this bottom water is formed in the western Weddell Sea, most likely in interaction with the Larsen C Ice Shelf. By applying an Optimum Multiparameter Analysis (OMP) using temperature, salinity, and noble gas observations (helium isotopes and neon), we obtained mean glacial melt-water fractions of about 0.1% in the bottom water. On sections across the Weddell Gyre farther north, melt-water fractions are still on the order of 0.04%. Using chlorofluorocarbons (CFCs) as age tracers, we deduced a mean transit time between the western source and the bottom water found on the slope toward the north (9±3 years). This transit time is larger and the inferred transport rate is small in comparison to previous findings. But accounting for a loss of the initially formed bottom water volume due to mixing and renewal of Weddell Sea Deep Water, a formation rate of 1.1±0.5 Sv in the western Weddell Sea is plausible. This implies a basal melt rate of 35±19 Gt/year or 0.35±0.19 m/year at the Larsen Ice Shelf. This bottom water is shallow enough that it could leave the Weddell Basin through the gaps in the South Scotia Ridge to supply Antarctic Bottom Water. These findings emphasize the role of the western Weddell Sea in deep- and bottom-water formation, particularly in view of changing environmental conditions due to climate variability, which might induce enhanced melting or even decay of ice shelves.

Introduction

Antarctic Bottom Water (AABW) is the coldest and densest water mass in the world ocean. More than 60% of it is formed in the Atlantic sector of the Southern Ocean, i.e., in the Weddell Sea (Orsi et al., 1999), making it a key region for the deep and cold branch of the global thermohaline circulation. At the southern and western margins of the Weddell Basin, the precursors of AABW, Weddell Sea Deep and Bottom Water (WSDW and WSBW), are produced by interaction of warmer mid-depth and surface water masses with different shelf water types (Fig. 3). The latter are partly generated by interaction with the Antarctic ice shelves and are, apparently, extremely sensitive to climate variability, which affects sea-ice conditions and might alter basal melt rates or even induce ice-shelf decay.

High-Salinity Shelf Water (HSSW; salinities typically above 34.7 and potential temperature θ near the surface freezing point Tf≈−1.9 °C) is formed by brine rejection during sea-ice production on the broad shelf regions of the southwestern Weddell Sea during winter. It can become dense enough to reach the bottom, as reported from the western Weddell Sea (Gordon, 1998). Moreover, HSSW can mix with Warm Deep Water (WDW; with θ>0 °C, advected into the Weddell Basin from the Antarctic Circumpolar Current) and Winter Water (WW; a remnant of the winter mixed layer), to produce Weddell Sea Bottom Water (WSBW; θ<−0.7 °C). This process is often referenced to as the Foster–Carmack process (e.g., Foster and Carmack, 1976).

Concurrently, HSSW, which flows into the caverns below the ice shelves, melts the ice from below (enhanced hydrostatic pressure at the base lowers the freezing temperature) or at its front. The blend of glacial melt water and HSSW, called Ice Shelf Water (ISW; with θ below surface freezing temperature), mixes further with WDW or mWDW (i.e., WDW modified by WW) to produce WSBW as well. That process is usually referred to as the Foldvik process (e.g., Foldvik et al., 1985). Further mixing with ambient water masses (Foster and Carmack, 1976) or additions from external sources (Hoppema et al., 2001) renew Weddell Sea Deep Water (WSDW; −0.7 °C<θ<0 °C).

WSBW is known to be formed in the southern Weddell Sea, partly in interaction with the Filchner–Ronne Ice Shelf (FRIS). Based on observations in front of the FRIS, Foldvik et al. (2004) described plumes of ISW flowing out from the Filchner Depression on distinct pathways northwards down the slope into the deep Weddell Basin, or following the continental slope westward. Using current meter data they estimate an outflow of cold water (θ<−1.9 °C) of 1.6±0.5 Sv (1 Sv=106 m3/s). Due to further entrainment of adjacent water, they deduce a WSBW formation rate of 4.3±1.4 Sv (θ<−0.8 °C). However, the characteristic cold temperature of the ISW outflow is eroded rapidly when it leaves the Filchner Depression, as can be seen on sections across the slope west of the Filchner Depression. Further to the west, on the broad shelf in front of the Ronne Ice Shelf, HSSW is the dominating water mass. Weiss et al. (1979) and Nicholls et al. (2003) observed ISW also at the Ronne Ice Front, but from this source ISW could not be traced further northward into the Weddell Basin.

Still under consideration is the capability of the western ice shelf areas along the Antarctic Peninsula, particularly the Larsen Ice Shelf (LIS), to contribute to WSBW formation. Fahrbach et al. (1995) report WSBW formation by mixing of deep water with a flow originating from the shelf in front of the LIS. Weppernig et al. (1996) found evidence that this water contains significant fractions of ISW, possibly of LIS origin. Mensch et al. (1998) calculated a total WSDW and WSBW formation (θ<0 °C) of 5 Sv originating from the western Weddell Sea. Nicholls et al. (2004) report ISW formation in the western Weddell Sea by basal ice shelf melting induced by modified WDW, pre-conditioned during winter, but not by HSSW. The issue of glacial melt-water contributions to WSBW formation in the western Weddell Sea is of special interest, considering the decay of Larsen A and B in 1998 and 2002, respectively, (caused by increased basal melt rates, Shepherd et al., 2003) and in view of climate variability and increasing deep and bottom water temperatures as reported by Robertson et al. (2002), Schröder et al. (2002), and Fahrbach et al. (2004).

Useful tools to identify glacial melt water (or ISW) are the low-solubility noble gases helium (He) and neon (Ne) (Schlosser, 1986). Atmospheric air with a constant composition of these noble gases is trapped in the ice matrix during formation of the meteoric ice. Due to the enhanced hydrostatic pressure at the base, these gases are completely dissolved in the water, when the shelf ice is melting from below. This leads to an excess of Δ4He=1060% and ΔNe=770% in pure glacial melt water (Δ stands for excess over an air–water solubility equilibrium) (Hohmann et al., 2002). With an accuracy of 0.5% for 4He measurements, melt-water fractions of 0.05% are detectable by this method.

In ISW observed within the Filchner Depression, the helium excess was about 20% (Schlosser et al., 1990), equivalent to a fraction of roughly 1.4% pure glacial melt water (assuming a background Δ4He in the adjacent water masses of about 5%). In front of the Filchner Depression, the 4He excess was still as high as 10% (glacial melt water ≈0.5%). From Ice Station Weddell in 1992 (subsequently referred to as ISW-1), drifting east of the Antarctic Peninsula, Weppernig et al. (1996) used 4He and oxygen isotope observations to estimate the fraction of ISW in WSBW (θ<−0.7 °C). They found 15% ISW, originating from the Filchner Ronne Ice Shelf, at the southern end (72°S) of the drift track, generally decreasing as the drift track progresses northward. But two anomalies of elevated ISW (glacial melt water) fractions were detected along the track. These were observed in the vicinity of Larsen C Ice Shelf (68.5°S) and at the northern end of the LIS. Based on the previously reported ISW outflow of 1 Sv from the Filchner Depression (Foldvik et al., 1985). Weppernig et al. (1996) estimated a WSBW formation rate of 5 Sv, including 10% ISW of Filchner Ice Shelf origin.

Particularly at mid-ocean ridges of the deep Pacific, additional 3He (and less 4He) is released into the deep water, resulting in an enhanced 3He/4He ratio of about eight times the atmospheric ratio (Well et al., 2001). These waters contribute to Circumpolar Deep Water (Well et al., 2003), and hence to WDW, which then has a maximum 3He/4He ratio and slightly enhanced 4He values. Neon provides complementary information (Rodehacke et al., 2006). In contrast to helium, neon has no internal sources (Well and Roether, 2003) other than glacial melt water, and has minimum values in WDW (compare Fig. 5B and C, and the gray arrow in Fig. 2 pointing to the 4He shift off the atmospheric 4He/Ne ratio particularly in the WDW range that originates from Circumpolar Deep Water).

The chlorofluorocarbons CFC-11 and CFC-12 are age tracers due to their temporal evolution in the atmosphere and in the ocean surface waters. They allow estimating transit times of recently ventilated waters in the ocean interior, depicting the transport of these water masses (Weiss et al., 1985; Bullister, 1989).

There is only limited knowledge about the western Weddell Sea and, particularly, the LIS concerning WSBW formation processes and rates. One aim of the ISPOL experiment (R.V. Polarstern cruise ANT XXII/2, October 2004–January 2005) was to find evidence for glacial melt water originating from LIS. A comprehensive description of the ISPOL cruise, the drift track and the ice conditions, is given by Dieckmann et al. (2007); see also Hellmer et al., 2006, Hellmer et al., 2008. Absy et al. (2008) focus on the hydrography of the upper water column and a comparison with ISW-1. Here we analyze the formation of WSBW and quantify glacial melt-water fractions by using hydrographic and tracer observations. We find clear evidence of an ISW/glacial melt-water source at the Antarctic Peninsula—most definitely related to the LIS. This water can be traced further downstream along the slope towards the northern tip of the Antarctic Peninsula between 1000 m and the base of the slope.

In the following sections, we describe the sampling and measurement of the presented data and the applied methods. Subsequently, we analyze these observations and show how they relate to previous observations up- and downstream of the ISPOL investigation area. Thereafter, we deduce the fractions of the contributing water masses and the formation and export rates of WSBW.

Section snippets

Data and methods

The location of the ISPOL experiment is shown in Fig. 1A and B. Heavy sea-ice conditions during November prevented R.V. Polarstern to proceed further southwest and close to the continental shelf. A canyon located further south (69.5°S), suspected to be a possible pathway of locally formed ISW, was missed. Another canyon further towards the northwest (67°S) was not found, presumably due to unreliable charts; CTD observations differed partly by more than 200 m from the charts (Dieckmann et al.,

Observations

Fig. 4 shows vertical sections of 4He, δ3He, and CFC-12 along the main ISPOL drift track (about parallel to the meridional θ and S sections shown in Absy et al., 2008). The 4He section appears to be more scattered than those of δ3He and CFC-12, which is due to the lower dynamic range of 4He. However, the 4He section shows increasing values towards the bottom with a maximum in the WSBW layer (θ<−0.7 °C) several 10s of meters above the bottom, indicating large glacial melt-water contributions. The

Shelf water contributions to deep and bottom water observed in western Weddell Sea

For WSDW and WSBW formation, we account for four parent source-water masses: WW, WDW, HSSW, and glacial melt water (see Fig. 3). We do not account for mWDW, a mixture of WW and WDW, nor ISW, a mixture of HSSW and glacial melt water. We employ an OMP using θ, S, 4He, Ne, and the 3He/4He ratio (i.e., δ3He) to calculate the relative fraction of each of the four source-water masses in the deep and bottom water observed at the ISPOL site. We define the source-water masses within limits of

Conclusions

One aim of the ISPOL experiment (R.V. Polarstern cruise ANT XXII/2, October 2004–January 2005) was to find evidence for deep- and bottom-water formation in the hardly accessible area of the western Weddell Sea. Deep- and bottom-water formation processes and rates, particularly the role of the ice shelves at the Antarctic Peninsula and contributions of glacial melt water as well as basal melt rates, are rather unknown. Previous studies (Fahrbach et al., 1995; Weppernig et al., 1996; Mensch et

Acknowledgments

We thank the master and the crew of R.V. Polarstern, the helicopter team, and the ISPOL scientific party for excellent and fruitful cooperation. Special thanks to K. Bulsiewicz, G. Fraas, W. Plep, J. Sültenfuß, and their student assistants for their excellent work in our CFC and noble gas lab. Further thanks to A. Wisotzki for analysis and assessment of the hydrographic data obtained during ISPOL and to K.-P. Lieckfeldt and M. Schodlok for their assistance during the noble gas and CFC sampling.

References (56)

  • R. Well et al.

    An additional deep water mass in Drake Passage as revealed by 3He data

    Deep-Sea Research I

    (2003)
  • R.F. Weiss et al.

    Geochemical studies of the Weddell Sea

    Deep Sea Research, Part A

    (1979)
  • Absy, J.M., Schröder, M., Muench, R., Hellmer, H.H., 2008. Early summer thermohaline characteristics and mixing in the...
  • P. Beining et al.

    Temporal evolution of CFC 11 and CFC 12 concentrations in the ocean interior

    Journal of Geophysical Research

    (1996)
  • J.L. Bullister

    Chlorofluorocarbons as time-dependent tracers in the ocean

    Oceanography

    (1989)
  • K. Bulsiewicz et al.

    A capillary-column chromatographic system for efficient chlorofluoromethane measurement in ocean waters

    Journal of Geophysical Research

    (1998)
  • G. Dieckmann et al.

    The expeditions ANTARKTIS-XXII/1 and XXII/2 of the research Vessel “Polarstern” in 2004/2005

  • E. Fahrbach et al.

    Formation and discharge of deep and bottom water in the northwestern Weddell Sea

    Journal of Marine Research

    (1995)
  • E. Fahrbach et al.

    Flow of bottom water in the northwestern Weddell Sea

    Journal of Geophysical Research

    (2001)
  • E. Fahrbach et al.

    Decadal-scale variations of water mass properties in the deep Weddell Sea

    Ocean Dynamics

    (2004)
  • Foldvik, A., Gammelsrød, T., Tørresen, T., 1985. Circulation and water masses on the southern Weddell Sea shelf. In:...
  • A. Foldvik et al.

    Current measurements near Ronne Ice Shelf: implications for circulation and melting

    Journal of Geophysical Research

    (2001)
  • A. Foldvik et al.

    Ice Shelf Water overflow and bottom water formation in the southern Weddell Sea

    Journal of Geophysical Research

    (2004)
  • T. Gammelsrød et al.

    Distribution of water masses on the continental shelf in the southern Weddell Sea

  • Gordon, A.L., 1998. Western Weddell Sea thermohaline stratification, In: Jacobs, S.S., Weiss, R.F. (Eds.), Ocean, Ice,...
  • A.L. Gordon et al.

    Deep and bottom water of the Weddell Sea's western rim

    Science

    (1993)
  • H.H. Hellmer

    Impact of Antarctic ice shelf basal melting on sea ice and deep ocean properties

    Geophysical Research Letters

    (2004)
  • H.H. Hellmer et al.

    Sea ice feedbacks observed in the western Weddell Sea

    EOS

    (2006)
  • Cited by (76)

    • Ice retreat in the eastern Antarctic Peninsula region: the application of diatoms for understanding climate change

      2020, Marine Micropaleontology
      Citation Excerpt :

      Both embayments are sites of seasonal primary production at levels comparable to that observed from other continental shelf sites around Antarctica, however, in contrast to the western Antarctic Peninsula shelf, productivity increases along an offshore gradient (Cape et al., 2014). The physical oceanography of the Weddell Sea (see review in Vernet et al., 2019), a region known for its importance in the formation of oceanic bottom waters (Gordon, 1991; Comiso and Gordon, 1998), has been the focus of studies extending back to the mid-1970s (e.g., Carmack, 1974; Carmack and Foster, 1975; Foster and Carmack, 1976; Killworth, 1977), continuing to the present (e.g., Absy et al., 2008; Gordon et al., 2007; Huhn et al., 2008; Nicholls et al., 2004; Nicholls et al., 2008; Thompson and Heywood, 2008; Gordon et al., 2010; Meredith et al., 2011; van Caspel et al., 2015; Vernet et al., 2019; Gordon et al., 2020). Bottom water formation occurs around the margin of the Weddell Sea, where cold katabatic winds flow off the ice sheet.

    • Three decades of deep water mass investigation in the Weddell Sea (1984–2014): Temporal variability and changes

      2018, Deep-Sea Research Part II: Topical Studies in Oceanography
      Citation Excerpt :

      Considering the Case B applied here, the SWT indices (Table 2) were defined using the WOD13 data available nearby the western and southern continental shelf and shelf-break of the Weddell Sea. This follows a previous investigation of the water mass properties executed by Huhn et al. (2008) to better define the SWT indices for HSSW and ISW. Finally, only one SWT was used to represent each of the water masses considered, independently of the mixing schemes (Fig. 2; Table 2).

    • On the ventilation of Bransfield Strait deep basins

      2018, Deep-Sea Research Part II: Topical Studies in Oceanography
      Citation Excerpt :

      It is surprising that along the WS shelf break and thus in the BS deep basins the GMW coming from FRIS has a larger concentration than the GMW from LIS. Different studies (e.g., Fahrbach et al., 1995; Gordon et al., 2001; Schröder et al., 2002; Absy et al., 2008; Huhn et al., 2008; van Caspel et al., 2015) suggest that dense waters formed in the southwestern WS cover the deeper parts of the continental slope while the upper slope is dominated by dense waters formed on the western Weddell Sea continental shelf. The belief that the western WS is a source of dense waters is based on observations done near the tip of the Antarctic Peninsula (Fahrbach et al., 1995; Schröder et al., 2002; van Caspel et al., 2015) or on the continental slope in front of LIS (Gordon et al., 2001; Absy et al., 2008; Huhn et al., 2008), but there are no irrefutable observations of dense water formation on the continental shelf off LIS. van Caspel (2016) (

    View all citing articles on Scopus
    1

    Now at: Zentrum für Marine und Atmosphärische Wissenschaften, Hamburg, Germany.

    2

    Now at: Jet Propulsion Laboratory, California Institute of Technology, Pasadena, CA, USA.

    View full text