Australia: The Land Where Time Began

A biography of the Australian continent 

Antarctic Bottom Water Produced by intense formation of Sea-Ice in the Cape Darnley Polynya

The cold, dense water forming the abyssal layer of the global ocean is the Antarctic Bottom Water (AABW). The formation of the Antarctic Bottom Water performs a key role in the circulation of the oceans. Dense shelf water (DSW) sinking to depth at 3 locations around Antarctica, that have previously been documented, forms this bottom water. According to Ohshima et al. another site of bottom water formation has been identified in hydrographic and tracer data, though the formation site is not well constrained. In this article Ohshima et al. identify the Cape Darnley Polynya, 65^o - 69^o E, as a site of dense shelf water formation and its subsequent transformation into bottom water. They collected the data by the use of moorings and attaching instruments to elephant seals, Mirounga lionina. The Cape Darnley Polynya differs from the other sources of Antarctic bottom water that were identified previously as it is driven by the salt flux, the salt being released by sea ice formation, the other sites requiring the presence of an ice shelf or a large storage volume. Ohshima et al. estimate that about 0.3-0.7 x 10⁶ m³/s of dense shelf water that is produced by the Cape Darnley polynya is transformed into Antarctic Bottom Water. The Cape Darnley Bottom Water, the mass of water transformed, amounts to 6-13% of the circumpolar total.

The Antarctic bottom water (AABW) comprises 30-40 % of the mass of the global ocean (Johnson, 2008), and the production is a major component of the global overturning circulation ((Johnson, 2008; Orsi, Johnson & Bullister, 1999; Marshall & Speer, 2012), representing heat and possibly CO₂ (Sigman & Boyle, 2000). Originating as dense shelf water, AABW forms on the continental shelf by combinations of brine rejection as ice forms and ocean/ice shelf interactions, that vary regionally. The DSW can mix down the continental slope with ambient water masses to produce AABW, providing it has enough negative buoyancy and a pathway across the shelf break (Baines & Condie, 1998). Around Antarctica there were 3 main regions of AABW production known at the time this study was carried out - the Weddell Sea (Foster & Carmack, 1976; Gordon et al., 1993; Fahrbach et al., 2001; Foldvik et al., 2004), the Ross Sea (Jacobs et al., 1970; Whitworth & Orsi, 2006; Gordon et al., 2009), and off the Adéle Coast (Rintoul, 1998; Williams et al., 2008; Williams et al., 2010). Large continental embayments in the Ross Sea and the Weddell Sea associated with large continental ice shelves were mechanisms that were considered to be necessary to generate enough negative buoyancy in the local DSW for AABW to be formed (Foldvik et al., 2004; Jacobs et al., 1970). When the Adélie Land Bottom Water was linked directly to polynyas (areas of thin ice where there is increased production of sea-ice and the associated rejection of brine) on the coast this paradigm was broken. There are many polynya regions, especially in East Antarctica, but it was proposed that the storage capacity of the continental shelf in this polynya region was the reason this appeared to be the only polynya region where there was enough DSW to form AABW (Williams et al., 2008; Williams et al., 2010).

According to Ohshima et al. there was another independent variety of AABW that had been identified previously in the Weddell-Enderby Basin from offshore properties that had been obtained from hydrographic and tracer studies (Jacobs & Georgi, 1977; Mantisi et al., 1991; Meredith et al., 2000; Hopperma et al., 2001; Orsi & Whitworth, 2005; Meijers et al., 2010) but there was a problem in terms of its existence and DSW source. The most likely candidate appeared to the Prydz Bay region (71^o-80^o E) (Middleton & Humphries, 1989; Nunes Vaz & Lennon, 1996; Yabuki et al., 2006), as it had a large continental embayment and it is where the Amery Ice Shelf is located, making it similar in appearance to the AABW regions in the Weddell Sea and the Ross Sea. Ohshima et al. say the problem with this suggestion is that the results from a number of ship-board studies were inconclusive, the export and downslope transport of DSW never being observed in this region.

Estimates of sea-ice production made from the data received from the most recent satellite studies suggests, according to Ohshima et al. that the Cape Darnley polynya (CDP), to the northwest of the Emery Ice Shelf, has the second highest production of ice after the Ross Sea Polynya (Tamura et al., 2008). This satellite analysis was the motivator for an extensive Japanese mooring program that was conducted in 2008-2009 as part of the International Polar Year, aimed at proving the production of AABW from the CDP. In this article Ohshima et al. report the results of their moorings and observations, that were successful, of new AABW on the continental slope off the Cape Darnley Polynya, also reporting confirmation that the Cape Darnley Polynya is the source of the Dense Shelf Water for this AABW, which was based on the instrumental data obtained by use of elephant seals to carry instruments into the depths. This method of obtaining data has, according to Ohshima et al. , become an important source of hydrographic profiles around the margins of Antarctica (Fedak, 2006; Charrassin et al., 2008; Williams et al., 2011) in regions/seasons when the gathering of hydrographic data is logistically difficult to obtain.

Production of sea-ice in the polynya

Ohshima et al. have used a data set from high-resolution satellite data to enhance the understanding of CDP obtained from previous studies (Tamura & Oshima, 2008; Massom et al., 2008; Fraser et al., 20012). The extent of the CDP, >10⁴ km², is clearly seen in a typical image from synthetic aperture radar, the sea ice that is newly formed appears as white streaks (high radar backscatter). Based on advanced Microwave Scanning Radiometer-EOS (AMSR-E) data an estimate of the annual ice production is shown to be extremely high, 195 ± 71 km³, with local rates of production above 10 m/yr. A grounded iceberg tongue (Fraser et al., 20012) is shown as a dark zone with white patches, just to the east of the polynya. With the westward projection of the prevailing ocean currents and dominant winds offshore, ice that is newly formed accumulates on the eastern side of this barrier and advected away to the west. This mechanism has made the polynya the second largest in area in Antarctica, and has a large ice flux associated with it.

Production of new AABW off the polynya

Ohshima et al. deployed 4 moorings at various locations, 2 of which were in the Daly and Wild canyons, that had been predicted to be the path taken by the DSW/AABW, M1 at a depth of 2,923 m and M3 at 2,608 m, and over the continental slope M2 at 1,439 m and M4 at 1,824 m. They placed M3 at the centre of Wild Canyon just off the CDP and it was this mooring that detected the most prominent of the downslope currents. Multi-beam bathymetry measurements during deployment showed that the mean flow was directed downslope, and there was a prominent bottom intensification. All stations, up to 224 m from the bottom, detected an abrupt signal of colder, less saline, denser water that became dominant after June, and the current speeds increased across this layer in conjunction with the cold, dense signal. The speeds of downslope currents of up to 0.5 m/s fluctuated coherently with these water properties. At periods of 4 and 5 days strong peaks were revealed by spectral analysis. It has been estimated by linear extrapolation that the thickness of the cascading AABW layer increased by up to about 400 m from May to October and remained above 300 m until January. With these properties (neutral density γⁿ >28.27 kg/m³ ) at this depth range (>2,500 m), these observations represent new AABW, which is in the range for the Weddell Sea (Orsi et al., 1999).

Ohshima et al. suggest that gravity currents drive the observed downslope flows with properties of cold, dense water, and strong bottom intensification. Based on satellite data, analysis of a salinity budget suggested that following 2 months in which there were high levels of sea ice production in the CDP the salinity, and hence the density, of the shelf water reaches a high enough level to begin descending the slope, down which it is channelled by the canyons. At M3 the fluctuating signal is suggested  by Ohshima et al. to probably originate from a periodic outflow resulting from baroclinic instability that is associated with the front between the DSW and the water offshore that is less dense (Matsumura & Hatsumi, 2010). The DSW can then descend in the form of an eddy or plume (Baines & Condie, 1998; Matsumura & Hatsumi, 2010; Darelius et al., 2009; Wang et al., 2008). Thermobaricity could aid the cold dense water to descend to even greater depths^9,34.(Foldvik et al., 2004; Budillon, 2011). The production, to varying degrees, of new AABW was also reported from moorings M1, M2 and M4. At moorings M1 and M2 cold and dense water was detected following shortly after the onset of the active production of sea ice. The DSW from the CDP is indicated by the data to be advected to the west along the slope, M2, and also the transport of new AABW down Daly Canyon, M1. M4 also reported a cold and dense signal that Ohshima et al. suggest possibly indicates that some DSW was formed upstream of the CDP, though as the near-bottom signal is much thinner suggests that the contribution of this upstream DSW is minor compared to that from the CDP.

Elephant seals with instrument packs attached revealed DSW

Ohshima et al. attached instruments to southern elephant seals, Mirounga leonina, that included CTD sensors to detect conductivity, temperature and depth, and the results revealed the spatial distribution and seasonal evolution of the properties of DSW off Cape Darnley on the continental shelf and in Prydz Bay in the east. Ohshima et al. used a set of delayed mode techniques (Roquet et al., 2011) to post-process hydrographic data, and the results yielded enough accuracies to characterise the DSW, about 0.03^oC, and a salinity of 0.05 or more. A high salinity DSW over the region of the CDP, that is revealed by the spatial distribution, is consistent with a high level of ice production. Ohshima et al. have identified 6 regional subsets covering the continental shelf in this region, and the most saline DSW they found was in the inshore region of the continental shelf to the west of Cape Darnley, with actively forming DSW salinities from late June to early July and remnant DSW salinities >34.8 for January-February. According to Ohshima et al. for the shelf waters around Antarctica these are among the most saline, and hence the most dense (Williams et al., 2008; Williams et al., 2010; Yabuki et al., 2006; Gordon et al., 2001; Orsi & Wiederwohl, 2009) the region overlapping with the high rate of ice production of >5 m/yr. In contrast a lower salinity variety, 34.5-34.6, that is observed over the shelf at Prydz Bay between March and mid-June.

Depths of as much as 1,800 m were reached by several of the seals carrying instrument packs foraging on the continental slope, the data they returned included very rare measurements of wintertime overflowing DSW, aka modified Shelf Water (mSW), in a region just off the CDP and to the southeast of M3 between 67.5^o and 71.0^oE. The shift that is clear here to bottom values that are more saline and denser (>34.64) from east to west, in particular west of 68.5^oE, is the key feature. The main source of the new AABW that cascades down the canyons is, together with the results from the moorings, confirmed by these results to be waters of high salinity from the CDP region. Though it is likely that some weak contribution or pre-conditioning by DSW of low salinity exported from the Prydz Bay has not been quantified at this time.

Implications of the production of the Cape Darnley Bottom Water

According to Ohshima et al.  their discovery of the production of new AABW offshore from the CDP is a major missing piece of a puzzle that has continued for 30 years concerning the source of AABW that was recently ventilated in the Weddell-Enderby Basin. Ohshima et al. propose the name the Cape Darnley Bottom Water (CDBW) for this AABW. The DSW that is produced in the CDP descends down the Wild Canyon and the Daly Canyon and it is converted to CDBW by mixing with overlying Circum Polar Deep Water (CDW), eventually becoming part of the AABW in the Weddell Sea, the Weddell Sea Deep Water (WSDW).

Ohshima et al. investigated the ventilation rate of the DSW, the volume flux of the DSW exported from the continental shelf that ultimately produces the CDBW and the Wild Canyon, 0.52 ± 0.26 Sv, where Sv = 10⁶ m³/s. Ohshima et al.  propose that 0.3-0.7 Sv of the DSW is ventilated in this region, based on their estimate of annual transport of CDBW down Wild Canyon, in addition to another estimate using a salinity Ohshima et al. say this represents about 6-13% of the circumpolar contribution, that was based on the total ventilation rate of 5.4 Sv for DSW, determined from chlorofluorocarbon data (Williams et al., 2011), and is the same order of magnitude of transport that has been reported from Adélie Land (Williamd et al., 2008) Ohshima et al. suggest this significant injection, surface-to-bottom, is consistent with the concentrations of chlorofluorocarbon and oxygen near the bottom offshore from Cape Darnley and to the west of the cape (Orsi & Whitworth, 2005). To determine the volume of WSDW renewed through CDBW the mean property of the AABW in the Atlantic sector² is used as it is close to that of DSDW. The contribution of CDBW was estimated to be 0.65-1.5 Sv by using the ventilation rate of the DSW estimated above the mixing ratio of the CDBW and CDW to this mean property, the estimate of the contribution of CDBW is 0.65-1.5Sv. Several estimates have been made (Orsi et al., 1999; Meredith et al., 2001; Naveira Garabato, et al., 2002) for the total production of AABW in the Atlantic sector, in the range of about 3-10 Sv. The contribution of the CDBW is about 13-30% of the production of AABW in the Atlantic, taking the value of 4.9 Sv.

Ohshima et al. note that, in comparison to the other AABW regions of formation, the overflow periodicity is similar to what has been reported in Weddell (Darelius et al, 2009) and Ross (Budillon et al., 2011) Seas, and the thickness of the AABW layer is comparable to that in the Ross Sea (Whitworth & Orsi, 2006; Gordon et al., 2009), though there is a difference in that tidal influence is much weaker here, <0.5 m/s, relative to the Ross Sea (Whitworth & Orsi, 2006; Gordon et al., 2009) Ohshima et al. say that it is demonstrated conclusively at Cape Darnley that AABW can be produced as the result of sea ice production in a polynya in a relatively narrow section of continental shelf that has limited storage capacity of DSW. They also speculate that there could possibly be more regions where similar polynya-production of ice forms AABW, in particular, in East Antarctica (Williams et al., 2010; Tamura et al., 2008) Cape Darnley ice barrier (ice tongue) blocks the westward advection of sea ice leading to high ice production in the CDP. The mechanism involved is similar to that of the Mertz Glacier Tongue on the Mertz Glacier and Drygalski Ice Tongue on the Terra Nova polynyas. It was demonstrated in 2010 that calving of the Mertz Glacier Tongue that when such a barrier undergoes rapid change (collapse) the result can be a pronounced change in the production of sea ice and AABW. Any significant change to the ice barrier would impact strongly on the production of CDBW, though during the AMSR-E period, 2003-2010, the iceberg tongue and the production of sea ice were relatively stable in the CDP.

This study has shown that the CDBW provides an important contribution to WSDW, a major part of the AABW that drives the lower limb of the meridional overturning circulation (MOC) in the Atlantic sector (Meredith et al., 2008). It has been suggested that the MOC could be weakening as a result of warming of the WSDW that has been reported to have been occurring since the 1980s (Fahrbach et al.,, 2011), with possible contraction of its volume (Purkey & Johnson, 2012). It has also been indirectly suggested by sediment core records taken around the CDP that there is a millennium-scale variability of the production of local AABW (Harris, 2000). Ohshima et al. suggest it is vital that the CDBW is incorporated into the global assessment of the MOC, which is a key component of the climate system. If this is done it will improve numerical simulations that predict its response to long-term climate change.

Sources & Further reading

1. Ohshima, Kay I., Yasushi Fukamachi, Guy D. Williams, Sohey Nihashi, Fabien Roquet, Yujiro Kitade, Takeshi Tamura, et al. "Antarctic Bottom Water Production by Intense Sea-Ice Formation in the Cape Darnley Polynya." Nature Geosci 6, no. 3 (03//print 2013): 235-40.