Bottom Water Formation Research Articles

Distribution of anthropogenic CO2 in the Pacific Ocean

This work presents an estimate of anthropogenic CO2 in the Pacific Ocean based on measurements from the WOCE/JGOFS/OACES global CO2 survey. These estimates used a modified version of the ΔC* technique. Modifications include a revised preformed alkalinity term, a correction for denitrification, and an evaluation of the disequilibrium terms using an optimum multiparameter analysis. The total anthropogenic CO2 inventory over an area from 120°E to 70°W and 70°S to 65°N (excluding the South China Sea, the Yellow Sea, the Japan/East Sea, and the Sea of Okhotsk) was 44.5 ± 5 Pg C in 1994. Approximately 28 Pg C was located in the Southern Hemisphere and 16.5 Pg C was located north of the equator. The deepest penetration of anthropogenic CO2 is found at about 50°S. The shallowest penetration is found just north of the equator. Very shallow anthropogenic CO2 penetration is also generally observed in the high‐latitude Southern Ocean. One exception to this is found in the far southwestern Pacific where there is evidence of anthropogenic CO2 in the northward moving bottom waters. In the North Pacific a strong zonal gradient is observed in the anthropogenic CO2 penetration depth with the deepest penetration in the western Pacific. The Pacific has the largest total inventory in all of the southern latitudes despite the fact that it generally has the lowest average inventory when normalized to a unit area. The lack of deep and bottom water formation in the North Pacific means that the North Pacific inventories are smaller than the North Atlantic.

Importance of the Gulf of Aqaba for the formation of bottom water in the Red Sea

Conductivity‐temperature‐depth tracer and direct current measurements collected in the northern Red Sea in February and March 1999 are used to study the formation of deep and bottom water in that region. Historical data showed that open ocean convection in the Red Sea can contribute to the renewal of intermediate or deep water but cannot ventilate the bottom water. The observations in 1999 showed no evidence for open ocean convection in the Red Sea during the winter 1998/1999. The overflow water from the Gulf of Aqaba was found to be the densest water mass in the northern Red Sea. An anomaly of the chlorofluorocarbon component CFC‐12 observed in the Gulf of Aqaba and at the bottom of the Red Sea suggests a strong contribution of this water mass to the renewal of bottom water in the Red Sea. The CFC data obtained during this cruise are the first available for this region. Because of the new signal, it is possible for the first time to subdivide the deep water column into deep and bottom water in the northern Red Sea. The available data set also shows that the outflow water from the Gulf of Suez is not dense enough to reach down to the bottom of the Red Sea but was found about 250 m above the bottom.

A coupled climate model simulation of the Last Glacial Maximum, Part 1: transient multi-decadal response

A coupled atmosphere–ocean–sea ice–land surface climate system model developed at the Canadian Centre for Climate Modelling and Analysis (CCCma) is used to investigate the response to glacial ice-sheet topography and decreased CO2 representative of the last glacial maximum (LGM, roughly 21,000 years before present). Imposing these glacial boundary conditions leads to a strong and rapid climate adjustment, first in the atmosphere, and then in the ocean. We describe the model and boundary conditions and analyze the initial transient response. In a subsequent paper we analyze the near-equilibrium response. During the transient phase, surface air temperature rapidly decreases with time. The rapid cooling over the Laurentide and Fennoscandian ice sheets introduces a strong inter-hemispheric asymmetry. Sea surface temperature also decreases and after 80 years of integration, ocean surface temperature has decreased by 3 °C and global surface air temperature by 5 °C. The expected overall cooling is contrasted by marked regions of unexpected localized ocean surface warming and the retreat of sea ice in the northern North Atlantic and in the high latitude Southern Ocean. This strong transient behavior, associated with more vigorous oceanic convection at these high latitudes, represents the localization of the initial response in areas remote from the direct forcing change. The strength of the North Atlantic (Southern Ocean) overturning stream function, associated with the formation of North Atlantic Deep Water (Antarctic Bottom Water), increases from 12 to 36 Sv (2 to 26 Sv). There is a marked increase in the Antarctic Circumpolar Current transport as well (from 80 to 140 Sv), even though the wind stress decreases over the Southern Ocean, presumably as a consequence of the enhanced bottom pressure torque due to an increase in the density and the formation of Antarctic Bottom Water. Both precipitation and evaporation decrease on average with time and the change in the hydrological cycle modifies sea surface salinity (SSS). SSS increases in the Arctic Ocean due to a decrease in river runoff while it decreases in the North Atlantic due to an increase in river runoff and an increase in precipitation minus evaporation. The initial (and near equilibrium) climate response to LGM forcing in the fully coupled atmosphere–ocean climate model is quite different from that in the associated atmosphere-mixed layer ocean model due to very strong oceanic feedbacks. This is especially evident where oceanic convection is active. The anomalously warm surface temperature, especially in the Southern Ocean, and the marked increase in the overturning circulation are transient features of the response to imposed LGM boundary conditions which moderate or reverse as the system approaches equilibrium (as described in the companion paper). Other differences with the mixed layer simulation include a considerably colder tropical Pacific and the development of a La Nina-like response both of which are absent in the mixed layer simulations.

Correction to “A model study of ocean circulation beneath Filchner‐Ronne Ice Shelf, Antarctica: Implications for bottom water formation” by Adrian Jenkins and David M. Holland

Correction to “A model study of ocean circulation beneath Filchner‐Ronne Ice Shelf, Antarctica: Implications for bottom water formation” by Adrian Jenkins and David M. Holland

Late Quaternary variability of ocean circulation in the southeastern South Atlantic inferred from the terrigenous sediment record of a drift deposit in the southern Cape Basin (ODP Site 1089)

During Leg 177 of the Ocean Drilling Program, an expanded sequence of Pliocene to Holocene calcareous muds was recovered at Site 1089 on a drift deposit in the southern Cape Basin (SE South Atlantic). The reconstruction of detrital sources and modes of sediment transport gives insight into the operational modes of regional current systems in response to climate variability over the last 590 kyr, as inferred from sedimentological and mineralogical parameters of the terrigenous sediment fraction. Terrigenous sediments mainly originate from African sources with minor contributions from distant southern sources (South America and Antarctica) and are supplied by circumpolar water masses, North Atlantic Deep Water (NADW), and surface currents of the Agulhas Current. Changes in clay mineralogy as tracers of deep and shallow ocean circulation, best displayed by variations in quartz/feldspar ratios and kaolinite/chlorite ratios of clay, reflect both the northward displacement of NADW injection into the Antarctic Circumpolar Current and a weakening of Agulhas Current leakage from the Indian Ocean around South Africa to the South Atlantic during glacial stages, sub-stages, and stadials. Modifications of these regional current patterns are consistent with perturbations in global conveyor circulation and climate variability on Milankovitch and sub-Milankovitch time scales. Elevated mass-accumulation rates of terrigenous matter generally document high particle fluxes and focusing effects by bottom-current action throughout the late Quaternary. Current sorting and coarsening of terrigenous mud, independently of its source signals, prevails during interglacial periods and is linked to a stronger flow of Antarctic Bottom Water and the invigoration of deep contour currents in response to long-term changes (100-kyr cyclicity) in Antarctic ice-sheet dynamics, high-amplitude fluctuations in global sea level, and increased bottom-water formation.

On the Role of Sea Ice and Convection in a Global Ocean Model

Abstract An earlier estimate regarding the possible impact of sea ice on deep-ocean water mass properties and the global thermohaline circulation in a coupled sea ice–ocean general circulation model (OGCM) is updated. Compared to the earlier application, the main upgrade is a subgrid-scale plume-convection parameterization that replaces conventional grid-cell-wide convective adjustment. The different treatment of convection leads to some noticeable differences in some of the repeated sensitivity experiments. For example, in an experiment where sea ice salinity is assumed to be that of the upper ocean, thus neglecting the primary effect of sea ice formation and melting on the ocean's buoyancy forcing, Antarctic Bottom Water formation comes essentially to a halt, the global deep-ocean properties and thermohaline circulation thus being almost solely determined by North Atlantic Deep Water. The much weaker impact in the earlier estimate turns out to be mainly due to excessive open-ocean convection in the Sout...

On the near-bottom variability in the northwestern Weddell Sea

The thermohaline data from the first Brazilian hydrographic cruise to the northwestern Weddell Sea (AR XVIII) revealed significant near-bottom changes in water-column properties over seasonal and interannual time scales. Favorable ice conditions in 2000 allowed a dense station coverage of the area including the main pathways for Weddell Sea deep and bottom waters. The new results are compared with the data from the 1998 German cruise ANT XV/4 and other historical data. A warming of the bottom layer was discovered that was more attributable to short-term seasonal or interannual fluctuations in the formation of cold bottom water than to a long-term trend. There was both consistency and variation between and within seasons. Invariant bottom-water characteristics were observed in different seasons (summer/winter), and variable bottom-water characteristics were observed in the same season (summer) at the same locations. This reduces the possibility of a dominant seasonal effect. Instead, we propose that the intermittent behavior of small cold-water sources along the Weddell Sea's periphery causes the variability measured in the deep northwestern Weddell Sea. The observed variability has consequences for the water-mass export across the South Scotia Ridge, as the absence of the fresher/lighter Weddell Sea Bottom Water south of South Orkney Plateau during AR XVIII might be linked to a reduced ventilation of the deep Scotia Sea. The results of this study show the need for ongoing efforts for establishing a long-term monitoring of this region with global importance.

Sediment denitrification in the Gulf of Mexico zone of hypoxia

The largest zone of anthropogenic bottom water hypoxia in the Western Hemisphere occurs seasonally in the northern Gulf of Mexico between the Mississippi River delta and the coast of eastern Texas. This zone of hypoxia reaches its greatest extent in the summer months and is a conse- quence of seasonal stratification of the water column com- bined with the decomposition of organic matter derived from accelerated rates of primary production. This enhanced pro- ductivity is driven primarily by the input of inorganic nitrogen from the Mississippi River. There are 3 likely sinks for fixed nitrogen within this zone of hypoxia: sequestration in the sediment, dispersion and dilution into the Gulf of Mexico, and denitrification. We assessed potential denitrification rates at 7 stations in the zone of hypoxia during the summer of 1999. Those data are compared with bottom water nitrate, ammo- nium and dissolved oxygen (DO) concentrations. No denitri- fication was observed in the water column. Denitrification potential rates in the surface sediments were unexpectedly low and ranged between 39.8 and 108.1 µmol m -2 h -1 . The highest rates were observed at stations with bottom water DO concentrations between 1 and 3 mg l -1 . Denitrification activity was significantly lower at stations where DO was lower than 1 mg l -1 or higher than 3 mg l -1 . Nutrient data for these sta- tions demonstrate that as anoxia is approached, the dominant species of nitrogen shifts from nitrate to ammonium. The shift in nitrogen species suggests competition between microbial populations in the sediment community. The lower denitrifi- cation rates at stations with bottom water DO <1 mg l -1 may be due to nitrate limitation or an increase in the competitive advantage of microorganisms capable of dissimilatory nitrate reduction to ammonium (DNRA). Suppression of denitrifica- tion at low DO by any mechanism will increase the residence time of bioavailable nitrogen. This trend could act as a posi- tive feedback mechanism in the formation of hypoxic bottom waters.

Long-term temperature trends in the deep waters of the Weddell Sea

Warming of the deep water in the Weddell Sea has important implications for Antarctic bottom water formation, melting of pack ice, and the regional ocean–atmosphere heat transfer. In order to evaluate warming trends in the Weddell Sea, a historical data set encompassing CTD and bottle data from 1912 to 2000 was analyzed for temporal trends in the deep water masses: warm deep water (WDW) and Weddell Sea deep water (WSDW). The coldest WDW temperatures were primarily associated with the Weddell Polynya of the mid-1970s. Subsequent warming occurred at a rate of B0.01270.0071 Cy r � 1 from the 1970s to 1990s. This warming was comparable to the global, average surface water warming observed by Levitus et al. (Science 287 (2000) 2225), to the warming of the WSBW in the central Weddell Sea observed by Fahrbach et al. (Filchner–Ronne Ice Shelf Program, Report No. 12, Alfred-Wegener-Institut, Bremerhaven, Germany, 1998a, p. 24), and to the surface ice temperature warming from 1970 to 1998 in the Weddell Sea observed by Comiso (J. Climate 13 (2000) 1674). The warming was not compensated by an increase in salinity, and thus the WDW became less dense. The location of the warmest temperature was displaced towards the surface by B200 m from the 1970s to the 1990s. Although the average WSDW potential temperatures between 1500 and 3500 m were warmer in the 1990s than in the 1970s, high variability in the data prevented identification of a well-defined temporal trend. r 2002 Elsevier Science Ltd. All rights reserved.

Multinational effort studies differences among arctic ocean models

The Arctic Ocean is an important component of the global climate system. The processes occurring in the Arctic Ocean affect the rate of deep and bottom water formation in the convective regions of the high North Atlantic and influence ocean circulation across the globe. This fact is highlighted by global climate modeling studies that consistently show the Arctic to be one of the most sensitive regions to climate change. But an identification of the differences among models and model systematic errors in the Arctic Ocean remains unchecked, despite being essential to interpreting the simulation results and their implications for climate variability. For this reason, the Arctic Ocean Model Intercomparison Project (AOMIP), an international effort, was recently established to carry out a thorough analysis of model differences and errors. The geographical focus of this effort is shown in Figure 1.

Causes of long-term changes in Aegean sea deep water

Recent observations of newly formed deep water in the Aegean sea prompts this analysis of whether such deep water could be formed locally by the combination of an increase in net evaporation and wintertime water mass transformation. River diversion projects in Russia and Egypt since the 1950s have effectively increased the amount of net evaporation over the eastern Mediterranean basin. Historical profiles show that low salinity intermediate waters separated the deep and upper waters in the Aegean in 1961–1962. Within a simple mixed layer model, imposing a small net evaporation of 10 cm·yr –1 on the observed hydrographic conditions in March 1962 results in the gradual erosion of the low salinity intermediate waters. After 25 years, the low salinity intermediate waters are absent in agreement with observations made in September 1987. Continuing to force the model with the small net evaporation and with monthly heat and freshwater exchange anomalies from March 1987, new deep water could have been formed as early as March 1988. In the model, major deep water formation events occurred during the severe winters of 1991–1992 and 1992–1993 resulting in the formation of saltier and notably colder deep waters. The effective increase in net evaporation slowly increases the salinity and decreases the stratification in the Aegean sea until a severe winter leads to deep convection and new bottom water formation.

Collapse and reorganisation of the Southern Ocean overturning under global warming in a coupled model

This study investigates the long‐term behaviour of the thermohaline circulation (THC) in the CSIRO climate model, under a scenario of transient increase of atmospheric (equivalent) CO2 concentration followed by a perpetual stabilisation at triple the initial level (3 × CO2). The North Atlantic Deep Water Formation (NADWF) declines substantially and the Antarctic Bottom Water Formation (AABWF) essentially ceases by the time of CO2 tripling. During the subsequent millennium of stable 3 × CO2, NADWF recovers slightly but the AABWF shows no sign of returning and the residual deep overturning dies away. Accelerating the convergence to equilibrium of the deep ocean under the 3 × CO2 condition, the global THC eventually reaches a near‐stable state with the entire ocean warming by about 7°C, NADWF fully recovered and AABWF partly re‐established. This result shows a possible new quasi‐equilibrium of the ocean under long‐term global warming induced by the anthropogenic CO2 increase.

Late Pleistocene deep-water circulation in the subantarctic eastern Atlantic

We present three new benthic foraminiferal δ 13C, δ 18O, and total organic carbon time series from the eastern Atlantic sector of the Southern Ocean between 41°S and 47°S. The measured glacial δ 13C values belong to the lowest hitherto reported. We demonstrate a coincidence between depleted late Holocene (LH) δ 13C values and positions of sites relative to ocean surface productivity. A correction of +0.3 to +0.4 [‰ VPDB] for a productivity-induced depletion of Last Glacial Maximum (LGM) benthic δ 13C values of these cores is suggested. The new data are compiled with published data from 13 sediment cores from the eastern Atlantic Ocean between 19°S and 47°S, and the regional deep and bottom water circulation is reconstructed for LH (4–0 ka) and LGM (22–16 ka) times. This extends earlier eastern Atlantic-wide synoptic reconstructions which suffered from the lack of data south of 20°S. A conceptual model of LGM deep-water circulation is discussed that, after correction of southernmost cores below the Antarctic Circumpolar Current (ACC) for a productivity-induced artifact, suggests a reduced formation of both North Atlantic Deep Water in the northern Atlantic and bottom water in the southwestern Weddell Sea. This reduction was compensated for by the formation of deep water in the zone of extended winter sea-ice coverage at the northern rim of the Weddell Sea, where air–sea gas exchange was reduced. This shift from LGM deep-water formation in the region south of the ACC to Holocene bottom water formation in the southwestern Weddell Sea, can explain lower preformed δ 13C DIC values of glacial circumantarctic deep water of approximately 0.3‰ to 0.4‰. Our reconstruction brings Atlantic and Southern Ocean δ 13C and Cd/Ca data into better agreement, but is in conflict, however, with a scenario of an essentially unchanged thermohaline deep circulation on a global scale. Benthic δ 18O-derived LGM bottom water temperatures, by 1.9°C and 0.3°C lower than during the LH at deepest southern and shallowest northern sites, respectively, agree with the here proposed reconstruction of deep-water circulation in the eastern South Atlantic Ocean.

Warming and structural changes in the east (Japan) Sea: A clue to future changes in global oceans?

The East (Japan) Sea has been in a warming trend during the last more than 40 years: a 0.1–0.5 degree of warming in the upper 1000 meters. The warming is accompanied by the deepening of the oxygen minimum layer by more than 1000 meters. The analysis of chemical tracers such as dissolved oxygen and CFCs clearly implies that the warming is associated with changes in deep water structures in the area, resulted from a replacement of the past bottom‐water formation with an intermediate water formation in recent time. This shift has a remarkable resemblance to that anticipated to the ocean conveyor‐belt system in coming century associated with recent global warming. In considering a rapid turn‐over time of time scale less than 100 years, the East Sea may serve as a natural laboratory for global changes in the future.

On the origin of the deep CFC maximum in the Eastern Weddell Sea—Numerical model results

CFC tracer observations indicate that Prydz Bay in the Indian sector of the Southern Ocean is a region of deep and bottom water formation. Results of a circumpolar ocean circulation model which includes CFC, an age tracer, and numerical floats indicate Prydz Bay as being a convection site which contributes to the reservoir of freshly ventilated waters in the Weddell Sea. In contrast to the newly formed Weddell Sea Bottom Water, captured near bottom, water masses of Prydz Bay origin spread on horizons which pass the ridges confining the Weddell Sea, therefore, contributing directly to the ventilation of the global abyss.

Numerical evidence against reversed thermohaline circulation in the warm Paleocene/Eocene ocean

The question of whether deep water formation might have occurred in subtropical latitudes in the early Cenozoic is examined through use of a global ocean model forced by mixed boundary conditions. Zonal mean surface temperatures and wind stresses are derived from an atmospheric general circulation model (AGCM) simulation of the warm Paleocene/Eocene boundary interval (∼55 Ma) and are held constant for a series of sensitivity tests. The control case for moisture flux (evaporation minus precipitation,E‐P), also derived from the AGCM, is perturbed so that the subtropical evaporation increases and high‐latitude precipitation increases. A dramatic response is seen in the temperature and salinity structure of the model ocean, but the perturbation does not result in deep convection in subtropical latitudes. In all cases, bottom water is formed in the southern high latitudes, and the global meridional overturning is characterized by a strongly asymmetric circulation. No multiple equilibria have been found for any particularE‐Pconfiguration. In the most extreme case (5 times the controlE‐P) the model oscillates between meridional overturning circulation “on” and “off”. Shorter‐lived thermohaline slowing and reinvigoration are observed as a transient response under less extremeE‐Pperturbations. Despite the high evaporation implied in the perturbation experiments, mean mixed layer salinities in the subtropics do not rise much above the control case because of efficient removal of salt (and heat) through deepened subduction beneath the subtropical gyres. The sensitivity of the results to the parameterization of continental runoff and the specified diapycnal mixing coefficient (Kν) are also examined. Distributing runoff purely zonally, rather than globally, has approximately the same effect as a 50% increase in the strength of the hydrologic cycle. DecreasingKνto 0.3 cm2s−1from the standard value of 1.0 cm2s−1increases the sensitivity to an increased hydrologic cycle considerably, but in no case does low‐latitude deep water formation occur, indicating that subtropical bottom water formation is implausible in a model with some degree of realism. These experiments support changes in moisture flux as a mechanism for ocean warming (largely in the thermocline through intermediate water depths), but the process involved is deepened subtropical subduction and not subtropical deep water formation.

How iceberg calving and grounding change the circulation and hydrography in the Filchner Ice Shelf‐Ocean System

The formation of bottom water in the southern Weddell Sea is strongly influenced by the flow of Ice Shelf Water (ISW) out of the Filchner‐Ronne Ice Shelf cavity. The breakout of three giant icebergs in 1986 and their grounding on the shallow Berkner Bank modified the circulation and water mass formation in the Filchner Trough and the adjacent sea areas. Hydrographic measurements along the Filchner Ice Shelf front, carried out with RV Polarstern in 1995, show significant changes in the water mass characteristics and flow patterns in the Filchner Trough in comparison to measurements from the early 1980s. Changes in the trough will affect the flow over the sill to the deep Weddell Abyssal Plain. We combine a three‐dimensional ocean circulation model with conductivity‐temperature‐depth and stable isotope measurements to investigate the details of the circulation in front of and beneath the Filchner Ice Shelf. We assess the impact of stranded icebergs and a more southerly ice shelf front position caused by a 1986 iceberg calving event on the circulation and observed water mass properties. Results indicate variations of the flow pattern in the Filchner Trough and on Berkner Bank, where High‐Salinity Shelf Water, the feedstock for ISW, is produced. The calving and grounding impacts illustrate the sensitivity of the ice shelf‐ocean system to perturbations in local bathymetric settings.

China's Development Could Lead to Bottom Water Formation in the Japan/East Sea

Abstract Using hydrographic data and box models, it is shown that the presently discussed diversion of rivers such as the Yellow or the Yangtze for agricultural use is likely to cause the renewal of Bottom Water formation in the Japan/East Sea. Such formation was common (near the Siberian coast) in the 1930s, 1940s, and 1950s, but subsided since that time due to a warming trend (accompanied by a decreased salinity due to the melting of ice). Since a diversion of freshwater is analogous to evaporation, a (diversion induced) increase of salinity is expected and the increase is large enough to allow Bottom Water formation even at the present–day cooling rates. Even a modest diversion of merely 3000 m3 s–1 (which is 10% of the total freshwater flux) will probably cause Bottom Water formation at a rate of roughly 750 000 m3 s–1. This is the first study that predicts anthropogenic reversal of an existing vertical structure in a semienclosed sea.

Oxygen and carbon stable isotope tracers of Weddell Sea water masses: new data and some paleoceanographic implications

Stable oxygen isotopic composition of sea water and stable carbon isotopes of dissolved inorganic carbon (DIC) on the continental shelf in the southern Weddell Sea are presented. Using the stations sampled during the summer 1995 two sections can be constructed, one closely parallel to the ice shelf edge and the other perpendicular to the upper continental slope. Generally, δ 18O values clearly separate between different shelf water masses depending on the content of meteoric meltwater added during melting of glacial ice. Extrapolation of the mixing line between the cores of High Salinity Shelf Water (HSSW) and supercooled Ice Shelf Water (ISW) reveals δ 18O values of the glacial ice of −27‰, whereas extrapolation of the mixing line between the δ 18O values of the most-saline HSSW and lowest temperature ISW results in δ 18O values of −34‰ for glacial ice. These values point to an origin of meltwater from below the ice shelf, where ice is less depleted in 18O, since deep beneath the ice shelf close to the grounding line, values may reach −40‰. If values between −34 and −27‰ are used as δ 18O end member values for glacial ice, the amount of meltwater from the ice shelf that adds to the formation of ISW off the Filchner–Ronne Ice Shelf ranges from 0.2 to 0.8%, in agreement with previous studies based on δ 18O and 4He. Carbon isotopic fractionation due to gas exchange between the atmosphere and the ocean at cold temperatures results in Δ δ 13C DIC values of 0.20±0.17‰ for Weddell Sea Deep Water, the water mass that ventilates the global abyssal ocean, typically defined as Antarctic Bottom Water (AABW). This confirms the low end of the range estimated previously (0.2–0.4‰), and thus corroborates the dominance of biology in shaping the deep and bottom water δ 13C signal. It has been hypothesized that different modes of glacial/interglacial Antarctic bottom water formation may be separated by different stable isotopic compositions of deep-sea foraminiferal calcite. Here I show that differences between Δ δ 13C and δ 18O values of HSSW and ISW, both of which contribute to bottom water formation today, are too small to be resolved in deep and bottom water masses. Therefore, glacial/interglacial changes in relative proportions of these water masses in Antarctic deep and bottom water cannot be separated by stable isotopes of fossil benthic foraminiferal calcite.

The sources of Antarctic bottom water in a global ice–ocean model

Two mechanisms contribute to the formation of Antarctic bottom water (AABW). The first, and probably the most important, is initiated by the brine released on the Antarctic continental shelf during ice formation which is responsible for an increase in salinity. After mixing with ambient water at the shelf break, this salty and dense water sinks along the shelf slope and invades the deepest part of the global ocean. For the second one, the increase of surface water density is due to strong cooling at the ocean–atmosphere interface, together with a contribution from brine release. This induces deep convection and the renewal of deep waters. The relative importance of these two mechanisms is investigated in a global coupled ice–ocean model. Chlorofluorocarbon (CFC) concentrations simulated by the model compare favourably with observations, suggesting a reasonable deep water ventilation in the Southern Ocean, except close to Antarctica where concentrations are too high. Two artificial passive tracers released at surface on the Antarctic continental shelf and in the open-ocean allow to show clearly that the two mechanisms contribute significantly to the renewal of AABW in the model. This indicates that open-ocean convection is overestimated in our simulation. Additional experiments show that the amount of AABW production due to the export of dense shelf waters is quite sensitive to the parameterisation of the effect of downsloping and meso-scale eddies. Nevertheless, shelf waters always contribute significantly to deep water renewal. Besides, increasing the P.R. Gent, J.C. McWilliams [Journal of Physical Oceanography 20 (1990) 150–155] thickness diffusion can nearly suppress the AABW formation by open-ocean convection.