Deep drilling of glaciers: Russian projects in the Arctic (1975-1995)

The data collection "Deep Drilling of Glaciers: Soviet-Russian projects in Arctic, 1975-1995" was collected by the following basic considerations: - compilation of deep (>100 m) drilling projects on Arctic glaciers, using data of (a) publications; (b) archives of IGRAN; (c) personal communication of project participants; - documentation of parameters, references. Accuracy of data and techniques applied to determine different parameters are not evaluated. The accuracy of some geochemical parameters (up to 1984 and heavy metalls) is uncertain. Most reconstructions of ice core age and of annual layer thickness are discussed; - digitizing of published diagrams (in case, when original numerical data were lost) and subsequent data conversion to equal range series and adjustment to the common units. Therefore, the equal-range series were calculated from original data or converted from digitized chart values as indicated in the metadata. For the methodological purpose, the equal-range series obtained from original and reconstructed data were compared repeatedly; the systematic difference was less then 5-7%. Special attention should be given to the fact, that the data for individual ice core parameters varies, because some parameters were originally measured or registered. Parameters were converted in equal-range series using 2 m steps; - two or more parameter values were determined, then the mean-weighted (i.e. accounting the sample length) value is assigned to the entire interval; - one parameter value was determined, measured or registered independently from the parameter values in depth intervals which over- and underlie it, then the value is assigned to the entire interval; - one parameter value was determined, measured or registered for two adjoining depth intervals, then the specific value is assigned to the depth interval, which represents >75% of sample length ; if each of adjoining depth intervals represents <75% of sample length, then the correspondent parameter value is assigned to both intervals of depth. This collection of ice core data (version 2000) was made available through the EU funded QUEEN project by S.M. Arkhipov, Moscow.

Stable isotope ratios and foraminiferal abundance of sediment cores from the Western Iberian Margin

Rapid climate changes at the onset of the last deglaciation and during Heinrich Event H4 were studied in detail at IMAGES cores MD95-2039 and MD95-2040 from the Western Iberian margin. A major reorganisation of surface water hydrography, benthic foraminiferal community structure, and deepwater isotopic composition commenced already 540 years before the Last Isotopic Maximum (LIM) at 17.43 cal. ka and within 670 years affected all environments. Changes were initiated by meltwater spill in the Nordic Seas and northern North Atlantic that commenced 100 years before concomitant changes were felt off western Iberia. Benthic foraminiferal associations record the drawdown of deepwater oxygenation during meltwater and subsequent Heinrich Events H1 and H4 with a bloom of dysoxic species. At a water depth of 3380 m, benthic oxygen isotopes depict the influence of brines from sea ice formation during ice-rafting pulses and meltwater spill. The brines conceivably were a source of ventilation and provided oxygen to the deeper water masses. Some if not most of the lower deep water came from the South Atlantic. Benthic foraminiferal assemblages display a multi-centennial, approximately 300-year periodicity of oxygen supply at 2470-m water depth. This pattern suggests a probable influence of atmospheric oscillations on the thermohaline convection with frequencies similar to Holocene climate variations. For Heinrich Events H1 and H4, response times of surface water properties off western Iberia to meltwater injection to the Nordic Seas were extremely short, in the range of a few decades only. The ensuing reduction of deepwater ventilation commenced within 500-600 years after the first onset of meltwater spill. These fast temporal responses lend credence to numerical simulations that indicate ocean-climate responses on similar and even faster time scales.

Barium in seawater and sediments from the Artic Ocean

As part of ongoing circulation studies in the Arctic, seawater samples for dissolved Ba concentrations were obtained during Sep.-Oct., 1992 at several locations in the Bering Strait, Eastern Chukchi and Southern Beaufort Seas. The results reveal a dynamic rang (10 to 150 nmol/kg) for this element in the Arctic equal to or greater than that in combined Atlantic, Indian and Pacific oceans. Lowest levels are observed in surface waters, with values tending to decrease northwards in the direction of currents generally flowing frorn the Bering Strait along the Alaskan coast. Low surfacc concentrations tend to be accompanied by relatively enriched near bottom levels. On the basis of these spatial distributions, hydrographic observations and a knowledge of its behavior in other marine settings, it appears that Ba can be significantly depleted from surface waters as a result of the highly seasonal biological aclivities over Arctic marginal shelves. Removal at the surface is counteracted to some extent by regeneration at depth or in the sediments and by riverine inputs. The biologically related drawdown is likely to enhance the contrast between 'background' surface Ba levels in the Arctic and waters imprinted by regeneration and/or rivers, These preliminary findings suggest that Ba holds particular promise for tracing river waters and the ventilation of halodine waters hy laterally sinking brines produced during ice formation over the shelves.

Physical and chemical oceanography in the Arctic Ocean

Data from sections across the Eurasian Basin of the Arctic Ocean occupied by the German Research Vessel Polarstern in 1987 and by the Swedish icebreaker Oden in 1991 are used to derive information on the freshwater balance of the Arctic Ocean halocline and on the sources of the deep waters of the Nansen, Amundsen and Makarov basins. Salinity, d18O and mass balances allow separation of the river-runoff and the sea-ice meltwater fractions contained in the Arctic halocline. This provides the basis for tracking the river-runoff signal from the shelf seas across the central Arctic Ocean to Fram Strait. The halocline has to be divided into at least three lateral regimes: the southern Nansen Basin with net sea-ice melting, the northern Nansen Basin and Amundsen Basin with net sea-ice formation and increasing river-runoff fractions, and the Canadian Basin with minimum sea-ice meltwater and maximum river-runoff fractions and water of Pacific origin. In the Canadian Basin, silicate is used as a tracer to identify Pacific water entering through Bering Strait and an attempt is made to quantify its influence on the halocline waters of the Canadian Basin. For this purpose literature data from the CESAR and LOREX ice camps are used. Based on mass balances and depending on the value of precipitation over the area of the Arctic Ocean the average mean residence time of the river-runoff fraction contained in the Arctic Ocean halocline is determined to be about 14 or 11 years. Water column inventories of river-runoff and sea-ice meltwater are calculated for a section just north of Fram Strait and implications for the ice export rate through Fram Strait are discussed. Salinity, tritium, 3He and the d18O ratio of halocline waters sampled during the 1987 Polarstern cruise to the Nansen Basin are used to estimate the mean residence time of the river-runoff component in the halocline and on the shelves of the Arctic Ocean. These estimates are done by comparing ages of the halocline waters based on a combination of tracers yielding different time information: the tritium 'vintage' age which records the time that has passed since the river-runoff entered the shelf and the tritium/3He age which reflects the time since the shelf waters left the shelf. The difference between the ages determined by these two methods is about 3 to 6 years. Correction for the initial tritium/3He age of the shelf waters (about 0.5 to 1.5 years) yields a mean residence time of the river-runoff on the shelves of about 3.5 ± 2 years. Comparison of the 18O/16O ratios of shelf water, Atlantic water and the deep waters of the Arctic Ocean indicate that the sources of the deep and bottom waters of the Eurasian Basin are located in the Barents and Kara seas.

Physicochemical characteristics and protist occurrence at stations sampled in 2007/2008 in the eastern Beaufort Sea

This study documents, for the first time, the abundance and species composition of protist assemblages in Arctic sea ice during the dark winter period. Lack of knowledge of sea-ice assemblages during the dark period has left questions about the retention and survival of protist species that initiate the ice algal bloom. Sea-ice and surface water samples were collected between December 27, 2007 and January 31, 2008 within the Cape Bathurst flaw lead, Canadian Beaufort Sea. Samples were analyzed for protist identification and counts, chlorophyll (chl) a, and total particulate carbon and nitrogen concentrations. Sea-ice chl a concentrations (max. 0.27 µg/l) and total protist abundances (max. 4 x 10**3 cells/l) were very low, indicating minimal retention of protists in the ice during winter. The diversity of winter ice protists (134 taxa) was comparable to spring ice assemblages. Pennate diatoms dominated the winter protist assemblage numerically (averaging 77% of total protist abundances), with Nitzschia frigida being the most abundant species. Only 56 taxa were identified in surface waters, where dinoflagellates were the dominant group. Our results indicate that differences in the timing of ice formation may have a greater impact on the abundance than structure of protist assemblages present in winter sea ice and at the onset of the spring ice algal bloom.

Geochemistry of gas hydrates of DSDP Hole 84-570

During Deep Sea Drilling Project (DSDP) Leg 84 a core 1 m long and 6 cm in diameter of massive gas hydrate was unexpectedly recovered at Site 570 in upper slope sediment of the Middle America Trench offshore of Guatemala. This core contained only 5-7% sediment, the remainder being the solid hydrate composed of gas and water. Samples of the gas hydrate were decomposed under controlled conditions in a closed container maintained at 4°C. Gas pressure increased and asymptotically approached the equilibrium decomposition pressure for an ideal methane hydrate, CH4.5-3/4H2O, of 3930 kPa and approached to this pressure after each time gas was released, until the gas hydrate was completely decomposed. The gas evolved during hydrate decomposition was 99.4% methane, ~0.2% ethane, and ~0.4% CO2. Hydrocarbons from propane to heptane were also present, but in concentrations of less than 100 p.p.m. The carbon-isotopic composition of methane was -41 to -44 per mil, relative to PDB standard. The observed volumetric methane/water ratio was 64 or 67, which indicates that before it was stored and analyzed, the gas hydrate probably had lost methane. The sample material used in the experiments was likely a mixture of methane hydrate and water ice. Formation of this massive gas hydrate probably involved the following processes: (i) upward migration of gas and its accumulation in a zone where conditions favored the growth of gas hydrates, (ii) continued, unusually rapid biological generation of methane, and (iii) release of gas from water solution as pressure decreased due to sea level lowering and tectonic uplift.

Middle Miocene stable carbon and oxygen isotope record of foraminifera of ODP Hole 120-747A

Paleoceanographic variability at southern high latitude Ocean Drilling Program (ODP) Site 747 was investigated in this study through the interval which spans the Middle Miocene Climate Transition (MMCT). Between 15.0 and 12.2 million years ago (Ma), foraminiferal d18O records derived from both benthic (Cibicidoides spp.) and planktonic taxa (Globorotalia praescitula and Globigerina bulloides) reveal a history of changes in water column thermal and salinity structure and a strong imprint of seasonality. Prior to the MMCT, in the interval between 14.35 and 13.9 Ma, G. bulloides displays relatively high d18O values similar to those of G. praescitula, interpreted to indicate weakening of the thermocline and/or increased seasonality with cooler early-spring and/or late-fall temperatures. Following this interval, G. bulloidesd18O values diverge significantly from benthic and G. praescitula values, with G. bulloides values remaining relatively low for at least 600 kyr following the benthic foraminiferal d18O shift during the MMCT at ~13.9 Ma. This divergence in d18O records occurs in direct association with the Mi3 cooling and glaciation event and may suggest: (1) a strengthening of the vertical temperature gradient, with greater cooling of deep waters than surface waters, (2) changes in the depth habitat of G. bulloides, (3) changes in the dominant season of G. bulloides calcification, (4) modification of surface-water d18O values in association with enhanced sea-ice formation, (5) increased surface-water carbonate ion concentration, and/or (6) a significant decrease in surface-water salinity across the MMCT. The first of these possible scenarios is not likely, particularly in light of recent Mg/Ca evidence for significant surface-water cooling in the Southern Ocean associated with the MMCT. Of the remaining possibilities, we favor a change in surface salinity to explain the observed trends in d18O values and hypothesize that surface salinity may have decreased by up to 2 salinity units at ~13.9 Ma. In this scenario, the development of a lower-salinity Antarctic surface layer coincided with regional cooling of both surface and deep waters of the Southern Ocean during the Mi3 glaciation of East Antarctica, and contributed into the dominance of Neogloboquadrina spp. between 13.8 and 13.2 Ma. Additionally, the distinct patterns observed in planktonic foraminiferal d18O records spanning the MMCT correspond with changes in the vertical d13C gradient between planktonic and benthic foraminiferal records and major changes in planktonic foraminiferal assemblages at Site 747, providing further evidence of the environmental significance of this climatic transition.

Tab. 3: Concentrations of selected chemical species at various depths in Ace Lake

This article reviews the history, chemical stratification, biology and biogeochemistry of Ace Lake, which is one of the many marine-derived meromictic (permanently stratified) lakes in the Vestfold Hills, Eastern Antarctica. The lake has an area of 18 ha, a maximum depth of 25 m, and a salinity range from 7 to 43 g l**-1. The lake mixes to a depth of 7 m in late winter as a result of brine freeze out during ice formation. Deeper mixing is precluded by a sharp halocline. The water beneath 12 m is permanently anoxic, The lake was formed approximately 10,800 yr BP as the polar ice cap melted. Sea level rise 7,800 yr BP resulted in invasion of seawater into the initially freshwater lake. Subsequently, sea level dropped, and the now saline lake became isolated from the ocean. The biota of the lake was derived from species trapped when the connection between the lake and the ocean was cut off. The oxic zone above 12 m supports a relatively simple community which includes microbial mats, four major species of phytoplankton (including a picocyanobacterium), two copepod species, and a variety of heterotrophic flagellates and ciliates. The anoxic zone contains populations of photosynthetic sulfur, sulfate reducing, fermentative and methanogenic bacteria, which combine to remineralise organic carbon which sediments from the upper waters. Research on the physics, biology and chemistry of Ace Lake has contributed significantly to knowledge of Antarctic meromictic lakes.

Stable isotope record of foraminifera from the subarctic Pacific

Benthic (Uvigerina spp., Cibicidoides spp., Gyroidinoides spp.) and planktonic (N. pachyderma sinistral, G. bulloides) stable isotope records from three core sites in the central Gulf of Alaska are used to infer mixed-layer and deepwater properties of the late glacial Subarctic Pacific. Glacial-interglacial amplitudes of the planktonic delta18O records are 1.1-1.3 per mil, less than half the amplitude observed at core sites at similar latitudes in the North Atlantic; these data imply that a strong, negative deltaw anomaly existed in the glacial Subarctic mixed layer during the summer, which points to a much stronger low-salinity anomaly than exists today. If true, the upper water column in the North Pacific would have been statically more stable than today, thus suppressing convection even more efficiently. This scenario is further supported by vertical (i.e., planktic versus benthic) delta18O and delta13C gradients of >1 per mil, which suggest that a thermohaline link between Pacific deep waters and the Subarctic Pacific mixed layer did not exist during the late glacial. Epibenthic delta13C in the Subarctic Pacific is more negative than at tropical-subtropical Pacific sites but similar to that recorded at Southern Ocean sites, suggesting ventilation of the deep central Pacific from mid-latitude sources, e.g., from the Sea of Japan and Sea of Okhotsk. Still, convection to intermediate depths could have occurred in the Subarctic during the winter months when heat loss to the atmosphere, sea ice formation, and wind-driven upwelling of saline deep waters would have been most intense. This would be beyond the grasp of our planktonic records which only document mixed-layer temperature-salinity fields extant during the warmer seasons. Also we do not have benthic isotope records from true intermediate water depths of the Subarctic Pacific.

Physical oceanography and hydrochemistry from the Laptev Sea, Arctic Ocean

Combined d18O/salinity data reveal a distinctive water mass generated during winter sea ice formation which is found predominantly in the coastal polynya region of the southern Laptev Sea. Export of the brine-enriched bottom water shows interannual variability in correlation with atmospheric conditions. Summer anticyclonic circulation is favoring an offshore transport of river water at the surface as well as a pronounced signal of brine-enriched waters at about 50 m water depth at the shelf break. Summer cyclonic atmospheric circulation favors onshore or an eastward, alongshore water transport, and at the shelf break the river water fraction is reduced and the pronounced brine signal is missing, while on the middle Laptev Sea shelf, brine-enriched waters are found in high proportions. Residence times of bottom and subsurface waters on the shelf may thereby vary considerably: an export of shelf waters to the Arctic Ocean halocline might be shut down or strongly reduced during "onshore" cyclonic atmospheric circulation, while with "offshore" anticyclonic atmospheric circulation, brine waters are exported and residence times may be as short as 1 year only.

Sedimentology and geochemistry of core GIK23258-2 in the Barents Sea

At the western continental margin of the Barents Sea, 75°N, hemipelagic sediments provide a record of Holocene climate change with a time resolution of 10-70 years. Planktic foraminifera counts reveal a very early Holocene thermal optimum 10.7-7.7 kyr BP, with summer sea surface temperatures (SST) of 8°C and a much enhanced West Spitsbergen Current. There was a short cooling between 8.8 and 8.2 kyr BP. In the middle and late Holocene summer, SST dropped to 2.5°-5.0°C, indicative of reduced Atlantic heat advection, except for two short warmings near 2.2 and 1.6 kyr BP. Distinct quasi-periodic spikes of coarse sediment fraction (with large portions of lithic grains, benthic and planktic foraminifera) record cascades of cold, dense winter water down the continental slope as a result of enhanced seasonal sea ice formation and storminess on the Barents shelf over the entire Holocene. The spikes primarily cluster near recurrence intervals of 400-650 and 1000-1350 years, when traced over the entire Holocene, but follow significant 885-/840- and 505-/605-year periodicities in the early Holocene. These non-stationary periodicities mimic the Greenland-[Formula: See Text]Be variability, which is a tracer of solar forcing. Further significant Holocene periodicities of 230, (145) and 93 years come close to the deVries and Gleissberg solar cycles.

Water mass evolution of the Greenland Sea since late glacial times

Four sediment cores from the central and northern Greenland Sea basin, a crucial area for the renewal of North Atlantic deep water, were analyzed for planktic foraminiferal fauna, planktic and benthic stable oxygen and carbon iso- topes as well as ice-rafted debris to reconstruct the environ- mental variability in the last 23 kyr. During the Last Glacial Maximum, the Greenland Sea was dominated by cold and sea-ice bearing surface water masses. Meltwater discharges from the surrounding ice sheets affected the area during the deglaciation, influencing the water mass circulation. During the Younger Dryas interval the last major freshwater event occurred in the region. The onset of the Holocene interglacial was marked by an increase in the advection of Atlantic Wa- ter and a rise in sea surface temperatures (SST). Although the thermal maximum was not reached simultaneously across the basin, benthic isotope data indicate that the rate of overturn- ing circulation reached a maximum in the central Greenland Sea around 7ka. After 6-5ka a SST cooling and increas- ing sea-ice cover is noted. Conditions during this so-called "Neoglacial" cooling, however, changed after 3 ka, probably due to enhanced sea-ice expansion, which limited the deep convection. As a result, a well stratified upper water column amplified the warming of the subsurface waters in the central Greenland Sea, which were fed by increased inflow of At- lantic Water from the eastern Nordic Seas. Our data reveal that the Holocene oceanographic conditions in the Green- land Sea did not develop uniformly. These variations were a response to a complex interplay between the Atlantic and Polar water masses, the rate of sea-ice formation and melting and its effect on vertical convection intensity during times of Northern Hemisphere insolation changes.

Stable isotopes, Mg/Ca-ratios, and biogenic opal of five sediment cores from the subarctic northwest Pacific

Based on models and proxy data it has been proposed that salinity-driven stratification weakened in the subarctic North Pacific during the last deglaciation, which potentially contributed to the deglacial rise in atmospheric carbon dioxide. We present high-resolution subsurface temperature (TMg/Ca) and subsurface salinity-approximating (d18Oivc-sw) records across the last 20,000 years from the subarctic North Pacific and its marginal seas, derived from combined stable oxygen isotopes and Mg/Ca ratios of the planktonic foraminiferal species Neogloboquadrina pachyderma (sin.). Our results indicate regionally differing changes of subsurface conditions. During the Heinrich Stadial 1 and the Younger Dryas cold phases our sites were subject to reduced thermal stratification, brine rejection due to sea-ice formation, and increased advection of low-salinity water from the Alaskan Stream. In contrast, the Bølling-Allerød warm phase was characterized by strengthened thermal stratification, stronger sea-ice melting, and influence of surface waters that were less diluted by the Alaskan Stream. From direct comparison with alkenone-based sea surface temperature estimates (SSTUk'37), we suggest deglacial thermocline changes that were closely related to changes in seasonal contrasts and stratification of the mixed layer. The modern upper-ocean conditions seem to have developed only since the early Holocene.

Sedimentology at the continental margin of the southern Weddell Sea

During four expeditions with RV "Polarstern" at the continental margin of the southern Weddell Sea, profiling and geological sampling were carried out. A detailed bathymetric map was constructed from echo-sounding data. Sub-bottom profiles, classified into nine echotypes, have been mapped and interpreted. Sedimentological analyses were carried out on 32 undisturbed box grab surface samples, as well as on sediment cores from 9 sites. Apart from the description of the sediments and the investigation of sedimentary structures on X-radiographs the following characteristics were determined: grain-size distributions; carbonate and Corg content; component distibutions in different grain-size fractions; stable oxygen and carbon isotopes in planktic and, partly, in benthic foraminifers; and physical properties. The stratigraphy is based On 14C-dating, oxygen isotope Stages and, at one site, On paleomagnetic measurements and 230Th-analyses The sediments represent the period of deposition from the last glacial maximum until recent time. They are composed predominantly of terrigenous components. The formation of the sediments was controlled by glaciological, hydrographical and gravitational processes. Variations in the sea-ice coverage influenced biogenic production. The ice sheet and icebergs were important media for sediment transport; their grounding caused compaction and erosion of glacial marine sediments on the outer continental shelf. The circulation and the physical and chemical properties of the water masses controlled the transport of fine-grained material, biogenic production and its preservation. Gravitational transport processes were the inain mode of sediment movements on the continental slope. The continental ice sheet advanced to the shelf edge and grounded On the sea-floor, presumably later than 31,000 y.B.P. This ice movement was linked with erosion of shelf sediments and a very high sediment supply to the upper continental slope from the adiacent southern shelf. The erosional surface On the shelf is documented in the sub-bottom profiles as a regular, acoustically hard reflector. Dense sea-ice coverage above the lower and middle continental slope resulted in the almost total breakdown of biogenic production. Immediately in front of the ice sheet, above the upper continental slope, a <50 km broad coastal polynya existed at least periodically. Biogenic production was much higher in this polynya than elsewhere. Intense sea-ice formation in the polynya probably led to the development of a high salinity and, consequently, dense water mass, which flowed as a stream near bottom across the continental slope into the deep sea, possibly contributing to bottom water formation. The current velocities of this water mass presumably had seasonal variations. The near-bottom flow of the dense water mass, in combination with the gravity transport processes that arose from the high rates of sediment accumulation, probably led to erosion that progressed laterally from east to West along a SW to NE-trending, 200 to 400 m high morphological step at the continental slope. During the period 14,000 to 13,000 y.B.P., during the postglacial temperature and sea-level rise, intense changes in the environmental conditions occured. Primarily, the ice masses on the outer continental shelf started to float. Intense calving processes resulted in a rapid retreat of the ice edge to the south. A consequence of this retreat was, that the source area of the ice-rafted debris changed from the adjacent southern shelf to the eastern Weddell Sea. As the ice retreated, the gravitational transport processes On the continental slope ceased. Soon after the beginning of the ice retreat, the sea-ice coverage in the whole research area decreased. Simultaneously, the formation of the high salinity dense bottom water ceased, and the sediment composition at the continental slope then became influenced by the water masses of the Weddell Gyre. The formation of very cold Ice Shelf Water (ISW) started beneath the southward retreating Filchner-Ronne Ice Shelf somewhat later than 12,000 y.B.P. The ISW streamed primarily with lower velocities than those of today across the continental slope, and was conducted along the erosional step on the slope into the deep sea. At 7,500 y.B.P., the grounding line of the ice masses had retreated > 400 km to the south. A progressive retreat by additional 200 to 300 km probably led to the development of an Open water column beneath the ice south of Berkner Island at about 4,000 y.B.P. This in turn may have led to an additional ISW, which had formed beneath the Ronne Ice Shelf, to flow towards the Filcher Ice Shelf. As a result, increased flow of ISW took place over the continental margin, possibly enabling the ISW to spill over the erosional step On the upper continental slope towards the West. Since that time, there is no longer any documentation of the ISW in the sedimentary Parameters on the lower continental slope. There, recent sediments reflect the lower water masses of the Weddell Gyre. The sea-ice coverage in early Holocene time was again so dense that biogenic production was significantly restricted.

Sedimentological investigations and age model on sediment core PS2561-2

Benthic d13C values (F. wuellerstorfi), kaolinite/chlorite ratios and sortable silt median grain sizes in sediments of a core from the abyssal Agulhas Basin record the varying impact of North Atlantic Deep Water (NADW) and Antarctic Bottom Water (AABW) during the last 200 ka. The data indicate that NADW influence decreased during glacials and increased during interglacials, in concert with the global climatic changes of the late Quaternary. In contrast, AABW displays a much more complex behaviour. Two independent modes of deep-water formation contributed to the AABW production in the Weddell Sea: 1) brine rejection during sea ice formation in polynyas and in the sea ice zone (Polynya Mode) and 2) super-cooling of Ice Shelf Water (ISW) beneath the Antarctic ice shelves (Ice Shelf Mode). Varying contributions of the two modes lead to a high millennial-scale variability of AABW production and export to the Agulhas Basin. Highest rates of AABW production occur during early glacials when increased sea ice formation and an active ISW production formed substantial amounts of deep water. Once full glacial conditions were reached and the Antarctic ice sheet grounded on the shelf, ISW production shut down and only brine rejection generated moderate amounts of deep water. AABW production rates dropped to an absolute minimum during Terminations I and II and the Marine Isotope Transition (MIS) 4/3 transition. Reduced sea ice formation concurrent with an enhanced fresh water influx from melting ice lowered the density of the surface water in the Weddell Sea, thus further reducing deep water formation via brine rejection, while the ISW formation was not yet operating again. During interglacials and the moderate interglacial MIS 3 both brine formation and ISW production were operating, contributing various amounts to AABW formation in the Weddell Sea.