North Atlantic Deep Water Production Research Articles

The Southern Ocean surface between Marine Isotope Stages 6 and 5d: Shape and timing of climate changes

Summer sea surface temperature (SSST) and sea-ice extent, obtained at millennial to centennial resolution, are presented for seven cores recovered in the Atlantic and western Indian Antarctic Zone, from Termination II and the last interglacial s.s., i.e. Marine Isotope Substage (MIS) 5e. The diatom-based SSST estimates show that during the late MIS 6 intensified export of cold waters from the Weddell Sea resulted in SSSTs close to 0°C in the central and eastern Atlantic sector. Due to northward expansion of the winter sea-ice edge by 3–5° latitude the sea-ice seasonality was reduced. Such conditions have great potential for efficient year-round reduction in air–sea gas exchange and glacial CO 2 draw down. The warming at Termination II started at 132–131 ka, marked by rapid retreat of the sea-ice edge. The SSST increased within a few thousands of years to reach maximum interglacial summer temperatures ranging between 2.5 and 3.5°C, in the earliest MIS 5e (128–125 ka). Besides a distinct cold reversal around 129.5 ka, the warming was punctuated by short-lived temperature rebounds with amplitudes of 1–1.5°C, occurring at 200–300-yr frequency, thus in the range of the periodicity of solar input. Increased Chaetoceros spp. abundance suggests meltwater events as the primary cause of the coolings. The SSST optimum, characterised by strongly enhanced biogenic sedimentation rates, ended at ca. 125 ka, with the full re-establishment of the global conveyor belt and beginning of the Northern Hemisphere climatic optimum. At this time SSSTs decreased stepwise and stabilised on values slightly warmer than today, showing in the following period of MIS 5e only low amplitude oscillations. During MIS 5d a northward expansion of the sea-ice field took place, but SSSTs only slightly colder than modern ones and a weakened meridional temperature gradient indicate that the Southern Ocean did not fall in a true glacial mode. The tight correlation between Southern Ocean SSST and winter sea-ice records, and Vostok atmospheric temperature and CO 2 records support the idea that the Antarctic winter sea-ice and Southern Ocean temperature exert a primary control on the atmospheric–ocean CO 2 exchanges and hence on global climate. Reduction of the Antarctic sea-ice field and increase of SSST at the onset of Termination II could have been triggered by precessional changes influencing southern high latitude summer insolation. Positive feedback mechanisms, reduction of surface albedo and release of CO 2 into the atmosphere, and changes in global ocean circulation, affected by a collapse or strong reduction in North Atlantic Deep Water (NADW) production, could have induced further reinforcement of the Southern Ocean warming.

Gulf Stream and Western Boundary Undercurrent variations during MIS 10–12 at Site 1056, Blake-Bahama Outer Ridge

The stable isotopic composition of two planktonic foraminifer species ( Globigerinoides sacculifer and Neogloboquadrina dutertrei) and two benthic foraminifer species ( Cibicidoides wuellerstorfi and Uvigerina peregrina) was measured at sub-orbital resolution through the marine isotope stages (MISs) 10, 11, and 12 (345–460 ka) at Site 1056 on the Blake-Bahama Outer Ridge. Planktonic foraminifers were counted for the interval 405–450 ka at 2–4-kyr resolution. Site 1056 (32°29′N, 76°20′W) is located on the continental slope (water depth: 2167 m) beneath the Gulf Stream. The average rate of sediment accumulation through the interval is 11.4 cm/kyr, but sediment accumulation is much more rapid during glacial intervals (15–17 cm/kyr). The decline in percent carbonate during glacial intervals, and its rise during interglacials, indicates that the increased sediment supply is of terrigenous origin. Low carbonate values and low benthic δ 13C, which are both associated with a weak Western Boundary Undercurrent and low North Atlantic Deep Water production, persist for 6 kyr after the benthic δ 18O record indicates that ice volume has begun to decrease. Recovery of carbonate and benthic δ 13C values is abrupt and rapid. Millennial-scale variation (∼3–4 kyr) is apparent in the glacial intervals of the planktonic δ 18O records and is more pronounced in a Δδ 18O record, which represents the temperature range in the photic zone. Semi-precessional (10–12-kyr) cycles are apparent in the planktonic Δδ 13C record. The millennial-scale cycles are largely caused by an increase in G. sacculifer δ 18O and represent surface warming. They are interpreted as representing periodic increases in westward intensification of the gyre. The semi-precessional cycles are driven by changes in the N. dutertrei δ 13C and represent fluctuations in the Gulf Stream itself and therefore likely have a tropical origin. Planktonic foraminifer census counts did not show an expected response to one of the largest glacial/interglacial transitions of the late Pleistocene. The most obvious change was an increase in faunal diversity during MIS 12.2, the interval of maximum δ 18O values. This suggests that cool slope water and warm subtropical gyre water penetrated a more sluggish Gulf Stream with greater frequency at this time. The millennial-scale maxima in the Δδ 18O record are accompanied by decreases in diversity, which is consistent with the interpretation of surface warming during these events.

It's All in the Crust

CLIMATE SCIENCE The intensity of North Atlantic Deep Water (NADW) production is an important determinant of global ocean thermohaline circulation and climate. It is affected over long periods by large-scale processes such as glaciation and continental drift, but different proxies (a substitute parameter that is easily measured) can provide conflicting evidence about its strength. To help resolve these inconsistencies and to construct a more detailed record of NADW since the mid-Miocene, Frank et al. analyzed the Nd-and Pb-isotopic compositions of ferromanganese nodule crusts from the Southern ocean. Furthermore, by combining them with analyses of other nodules from the North Atlantic basin, they reconstructed the rate of NADW production over the last 14 million years. This method has two major advantages: First, Nd and Pb are not subject to biological fractionation, as are other commonly used water-mass proxies like carbon isotopes or Cd/Ca; and second, ferromanganese nodules can be dated accurately with Be-isotopic determinations. They find that there was a continuous and strong export of NADW (or a precursor of it) into the Southern Ocean between 14 and 3 million years ago and that over the past 3 million years, since the start of Northern Hemisphere glaciation, a progressive reduction of about 25% has occurred. — HJS Paleoceanography 17 , 10.1029/2000PA000606 (2002).

North Atlantic Deep Water export to the Southern Ocean over the past 14 Myr: Evidence from Nd and Pb isotopes in ferromanganese crusts

The intensity of North Atlantic Deep Water (NADW) production has been one of the most important parameters controlling the global thermohaline ocean circulation system and climate. Here we present a new approach to reconstruct the overall strength of NADW export from the North Atlantic to the Southern Ocean over the past 14 Myr applying the deep water Nd and Pb isotope composition as recorded by ferromanganese crusts and nodules. We present the first long‐term Nd and Pb isotope time series for deep Southern Ocean water masses, which are compared with previously published time series for NADW from the NW Atlantic Ocean. These data suggest a continuous and strong export of NADW, or a precursor of it, into the Southern Ocean between 14 and 3 Ma. An increasing difference in Nd and Pb isotope compositions between the NW Atlantic and the Southern Ocean over the past 3 Myr gives evidence for a progressive overall reduction of NADW export since the onset of Northern Hemisphere glaciation (NHG). The Nd isotope data allow us to assess at least semiquantitatively that the amount of this reduction has been in the range between 14 and 37% depending on location.

Carbonate preservation records of the past 3 Myr in the Norwegian–Greenland Sea and the northern North Atlantic: implications for the history of NADW production

Carbonate preservation records from a number of drill sites in the North Atlantic and adjacent Norwegian–Greenland Sea (NGS) are used to reconstruct variations in North Atlantic Deep Water (NADW) production over the past 3 Myr. Before the initiation of major Northern Hemisphere glaciation, good carbonate preservation was recorded in the North Atlantic and the NGS supporting the superconveyor model of accelerated NADW formation in the late Pliocene. After the inception of main Northern Hemisphere glaciation, NADW formation in the NGS was blocked-off during the period 2.8–1.9 Ma. Carbonate was only badly preserved due to (1) low production of carbonate shells in surface waters, (2) sluggish renewal of deep waters induced by a rather stable sea-ice cover, and/or (3) production of carbonate-corrosive dense brines during sea-ice formation (e.g. sea ice dissolution mode). In contrast, contemporaneous good preservation in the adjacent North Atlantic indicates efficient NADW production. During the middle Matuyama (1.9–1.4 Ma), the first intrusions of the Proto-Norwegian Current into a narrow corridor in the southeastern NGS are evidenced by much better carbonate preservation. A decrease both in carbonate content and preservation towards the north and west of this corridor indicates the proximity of the polar front and gives evidence that NADW production was efficiently triggered by the Atlantic water entrainment mode. During the past 1.2 Myr carbonate preservation patterns in the NGS clearly reflect major global events like the Mid-Pleistocene Transition and the mid-Brunhes dissolution event. The onset of the Mid-Pleistocene Transition at 1.2 Ma is characterized by a complete shift to higher carbonate contents in the southeastern NGS and Labrador Sea. Overall, good preservation during both interglacials and glacials is only interrupted by high-frequency, short-term dissolution spikes, which were induced by ice sheet collapse and development of extensive meltwater lids. As a consequence, NADW was produced nearly continuously during glacials in the Nordic Seas. However, due to its lower density it was entrained into intermediate water levels in the North Atlantic and, thus, enforced the intermediate water circulation loop, whereas a decrease in lower-NADW production is observed contemporaneously.

The early Matuyama Diatom Maximum off SW Africa: a conceptual model

An important discovery during Ocean Drilling Program Leg 175, when investigating the record of upwelling off Namibia, was the finding of a distinct Late Pliocene diatom maximum spanning the lower half of the Matuyama reversed polarity chron (MDM, Matuyama Diatom Maximum) and centered around 2.6–2.0 Ma. This maximum was observed at all sites off southwestern Africa between 20°S and 30°S, and is most strongly represented in sediments of Site 1084, off Lüderitz, Namibia. The MDM is characterized by high biogenic opal content, high numbers of diatom valves, and a diatom flora rich in Southern Ocean representatives (with Thalassiothrix antarctica forming diatom mats) as well as coastal upwelling components. Before MDM time, diatoms are rare until ca. 3.6 Ma. After the MDM, in the Pleistocene, the composition of the diatom flora points to increased importance of coastal upwelling toward the present, but is accompanied by a general decrease in opal and diatom deposition. Here we present a simple conceptual model as a first step in formalizing a possible forcing mechanism responsible for the record of opal deposition in the upwelling system off Namibia. The model takes into account Southern Ocean oceanography, and a link with deepwater circulation and deepwater nutrient chemistry which, in turn, are coupled to the evolution of North Atlantic Deep Water (NADW). The model proposes that between the MDM and the Mid-Pleistocene climate revolution, opal deposition off Namibia is not directly tied to glacial–interglacial fluctuations (as seen in the global δ 18O record), but that, instead, a strong deepwater link exists with increased NADW production (as seen in the deepwater δ 13C record) accounting for higher supply of silicate to the thermocline waters that feed the upwelling process. The opal record of Site 1084 shows affinity to eccentricity on the 400-kyr scale but not for the 100-kyr scale. This points toward long-term geologic processes for delivery of silica to the ocean.

Effects of glacial meltwater in the GISS coupled atmosphereocean model: 1. North Atlantic Deep Water response

Varying magnitudes of freshwater input through the St. Lawrence are added to different versions of the GISS coupled atmosphere‐ocean model. The results show a generally linear response in percentage North Atlantic Deep Water (NADW) reduction to the volume of water added without any obvious threshold effects. When the estimated best guess freshwater input for the Allerod‐Younger Dryas interval is added, only small reductions in NADW production occur in less than a century, with complete cessation requiring more than 300 years. If the ocean circulation was weaker in the Allerod preceding the Younger Dryas, somewhat larger percentage reductions occur, and it would take 150–200 years for a complete shutdown. The freshening is aided by a positive feedback associated with the reduction in evaporation relative to precipitation in the North Atlantic, amounting to some 20–30% of the added freshwater runoff. When the freshwater input is stopped after complete NADW shutdown, NADW production does not resume. With complete NADW cessation, cooling eventually occurs throughout the Northern Hemisphere, including regions where it is not thought to have arisen during the Younger Dryas, and the globe cools by 1°C due primarily to the sea ice albedo feedback; while in many regions the observed cooling is reproduced, the cooling in Greenland is much less than is estimated. With only 50% NADW cessation, cooling is primarily around the North Atlantic. No rapid warming is seen in response to freshwater cessation. The sensitivity displayed here for NADW reduction due to freshwater input from the St. Lawrence is roughly similar to that associated with increased precipitation accompanying global warming in the next century, so a similar sensitivity may arise with Younger Dryas freshwater inputs from other suspected locations (Greenland, Europe).

Deep flow variability under apparently stable North Atlantic deep water production during the Last Interglacial of the subtropical NW Atlantic

The environmental history of the last interglacial period is equivocal, since its climate has been reported to be anything from unstable, with at least one significant cold event, to generally stable and comparable to the Holocene. Here the activity of the southerly Deep Western Boundary Current (DWBC) in the ∼130–110 ka interval is studied at two sites (3481 and 4760 m) in the subtropical NW Atlantic presently bathed by Lower North Atlantic Deep Water (LNADW). Proxies for both bottom water chemistry (δ13C) and near‐bottom flow vigor ( , the mean size of the “sortable silt” fraction) are employed. The δ13C data indicate the interval of peak interglaciation when LNADW bathed both sites studied is followed by the development of a vertical water mass gradient between them, with nutrient‐rich waters of southern origin flooding the deeper parts of the area. After ∼113 ka this gradient disappears as the deeper, low δ13C waters homogeneously invade the whole of the depth range investigated. The data imply these events are likely to be associated with the vertical migration of the flow axis of the DWBC. Furthermore, they also reveal unexpected fluctuations in near‐bottom flow activity during the period of minimum ice volume which are tentatively ascribed to small changes in the convective activity of the Nordic Seas. Alternatively, DWBC flow fluctuations may be controlled by processes occurring farther downstream from the areas of deep water formation, therefore not requiring changes in the production rates of LNADW. In either case, this series of small fluctuations during the peak of the last interglaciation may ultimately be linked to the insolation‐related evolution of latitudinal sea surface temperature and salinity gradients in the North Atlantic.

Palaeoceanographical changes recorded by Cenozoic deep‐sea ostracod assemblages from the South Atlantic and the Southern Ocean (ODP Sites 1087 and 1088)

Cenozoic palaeoceanography of the SE Atlantic and Southern Oceans has been investigated using Late Eocene/Early Oligocene to Quaternary ostracod assemblages from 49 samples of ODP Sites 1087 and 1088. Although the overall abundance of ostracods is relatively low (means of 17 and 49 specimens per sample at Sites 1087 and 1088, respectively) and there is an apparently high level of endemism (ranging from 50% to 80% at Sites 1087 and 1088), three major changes in the faunal assemblages are identified at Site 1087 (denoted A, B and C) and two at Site 1088 (denoted B′ and C′). The assemblage boundaries, detected on the basis of stepwise changes in the abundance, diversity, dominance, endemism, faunal turnover and relative abundance of common taxa, coincide broadly with previously identified, ostracod-based palaeoceanographical ‘events’ discussed by Benson and co-workers over the last two decades. The data do not extend sufficiently far back to record the initiation of Assemblage A, but the faunal change between Assemblages A and B, marked by a decline in abundance, species diversity and faunal turnover, occurs within the Middle Miocene (NN5-6). It coincides with a previously documented palaeoceanographical ‘event’ at 16–14 Ma which, we suggest, may be related to the initiation of North Atlantic Deep Water (NADW) production and/or an expansion of the East Antarctic ice sheet. Assemblage B′ is subdivided into the two Sub-assemblages B′1 and B′2 mainly on the basis of an increase in diversity, a peak in faunal turnover and a drop in the relative abundance of the genus Krithe in early Late Miocene time (NN9, c. 10.5 Ma). The B′1/B′2 Sub-assemblage boundary cannot be related to any previously documented faunal change in deep-sea ostracods. Changes associated with the boundaries between Assemblages B and C, and B′ and C′, which we believe to be synchronous, both include a decrease in diversity and abundance. In addition, two strong turnover peaks occur near the B′/C′ boundary at Site 1088. The B/C and B′/C′ boundaries coincide with a previously documented midPliocene ‘event’ (3.5 Ma) (NN15-16) which may be linked to putative closure of the Straits of Panama and increased production of NADW, the latter in turn leading to increased production of Antarctic Intermediate Water (AAIW) and Antarctic Bottom Water (AABW). Alternatively, fiuctuations in size of the Antarctic ice sheet during possible Pliocene warm periods could indirectly be responsible for the observed midPliocene faunal changes.

Paleoclimate and Amerindians: evidence from stable isotopes and atmospheric circulation.

Two Amerindian demographic shifts are attributed to climate change in the northwest plains of North America: at approximately 11,000 calendar years before present (yr BP), Amerindian culture apparently split into foothills-mountains vs. plains biomes; and from 8,000-5,000 yr BP, scarce archaeological sites on the open plains suggest emigration during xeric "Altithermal" conditions. We reconstructed paleoclimates from stable isotopes in prehistoric bison bone and relations between weather and fractions of C(4) plants in forage. Further, we developed a climate-change model that synthesized stable isotope, existing qualitative evidence (e.g., palynological, erosional), and global climate mechanisms affecting this midlatitude region. Our isotope data indicate significant warming from approximately 12,400 to 11,900 yr BP, supporting climate-driven cultural separation. However, isotope evidence of apparently wet, warm conditions at 7,300 yr BP refutes emigration to avoid xeric conditions. Scarcity of archaeological sites is best explained by rapid climate fluctuations after catastrophic draining of the Laurentide Lakes, which disrupted North Atlantic Deep Water production and subsequently altered monsoonal inputs to the open plains.

Tropical Atlantic coral oxygen isotopes: glacial–interglacial sea surface temperatures and climate change

Abstract We have generated a detailed oxygen isotope time-series from the fossil coral reefs from offshore Barbados. The Barbados coral-based record is a unique paleoceanographic data set with an equivalent sedimentation rate in excess of 600 cm/kyr where not only is the annual signal uniquely preserved but seasonality as well. Oxygen isotope values during the late glacial and LGM (16–20 14C kyrs; 19–24 calendar kyrs) are ∼2.3‰ heavier than corresponding living specimens, and indicate a regional cooling on the order of 4.5°C. There is also an isotopic expression of the Bolling-Pre-Boreal climate oscillation, with values reflecting a cooling during the Younger Dryas. Seasonality, a key diagnostic of the state of the climate system, remained the same or slightly less than present and indicates that the observed coolings were a change in the mean state of the western tropical Atlantic. Pan-tropic cooling during the last glaciation is best explained by a change in the radiative balance of the tropics. Variable tropical sea surface temperatures during climate oscillations such as the Younger Dryas challenge the paradigm that climate change only cascades from variations in North Atlantic deep water production.

Changes in North Atlantic deep‐sea temperature during climatic fluctuations of the last 25,000 years based on ostracode Mg/Ca ratios

We reconstructed three time series of last glacial‐to‐present deep‐sea temperature from deep and intermediate water sediment cores from the western North Atlantic using Mg/Ca ratios of benthic ostracode shells. Although the Mg/Ca data show considerable variability (“scatter”) that is common to single‐shell chemical analyses, comparisons between cores, between core top shells and modern bottom water temperatures (BWT), and comparison to other paleo‐BWT proxies, among other factors, suggest that multiple‐shell average Mg/Ca ratios provide reliable estimates of BWT history at these sites. The BWT records show not only glacial‐to‐interglacial variations but also indicate BWT changes during the deglacial and within the Holocene interglacial stage. At the deeper sites (4500‐ and 3400‐m water depth), BWT decreased during the last glacial maximum (LGM), the late Holocene, and possibly during the Younger Dryas. Maximum deep‐sea warming occurred during the latest deglacial and early Holocene, when BWT exceeded modern values by as much as 2.5°C. This warming was apparently most intense around 3000 m, the depth of the modern‐day core of North Atlantic deep water (NADW). The BWT variations at the deeper water sites are consistent with changes in thermohaline circulation: warmer BWT signifies enhanced NADW influence relative to Antarctic bottom water (AABW). Thus maximum NADW production and associated heat flux likely occurred during the early Holocene and decreased abruptly around 6500 years B.P., a finding that is largely consistent with paleonutrient studies in the deep North Atlantic. BWT changes in intermediate waters (1000‐m water depth) of the subtropical gyre roughly parallel the deep BWT variations including dramatic mid‐Holocene cooling of around 4°C. Joint consideration of the Mg/Ca‐based BWT estimates and benthic oxygen isotopes suggests that the cooling was accompanied by a decrease in salinity at this site. Subsequently, intermediate waters warmed to modern values that match those of the early Holocene maximum of ∼7°C. Intermediate water BWT changes must also be driven by changes in ocean circulation. These results thus provide independent evidence that supports the hypothesis that deep‐ocean circulation is closely linked to climate change over a range of timescales regardless of the mean climate state. More generally, the results further demonstrate the potential of benthic Mg/Ca ratios as a tool for reconstructing past ocean and climate conditions.

Changes in the Bathymetry and Volume of Glacial Lake Agassiz Between 11,000 and 9300 14C yr B.P.

The volume and surface area of glacial Lake Agassiz varied considerably during its 4000-year history. Computer models for seven stages of Lake Agassiz were used to quantify these variations over the lake's early history, between about 11,000 and 9300 14C yr B.P. (ca. 13,000 to 10,300 cal yr B.P.). Just after formation of the Herman strandlines (ca. 11,000 14C yr B.P.), the volume of Lake Agassiz appears to have decreased by >85% as a consequence of the abrupt rerouting of overflow to its eastern outlet from its southward routing into the Mississippi River basin. This drainage released about 9500 km3 of water into the North Atlantic Ocean via the Great Lakes and Gulf of St. Lawrence. Following closure of this eastern routing of overflow, the lake reached its maximum size at about 9400 14C yr B.P. with an area of >260,000 km2 and a volume of >22,700 km3. A second major reduction in volume occurred shortly after that, when its volume decreased >10% following the opening of the Kaiashk outlet to the east into the Great Lakes, and 2500–7000 km3 of water was released into the North Atlantic Ocean. These discharges may have affected ocean circulation and North Atlantic Deep Water production.

Increased Pliocene North Atlantic Deep Water: Cause or consequence of Pliocene warming?

Raymo et al. [1996] suggested that the mid‐Pliocene (∼3 Ma) warm period was associated with increased North Atlantic Deep Water (NADW) production. Is this circulation change a cause or consequence of Pliocene warming? We test the hypothesis that increased strength of NADW was a consequence of the warming around Antarctica affecting deep Antarctic outflow. A sensitivity experiment with an ocean general circulation model with Pliocene surface conditions changed only over the Southern Ocean (SO) indicates that warmer temperatures around Antarctica result in lower rates of sea ice formation and SO deep water outflow. The decreased abyssal density gradient in the SO directly leads to about a 20% increase in NADW outflow at 30°S, a 10% increase in NADW overturning in the subpolar North Atlantic, and a 20% increase in poleward heat transport in the North Atlantic. We postulate that the largest initial Pliocene climate change was in the SO because the greater sea ice area in this region is more sensitive to inferred slightly higher CO2 levels in the mid‐Pliocene.

Millennial-scale instability of the antarctic ice sheet during the last glaciation

Records of ice-rafted detritus (IRD) concentration in deep-sea cores from the southeast Atlantic Ocean reveal millennial-scale pulses of IRD delivery between 20,000 and 74,000 years ago. Prominent IRD layers correlate across the Polar Frontal Zone, suggesting episodes of Antarctic Ice Sheet instability. Carbon isotopes (delta(13)C) of benthic foraminifers, a proxy of deepwater circulation, reveal that South Atlantic IRD events coincided with strong increases in North Atlantic Deep Water (NADW) production and inferred warming (interstadials) in the high-latitude North Atlantic. Sea level rise or increased NADW production associated with strong interstadials may have resulted in destabilization of grounded ice shelves and possible surging in the Weddell Sea region of West Antarctica.

Reduced North Atlantic Deep Water flux to the glacial Southern Ocean inferred from neodymium isotope ratios

The global circulation of the oceans and the atmosphere transports heat around the Earth. Broecker and Denton suggested that changes in the global ocean circulation might have triggered or enhanced the glacial-interglacial cycles. But proxy data for past circulation taken from sediment cores in the South Atlantic Ocean have yielded conflicting interpretations of ocean circulation in glacial times--delta13C variations in benthic foraminifera support the idea of a glacial weakening or shutdown of North Atlantic Deep Water production, whereas other proxies, such as Cd/Ca, Ba/Ca and 231Pa/230Th ratios, show little change from the Last Glacial Maximum to the Holocene epoch. Here we report neodymium isotope ratios from the dispersed Fe-Mn oxide component of two southeast Atlantic sediment cores. Both cores show variations that tend towards North Atlantic signatures during the warm marine isotope stages 1 and 3, whereas for the full glacial stages 2 and 4 they are closer to Pacific Ocean signatures. We conclude that the export of North Atlantic Deep Water to the Southern Ocean has resembled present-day conditions during the warm climate intervals, but was reduced during the cold stages. An increase in biological productivity may explain the various proxy data during the times of reduced North Atlantic Deep Water export.

Evidence of 550-year and 1000-year cyclicities in North Atlantic circulation patterns during the Holocene

Sedimentological colour data obtained from North Atlantic deep-sea sediment core NEAP15K pro vide a continuous climate record for the last 11500 yr with a mean sampling interval of <15 yr. Variations in sediment lightness, a proxy for changes in North Atlantic Deep Water (NADW) circulation, are characterized by high-frequency fluctuations with periodicities equivalent to 550 and 1000 yr, as well as a <1600-yr cyclicity comparable to that reported in other studies of Holocene climate. Cross-spectral analysis suggests that the 550-yr and 1000-yr cyclical variations in NADW circulation were coherent with fluctuations in atmospheric con ditions over Greenland, where both these periodicities have been identified previously in the Holocene GISP2 d18O record (Stuiver et al., 1995). A similar statistical comparison provides evidence that high-frequency (550-yr) variability in NADW circulation patterns may have been linked to short-term variations in atmospheric 14C values during the Holocene. In general, NADW production was lower and 14C levels were higher when colder climatic conditions prevailed. These results imply that the pattern of NADW production has been a significant factor affecting centennial-to millennial-scale Holocene climate variability in the North Atlantic region.

Millennial-scale changes in North Atlantic Deep Water circulation during marine isotope stages 11 and 12: Linkage to Antarctic climate

The time interval represented by marine isotope stages (MIS) 11 and 12 (ca. 360–470 ka) may contain the most extreme glacial and interglacial climate conditions of the late Pleistocene. Sediments from the Bermuda Rise (western North Atlantic) provide clues to the nature of climate variability during this period. Our geochemical records indicate that millennial-scale climate instability and associated changes in North Atlantic Deep Water (NADW) production occurred during both interglacial MIS 11 and glacial MIS 12. Stage 12 is punctuated by a series of interstadial events that occurred at a 5–6 k.y. periodicity, and the occurrence of ice-rafted debris at various times during stage 12 indicates that icebergs were present at least as far south as 34°N during this glacial period. Within the limits of our correlation, the atmospheric temperature changes recently reported for the Vostok ice core for the stage 11 time period appear to be represented by coeval changes in NADW flow. Specifically, warming in Antarctica is associated with increased production of NADW.

The structure of Termination II (penultimate deglaciation and Eemian) in the North Atlantic

A study of the 140–100 ka interval in core T90-9P from the North Atlantic (45° N, 25° W), based on analysis of oxygen and carbon isotope records from planktonic and benthonic foraminifera, and from the bulk sediment fine fraction facilitates a detailed paleoceanographic reconstruction of the penultimate deglaciation (Termination II), and of the Eemian interglacial (δ 18O stage 5e). The first step of Termination II was characterised by low productivity and a mixed water column, which was a remnant of glacial conditions. A 3 ka period of relatively stable conditions, with a stratified water column (‘Termination II pause’), occurred half-way through Termination II, and preceeded a second and more rapid climatic shift. The end of the deglaciation (Eemian maximum, i.e. isotopic event 5.53) initiated the establishment of strong, seasonal, water column stratification. North Atlantic Deep Water (NADW) production remained low during the complete glacial–interglacial transition. After the Eemian maximum, NADW prodution was restored, and bottom waters remained quite stable during the course of the Eemian, while surface waters gradually cooled in the second half of the stage. A short surface water cooling event accompanied by a reduced seasonal water column stratification and nutrient instability occurred at approximately 117 ka BP.

North Atlantic Intermediate Waters in the Late Pliocene to Early Pleistocene

We generated benthic isotope records from Ocean Drilling Program (ODP) site 981 on the Feni drift (2173 m water depth) and from ODP site 983 on the Gardar drift (1983 m water depth) to examine the interaction between North Atlantic Deep Water (NADW) and Glacial North Atlantic Intermediate Water (GNAIW) formation from 2.0 to 1.4 Ma. We find NADW at both sites during interglacial periods, and a mix of NADW and Southern Ocean water at the Feini drift during most glacial periods. Prior to 1.7 Ma we find no evidence ofr GNAIW at the Gardar drift site. Instead, glacial Gardar drift δ13C values are as low or lower than values for all other sites in the North Atlantic and reflect continued glacial overflow from the Nordic seas. After 1.7 Ma Gardar drift δ13C values increase and suggest that there was GNAIW at the Gardar drift site during some glacial intervals. Overall, we find that NADW and GNAIW production changed around 1.7 Ma in concert with changes in sea surface temperature and salinity and in the Earth's obliquity cycle.