Stable isotope and sedimentological (clay mineral, X-ray diffraction) profiles along core M15612 from the Mid-Atlantic Ridge at 44°N, 26°W, 3050 m water depth are used to monitor surface and deep water properties in association with North Atlantic meltwater events during the last glacial-interglacial cycle i.e., the past 140,000 years. The core position is close to the glacial polar front and its depth position at the deep hydrographic front between glacial deep and bottom waters. As such, our profiles allow to monitor even fine-scale changes in surface and deep hydrography both on glacial-interglacial time scales and on millennial-scales. Distinct negative 18O anomalies punctuate the planktonic isotope records and correlate with Heinrich events. Anomalies are strongest in the N. pachyderma isotope record. Systematic changes in the 18O offset between G. bulloides and G. inflata signify changes in mid-latitude thermocline structure. IRD layers in core M15612 contain varying amounts of quartz, plagioclase, dolomite, and minor amounts of mica. Clay mineral assemblage consists mainly of illite, kaolinite and chlorite. Detrital carbonate contributes between 2% and 13% to the total number of detrital fragments in the IRD layers and supports the contention that the layers are the southern equivalents of the northern North Atlantic's Heinrich events. 18O anomalies during Heinrich events are also seen in benthic 18O. In conjunction with subtle negative benthic 13C anomalies, the data document a stronger contribution of a O-18 depleted, nutrient-rich water mass during the Heinrich events. This pattern is consistent with an enhanced advection of AABW during times of meltwater-induced convection slow-down in the northern North Atlantic. Superimposed on the fine-scale 13C patterns is a longer-term trend with positive 13C values around 1-1.2 permil PDB during stages 1, 3, and 5, and depleted values of 0-0.2 permil PDB during glacial stages 2, 4, 6. The 13C amplitude of more than 1 permil indicates changes between northern source (NADW) and southern source (AABW) water masses at this site.
Coarse lithic content, planktonic foraminifera oxygen isotope and magnetic susceptibility records have been obtained from a series of sediment cores in the Irminger Basin. The results have been closely compared to Norwegian sea and North Atlantic mid-latitude sediment cores. The records give evidence of numerous layers rich in ice rafted debris which correspond to periods of millennial scale increases in iceberg calving rates. The synchroneity of the different iceberg surges has been tested by radiocarbon dating and by correlating magnetic susceptibility records. The results indicate that whereas the Heinrich events correspond to synchronous surges of the different northern hemisphere ice caps, the millennial scale ice sheet instabilities have a more regional signature and do not occur throughout the Nordic regions at the same time. Furthermore, the impact of these events on the sea surface hydrology is different, the coarse lithic layers are not associated with such important meltwater signatures as those observed during the Heinrich events. Thus, our conclusions point to two different types of ice sheet instabilities. The first process, which operates on millennial scale, corresponds to variations of iceberg production in the coastal regions and ice shelves in the Nordic regions. The absence of large anomalies of the oxygen isotope ratio recorded by planktonic foraminifera Neogloboquadrina pachyderma left coiling could result from slower iceberg melting rates, as a result of lower sea surface temperatures, or different isotopic composition of icebergs. Indeed, ice which accumulates along the coastal regions and within the ice shelves has a global oxygen isotope ratio less fractionated than that deposited in the inner regions of the continental ice sheet. The second process corresponds to iceberg surges from the massif continental ice sheets, such as the Laurentide, and occurs at a slower pacing every 5 to 10 ka corresponding to the Heinrich events. The synchronous surges from all the northern hemisphere ice sheet at the time of the Heinrich events attest of a strong link between the millennial scale ice sheet instabilities which operates in the Nordic regions and the Laurentide ice sheet surges.
Four new paleo-ceanographic records with an ultrahigh resolution of 20-80 yr over MIS 2-3 define the various processes linked to Dansgaard-Oeschger (D-O) cycles on a precise time scale and thus help to better constrain the origin of Heinrich and D-O events. The records stem from a core transect extending from the Irminger across the Icelandic to the northeastern Norwegian Sea, from 59°-72°N. The age control is based on both tuning to the GISP2 ice record and altogether >300 AMS 14C ages.
The records display regional differences in short-term SST variability and iceberg-derived meltwater in-cursions that suggest differential features in North Atlantic paleoceanography, which may have driven the climatic variability with dominant periodicities of 2900, 1480, and 1100 years: 1) Such as the big Heinrich events the other cold episodes of the D-O cycles are marked by prominent IRD spikes in the Irminger and Icelandic Seas. Accordingly, we see no basic, but only gradual differences between the origin of Heinrich events and that of other cold phases. 2) We surmise that the iceberg flotillas that finally blocked the Irminger Current and induced the numerous cold episodes of the D-O cycles via throttling the Atlantic thermohaline circulation (THC) largely originated from East Greenland such as during Heinrich event (H) 3. This conclusion is based on both the systematic decrease in IRD concentrations from the northeastern Denmark Strait to the Irminger Sea by a factor of five and the persistent input of hematite stained quartz during the cold episodes. 3) Similar to the D-O cycles in the Greenland ice records the 1480-yr cycles in the marine records show a distinct asymmetry, with rapid jumps from peak cold to moderately warm climate, a subsequent gradual deterioration of climate ending with a jump back to peak cold. Accordingly, the IRD spikes lead the maximum surface-water stratification by about 300 yr and the SST maxima by 460 yr. 4) The various, not strictly periodic Heinrich events were induced by surges from different ice sheets: H 4-6 were linked to a major meltwater flux in the East Greenland/ East Iceland currents, opposed to the inferior meltwater signals found in the eastern Norwegian Sea. Thus glacial surges from the East Greenland margin may have played an important role in throttling the THC, in addition to the great surges coming from Labrador. H 3 is the only case with a significant meltwater signal in the Irminger Sea. H 1 shows the most extreme 18O meltwater signal southwest of the Barents shelf, suggesting a great ice surge from the Barents shelf, that has largely controlled the major disturbancies in THC - on top of the effect of meltwater supply from Labrador to the mid latitudes.
Recent studies have shown very rapid oscillations of the air temperature over Greenland (Dansgaard-Oeschger events) during the Last Glacial Period. Such rapid variations have been also recognised in the abundance and assemblages of planktic fauna and in their oxygen isotope compositions, both are correlated with the DOE. These data imply paleocirculation changes which must have been recorded in the detrital fraction of the sediment. To characterise the sedimentary response during rapid climatic oscillations (DOE or Heinrich events) and the associated changes in the North Atlantic oceanic circulation, a mineralogical and geochemical study has been performed on the IMAGES core MD 95-2009 located in the Faeroe-Shetland channel, a key area in the formation of the North Atlantic Deep Water.
The stratigraphy was established using 18O analyses on both planktic and benthic foraminifera and also by correlation of the Magnetic Susceptibility with well dated cores from the same area. High resolution chemical analyses of the silicated fraction of the sediment and selected geochemical parameters indicate well defined oscillation of the chemical composition of the sediments. The A index (A= Al-(Na+K+2Ca)) represents the excess of aluminium in respect to the plagioclase-feldspar content of the rock and Ca/Na ratio will be used as a tracer of the nature of the sedimentary material. High Ca/Na ratios indicate the influence of basalt-derived material while low ratio will characterise the influence of mature derived continental material.
During cold periods (Heinrich and/or Stadial), the A index increases and the Ca/Na ratio decreases suggesting that the sedimentation is mainly under the influence of material derived from the continental crust. An opposite feature prevails during the interstadials which implies an input of basalt-derived material. Clay mineralogy shows that smectites are associated with coarser detrital material and are predominant during interstadials. Illites are associated with finer grained material and characterise the stadials and Heinrich events.
High resolution studies (mineralogical, isotopic, geochemical and magnetic) on IMAGES core MD 95-2009 demonstrate that, during the last glacial period and particularly on isotopic stage 3, this area recorded high frequency oscillations between two different reservoirs. The sediments corresponding to the stadials or Heinrich events are derived from a mature distal continental crust probably the Fennoscandian crust. During the interstadials, this area is clearly under the influence of proximal basaltic derived material, the Iceland-Faeroe islands. The sedimentary system reacts very rapidly to the air temperature and to the sea surface conditions.
Abrupt, massive discharge of icebergs during Heinrich events transported and deposited large quantities of ice rafted-debris (IRD) into the North Atlantic. Theories explaining the origins of the debris and its distribution generally point to the Laurentide ice sheet as the culprit behind the events. Doubt was cast by Grousset et al. (1993) and Revel et al. (1996), who, by measuring the Sr-Nd isotopic compositions on the bulk lithic fraction of the Heinrich Layers (HLs), found that the sediments originated not only from the Laurentide ice sheet, as heretofore envisioned, but also from other, non-Laurentide ice-sheets, including those from Europe. These doubts were clearly confirmed by analyzing the carbonate-free, coarse (>63 µm) lithic fraction of HLs, which can only be transported by ice-rafting (Snoeckx et al., 1999). We have recently analyzed at high resolution the carbonate-free, coarse (>63 µm) lithic fraction of HLs in a few cores from the northeastern Atlantic margin. Examination of their isotopic fingerprints reveal that the transporting icebergs were calved first from the European ice sheets (Fennoscandia, and British Isles). These data show a more distinctly European origin for IRD deposited in the earliest sections of Heinrich events H1, H2, H4 and H5. Then the Laurentide isotopic signal occurs on top of this early European signal, and replaces the initially pure European signal. Finally, the top of the HL layer again consists of pure European particles. This sequence is visible along the Portugese margin and in the Bay of Biscay. The Laurentide-derived icebergs did not reach the Norwegian Sea, where IRD were derived only from the Fennoscandian ice sheet and the surrounding volcanic islands (Iceland and Faeroes). Thus, the theory that the Laurentide ice sheet set the HL events into motion (Alley and MacAyeal, 1993) is challenged by the European origin of the early-HL deposits, followed by the Laurentide signature of the later, mid-HLs, followed by the return to IRD with a pure European signature in the same cores. This observation suggests that, not only did the European ice sheets release their icebergs prior to those from the Laurentide, they continued after the Laurentide ceased. A caveat to this conclusion is that it assumes that the time required for icebergs of Laurentide origin to exit the Labrador Sea and traverse the Atlantic to the coring sites is significantly shorter than the time required to accumulate the IRD of European origin at the bottom of the HL sequence. Mechanisms by which the release of European icebergs could trigger the Laurentide surge are discussed.
Alley Rand MacAyeal D, Paleoceanography, 9, 503-511, (1993).
Grousset FE et al, Paleoceanography, 8, 175-192, (1993).
Revel Met al, Paleoceanography, 11, 95-113, (1996).
Snoeckx Het al, Marine Geology, (in press)
Sr and Nd isotopic compositions have been measured by thermal ion mass-spectrometry on 34 sediment samples of the lithic, carbonate free fraction (< 63 µm) from the marine arctic sediment core PS 1533 taken by the Polarstern on the ARK IV/3 expedition 1987. Core location is north of Spitzbergen on the Yermak Plateau at 82° 01,6' N and 15°11,4' E in 2030 m water depth. The core samples cover the last 140 ka (Eisenhauer et al., 1994) and were analysed to investigate whether temporal variations of the Sr and Nd isotope compositions correlate with climatic changes in the Arctic region. In addition 24 samples of modern surface sediments from Arctic shelf regions (Barents Sea, Novaya Zemlya, Franz Josef Land) and sediment laden sea ice (Barents Sea, Laptev Sea) have been analysed. These complementary isotopic data were measured to characterise possible sediment source regions and to trace particle pathways of sediments in the modern and ancient Arctic Ocean.
The core 87Sr/86Sr-ratios and Sr-concentrations vary between 0,717-0,740 and 98-143 ppm, respectively, <epsilon>Nd-values and Nd-concentrations between -9,3 to -15 and 24-33 ppm, respectively. The Sr- and Nd-isotopic results of the core sediments show strong climatic correlation with glacial and interglacial phases. During glacial oxygen isotope stages 2, 4 and 6 the 87Sr/86Sr-ratio reaches maxima at the beginning, then falling to deep minima in the glacial terminations I and II. After glacial/interglacial boundaries the 87Sr/86Sr-ratio increases remarkably (> 0,005) to the following interglacial period. The <epsilon>Nd-value shows inverse behaviour. During isotopic stages 1, 3 and 5 the 87Sr/86Sr- and <epsilon>Nd-values show only small variations around the core average values of 0,72796 and -12,5 respectively.
The strong temporal variations of Sr- and Nd-isotope-ratios in the core sediments can be explained by drastic climatically induced changes in sedimentation regime at the Eurasian continental margin leading to a change in sediment provenance. Isotopic data show two main sources for sediment provenance: Eurasian shelf sediments (87Sr/86Sr-ratios of 0,713-0,719 and <epsilon>Nd-values of -8,5 to -13,5) and Proterozoic or Palaeozoic bedrock from Svalbard (87Sr/86Sr-ratios of 0,742-0,780 and <epsilon>Nd-values of -18,6 to -24,6 (Johansson et al., 1995)).
The sedimentation during interglacials is dominated by fine grained ice rafted sediment of sediment laden sea ice, formed by suspension freezing on wide and shallow Eurasian Shelf areas. During glacial periods with a drop of sea level around 120 m the shelf regions of the Barents, Laptev and Eastsiberian Sea fell dry, strongly affecting sea ice formation causing a switch to another transport process and sediment source.
During glacials sedimentation became dominated by ice rafted detritus (IRD) of calving icebergs from the advancing glaciers of the Barents Ice Sheet. At first this IRD probably consisted mainly of eroded Proterozoic or Palaeozoic bedrock from glaciated continental areas especially Svalbard explaining higher 87Sr/86Sr-ratios and lower <epsilon>Nd-values in the core sediments at the glacial onset. During the glacial the Barents Ice Sheet extended onto the shelf reaching the shelf edge at glacial maximum. The glaciers incorporated much of the unconsolidated shelf sediments, formerly deposited by suspension transport from the Siberian rivers and melting of sediment laden sea ice. Therefore a huge amount of IRD input from shelf sediments by icebergs took place during the rapid decay of the mainly shelf based parts of the Barents Ice Sheet in the glacial termination explaining the similarity of Sr- and Nd-isotopic composition of these core sediments to modern sea ice and shelf sediment samples.
Eisenhauer A, Mangini A, Botz R, Walter P, Beer J, Bonani G, Suter M, Hoffmann HJ & Wölfli W, Earth. Planet. Sci. Lett., 124, 171-184, (1994).
Johansson A, Gee DG, Björklund L & Witt-Nilsson P, Geol. Mag., 132, 303-320, (1995).
In the SW-Atlantic, east of Drake Passage, terrigenous detritus is supplied from South America by rivers and wind, and by ice from Antarctica. Bottom water currents are important for the distribution of sediment. To which extent the sources of the sediments and transport mechanisms (e.g. aeolian vs. bottom water currents) have changed through time remains to be resolved. Here we place constraints on these aspects by combining radiogenic isotopic and mineralogical data for sediment samples from two cores in the Scotia Sea, the seafloor in the SW-Atlantic, the Weddell Sea, and the SE-Pacific. The data represent the sedimentary record of the last 150 ka. We can make this assumption because the magnetic susceptibility curve can be related to the standard 18O-stratigraphy in carbonate-bearing cores fromadjacent areas. This permits the distinction between glacial and interglacially deposited sediment.
Sr- and Nd-isotopic compositions of the bulk sediment samples reveal significant regional (<epsilon>Nd -1 to -10) and temporal (<epsilon>Nd -3 to -8) differences indicating variable contributions from two major age provinces. The <epsilon>Nd-values in the core samples are strongly correlated with their magnetic susceptibility indicating the predominance of young material (high <epsilon>Nd-value) during glacial periods. There is an increase of the young component towards the South American continent.
Nd isotopic compositions of separated silt and clay samples provide more details on the sedimentary history than the bulk sediments: The <epsilon>Nd-values of the silt samples (<epsilon>Nd -3 to -4) are invariant with time and consistent with a single source - the South American crust. Those of the clay minerals (<epsilon>Nd -3 to -7) mimic the variations recorded in the bulk sediments. The isotopic compositions of clay and silt samples differ most pronounced for interglacial cycles (up to three <epsilon>Nd-units) and they converge during glacial cycles to similar values.The isotopic variations of the clay samples are correlated with the detrital smectite/chlorite ratio, a fingerprint for material from the Pacific (source of smectite) and Argentina (source of chlorite).
The combined isotopic and mineralogical data indicate for glacial cycles a sediment with chlorite and silt (high <epsilon>Nd-values) of Argentinian provenance; it is transported by wind and/or bottom water currents. In contrast, sediment samples of interglacial cycles are characterized by a higher proportion of smectite (low <epsilon>Nd-values) from the SE-Pacific, transported with the Antarctic Circumpolar Current.
Illite-vermiculite mixed-layered minerals (IVML) are characterized by d-spacing of 12A on air-dried and glycol solvated X-ray diagrams. IVML are uncommon in sediments, excepted in Pleistocene sediments of the North Atlantic ocean where they constitute a good indicator of detrital sources. IVML occur (5 to 35%) at shallow depth in Labrador shelf sediments but their abundance decreases to traces at greater water-depth (Fagel et al., 1996). IVML constitute 5-25% of the clay mineral association of Pleistocene sediments at 2698 m off New-Jersey (ODP Site 905A, Deconinck and Vanderaveroet, 1996), and they represent respectively 15 to 18% of the clay mineral fraction in cores SU90-11 and SU90-12 in the northwestern Atlantic basin. According to geochemical data (<epsilon>Nd = -23 and 87/86Sr = 0.736) and to their geographical distribution, IVML are thought to be formed by moderate weathering processes during interglacial in Northeastern Canada and are transported by run-off to the Labrador shelf area. Mineralogical studies indicate IVML are more abundant in sediment during interglacial periods than during glacial. This is confirmed by spectral analysis: IVML supplies in the northwest Atlantic basin are controlled by low- to middle-latitude processes (1/23 kyr-1), and the maximum IVML fluxes are about in phase with the maximum winter insolation (Bout-Roumazeilles et al., 1997). The Heinrich events, corresponding to iceberg discharge, are also enriched in IVML (35%) in cores SU90-11 and SU90-12. Ice-flow during fragmentation of ice caps and release of icebergs eroded surface sediments of the Labrador continental shelf. Then the enhanced supply in IVML probably results from the erosion of the shelf during formation of a dense nepheloid layer flowing at an intermediate water-depth (Bout-Roumazeilles et al., in press). Moreover, IVML have been very recently reported from the northeastern Atlantic basin. They constitute 10 to 30% of the clay mineral association in Pleistocene sediments on the Spitzberg continental area (ODP site 986). In core MD95209, located near Feroes Island, IVML are also present, characterizing the Heinrich events. In these two cases, the IVML supply appear to be linked to ice-sheet dynamics.Here we will discuss the occurrence of the IVML in the North Atlantic ocean and their use as tracers of ice-sheet dynamics.
Bout-Roumazeilles V, Debrabant P, Labeyrie L, Chamley H & Cortijo E, Paleoceanography, 12, 671-686
Fagel N, Robert C & Hillaire-Marcel C, Marine Geology, 130, 19-28
Deconinck JF & Vanderaveroet P, Proc. ODP, Sci. Res, 150, 147-170
In a deep-sea core collected within the oxygen-minimum zone (OMZ) off Pakistan (136 KL, 23°07'N, 66°3°'E; 568 m WD) we analyzed several organic geochemical parameters in order to reconstruct past changes of primary productivity and redox-conditions during the last 65 kyr. These results can be compared with similar measurements that we performed in a second core from the equatorial Indian Ocean (MD 900963, 05°04'N, 73°53'E; 2450 m water depth) which overlying water column is well oxygenated today. For the reconstruction of the primary productivity we measured the concentration of dinosterol, brassicasterol (136 KL) and the total of C37-alkenones (tot-C37; MD900963), respectively, which can all be used as qualitative proxies for primary productivity in the Arabian Sea as previously shown by Schubert et al. (1998). This interpretation is strengthened by the excellent correlation of the tot-C37 record with two independent foraminifera distributions (Beaufort et al., 1997; Cayre et al. in press). The total organic carbon (TOC) record of core 136 KL is not well correlated with the productivity parameter (dinosterol, brassicalsterol), it is rather positively correlated with the C35/C31-n-alkane-ratio which has been shown to be enriched under suboxic conditions (Schulte, 1997, Schulte et al., 1998) and thus reflects changes of preservation due to suboxic conditions prevailing in the OMZ. By contrast at the site of core MD900963 the deep water remained well oxygenated during the last 330 kyr (C35/C31 n-alkane ratio lower than 0.5) and thus preferred preservation due to suboxic conditions can be excluded. The excellent correlations in core MD900963 between TOC, tot-C37 concentration and TF-palaeoproductivities rather indicate that the variations of these parameters are controlled by primary productivity fluctuations. Similar conditions were found by Schubert et al. (1998) at a site in the northern Arabian Sea close to core 136 KL.A timescale is available fore core 136 KL which is based on AMS 14C ages and 18O variations (Schulz et al., 1998). The low C35/C31 values (< 0.5) during the Heinrich events suggest that OMZ was much weaker while during the Dansgaard-Oeschger (D-O) events the OMZ was well established (C35/C31-n-alkane-ratio higher than 0.6). Furthermore, the accumulation of organic carbon was high during D-O events, which could be due to better preservation under suboxic conditions, but very low during the Heinrich events.Today the Arabian Sea OMZ is maintained both by high productivity and by advection from the south of oxygen-depleted intermediate water masses. The partial decoupling observed in core 136 KL between OMZ and paleoproductivity proxies suggests that observed OMZ intensity should also be attributed to oceanic circulation changes. This interpretation of the palaeorecord is different from that by Schulz et al. (1998) who claimed that the variability of TOC in core 136 KL is only due to monsoonal driven fluctuations of primary productivity.
Beaufort L, Lancelot Y, Camberlin P, Cayre O, Vincent E, Bassinot F, Labeyrie L, Science, 278, 1451-1454, (1997).
Cayre O, Beaufort L, Vincent E, Quaternary Science Reviews, in press
Schubert C, Villanueva J, Calvert SE, Cowie GL, von Rad U, Schulz H, Berner U, Erlenkeuser H, Nature, 394, 563-566, (1998).
Schulte S, Ph. D. Thesis, University of Oldenburg, Germany, (1997).
Schulte S, Rostek F, Bard E, Mineralogical Magazine, 62A, 1356-1357, (1998).
Schulz H, von Rad U, Erlenkeuser H, Nature, 393, 54-57, (1998).
During the last years Dansgaard-Oeschger scale climate variability has been documented from many different sites all over the globe but the forcing of these cyclic variations is still one of the most important open questions in the recent paleoclimate research. High resolution (<10 cm/1000 years) sediment cores from the Arabian Sea monsoon system show a very similar pattern of millennial scale oscillations during the last 75 ka. The cores 70KL, 64KL and 57KL were retrieved from water depth between 3200 m and 3800 m. 18O, AMS14C, millimetre scale magnetic susceptibility and reflective colour spectroscopy measurements allow a detailed correlation of the three cores. Core 70KL had been sampled in addition in 1 cm intervals for geochemical trace element analysis. The results show that all interstadials of the oxygen isotope stadiums 3 and 4 reveal high productivity in the Arabian Sea, indicated by high CaCO3 and Corg values as well as high Ba/Al-ratios. These intervals are also characterized by low clastic input. Minerogenic sedimentation in the Arabian Sea area is mainly dominated by eolian dust flux via northwesterly winds which were stronger during glacial periods. Intervals with high magnetic susceptibility values, representing higher minerogenic contributions, and low productivity almost coincide with North Atlantic Heinrich Events and other stadials.Thus, Dansgaard-Oeschger scale variations in CaCO3-content and the magnetic susceptibility in the Arabian Sea sediments seem to be connected with the stadial/interstadial alternation in the Greenland Ice cores. This higher frequency oscillations are superimposed on five longer term humid intervals which appear to be associated to five temperature maxima recorded in the Antarctic Vostok ice core between 75 and 30 ka BP. The observed pattern is not necessary a response to northern or southern climate change, but may be also induced from other regions. However, our low latitude data provide a new perspective in the discussion about leads and lags between the hemispheres by exposing equivalents to climatic aspects known from the high northern and southern latitudes.
Core MD952042, retrieved from the Iberian margin (37°48N-10°10'W, Tagus abyssal plain, 3146 m water depth) during the Marion Dufresne-IMAGES I cruise, provides a high-resolution paleoenvironmental record covering the last climatic cycle. Marine proxy data from this core (planktonic foraminifera, coccoliths, carbon and oxygen isotopes, magnetic properties, IRD) have already identified an alternation of warm/cold phases and, in particular, the six Heinrich layers of the last glacial period. High resolution pollen and dinocyst records from this core allow us, for the first time, a detailed direct correlation between the Heinrich events and the continental environments. Pollen data suggest a succession of temperate-humid/ cold-dry phases on land (interstadials/stadials) between 70 and 10 ky BP. During the interstadials, pionners trees and shrubs such as Cupressaceae, Betula and Hippophae lead to the establishment of a deciduous and evergreen Quercus open forest. This forest is associated with the extension of heathlands (Ericaceae) indicating an increase in temperatures and precipitations. Pollen assemblages from the Heinrich layers indicate, in turn, a steppic vegetation dominated by Artemisia, Chenopodiaceae and Ephedra. Several ocean cold periods not contemporaneous with the Heinrich events are also characterized by a high content of steppics reflecting intermediate cold and dry phases on land. These phases may correlate with the stadials of the Dansgaard-Oeschger cycles. The pollen record corresponding to both types of cold events are very similar making difficult the distinction of major and minor cold oscillations on the continent. Dinocyst assemblages also exhibit a great variability over the sequence. In contrast with the vegetation, dinocyst response is not homogeneous when comparing the different stadial phases. Heinrich event 4 is dominated by peaks of Brigantedinium spp, Bitectatodinium tepikiense and Lingulodinium machaerophorum while the previous cold phase is characterized by Nematosphaeropsis labyrinthus and Bitectatodinium tepikiense.
The Alboran Sea is the westernmost basin of the Mediterranean Sea, the only one connected to the Atlantic Ocean and, therefore, the hydrography of the area is controlled by the continuous water mass interchange between N Atlantic and Mediterranean. North Atlantic Surface Water enters in the Alboran Sea to compensate the excess of evaporation of the Mediterranean Sea. The IMAGES piston core MD95-2043 (36o 08.598 N; 2o 37.269; 1841 m water depth) was recovered from one of the deepest areas of the Alboran Sea. Our study concerns to the top 16 m of this core which is a continuous record of the last 52 ky and characterized by very high sedimentation rates (average 30 cm/ky). The analyzed sampling interval (4-6 cm) provide an average temporal resolution of about 200 years.
Sea surface temperatures (SST) have been obtained from the relative composition of C37 unsaturated alkenones using the calibration from Müller et al.(1998). Experimental conditions are described in Villanueva et al. (1997). These measurements show relatively stable temperatures during the Holocene and abrupt changes during the glacial period. SST oscillations exhibit a good coherence with the Dansgaard-Oeschger cycles defined in the Greenland ice cores (Dansgaard et al., 1993). The coefficient correlation between the Alboran-SST and the GISP2-18O is r=0.88. Heinrich events from 1 to 5 are very well defined in the Alboran Sea by thermal drops of about 4oC. Occurrence of Neogloboquadrina pachyderma sinistral, an anomalous polar species for the Alboran Sea, is confined to the Heinrich events. The depletion of 18O from Globigerina bulloides during these events also suggests a decrease in the salinity of surface water as a consequence of the massive iceberg melting in the N Atlantic. The cold Dansgaard-Oeschger stadials are clearly differentiated from the Heinrich events first by the less intensive SST cooling and secondly by the lack of evidence of salinity dilution.
An alternative age model, based in the correction of the small misalignments between the Alboran-SST and the GISP2-18O records (well within dating uncertainties) provide a correlation coefficient of r=0.94. A close comparison of the GISP-Dansgaard-Oeschgerd interstadials with the Alboran-SST transitions, using this new age model, documents a recurrent pattern of marine warming and cooling following the atmospheric temperature changes.
Dansgaard W, Johnsen S, Clausen HB, Dahl-Jensen D, Gundestrup NS, et al, Nature, 364, 218-220, (1993).
Müller PJ, Kirst G, Ruhland G, von Storch I & Rosell-Melé A, Geochim. Cosmochim. Acta, 62, 1757-1772, (1998).
Villanueva J, Pelejero C & Grimalt JO, J. Chromatogr. A, 757, 145-151, (1997).
14C ages of planktonic foraminifera obtained by Accelerator Mass Spectrometry (AMS), must be corrected for the 14C difference between the atmosphere and the sea surface. The apparent 14C age of surface water strongly depends on mixing of surface and 14C depleted intermediate and deep waters. In the past, circulation changes may have strongly influenced the sea surface reservoir age. An evaluation of the 14C difference between the atmosphere and the sea surface waters has been derived for the Mediterranean sea during the last 18 kyrs. To compare the past and present sea surface reservoir age, we initially improved our knowledge of the modern reservoir age in the Mediterranean Sea. AMS 14C dates were therefore measured for 26 mollusk shells collected alive over the period AD 1837 to 1951 around the Mediterranean sea coasts. Results indicate a weighted-average reservoir age of 430 ± 110 yrs for the Mediterranean Sea basin, similar to that of the incoming Atlantic surface waters at the Gibraltar Strait. In order to evaluate the past reservoir age we have used volcanic ash layers which simultaneously settle down in continental and marine environments, and provide a good tool to quantify the atmosphere sea surface difference. Thus, we measured the 14C content of planktonic foraminifera contemporaneous to marine ash-layers in a high sedimentation deep sea core collected in the South Adriatic Sea (MD 90-917, 41°17N 17°37E depth 1010 m). This core contains several ash-layers, mainly originating from the Vesuvius and the Phlegrean Fields. Investigations of the 14C ages of paired terrestrial and marine deposits reveals that in the Mediterranean sea reservoir ages are similar to the modern one during the Holocene and the Younger Dryas, and reach about 700-800 years at the onset of the deglacial period. Several causes may be invoked which are linked to changes of Mediterranean Sea or North Atlantic ocean circulation.
The isotopic variations are commonly used in glaciologic analyses as indicator of past ice surface temperature. However such isotopic thermometers are also sensitive to other environnemental parameters. The influence of the temperature of the water source has been investigated by E. Boyle for the central Greenland. He showed that the discrepancy between temperatures obtained for the last Glacial from borehole thermometry and temperatures obtained with the isotopic thermometer calibrated by using the modern spatial 18O/temperature relation can be explained if the tropical sea surface water was 5°C cooler during the last glacial maximum than during the Holocene.
Here I speculate that short timed changes in tropical sea surface water temperature could also have affected the Antarctic isotopic records during the last glaciation. Indeed, a recent synchronization of Antarctic and Greenland ice cores done by Blunier et al. indicates that the development of the isotopic events observed in Antarctic ice cores around 37 and 45 kyr B.P. coincide with the Heinrich events H4 and H5. The amplitude of these isotopic events is consistent with a cooling of the water source of about 2.5°, according to the 18O / tropical surface water cooling relation proposed by Boyle.
I therefore suggest that these Antarctic isotopic events may reflect -at least partly- a cooling of the tropical source water, associated to the large ice discharges which occurred in the northern Atlantic during Heinrich events, and not Antarctic warmings. Such a scenario would give a simple explanation for the observed lag between isotopic signals of Greenland and Antarctica over the 47-25 kyr before present.
Boyle EA, Geophys. Res. Lett, 24, 273-276, (1997).
Blunier J, et al, Nature, 394, 739-743, (1998).
To understand the dynamics of the last deglaciation, it is important to determine the relative timing and the magnitude of the deglacial warming in the different parts of the world's ocean. Numerical models predict that a sudden melt-water discharge into the northern North Atlantic results in antiphased sea surface temperature (SST) changes in the Northern and Southern Hemispheres. If the models were correct, a slowdown of North Atlantic Deep Water formation - as observed during the Younger Dryas cold period - should be linked to a warming of the tropical-subtropical Atlantic due to a decreased northward heat transport. Here we present the first high-resolution alkenone SST record of the western tropical Atlantic covering the last 29,000 years. We show that the deglacial warming started 19,000 calendar years ago when SSTs were 3-4°C colder than present. During the Younger Dryas and also during Heinrich Event 1 indeed an asymmetrical pattern arose with warm tropical and cold North Atlantic SSTs consistent with a thermohaline circulation control of the interhemispheric temperature distribution.
We have studied two cores, BOFS 29K and 30K, from the subtropical Northeast Atlantic (20°31N 21°07W and 19°45N 20°44W respectively) using both micropaleontological and isotopic methods. The stratigraphic framework has been established using 18O, measured on the benthic species Cibicides wuellerstorfi and compared to the SPECMAP reference curve. For both cores, sampling intervals corresponding to a time resolution of better than 1Kyr were used. The period of time investigated covers the last climatic cycle. From thepoint of view of surface circulation, the core sites are located outside the main cell of Northwest African upwelling which is driven both by seasonal variations in the position and intensity of NE trade winds. From the point of view of intermediate circulation, both cores are in close proximity to the boundary between South Central Water and North Central Water. BOFS 29K and 30K were recovered at 4000 and 3500 meters water depth. They are at present bathed by NADW. Through time, the variability of this water mass could be estimated using 13C measurements on benthic foraminifera and the carbonate preservation index.
Sea Surface Temperatures (SST) reconstructed from planktonic foraminifera associations using transfer function range between 25°C and 16°C in summer and between 19°C and 11°C in winter. Rapid temperatures changes are observed during isotopic stages 3 and 2. The majority of these events show an amplitude of relative cooling of as much as 3°C. Some of these changes parallel similar events seen in the oxygen isotope record of the planktonic foraminifer species G. ruber and O. universa. Core 29K, located further north shows a higher variability during the early part of Stage 3 (58-40 kyr), whereas 30K stayed stable about 20°C during summer. During the later part of Stage 3 (40-24 kyr), both cores registered a general cooling, but in this case the more pronounced in 30K. The coldest episode that can be seen in both cores occurs after the Last Glacial Maximum as indicated by the heaviest value of 18O in Cibicides wullerstorfi. In fact, SST during the last glacial maximum is 4°C warmer than in during the subsequent cold, which is attributed to Heinrich event 1. Other cold periods could be also attributed to Heinrich events, but they are less pronounced than in UK37 -based SST records from more coastal upwelling cores. Our results using foraminiferal faunal SST and oxygen isotopes suggest an enhanced activity of the cold Canary Current during these cold events. We show that even in such a restricted area, palaeotemperature gradients may exist and can be explained by core locations: The more northerly core 29K was probably more regularly under the influence of increases in the intensity cold Canary Current during winter than the more southerly 30K.
Reconstructions of past changes in the deep ocean circulation are primarily based on measurements of the 13C and Cd/Ca ratios of fossil foraminiferal shells buried in marine sediments. They assume that changes of the foraminiferal ratios (which would document changes of the same ratios in ambient water) reflect mostly variations in the abundance of deep water masses, with a negligible contribution from air-sea gas exchange and organic matter cycling. This assumption has been very poorly investigated.
Here we examine the correlation between changes in North Atlantic Deep Water (NADW) and changes in the deep water 13C and PO4 content in a zonally averaged, circulation-biogeochemistry model. A total of 200 transient evolutions of NADW (treated as a color tracer), 13C, and PO4 are simulated, each characterized by the application of a different freshwater flux anomaly (FFA) at the surface in the North Atlantic. The FFA characteristics (volume, duration, and latitude of application) are chosen randomly between lower and upper limits based on paleoclimate estimates for the Heinrich events and the last deglaciation. The model responses to FFA range from insignificant changes in the Atlantic thermohaline circulation (THC), NADW, 13C, and PO4, to a complete THC collapse, drastic NADW (> 50%) and 13C reductions (< 0.5o/oo) and PO4 increase (> 0.5 mmol m-3). The distribution of isotopic and chemical anomalies in the deep Atlantic is generally bimodal due to a threshold in the convective activity in the northern North Atlantic. The higher mode is characterized by a complete THC collapse, strong north-south thermal antiphasing in the atmosphere, and correlations between (NADW,13C) and between (NADW,PO4) in the deep Atlantic which are always significant at the 95% probability level according to a non-parametric method. The lower mode, on the other hand, is characterized by a partial THC collapse, small north-south thermal antiphasing, and correlations between NADW and the oceanic tracers which are less significant at some locations in the Atlantic and Southern Ocean. Thus our model results support the interpretation of anomalies in benthic foraminiferal 13C and Cd/Ca documented in North Atlantic sediments in terms of changes in NADW abundance. They do suggest, however, that the relationship between the NADW and oceanic tracer anomalies may be discontinuous and may vary with latitude in the Atlantic basin.
MoBidiC is a zonally averaged climate model in which a quasi-geostrophic atmosphere is coupled to a three-basin (latitude-depth) ocean-sea-ice model. The main continental ice sheets are also represented. The atmospheric component of MoBidiC contains an interactive representation of the hydrological cycle based on a meridional transport of moisture and zonal redistribution of precipitation. The sea-ice model enables to represent explicitly the climatic effect of ice formation and ablation and their consequences on the ocean circulation. The interconnected ocean basins allow this model to simulate the main thermohaline circulation processes. MoBidiC is designed for long-term climate simulations and has a low computational cost. We present temperature, precipitation and ocean circulation patterns simulated for the present-day climate. Despite shortcommings in the representation of atmospheric dynamics and in the distribution of ocean salinity, the values obtained are realistic. Then we discuss a first paleoclimatic application of the model : the climate of the Holocene is simulated forcing the model with the variations of CO2 and insolation over that period.
Gallee H, van Ypersele J-P, Fichefet T, Tricot C & Berger A, J. Geophys. Res, 96, 13 139-13 161, (1991).
Gallee H, van Ypersele J-P, Fichefet T, Marsiat I, Tricot C & Berger A, J. Geophys. Res, 97, 15 713-15 740, (1992).
Hovine S & Fichefet T, Clim. Dyn, 10, 313-331, (1994).
Berger A, Loutre M-F, Gallee H, Clim. Dyn, 14, 615-629, (1998).
Cores spanning the last 10 ka raised from the outer part of the Norwegian Channel, northern North Sea and from a west Norwegian fjordsystem have been investigated. Age control is based on AMS dates on benthic and planktic foraminifera and 210Pb. All the studied intervals have sedimentation rates around 50 cm/1000 years or higher. Content of foraminifera and stable isotopes have been investigated in order to monitor changes in the surface and bottom water circulation through the Holocene. The results from the 'oceanic' core show several oscillations in bottom and surface waters during the Holocene which are related to changes in the inflow of Atlantic water into the Norwegian Sea. In comparison, the fjord record show a similar development indicating that the fjord archive faithfully record changes in the oceanic circulation system along the coast of Norway. Our findings indicate that the changes in the ocean circulation recorded through the Holocene occurred as a result of a weakening of the Norwegian Current which had a profound impact on the coastal climate. The recognition of some of these climatic oscillations in terrestrial records from Scandinavia and in other marine records from the North Atlantic region corroborate the increasing evidence of a close relationship between the intensity of the thermohaline circulation and NW European climate through the present interglacial.
The 8.2 ka (8.0-8.4 ka) event has been recorded at various places including ice cores from both poles and terrestrial and marine records from different latitudes. Here we report on early to mid Holocene (5 - 11 ka) variability of the Indian monsoon as recorded in the annually laminated core 63KA from the upper continental margin of Pakistan. The core was sampled for 18O (Globigerinoides ruber) at 10 - 30 years resolution with an age control established by 30 AMS dates on planktic foraminifera (Globigerinoides sacculifer), i.e. an dating resolution of approximately 100 years for the early Holocene. Around 8.3 ka precipitation and Indus river discharge, as recorded in 18O (G.ruber), dropped substantially from an early Holocene maximum (from 10.0 to 8.3 ka) within less than five decades. Monsoon activity remained mostly on an intermediate level until 6.3 ka, where it increased again. Prior to 8.3 ka large oscillations in the 190-240 yrs frequency band occurred. After the event, these oscillations disappeared or their amplitude was strongly diminished.
The Little Ice Age (LIA) is the most recent example in paleoclimate records of a relatively large, natural climate shift. At the time of that event, between early 1600 and late 1800 A.D., climate deteriorated over much of the globe. In the North Atlantic, the polar front shifted southeastward, sea ice nearly surrounded Iceland, and the ocean surface cooled up to 5 degrees centigrade. Climate mechanisms proposed for the LIA, which remain controversial, tend to place it in the context of an isolated, unique episode, justified by the widely held view that interglacial climates are stable.
Climate records from the North Atlantic spanning the Holocene and the Eemian, however, demonstrate that neither interglacial was stable. During both, the climate underwent a series of millennial time-scale oscillations. Accompanying each was a decrease in ocean surface temperatures, an increase in drifting ice, and an ocean surface circulation switch from a modern-like pattern to an altered one in which Greenland-Norwegian Sea surface waters were advected southward into the core of the North Atlantic Current.
The pacings of the interglacial climate events are statistically the same, ranging between 1,300 and 1,500 years. A statistically identical pacing also persists in North Atlantic records through the entire glaciation. Together, the glacial and interglacial event pacings have a mean value of 1,470 years (1 sigma = 600 years). Hence, the Holocene and Eemian climate shifts appear to be interglacial manifestations of a pervasive 1470-year climate oscillator (not demonstrably periodic) operating independently of the climate state. Rather than a unique episode, therefore, the LIA may have been the most recent cold phase of the climate oscillator. Indeed, the time step between the youngest cold phase identified in the marine records and the LIA is about 1,100 years, falling well within the range of variability of the 1470-year cycle. Moreover, hydrographic changes during its cold phases are not unlike those which occurred during the LIA.
If the LIA is part of a 1470-year climate oscillator, then it must be interpreted in the context of that oscillator. It is unlikely that increased explosive volcanism, which has been proposed for the LIA, can recur so regularly and frequently, and variability in solar output, also proposed for the LIA, is not known to have a 1470-year cycle. Perhaps, then, the explanation for the LIA lies in recurring, internal forcing of the Earth's ocean-atmosphere system. If that is true and given the persistence of the millennial time-scale oscillator, our present climate must be under the influence of that internal forcing.
Sedimentological, stable isotope (18O and 13C) and micropalaeontological analyses of NEAPACC core 15K (56°N, 28°W, 2848 m water depth) are used to examine surface and deep water circulation patterns in the northeast Atlantic Ocean during the Holocene. The stratigraphy of NEAP 15K is established using a combination of 14C dates, tephrochronology and planktonic and benthic 18O records. Average sedimentation rates during the Holocene are in the order of 40 cm/thousand years and, hence, are sufficient to provide detailed palaeoclimatic records capable of monitoring fluctuations in surface and deep water properties on millennial and century time scales. Spectrophotometer measurements of sediment lightness obtained at 5 mm intervals along the core have a temporal resolution of 10-50 years; other proxies have an average sample spacing of less than 200 years. Sediment lightness values exhibit a long-term increase through the Holocene, in excellent agreement with trends in calcium carbonate content (r=0.93). The close degree of similarity between these parameters suggests that downcore fluctuations are driven by the same, presumably climatic, mechanism. Superimposed on the general sediment lightness trend are a series of high frequency events. Although these smaller-scale variations in sediment lightness are less pronounced, they are several orders of magnitude greater than the instrumental precision. As calcium carbonate fluxes have remained relatively constant, these fluctuations in sediment lightness are believed to primarily reflect variations in the flux of fine-grained terrigenous material transported by deep water currents along the Gardar Drift. Variations in sediment lightness therefore provide a detailed record of North Atlantic Deep Water circulation intensity; an interval of enhanced NADW flow results in greater advection of terrigenous material, a reduction in calcium carbonate content and lower sediment lightness values. However, the 13C record of the planktonic foraminifer Globigerina bulloides also closely matches the Holocene patterns identified in the sediment lightness record, providing evidence of a direct linkage between surface ocean geochemistry in the subpolar North Atlantic and variations in the intensity of deep water flow over the last 11,500 years. Spectral analysis of the Holocene lightness record from NEAP 15K resolves prominent cyclicities centred at 1650 years, 1000 years, and 550 years. Cross-spectral analysis of the GISP2 18O and the NEAP 15K lightness records shows that fluctuations in North Atlantic circulation patterns and high latitude atmospheric temperatures are coherent at the 1000 year and 500 year frequencies during the Holocene.
Understanding the stability and variability of the current climates and natural environments is crucial for our ability to predict future natural climate variability. However, in order to assess variability it is important to obtain time series of ocean parameters that extend beyond the length of instrumental records. The longest time series with information on the strength of the Nordic heat pump comes from Ocean Weather Station M, but extend only 50 years back in time. Images site 2011 in the eastern Norwegian Sea provides a 7 m long open ocean Holocene section. The site is located at 1100 m water depth along the main axis of northward flowing warm Atlantic water. It is in an ideal position to monitor changes in the northward heat flux to northwestern Europe. The core is dated by AMS C-14 and Pb 210 isotope profiles. It has been studied at about 10-20 years resolution through the last 2500 years. Sea surface temperature (SST) variations are estimated by means of diatom transfer functions and isotopes. Deep water variability is analysed by benthic foraminifer carbon and oxygen isotopes. The records show clear evidence for late Holocene climatic events such as the Little Ice Age and the Medieval Warm Period. SST variability of 1-2° on timescales less than 100 years and a significant variability of intermediate waters are documented. Time-series analysis of the record will also be presented.
Quaternary Arctic Ocean sediments generally show very low sedimentation rates of a few cm or mm/ky. However, along the Arctic Ocean margins locations can be found, where sedimentation rates during the deglacial and Holocene were much higher, reaching 100 cm/ky in places. Although deposition was strongly dominated by fine-grained lithogenic particles, the sediments may contain variable, but generally low amounts of planktic foraminifers and other microfossils, which allow to establish an AMS-14C-based stratigraphy and to detect paleoenvironmental changes reflected in stable isotopes and microfossil abundances.
We will show at least two examples of high-resolution Arctic records. In a core from the Laptev Sea continental margin (78N, 135E, 985 m water depth) off the mouth of Lena river, deglacial changes in the freshwater outflow can be reconstructed from the oxygen isotope record of planktic foraminifers. Closely spaced 14C datings show that the sample resolution is about 20-50 years. The most conspicuous feature is a freshwater spike just before the onset of the Younger Dryas cool event. This spike can be traced in cores from the central Arctic and the western Fram Strait, which indicates that the freshwater may eventually have reached the Nordic Seas and influenced the thermohaline convection.
The second record comes from the western Yermak Plateau (81N, 2E, 1042 m water depth), where today Atlantic water inflow maintains open waters during summer. 14C datings show that the average sedimentation rate was >20 cm/ky during the Holocene and probably higher during the deglaciation. Although contents of coarse fraction and microfossils are generally very low in Holocene sediments, the abundance record of planktic foraminifers shows a distinct variability, which must have been generated by the strength of the Atlantic water inflow which is moving the position of the summer ice margin and influencing the transport of heat and moisture to the Arctic. The variability may correlate to the approx. 1500 yr cyclicity revealed in high-resolution paleoclimatic records from Greenland and the North Atlantic and can be considered as the first evidence ever that such short-term changes can be detected in Arctic Ocean sediments.
Our examples show that high-resolution records of paleoenvironmental and paleoclimatic changes can also be obtained from the Arctic Ocean. Whereas in general the Arctic Ocean ice cover may respond to hemispherical climatic developments, in cases the Arctic may even trigger strong climatic shifts.
Marine sediments off southern Chile are characterized by extremely high sedimentation-rates of primarily terrigenous material. Therefore, they provide the opportunity to study Holocene climate changes in the continental hinterland both with high time-resolution and robust age-control. The continental Holocene climatic history of southern Chile has been studied mainly by palynological and glaciological methods. These studies indicate significant climate changes regarding both temperature and precipitation which also affected the sedimentology of marine sediments from the adjacent continental margin.
Based on a preliminary stratigraphy, sedimentological data of sediment core GeoB 3313-1 (41°00´S; 74°27´W; water depth 852 m; core length 807 cm) from the southern Chilean upper continental slope suggest significant variations in the terrigenous sediment input during the last approximately 8,000 years. Clay mineralogy and physical properties (i.e. magnetic susceptibility) show intervals of increased input of material from Andean source rocks indicating more humid climatic conditions. This most likely points to a northward shift of the Southern Westerlies as the study area lies within the transition zone of the summer dry Mediterranean climate north of 37°S and year round humid conditions with heavy precipitation south of 42°S.
Our data add to results from mid-latitude and northern Chile (Klump et al., submitted; Lamy et al., in press; Lamy et al., 1998) indicating that the latitudinal position of the Southern Westerlies varied both on Milankovitch and millennial time-scales during the last glaciation and the Holocene.
Off southern Chile high sedimentation rates additionally allow spectral analyses of paleoclimatic proxies. Preliminary results suggest that Holocene climate cycles in southern Chile are dominated by 1,500 years cycles. This cycle has been observed in several Holocene records of the Northern Hemisphere and the tropics and is thought to be the pacemaker of rapid climate change. However the cause of the 1500-yr-cycle remains to be determined and is still highly speculative. Therefore, at the present point a global detection of the cycle in high resolution, well-dated paleoclimatic records, especially of the Southern Hemisphere, is essential.
Klump J, Lamy F, Hebbeln D & Wefer G, Geology, (submitted).
Lamy F, Hebbeln D & Wefer G, Quaternary Research, (in press).
Lamy F, Hebbeln D & Wefer G, Palaeogeography, Palaeoclimatology, Palaeoecology, 141, 233-251, (1998).
Today, the northern Arabian Sea is characterized by an expanded stable oxygen minimum zone (OMZ) and strong seasonal variability of monsoonal upwelling, i.e. high biological productivity in surface waters. During SONNE cruises SO-90 (1993) and SO-130 (1998) to the Pakistan continental margin 10 m-long kastencores and up to 20 m-long piston cores were taken from the OMZ off Karachi and off the Indus delta. The upper Holocene section (last 5000 years) of core 56ka (24°50.01'N; 65°55.01'E, 695 m water depth) reveals about 6 m of mm-laminated (varved), organic-carbon rich sediments reflecting seasonal changes of the monsoon circulation in the northeastern Arabian Sea. For high stratigraphic resolution studies in paleoclimate e.g. changes of the monsoon circulation during the Holocene we measured on core 56ka concentrations of TOC, isotope records of planktic foraminifers (G. ruber) and especially the alkenones which are produced by prymnesiophytes and were used to reconstruct paleo-sea-surface temperatures (SST). Our chronostratigraphy is based on varve counts verified by many AMS 14C ages (von Rad et al., Quat. Res., in press). Our alkenone records represents SST for the past 5000 years with a resolution of about 20 years. The alkenone data are compared with TOC and oxygen isotope values. Later, we will also study the peak glacial interval in detail, which is also represented by a varved facies, documenting a stable climate with strong OMZ conditions.
We studied several long coral cores from modern reefs in La Réunion, Mayotte and Madagascar. High resolution stable oxygen isotope data were calibrated with instrumental data and COADS data. Their annual variability reveals a slight warming at the end of this century. The longest record from Madagascar dates back to the year 1640 and records the Little Ice Age. Evidence for El Niño periodicities (3-5 yrs frequency) are seen by frequency analysis as well as for distinct years in the fact that the seasonal variability at Réunion and Madagascar is significantly reduced. The Holocene reef growth of the Western Indian Ocean reefs was studied on the base of several reef drillings. Six larger colonies were recovered in the cores from the Seychelles and Mauritius which were dated as 8.4, 7.9, 7.5, 6.8, 6.0 and 3.7 ky B.P.. In addition we analyzed a larger colony derived from a storm deposit with an age of 4.1 ky B.P.. High resolution stable isotope measurements (a sample every 2 or 4 weeks) were obtained on these time windows and compared to those obtained on modern coral colonies at the same sites. All the fossil Porites colonies are characterized by a slightly lower variability in 18O-values than their modern counterparts. The maximum amplitude of seasonal 18O-variation in the coral derived from the storm deposit is 1‰, in most of the years during the measured time span (18 years) the amplitude is ~ 0.8‰, which is equivalent to a temperature difference of ~ 4°C. Depleted 18O values in recent corals coincide with maximum SST in April, while enriched 18O values represent minimum SST occurring in August. The pattern of the annual variability of the 18O is linked to the typical asymetry of monsoonal climatic variability of this region. This pattern is also well documented in this late mid Holocene coral which also exhibits an ENSO rhythm (3-5 yrs frequency). Short term 13C enrichments occuring shortly after temperature minima may represent spawning events. Similar variabilities in the 18O-record (0.7-0.8%) was measured in Porites colonies from Mauritius (8,400 to 6,000 yrs old) and the Seychelles (7,500 to 3,700 yrs old). In modern Porites at the same sites, the annual variability in 18O-values averages 1.15 to 1.3%, reflecting SST differences of 5 to 6°C. The increase in annual variability during the past 8,400 years, suggests that climate variability has changed in the inter-decadal and inter-centennial timescale.
High resolution Last Glacial Maximum to Holocene palaeoenvironmental changes corresponding to a time interval of about 200 years are examined in the Myrtoon and Cretan basins in the South Aegean Sea, using the abundance variation in foraminifera and pollen in relation to oxygen isotope data and AMS 14C dating. Oxygen isotope data and foram abundance variation of Myrtoon basin show that from 18 ka BP to the Late Glacial / Holocene boundary at around 10.5 to 9 kyr BP the sea surface was cold except for a short interval of SST warming at about 14.3 kyr BP which is associated with the Termination 1A event. The disappearance of G. scitula and the change from high to low percentage occurrence of G. quinqueloba at about 13.5 kyr BP may indicate the transition from the Last Glacial Maximum to Late Glacial period. The Holocene started with a sharp warming of sea surface water, reaching present day level probably associated with a reduction in salinity. These changes coincide with changes in the foram assemblages i.e. Neogloboquadrina disappeared whilst G. bulloides, G. quinqueloba and the warm species are abundant. In both basins the sapropel sequences started at about 10 kyr BP about the same time with the sapropel formation in the other Aegean basins, which means that the environmental condition became suitable for the formation of sapropels in the Aegean about 800 years earlier than in the Adriatic, Ionian and Levantine basins. Within each sequence there are two high Corg content sublayers S -1a and S-1b separated by a lower Corg content sublayer indicating variation in the oceanographic and climatic condition prevailing in South Aegean Sea during the sapropel sequence formation. A combination of varying: (i) poor to zero oxygen supply in the bottom waters, caused by fresh water inputs by the outflowing rivers in to the Aegean Sea which later was supplemented and / or replaced by the overflowing of the Black Sea, and (ii) increased terrigenous and / or marine organic matter was responsible for the creation of disoxic to anoxic condition in the seabed, which lead to the formation of sapropel S - 1. The sapropel formation ended at 5.2 kyr BP. At about this time the sea surface temperature attained present day levels whilst the planktonic foram assemblages and their abundance is similar to present day ones. During the Last Glacial Maximum and Late Glacial periods the pollen source area was providing a suitable environment for the growing and survival of temperate tree taxa. The source area was apparently ice free and was characterized by sufficient humidity and not lethally low temperatures, probably only a few degrees colder than present.
Information about sedimentological, climatic, and oceanographic conditions at the time of deposition is recorded directly in the sediment colour of fresh marine/lacustrine sediment. Digital Sediment Colour Analysis (DSCA) provides a fast method that yields high-resolution information on these conditions, especially when integrated with other geological data. Since DSCA is a non-destructive method, it is easily applied to any kind of sedimentary archive. In the case of laminated sediments, laminae counting can be automated to provide a high-precision stratigraphy. We applied the method to colour records from ODP legs (Leg 146, Santa Barbara Basin, California and Leg 169S, Saanich Inlet, B.C., Canada, with conventional cores for the youngest, historical part of the record) and other localities. Results demonstrate the potential of DSCA to provide detailed insight into the high frequency variability of sedimentation and its forcing factors, e.g. El Niño, lunar-tidal influences, various solar influences.
Holocene to Late Pleistocene sediments in the Santa Barbara Basin, an anoxic basin off California, consist of alternating homogeneous and laminated intervals. The laminated intervals consist of sub-millimetre scale variations in clay, silt and diatom content, which reflect seasonal variation in productivity and runoff. Holocene sediments in Saanich Inlet, a fjord near Vancouver, are rhythmically laminated; individual couplets are 4 to 13 mm thick and have been shown to be annual deposits. The varves consist of alternating silty and diatomaceous mud. Thin lighter-coloured intervals, which are present in part of the laminae, are thought to represent flash floods. Laminae-scale colour variation in both records allowed for automated compilation of a continuous record with infra-annual resolution from digital images. For Saanich Inlet, time-scales could be generated from computerised laminae counts, which match well with radiometric dates. Time series analyses (evolutionary power spectra) show changing strength of variation in El Niño/ENSO frequencies and sunspot cycles in both areas during the Holocene. For example, the strength of the El Niño signal in the historical record in the Santa Barbara Basin was lower during warm periods then during cold periods like little Ice Age, with the 20th century as an exception.
During ODP Leg 167 aiming at the long-term reconstruction of the California current system a dedicated hole (E) has been drilled for high resolution paleoclimatic studies at Site 1017 60 km offshore Point Conception. The interval presented here spans the continuous uppermost four sections - dated by 14C AMS as the time frame of the last 25 ka. The average sample distance is 125 a.
Sea-surface temperatures (SSTs) have been reconstructed by converting the measured alkenone unsaturation (UK'37) using the equation by Prahl et al. (1988), which they calibrated in the northeast Pacific.
The late Glacial is characterized by SST changes between 11.6 and 13.9°C with a slightly increasing mean. An abrupt shift of +3.6°C to 16.8°C at 9.7 ka follows the beginning of the early Holocene warm period, which is well documented by pollen studies along the North American west coast (Heusser et al., 1980). After 7.5 ka SST values reached a lower plateau of 15-15.4°C until a final rise to 15.8-16.2°C at 3 ka. The uppermost five samples are not discussed in this context.
During the late glacial times carbonate and organic carbon contents vary rapidly between 2.2-11%w and 0.2-2.0%w, respectively. Age intervals of high carbonate contents exhibit elevated organic carbon percentages and coincide with warmer SSTs. A sharp increase to a carbonate maximum of 15.8%w beginning at 10.2 ka preceeds the surface water warming by 500 a. Steadily decreasing values reach a constant carbonate percentage of 7-8%w for the last 3 ka. Organic carbon contents increse continually during the Holocene up to 2.7%w.
The relationship of carbonate and organic carbon clearly distincts between different local carbonate preservation modes: The late Holocene is characterized by strong carbonate dissolution in contrast to late glacial times. The prominent carbonate maximum at 9.8 ka, which is found in many North Pacific sites, invokes changes in ocean circulation as ultimate cause (van Geen et al., 1996). During the progressive reduction of the height of the Laurentide Ice Shield the atmospheric pressure cells over the North Pacific according to modelling results (c.f. Amundson et al., 1996) have shifted northwards and established a dominant subtropical Pacific anticyclone. This explains the higher temperatures during the early Holocene warm period in the terrestrial realm (Heusser et al., 1985). The marine SST maximum seems to lag the atmospheric warming in the northeastern Pacific. After 7.5 ka the Aleutian low pressure cell has strengthened until 3 ka and resulted in intensified trade winds and hence upwelling and lower SSTs along the Caliofornia margin. Relative stable upwelling conditions at Point Conception during the last 3 ka are corrobated by constant 15N values.
Prahl FG, Muehlhausen LA & Zahnle DL, Geochim. Cosmochim. Acta, 52, 2303-2310, (1988).
Heusser CJ, Heusser LE & Streeter SS, Nature, 286, 702-704, (1980).
van Geen A, Fairbanks RG, Dartnell P, McGann M, Gardner JV & Kashgarian M, Paleoceanography, 11, 519-528, (1996).
Amundson R, Chadwick O, Kendall C, Wang Y & DeNiro M, Geology, 24, 23-26, (1996).
Heusser CJ, Heusser LE & Peteet DM, Nature, 315, 485-487, (1985).
The Adriatic sea is shallow sea oriented NW/SE and surrounded by mountainous chains. So that it is characterized by high terrigenous input and consequent very high sedimentation rate (e.g. more than 10 m in the last 10 ka). This feature represent an ideal condition to monitor the fossil assemblage fluctuations related to changes in either paleoceanography or climate. Here we present the result of a paleontologic study performed on nine sediment cores collected in the Central Adriatic Sea off Ancona. Within the chronostratigraphic framework provided by the ecozonal scheme based on planktic foraminifera for the Southern Adriatic Sea (Capotondi et al., in press), quantitative benthic foraminifers analyses revealed marked fluctuations in the assemblages related to changes in the paleoenvironment. In all the investigated cores the benthic microfauna shows a deepening upward succession related to the Holocene sea level rise. Eight significant changes in the foraminifer assemblage in some cases related to the most important climatic and paleoceanographic fluctuations (e.g.Younger Dryas, sapropel S1, Climatic Optimum) are recognized as well. The obtained data contribute to the reconstruction of the evolution of the basin during the last deglacial - interglacial interval. Finally, benthic assemblages fluctuations suggest that this records can be used as a local stratigraphic tool.
Capotondi L., Borsetti A., and Morigi C., Marine Geology, (in press).
The Mediterranean Sea, is a semienclosed marginal sea, that can be considered a natural laboratory for research in physical, chemical and biological oceanography and marine geology. This basin got its present morphology since the lower Pliocene and because of its unique regional conditions, the sedimentary record of regional and global variations in the climatic signal that affected it during the Pliocene Pleistocene time interval has been preserved in a excellent way. Mediterranean sediments are thus ideal for high resolution studies, either because of the excellent stratigraphic resolution due to the presence of ash layers, sapropel key beds and also to the excellent preservation of fossil signal. Moreover, these optimal conditions in the Adriatic sea are amplified by the high sedimentation rate Here we present the result of paleontologic analyses performed on the calcareous nannofossil assemblages of selected box cores and piston cores collected in the eastern Mediterranean and in the Adriatic sea, spanning the last 18 Ky. The study aimed to recognize fluctuations in the calcareous nannofossil assemblage, and the relationships between these fluctuations and the climatic changes that affected the basin (i.e. Last Glacial Maximum, Younger Dryas Climatic Optimum). A preliminar analysis shows important changes, characterized by the temporary disappearance or increase of different species of the assemblage across the studied interval. Furthermore, the sapropel S1 as previously observed by other Authors is characterized by a peculiar assemblage showing an increase in both F. profunda and Rhabdosphaera spp.
Core MD952014, taken at a depth of 2397 m, near the Gardar Ridge, south-west of Iceland, is investigated for variations in Ice Rafted Detritus content and distribution of planktonic foraminifera and ostracoda (benthic microfauna).
The planktonic foraminifera study shows a high variability of the isotopic stage 5 with a difference of six to eight degrees in temperature. A comparison of the planktonic foraminiferal based SST estimates with GRIP 18O signal indicates that stadials and interstadials 19 to 24 are recorded within core MD952014. However, intense warming episodes occur during cold isotopic sub-stage 5d and 5b. These warming episodes are characterised by higher SST than those recorded for sub-stage 5a.
Moreover, the marine record identifies iceberg discharges during cold sub-stage 5d and warm sub-stage 5c and 5a as well as during the 6/5e and 5a/4 transitions. These iceberg discharges are associated with decreases in SST and increased insolation, reflecting conditions similar to those reported for Younger Dryas.
Finally, during the warm periods of the isotopic stage 5, ostracods indicate a relatively intense deep circulation with sluggish episodes during iceberg discharges. Conversely, during cold periods, the deep circulation is sluggish with short phases of recurrent intensification.
High-frequency climatic oscillations (Dansgaard-Oeschger Cycles) first described from Greenlandic ice cores are evident in ice and sediment cores in both hemispheres. Here we present data from several sediment cores recoverd from the continental slope of the northern Bay of Bengal at water depth between 200 m and 2400 m. These cores recorded the changing fluviatile input from the large Ganges - Brahmaputra river system during the last 80.000 years. We used physical property measurements to correlate sediment cores over a depth transect of 200 km. Their oxygen isotope stratigraphy is based on the planctonic foraminifera Globigerina ruber white (50 - 200 years resolution) and AMS -14C ages. In addition, we established a high-resolution age model by correlating magnetic susceptiblity and 18O with dust and 18O measurements of the GRIP ice core. Sea surface temperature and salinity changes during the last 80.000 years were reconstructed using Uk37 and 18O data. We find Dansgaard-Oeschger cyclicity in the 18O signal as well as in the magnetic susceptibility and the geochemical records. The warm interstadials of the GRIP ice core correspond with light planctonic 18O values, implying a stronger SW-monsoon during interstadials. Very light 18O values at the end of the last glaciation point to an enormous fresh water input which might be caused by a combined effect of higher precipitation due to stronger monsoons and massive melting of the Tibetian ice sheet. At the long lasting interstadials (12, 8, 4, 2, 1), which are preceeded by Heinrich events, light 18O values and very high sedimentation rates can be attributed to an immense increase of the sediment transport capacity of the Ganges-Brahmaputra due to melting in the Himalya.
AMS14C-dated sediment cores from the Laptev Sea continental margin (Arctic Ocean) were studied in detail for changes in organic carbon composition and flux. The Holocene time interval represented in these cores is characterized by high sedimentation rates of 20 to 50 cm/ky which allows a high-resolution reconstruction of the depositional environment and its change in time and space. Based on total organic carbon content, organic-geochemical bulk parameters, biomarkers, maceral composition, and palynomorph assemblages distinct short-term variations in abundances of marine and terrigenous organic matter occur during the (post-glacial to) Holocene. Maximum terrigenous organic carbon supply culminating at about 8.7 - 9 ka (9500 - 10000 Calender years BP), is related to the post-glacial sea-level rise and changes in river discharge and coastal erosion. Preservation and short-term variability of marine organic carbon increasing near 8.5 ka (9000 Calender years BP), are interpretated as intensification and stronger variability of Atlantic-water inflow along the Eurasian continental margin.
During the CNR cruise performed in July 1979 (Project "Oceanografia e Fondi Marini") we collected two deep sea gravity cores from the Cefalù Basin (Core BS 79-33, Lat. 38° 15.7' N Long. 14° 01.8' E, Depth 1282 m and Core BS 79-22, Lat. 38° 23.1' N Long. 14°23.0' E, Depth 1449 m) in the southern Tyrrhenian Sea.As a semi-enclosed sea, the Mediterranean is affected by global palaeoclimatic changes, but with a strong overprint of local factors. Previous works on the Tyrrhenian area have demostrated that Circum-Tyrrhenian depressions (very close to the mainland), like the Cefalù Basin (close to the upper shelf of the Island of Sicily), are not so much disturbed by reworking as the central abyssal plain, and very often show an almost continuous geological record of the Late Quaternary. The high resolution micropaleontological sampling of the cores BS 79-33 and BS 79-22 provides new information on the palaeoceanographic and palaeoclimatic history of the Tyrrhenian Sea, focusing the attention on the last deglaciation and on the Holocene climatic fluctuations. Over 140 genera/species of benthonic foraminifera, mostly distinguished at species level, and 17 species of planktonic foraminifera, all belonging to the living fauna, were identified and counted in the >150 mm fraction. Radiocarbon chronology has been established through accellerator mass spectrometry (AMS) dating peaks of abundance of G. bulloides and G. ruber white from core BS 79-33. Oxygen and carbon isotope measurements have been performed on G. bulloides and on several other planktonic and benthonic species in both cores. The middle Holocene cooling is emphasised in both cores by rather high percentages (30-40%) of N. pachyderma r.c. and by a coeval disappearence of G. truncatulinoides l.c. The absence of this species is interpreted as an indication that during that time the southern Tyrrhenian Sea was cooler than today, but the water column was not so well mixed as other western Mediterranean areas of the same period (i.e. Ligurian Sea and Gulf of Lion). We also observed a strong increase in Chilostomella mediterranensis and Globobulimina affinis abundances (from 0 to 20% of the total benthonic foraminifera assemblages) which could be correlated to the Organic-Rich Layer (ORL) found in Alboran Sea by Zahn et al. in 1997 and dated 13.2-8.2 kyrs. Chilostomella and Globobulimina genera are considered to reflect dysoxic to anoxic environments and they are good indicators of high organic carbon content. Finally, the coccoliths record, based mainly on the relative abundance of Florisphaera profunda (a calcareous nannoplankton species presently living in the lower euphotic zone) will provide new data concerning the nutrient supply during the last 20 kyrs.
Variability of the continental and marine paleoenvironments has been recorded for the last 70 ka in the Western Mediterranean Sea. Analyses of the pollen and dinocysts proxies have been carried out from the ODP Leg 161 Site 976 (36°12'N., 4°18'W, 1108 m). Chronology has been assessed from oxygen isotope measurements on the planktonic foraminifera (Von Grafenstein et al., in press) and from correlations with neighbouring sites in Alboran Sea (Turon and Londeix, 1988).
Synchronous phase relationships are evidenced between the marine and continental proxies. We show the duality between semi desert (Artemisia, Chenopodiaceae and Ephedra) extensions and forest (Quercus) developments on the Western Mediterranean edges. From 70 to 20 ka, an abrupt climatic variability is observed. The semi desertic conditions on the continent are correlated with the development of the cold dinocysts Bitectatodinium tepikiense and cysts of Pentapharsodinium cf. dalei in the sea surface waters. Major fluctuations of the semi desert are contemporaineous with the time of large discharge of icebergs in the North Atlantic during the Heinrich events (Bond et al., 1993; Grousset et al., 1993). Minor ones may coincide with the Dansgaard/Oeschger events recorded in the ice cores (Dansgaard et al, 1993). During the LGM, humid climatic conditions are undisputable and followed by an increase in aridity during the Oldest Dryas.
Bond G, Broecker W, Johnsen S, McManus J, Labeyrie L, Jouzel & Bonani G, Nature, 365, 143-147, (1993).
Dansgaard W, Johnsen S, Clausen HB, Dahl-jensen D, Gundestrup NS, Hammer CU, Hvidberg CS, Steffensen JP, Sveinbjörnsdottir AE, Jouzel J & Bond G, Nature, 364, 218-220, (1993).
Grousset FE, Labeyrie L, Sinko JA, Cremer M, Bond G, Duprat J, Cortijo E & Huon S, Paleoceanography, 8, 175-192, (1993).
Turon JL & Londeix L, Bull. Centr. Explor. -Prod. Elf-Aquitaine, 12, 313-344, (1988).
Von Grafenstein R, Zahn R, Tiedemann R & Murat A, Proceedings of the Ocean Drilling Project. Scientific Results, 161, (1999,in press).
This presentation shows the objectives, structure, accessibility, and applications of the IMAGES database. IMAGES will provide an easy accessible and open to all database which is intended to improve international data synthesis. Crucial for the acceptance and trust in the database is the development of a good reference system which makes sure that data contributors are recognised and rewarded. Quality assessment of paleoceanographic data is also a major aspect of data management within IMAGES and a discussion on our approach will be presented. In addition, the poster will show some significant results of the programme and a summary of future plans.
The objectives of IMAGES are to quantify variability of the ocean on time scales of oceanic and cryospheric processes; to determine its sensitivity to identified internal and external forcing, and to identify its role in controlling atmospheric CO2. To characterise past oceanographic conditions, world-wide acquisition of high quality, preferably quantitative paleoceonographic data is needed. Therefore, IMAGES co-ordinates international coring cruises to obtain long and high resolution records. So far, four IMAGES supported coring cruises have been completed. The cores are being analysed in many different laboratories for various proxies in agreement with the IMAGES objectives. To optimise the international data acquisition and synthesis efforts, systematic data storage and data exchange is indispensable. Since IMAGES is affiliated to IGBP-PAGES and SCOR, data will be archived using protocols already established for other PAGES tasks and activities, and distributed over the Internet (http://www.images.cnrs-gif.fr).
The IMAGES database committee has set up a structure to archive: i) meta data (e.g., site information, references, methods); ii) raw data (e.g., physical properties, faunal counts, stable isotope values); iii) interpreted data (e.g., SST, SSS, productivity estimates), and iv) integrated data (e.g., SST maps for specific timeslices). Key to the project's success is that all these data components are subjected to a quality assessment. Such an assessment guarantees that different laboratories generate comparable, high quality data sets which can be integrated into, for example, numerical climate modelling experiments. For each discipline specialists have formulated a set of criteria to be applied to each data set. Criteria for stable isotopes, for example, include details on calibration, number of specimens analysed, and preparation methods used. A quality index can than be attached to each data set in the IMAGES database, so that no data set is a priori excluded for the next step of data interpretation and integration.
Estimates of past sea surface temperatures (SST) derived from the alkenone analysis provide a high resolution temperature record for the eastern South Atlantic (Angola Basin) over the last 25,000 calendar years BP. This SST record is especially focused on the Termination I, the transition period from the Last Glacial Maximum (LGM) into the Holocene, because this interval is one of the key periods for understanding the mechanisms of past global climate change with respect to interhemispheric teleconnections. The time scale was obtained by polynomial fitting the 14C ages calibrated according to Bard et al. (1997) into calendar years. The SST record obtained by applying the Uk´37 calibration for the eastern South Atlantic (Müller et al., 1998) shows a temperature increase in the order of about 4°C-5°C (17°C-22°C) between the LGM and the Holocene. The warming trend at the Termination I is interrupted by a "Younger Dryas"-like event for about 1 ky (12 cal. ky BP - 13 cal. ky BP), although the temperatures do not decrease again to almost glacial levels as is observed in North Atlantic records. This observation thus supports the Northern Hemisphere deglaciation pattern showing a mid-deglacial reversal with significant SST drop from 11,000 to 10,000 14C years BP also for the Southern Hemisphere. However, for the H1 event ( 15 cal. ky BP - 17 cal. ky BP), the South Atlantic record shows warming opposite to Northern Hemisphere records where significant cooling is observed for this period. Thus, the SST record from the eastern South Atlantic does not fully constrain the assumption of a bipolar sea saw (e.g. Broecker and Henderson, 1998) behaviour of Atlantic zonal circulation for Late Quaternary rapid climate events.
Bard E, Rostek F & Sonzogni C, Nature, 385, 707-710, (1997).
Broecker WS & Henderson M, Paleoceanography, 13(4), 352-364, (1998).
Müller PJ, Kirst G, Ruhland, G, von Storch I & Rosell-Melé A, Geochimica et Cosmochimica Acta, 62, 1757-1772, (1998).
The Little Ice Age is thought to have been a period with variable and cold climate with several glacier expansions. There are however some uncertainties in the timing of the interval and its geographical extent. The aim of the study is to document climate variability in the Norwegian Sea and the Denmark Strait during this time period. Two cores from the Kangerdlugssuaq Trough (93030 23A and 93030 19A), southeast Greenland shelf, and two cores from the Voring Plateau (MD95-2011 and JM96-948/2A) were investigated in the study. Paleoclimatic reconstructions are based on the definition of eight different assemblage factors and quantitative estimates of sea surface temperatures (SST) constructed from diatom transfer functions. The records show evidence of late Holocene climatic events such as the Little Ice Age and the Medieval Warm Period. The results from Voring Plateau show that the period from AD 700 - 1400 (the Medieval Warm Period) was characterized with relativly stable conditions and warm SSTs. An abrubt 1.5°C SST drop after AD 1400 represents the beginning of the Little Ice Age period that lasted until AD 1850. During this period there was a decrease in the influence of North Atlantic waters. The Kangerdlugssuaq Trough cores cover the last 300 years and show SST variations in the range of 2.5°C. Sea surface temperatures above todays level are documented during the Little Ice Age, implicating a greater influence of the Irminger Current during this period. A steady fall in SSTs from the 18th century and until today indicate a closer positioning of the East Greenland Front to the core sites. We suggest that the documentation of Little Ice Age can be explained by the eastward migration of the Polar Front and reduction in the inflow of North Atlantic waters into the Norwegian Sea, which was compensated by the increase in the strength of the westward flowing branch (Irminger Current). This scenario may have caused cooling in the eastern Norwegian Sea and warming in the Denmark Strait area.
We present detailed micropalaeontological, sedimentological and stable isotope data from sediment samples obtained from two long deep-sea cores, MD95-2039 and MD95-2040, recovered off the Portuguese Margin during the IMAGES MD101 scientific cruise, along with a third core, OMEXII-5K, recovered in the same oceanic sector. The cores contain distinct horizons with ice-rafted debris that represent the extreme extension of the well known Heinrich Events in the North Atlantic. Quartz is the dominant element in these terrigenous-rich layers along with detrital carbonate, volcanics, mafic minerals, basalt lithoclasts and mica. Other lithics derived from the nearby continental shelf, show up in significant amounts both within and outside Heinrich layers. SST was reconstructed using the modern analogue technique, and shows a good correlation with other independent proxies. During periods corresponding to the maximum abundance of sinistral Neogloboquadrina pachyderma there were significant drops in estimated SST, probably due to the input of cold meltwater from drifting icebergs, a conclusion supported by stable isotope evidence for reduced salinity. In all the cores SST minima within glacial stages 2, 3 and 4 and during the Younger Dryas in this sector of the Iberian Margin were between 5o and 12o C colder than the Holocene value of 19o C. At the "last glacial maximum" sea-surface temperatures were approximately as warm as today. At all IRD peaks estimated Summer SST fell to below 8oC; in the intervening interstadials temperature rose to near-interglacial levels.Stable isotopic measurements (18O, 13C) have been performed on planktonic (Globigerina bulloides) and benthic foraminifera (Uvigerina peregrina, Cibicidoides wuellerstorfi and Hoeglundina helegans). High resolution (150 yrs) planktic 18O records, % CaCO3 (J. Thompson, pers. comm.), faunal and lithological data for each core, allowed the development of a detailed multiproxy stratigraphic correlation between cores, through isotopic stages 1 to 4. 14C dating assisted us in generating a timescale by correlation with the central Greenland ice-core (GISP2). This enabled us to investigate offshore-inshore and latitudinal variations of temperature and salinity gradients over the last glacial cycle.
Within the framework of IMAGES program (International Marine Global Changes Study), a piston coring cruise (IMAGES IV-IPHIS III) has been organized in summer 1998 on board the R/V Marion Dufresne (IFRTP) in the Indonesian archipelago, the east of Philippines and the eastern China Sea. The aim of this cruise was to collect giant sediment records in high sedimentation rate areas (>10 cm/kyr) in order to obtain high resolution paleoclimatic records allowing the detailed reconstruction over the last climatic cycles of southeastern Asian monsoon and warm pool history, as well as the reconstruction of glacial-interglacial changes in heat and salt transfer from the Pacific to the Indian Ocean across the Indonesia archipelago. We intend in particular to reconstruct the Indonesian monsoon at the LGM (recent evidences suggest an increase in precipitation over Indonesia at this period); and to study possible links between southeast Asian monsoon and rapid climatic variability evidenced at high latitudes. Regarding the warm pool history, an important effort will be devoted to the reconstruction of the warm Kuroshio current activity and its effects on the climate of nearby continental areas.
During this cruise, 45 cores have been retrieved (mean length of about 35 m). These cores have been described on board, and measured for magnetic susceptibility, P-wave velocity, gamma-densimetry, and color reflectance. Additionally, surface water samples have been retrieved at almost all coring locations, and CTD profiles with samples have been performed along the water column at three sites in the Timor Sea. Three dredging operations took place in the Banda Sea for geodynamic activity reconstruction.
A preliminary survey of the results will be presented. In particular, we shall explore physical property records in terms of paleoclimatic variations. The interpretation of these preliminary IMAGES IV data will rest upon recent unpublished data obtained from shorter piston cores in the Indonesian and East China Sea areas during French-Indonesian cruises (SHIVA, 1990; BARAT, 1994).
Sr/Ca and trace elements, such as Mg, U, B, Ba, Mn and P, have been determined in Porites coral skeleton using TIMS and laser ablation ICP-MS (ArF excimer laser), in an attempt to retrieve a record of past sea surface temperature (SST) variations for the western tropical Pacific, mainly related to ENSO. Cores were collected near Kavieng (150.5°E, 2.5°S) in early 1997. Sr and U are positively correlated (r = 0.8), while Mg and B are anti-correlated with both Sr and U within a single laser sampling track. However, three parallel tracks, each separated by 1 cm, for the top 28 years of a Porites coral, show distinct Sr/Ca and trace element patterns, pointing to a strong dependence of the uptake of trace elements from seawater upon calcification processes. A laser ablation track, following almost perfectly the direction of growth as well as that of individual corallites, shows some correlation between Sr/Ca (or U/Ca) and SST, in particular high Sr/Ca (cool SSTs) associated with the El Niño events of 1972, 1982 and 1987. However, the correlation appears to degrade after 1992. Average Sr/Ca ratios are higher, and the annual range larger than expected from TIMS calibrations of Sr/Ca vs. SST established from Porites corals living at higher latitudes. This suggests that, in the absence of significant seasonal variation of water temperature, Sr/Ca may no longer be a reliable proxy for SST. As density bands, Sr/Ca, U/Ca and to a lesser extent Mg/Ca ratios carry a recognizable annual signal, these are more likely to reflect changes in the calcification rate driven by nutrient availability and annual changes of the wind regime from trade winds to westerlies.
The Northwest Atlantic Mid-Ocean Channel (NAMOC) is one of the longest deep-sea channels in the world. It extends from the Hudson Strait through the Labrador Sea to the Sohm Abyssal Plain of the western Atlantic Basin. The NAMOC flows downcurrent and is the main submarine channel for the transfer of sediments from the Labrador margin (Hesse at al. 1996). The Imarssuak Mid-Ocean Channel (IMOC) discharges sediments to the distal NAMOC from the Greenland - Iceland - Norwegian (GIN) seas. Sediments from these areas provide the best records of massive iceberg discharges, which correspond to "Heinrich Events". Heinrich Events are episodes of highly intensified iceberg drift in the North Atlantic during the last and penultimate glaciations, recognized in deep-sea cores by distinct layers of ice-rafted debris (Broecker et al. 1992) which were initiated by the repeated collapse and surge of the Laurentide Ice Sheet and other smaller European ice sheets. The NAMOC has apparently been scoured by numerous turbidity currents containing massive amounts of suspended bed-load during periods of maximum extension of the Laurentide ice sheet. As a result, different depositional facies are associated with Heinrich layers in the Labrador Sea (Hesse & Khodabakhsh, 1998).
Six cores were retrieved from below the confluence of the NAMOC and IMOC during the 1995 Marion Dufresne cruise MD101. At a length of 35 meters, core MD95-23 is one of the longest cores ever recovered from the Labrador Sea. High resolution radiography, content of the ice-rafting debris (IRD) and distribution of foraminifera suggest that MD95-23 contains up to nine Heinrich Events, while cores MD95-22, MD95-19 and MD95-18 contain five, two and four Heinrich Events, respectively. Core MD95-23 contains a similar number of ice-rafting events as core MD95-25, which was retrieved from the Orphan Basin of the Grand Banks of Newfoundland in the west Atlantic (Mudie et al. 1997). An accelerator mass spectrometric (AMS) 14C dating of 21,950 ± 940 ka was obtained from cold water dwelling Neogloboquadrina pachyderma (left coiling) at a depth of 14.27 m in c