Thermohaline instability in the North Atlantic during meltwater events: Stable isotope and ice-rafted detritus records from Core SO75-26KL, Portuguese Margin

PALEOCEANOGRAPHY, VOL. 12, NO. 5, PAGES 696-710, OCTOBER 1997 Thermohaline instability in the North Atlantic during meltwater events: Stable isotope and ice-rafted detritus records from core SO75-26KL, Portuguese margin Rainer Zahn, 1 Joachim Sch6nfeld, 1 Hermann-Rudolf Kudrass, 2 Myong-Ho Park, 1 Helmut Erlenkeuser, 3 and Pieter Grootes 3 Abstract. A benthic isotope record has been measured for core SO75-26KL from the upper Portuguese margin (1099 m water depth) to monitor the response of thermohaline overturn in the North Atlantic during Heinrich events. Evaluating benthic õ•80 in TS diagrams in conjunction with equilibrium õc fractionation implies that advection of Mediterranean outflow water (MOW) to the upper Portuguese margin was significantly reduced during the last glacial (< 15% compared to 30% today). The benthic isotope record along core SO75-26KL therefore primarily monitors variability of glacial North Atlantic conveyor circulation. The •4C-accelerator mass spectrometry ages of 13.54+.07 and 20.46+. 12 ka for two icerafted detritus (IRD) layers in the upper core section and an interpolated age of 36.1 ka for a third IRD layer deeper in the core are in the range of published •4C ages for Heinrich events H1, H2, and H4. Marked depletion of benthic õ•3C by 0.71.1%o during the Heinrich events suggests reduced thermohaline overturn in the North Atlantic during these events. Close similarity between meltwater patterns (inferred from planktonic õ•80) at Site 609 and ventilation patterns (inferred from benthic õ•3C) in core SO75-26KL implies coupling between thermohaline overturn and surface forcing, as is also suggested by ocean circulation models. Benthic õ•3C starts to decrease 1.5-2.5 kyr before Heinrich events H1 and H4, fully increased values are reached 1.5-3 kyr after the events, indicating a successive slowdown of thermohaline circulation well before the events and resumption of the conveyor's full strength well after the events. Benthic õ•3C changes in the course of the Heinrich events show subtle maxima and minima suggesting oscillatory behavior of thermohaline circulation, a distinct feature of thermohaline instability in numerical models. Inferrred gradual spin-up of thermohaline circulation after H 1 and H4 is in contrast to abrupt warming in the North Atlantic region that is indicated by sudden increases in Greenland ice core õ•80 and in marine faunal records from the northern North Atlantic. From this we infer that thermohaline circulation can explain only in part the rapid climatic oscillations seen in glacial sections of the Greenland ice core record. Introduction Sediments of the northern North Atlantic contain distinctive layers of ice-rafted debris (IRD) which are believed to have resulted from ice sheet instabilities that occurred episodically during the last glacial and triggered sudden surges of icebergs to the North Atlantic [Ruddiman, 1977; Heinrich, 1988; Broecker et al., 1992; Alley and MacAyeal, 1994; Broecker, 1994]. Negative planktonic foraminiferal /5•80 excursions in conjunction with increased abundances of polar planktonic foraminiferal species have been used to infer that these events were associated with cooling and freshening of the North Atlantic's surface waters north of 40øN [Bond et al., 1992, 1993; Keigwin and Lehman, 1994; Bond and Lotti, 1995]. On the basis of a conceptual model that invokes free ice sheet oscillations as a function of atmospheric and geothermal parameters, MacAyeal [1993] predicts that the Heinrich-IRD events were accompanied by a freshwater flux to 1GEOMAR, Research Center for Marine Geosciences, Kiel, Germany. 2Bundesanstalt fiir Geowissenschaften und Rohstoffe, Hannover, Germany. 3Leibniz-Labor fiir Altersbestimmung und Isotopenforschung, Christian-Albrechts-Universit•it, Kiel, Germany Copyright 1997 by the American Geophysical Union. Paper number 97PA00581. 0883-8305/97/97PA-00581 $12.00 the North Atlantic of the order of 0.16 Sv over a period of 250-500 years. The negative planktonic foraminiferal /5•80 response combined with decreased sea surface temperatures (SSTs) derived from transfer functions has been used to infer that the density of North Atlantic surface waters were lowered during these periods to an extent that vertical overturn spun down or was brought to a complete halt [Maslin, 1995]. This would be consistent with numerical simulations, which have shown that freshwater forcing of much shorter duration and smaller magnitude may cause convective instabilities that weaken or temporarily even terminate deep convection in the North Atlantic [Paillard and Labeyrie, 1994; Rahmstorf, 1994; 1995; Weaver and Hughes, 1994; Manabe and Stouffer, 1995]. IRD layers have also been documented outside the North Atlantic region of maximum IRD deposition, on the Portuguese and Moroccan continental margins [Kudrass and Thiede, 1970; Kudrass, 1973]. The occurrence of IRD there indicates a flow of icebergs from the north along the IberoMoroccan margin during the last glacial and deglacial. Recently, Lebreiro et al. [1996] reported on the occurrence of discrete IRD layers on Tore Seamount off Portugal. Oxygen isotope stratigraphy and mineral composition of these layers revealed that they correspond to Heinrich events 1, 2, 3, and 6. We use benthic isotope and IRD records from core SO7526KL from the southern Portuguese margin (Figure 1) to monitor the North Atlantic's ventilation response to the Heinrich events. Core SO75-26KL today is in the northward advection 696 ZAHN El' AL.: THERMOHALINE INSTABILITY DURING MELTWATER EVENTS 697 60ON .... • •' 40 ø 20 ø ß DSDP Site 609 • ß i , •' i 60ow 40 ø 20 ø 0 ø Figure 1. Position of R/V SONNE core SO75-26KL on the upper continental margin off Portugal. Ice-rafted detritus (IRD) and planktonic •5•80 records at DSDP Site 609 [Bond and Lotti, 1995] are used as reference for comparison with records from core SO75-26KL. path of Mediterranean Outflow Water (MOW) that flows at water depths between 700 and 1400 m along the western Iberian margin [Zenk and Armi, 1990]. Benthic •180 in conjunction with TS diagrams implies that the contribution of MOW to middepth water masses at the upper Portuguese margin was significantly reduced during the last glacial so that core SO7526KL was primarily under the influence of North Atlantic waters. Because the Portuguese margin is far outside the North Atlantic's main IRD belt, IRD fluxes there were considerably lower than in the open North Atlantic, and hemipelagic sedimentation continued during three IRD events found in core SO75-26KL. As a result, abundances of planktonic and benthic foraminifers remained high enough to allow for continuous stable isotope records and detailed 14C-accelerator mass spectrometry (AMS) dating across the IRD layers. The •4CAMS ages that have been determined for the two upper IRD layers and interpolated ages for the third IRD layer farther down in the core confirm that the IRD layers are coeval with Heinrich layers 1, 2, and 4. Sedimentation rates at the upper Portuguese margin are higher by a factor of 3 compared to the open North Atlantic and allow monitoring of ventilation changes during the Heinrich events at much higher resolution. This provides a unique opportunity to study the response of thermohaline overturn in the North Atlantic to the Heinrich events and associated meltwater events. Materials and Methods A benthic foraminiferal isotope record was measured for core SO75-26KL (37ø49.3'N, 09ø30.2'W, 1099 m water depth) from the upper Portuguese margin. The last glacialinterglacial transition and three IRD layers which occur along the core were sampled at 2 cm intervals to obtain a higher stratigraphic resolution and check for hydrographic and sedimentologic fine structure during these intervals. The rest of the core was sampled at 5-10 cm intervals. Stable isotope measurements were run on 1 to 21 specimens of Cibicidoides wuellerstorfi or C. pseudoungerianus. Within the IRD layers, isotope measurements were carried out on C. pseudoungerianus only, with a minimum of six specimens per isotope sample. The foraminiferal tests were picked from the size fractions >250 gm. In addition, a planktonic isotope record was measured using 25 specimens of Globigerina bulloides from the 315-400 gm size fraction. The planktonic isotope record was used to enhance stratigraphic control on the core. Prior to isotope analysis the foraminiferal shells were cracked open to release potential sediment fillings. They were then ultrasonically rinsed in methanol and transferred to a CARBO KIEL automated carbonate preparation device that is linked on-line to a FINNIGAN MAT 251 mass spectrometer. Long-term reproducibility was 0.08 %0 for fi•80 and 0.05 %0 for •13C as calculated from replicate analyses of an internal carbonate standard (Solnhofen limestone, 63-80 gm) that was routinely run at a ten-sample interval. The isotope data are referred to the Pee Dee belemnite (PDB scale). The •4C ages were determined via accelerator mass spectrometry (AMS) using the 3MV Tandetron system at the Leibniz-Labor of Kiel University [Nadeau et al., 1997]. Dated were 15 monospecific samples of G. bulloides containing 590 698 - ZAHN ET AL.: THERMO•INE INSTABILI'• DURING MELTWATER EVENTS Age (Cl&-•es. co=r. ) 0 4 8 12 I I i I .. -1 o. bullolclef Co•e 80'81-18, • 0 Po•ugueSe Margin • 1 2 loo11,. A1. [1989] •. •11o1•e• 0 100 200 300 400 500 600 700 • 3.13 • I - •L • • G• • / (&2 •-•) I I • • • • U A • ef S•. o •oo 200 200 •oo soo 6oo 700 •s l •r • >355• / g Sed• inte•late• 0 100 200 300 400 500 600 700 6O 40 -- 30 -- 0 100 200 300 4,00 depth in co•e ( Sedimentation Rate ! , ! ! 500 600 '700 Figure 2. (a) Age model developed for the last glacial/interglacial section of core SO75-26KL by correlating the planktonic õ1SO record with the continuously InC-accelerated mass spectrometry (AMS) dated planktonic õ]SO record from nearby core SU81-18 [Bard et al., 1989], and further •4C-AMS data deeper in the core. (b) Benthic õ•SO record. (c) The •4C-AMS ages obtained across two IRD horizons. Mean ages of 13.54_+.07 and 20.46+. 12 ka for these horizons and an interpolated age of 36.1 ka for the third IRD horizon deeper down the core correlate with the range of radiocarbon ages for Heinrich events H1, H2, and H4 measured at open North Atlantic core sites (see discussion in the text). (d) Age model for core SO75-26KL which implies that sedimentation rates varied between 15 and 55 cm kyr -•. to 1090 tests in the size fraction >250 pm and two shell fragments. All samples were ultrasonically rinsed in methanol. The shell fragments were treated with 10% H20 2 and dilute HC1 to remove organic material and carbonate dust. Ice-rafted detritus (IRD) was first counted from the size fraction >250 pm at 10 cm intervals. Three discrete IRD maxima were found at 127-143, 264-278, and 556-584 cm core depth. Detailed IRD countings were then carried out at 2 cm intervals across the IRD maxima to check for IRD variability within the maxima (Figure 2). To facilitate microscopic work, these countings were done on the size fraction >355 pm. Bulk volume of the samples was 45 cm 3 on average. IRD grains were counted from the total sample at the above given grain sizes, no split samples were used for the countings. Maximum countings reached 689 grains for IRD layer 3. The mineral composition of the IRD layers was determined by X ray powder diffraction (XRD) scans on IRD samples using a Siemens D 5000 automated diffractometer with incident and diffracted beam monochrometer (CuKa radiation at 25 mA and 40 kV; scanning angle was 20ø-50ø). The dhk I peaks were identified using the reference lists of Brindley and Brown [1984] and Bayliss et al. [1986]. Individual mineral percentages were determined from analog records using Biscaye's [1965] planimetry factors. Additional XRD scans were run on samples immediately below and above the IRD layers to obtain the mineralogy of the sediments which were deposited prior to and after the IRD events. Stratigraphy Age control on the last glacial-interglacial transition (100130 cm in core SO75-26KL)was obtained by correlating the ZAHN ET AL.: THERMOHALINE INSTABILITY DURING 1VIEI•A•R EVENTS 699 planktonic õ180 record with the planktonic õ180 record from a nearby core SU81-18 that has been dated by •4C-AMS in great detail [Bard et al., 1989]. A series of 17 •4C-AMS ages was measured farther down the core (Table 1). One sample at 174178 cm yields a •4C age (reservoir corrected; see Table 1) of 14.32 ka that fits well with the early phase of deglaciation documented by the initial decrease of benthic and planktonic õ•80 at this core depth (Figures 2a and 2b). Two samples bracketing the subtle benthic õ180 minimum around 400 cm core depth yield ages of 22.95 and 24.87 ka, somewhat younger than the age of 25.4 ka that has been estimated by Martinson et al. [1987] for oxygen isotope event 3.1. A coral fragment from 385 cm depth with an age of 24.26 ka (Table 1) supports the association with 3.1. A much more pronounced minimum occurs in the planktonic õ•80 record at this level, constraining late stage 3 isotope event 3.1 at 401.5 cm core depth. An additional age control point was obtained by assigning oxygen isotope event 3.13 of Martinson et al. [1987] to the õ•80 minimum at 651.5 cm in the benthic isotopic record (Table 2, Figure 2b). Beyond about 20 ka the Martinson et al. [1987] timescale is based on tuning to orbital parameters compared with UFFh dates. That is the age of 43.9 ka given b y Martinson et al. [1987] for oxygen isotope event 3.13 must be considered a calendar or calibrated age. Laj et al. [1996] suggest a correction of 2000 years at 40 ka (•4C), rapidly decreasing to zero around 47 ka (•4C), to convert the 14C timescale to calendar years. Thus we use a •4C age of 42 ka for oxygen isotope event 3.13 to ensure compatibility with the conventional •4C timescale that we use for core SO75-26KL. Detailed 14C-AMS Dating of IRD Layers I and 2 (Equivalent to Heinrich Events 1 and 2) Six •4C ages each were measured across IRD layers 1 (127143 cm) and 2 (264-278 cm) (Figure 3). Of the six 14C ages across IRD 1, the first two show an increase with depth, while the lower four form a plateau in the age-depth function and even decrease with increasing depth by 140 years. This decrease is, however, similar to the standard deviations and therefore statistically not significant. The closely similar ages at different core depths indicate increased sedimentation rates for this part of the core, which contains the heart of the IRD layer 1, between 130 and 142 cm core depth (Figure 3a). The three AMS dates for this interval give a mean age of 13.54+.07 ka (average depth 136 cm). The apparent age discontinuity between the IRD layer and overlying sediments is expected as a result of bioturbation whereby, after the rapid deposition of the IRD layer, younger tests of G. bulloides are mixed down to and into the upper reaches of the IRD layer, but few older tests are mixed up. This leads to artificially young dates for the upper boundary of the IRD layer [Manighetti et al., 1995; Trauth, 1995]. The same reasoning predicts an artificially high age for the bottom of the IRD layer due to lack of young tests being mixed down once rapid deposition of the IRD layer started. This is not observed as sample KIA 0007 at 147 cm shows the same age as the IRD layer. For our age model we use the average value of 13.54+_.07 ka at 136 cm (samples KIA 0004•0006, Table 1)for IRD layer 1 together with the other three measured ages (Table 2). 00000000000000000 00000000000000000 0000000000000 •'"• 0 ! ! ! ! ! i ! ! 0000000••••• 0000000000000000 00000000000000000 700 ZAHN ET AL.: THERMOHALINE INSTABILI'I• DURING MELTWATER EVENTS Table 2. Chronostratigraphic Control Points for Core SO75-26KL Core Depth, cm Age, ka a Age Marker Reference 1.5 3.0 b 14C-AMS Bard et al. [1989] 81.5 8.8 b 14C-AMS Bard et al. [1989] 121.5 12.7 b 14C-AMS Bard et al. [1989] 124.0 12.64 inC-AMS KIA 0002 c 130.0 12.91 inC-AMS KIA 0003 c 136.0 13.54 InC-AMS KIA 0004-0006 c'd 147.0 13.48 InC-AMS KIA 0007 c 176.0 14.31 InC-AMS KIA 0008 c 254.0 20.16 InC-AMS KIA 0009 263.0 19.75 InC-AMS KIA 0010 c 272.0 20.46 InC-AMS KIA 0011-0013 c'd 283.0 20.66 InC-AMS KIA 0014 c 321.0 21.82 InC-AMS KIA 0015 c 392.0 22.95 InC-AMS KIA 0017 c 415.0 24.87 InC-AMS KIA 0018 c 651.5 42.00 e 3.13 e Martinson et al. [1987] a All ages are corrected for reservoir age of 400 years. b Radiocarbon date from nearby core SU81-18. c KIA number is lab code listed in Table 1. e Oxygen isotope event; age is in 14C-kiloannums (see text). For the six AMS 14C dates across IRD 2 the age differences are likewise small, and because of the larger statistical uncertainties at this age, generally not statistically significant (Table 1, Figure 3b). The three dates from the heart of the IRD layer (265-279 cm, Figure 3b) give an age of 20.46_+0.12 ka at 272 cm depth for this episode of rapid accumulation. Again, the two overlying dates are younger, though the age difference to sample KIA 0009 at 20.16_+.2 ka is statistically insignificant. The average of 20.46_+0.12 ka (KIA 0011-0013, Table 1)for IRD layer 2 together with the other three dates is used in the age model (Table 2). The age model was calculated using a polynomial fit through the InC-AMS data (Figure 4). According to this model, IRD layers 1 and 2 lasted from 13.1 to 13.4 ka and from 20.2 to 20.6 ka. Mean sedimentation rates are 22 cm kyr -1 at core SO75-26KL. Sedimentation rates increase to 55 cm kyr -] and 32 cm kyr -• during IRD layers 1 and 2, respectively, suggesting rapid sedimentation during H1 and H2. Sevenfold to tenfold increases in sediment flux during Heinrich events have been inferred using 230Thex data from Heinrich layers in the open North Atlantic [Francois and Bacon, 1994; Thomson et al., 1995]. Smaller increases at our core site support the 14C J• ( kY ) l•a•er XltD >355gUa I g B•d. l•ad•er G.bulloJ. d.s I g B•d. 12.5 13 13.5 •4 0 I 2 3 4 S • 50 1•0 1•0 2•0 2•0 115 I I ...• ......... ,• ..... . .... l&O ..... • .• • - 1&5 • --+• -- 13.48 -- 14C J•es minus 400 Fr • 185c•-44 115 120 125 130 135 • l&0 lY 1&5 150 155 - 160 - 165 B 14C a,9e ( kY ) Nuad•r ZRD >3551.1m I g Bed. •= G. •11oides I g Bed. 19.5 20 20.5 21 0 I 2 • 0, 1•0 2•0 3•0 4•0 5?0 i i i = i i •50 •• ........... ] 246--17 290 1•C 250 255 260 265 • 270 J] 275 280 285 290 Figure 3. Age-depth function across (a) IRD layer I and (b) IRD layer 2. Within the range of error, 14C-AMS ages do not change from below the IRD layers into the IRD maxima and imply rapid sedimentation during both events. The right-hand panel shows the abundance distribution for G. bulloides that was used for 14C-AMS dating. See text for discussion. ZAHN ET AL.: THERMOHALINE INSTABILITY DURING MELTWATER EVENTS 701 • 0 3 i i i i i i i i i 0 5 10 15 20 25 30 35 4'0 4'5 1o 15 lO o 1.8- • - • 1.4, ol.0 0.6 0 5 10 15 20 25 30 35 4'0 4'5 50 Number IRD•i•355p.m [•i":ig •/'•i:•e•'u. iv H4' equiv. •il :•quiv. H2 I • I I I I i i 5 50 i I i S 10 15 20 25 30 35 4'0 4'5 50 age (ka) Figure 4. (a) Planktonic and (b) benthic 8180 records and (c) IRD abundances on age scales, and (d) benthic 813C record. The 15•3C record displays increased values during the last glacial. Negative anomalies correlate with Heinrich events 1, 2, and 4. We use this correlation to infer reduced thermohaline overturn in the subpolar North Atlantic during these events. contention that sediment flux from icebergs was reduced off Portugal as the site is outside the region of maximum iceberg flow. Using the polynomially fitted age model, total duration of HI and H2 is 300 and 400 years, respectively. These estimates depend on how good the age estimates for the top and bottom of the Heinrich layers are. Rapid sedimentation in conjunction with differential bioturbation likely make the top age too young and the bottom age too old. As we have discussed above, our detailed 14C-AMS dating shows some evidence for the first but not for the latter. Still, spacing of the 14C data across both IRD layers is not sufficient to estimate ages of IRD boundaries unambiguously. Using •4C-AMS data and age errors at face value, this would limit IRD 1 to 200-300 years, and IRD 2 to 300-600 years duration. These are still fairly wide ranges, but they fit to similar estimates derived from 230Thex data [Francois and Bacon, 1994; Thomson et al., 1995]. They are at the low side of estimates (250-1250 years) that have been derived from theoretical considerations of iceberg calving and sediment flux [MacAyeal, 1993; Dowdeswell et al., 1995]. Mineralogy of IRD Layers Off Portugal A distinct property of the Heinrich deposits in the North Atlantic is the elevated concentration of detrital carbonate and the presence of dolomite grains in the IRD layers [Bond et al., 1992; Bond and Lotti, 1995]. This has been related to limestone bedrocks in the Hudson Strait area and in Arctic Canada which where eroded by ice, transported by floating icebergs into the North Atlantic, and finally delivered to th• seafloor by melting of the icebergs. Lead, strontium, and neodymium isotope compositions of the IRD particles support this hypothesis in that these data point to the Canadian shield as the primary source for the lithie particles [Gwiazda et al., 1996; Revel et al., 1996]. Ice-rafted particles in IRD layer I off Portugal and Morocco consist of a great variety of rock types with 17% detrital carbonate and some striated specimens [Kudrass, 1973]. Our XRD scans show that primary IRD components are quartz, plagioclase, feldspar, and calcite (Figure 5). Since the XRD measurements were performed on bulk samples, most of the carbonate signal is likely of biogenic origin, that is, calcareous foraminiferal tests and nannoplank- 702 ZAHN ET AL.: THERMOHAL •tNEINST .AB.IL17• DURING MELTWATER EVENTS Heinrich Event 1 I I lO ' o • 1.0 ••l idolomite--_ • 100 x 60 ............... ............ I I I I o lo5 115 125 136 147 175 215 Heinrich Event 2 I I o 1.o o.o dolomite o 256 266 275 285 315 355 Depth in Core (cm) Depth in Core (cm) Heinrich Event 4 I do lomi t e t 100 ..... 80 60 40 F•t calcite .... • plagioclase • K-feldspar / quartz o 545 556 575 585 595 605 625 Depth in Core (cm) Figure 5. Clay mineral composition of IRD layers and sediments immediately above and below in core SO75-26KL Dolomite is present in small quantities in IRD layers 1 and 3 (equivalent to Heinrich events HI and H4), calcite is most likely of biogenic origin (foraminifera and nannoplankton). ton. Dolomite has been detected in minor quantities in samples from IRD layers 1 and 3 (equivalent to H1 and H4; Figure 5). After treatment with acetic acid the carbonate peaks disappeared from the diffractograms, but the intensity peak of dolomite was still observed, confirming the presence of dolomite in both IRD layers. Dolomite-bearing carbonate bedrock is widespread in the Laurentide domain, but small outcrops also exist in Ireland and Great Britain and along the western shores of Norway and Svalbard [Bond et al., 1992, Figure 1]. Thus the presence of dolomite cannot uniquely document a Laurentide versus Scandinavian source for the icebergs. However, our XRD data demonstrate that the mineralogy of the IRD deposits off Portugal is not different from those of the open North Atlantic Heinrich layers. In addition to the apparent close chronological correlation with the Heinrich layers in the North Atlantic, we use this as further evidence that the IRD layers in core SO75-26KL are derived from the same iceberg surges and ice sources that caused the North Atlantic Heinrich events. Heinrich Events and North Atlantic Thermohaline Instabilities Numerical modeling has shown that the North Atlantic's thcrmohalinc circulation constitutes a delicate balance between atmospheric forcing, northward heat transfer, and freshwater forcing [Paillard and Labeyrie, 1994; Weaver and Hughes, 1994; Rahmstorf, 1994, 1995]. From these simulations it was inferred that only small freshwater perturbations arc sufficient to trigger significant decreases and ultimately a complete shutdown of convection. Convective instability in the North Atlantic thus appears to be dependent not so much on the large-scale freshwater budget but more on local freshwater fluxes that are targeted at the immediate region of convcction [Rahmstorf, 1995]. During the last glacial, convection of dccp waters shifted from the Norwegian-Greenland Seas into the northern North Atlantic. As a result of the southward shift of convection and the associated reduction in thcrmic disequilibrium between surface waters and overlying atmosphere, evaporation rates were likely reduced leading to lower salinities and densities of surface waters and thus the ensuing reduction of convection rates in the North Atlantic [Boyle and Keigwin, 1987]. As a result, the core layer of deep water flow shifted in depth from 3000 m today to around 2000 m during the last glacial [Duplessy et al., 1988; $arnthein et al., 1994]. Only in the immediate vicinity of convection, that is, north of 45øN, did the influence of newly convected deep waters reach water depths similar to todays [$arnthein et aL, 1994]. Iceberg drift and associated meltwater flow during Heinrich events occurred close to sites of convection in the glacial North Atlantic. Paleoceanographic studies at northern North Atlantic sites have suggested that Heinrich events caused a significant drop of surface temperatures and salinities that also resulted in changes of surface circulation [Bond et al., 1993; Maslin, 1995; Maslin et al., 1995; $arnthein et al., 1995]. Enhanced meltwater flux in the course of the Heinrich events in this area thus likely induced convective instabilities that would have slowed down or even stopped thermohaline overturn. It has not been possible to test this hypothesis in detail because of the lack of continuous benthic isotope records across the IRD layers at type-core sites in northern North Atlantic that have been used to define the Heinrich events. Various studies provide evidence for a close coupling between decreased deep water production and variable meltwater discharge during the last deglaciation [Keigwin et al., 1991; Lehman and Keigwin, 1992]. However, these records do not reach far enough back in time to monitor thermohaline response to earlier Heinrich events, ar/d the high flux of IRD at coring sites from the northern North Atlantic wiped out the presence of benthic foraminifera because of dilution with IRD of the pelagic sediments in which benthic foraminiferal abundances are low. Thus sediment cores from outside the immediate area of maximum IRD deposition will provide the only chance to document thermohaline patterns during these meltwater events, provided that IRD deposition at these sites was small enough to ensure benthic foraminiferal abundances high enough for ZAh• ET AL.: THERMOHALINE INSTABILITY DURING lVlF. l JTWATER E•S 703 continuous benthic isotope measurements across the IRD layers. These sediment cores need to be taken from advection pathways that are linked to thermohaline overturn in the North Atlantic. Core SO75-26KL fulfills these requirements because (1) sedimentation rates are high, (2) IRD deposition was sufficiently low to leave enough benthic foraminifera for isotope analysis, and (3) the area is downstream from convection areas in the glacial North Atlantic. Core SO75-26KL as a Monitor of Thermohaline Patterns in the Northern North Atlantic At a water depth of 1099 m, core SO75-26KL today lies in the advection path of Mediterranean Outflow Water (MOW) that enters the North Atlantic with temperature-salinity (TS) values of 13øC/38.4 [Zenk, 1975] (Figure 6a). The density of this water is around 37.4 (o2=density on 200 dbar surface), that is, considerably higher than the density of 36.7 of North Atlantic Deep Water (NADW). Rapid mixing with less saline North Atlantic Central Water (NACW, TS=13ø/35.6 [Zenk, 1975]) and Labrador Sea Water (LSW, a component of upper NADW; TS=3ø/34.85 [Talley and McCartney, 1982], both flowing at the depth level of MOW in the North Atlantic, reduces the density of MOW so that it flows northward along the upper Portuguese margin at water depths of 700-1400 m (upper and lower core layers of MOW are at 750 and 1250 m [Zenk and Arrni, 1990]). TS values from hydrocasts near core SO7526KL are 11 øC/36.3 [Zahn, 1993]. Using end-member TS values of 13ø/38.4 for MOW entering the North Atlantic, 13ø/35.6 for NACW, and 3ø/34.9 for LSW, one arrives at a mixing ratio of 0.34:0.50:0.16 for MOW:NACW:LSW to generate in situ TS values of 11ø/36.3 at the site of core SO75-26KL (Figure 6a). That is, the middepth water mass at the upper Portuguese margin today consists of approximately 30% MOW and 70% middepth waters from the open North Atlantic, in good agreement with estimates derived from the "cascade box model" of Zenk [1975]. Following the concept of •using TS diagrams with equilibrium •5180 (•5 c) fractionation [Zahn and Mix, 1991] we determine the potential contribution of MOW to the site of core SO75-26KL during the last glacial maximum (LGM) (Figure 6). TS-õ c diagrams provide a valuable tool for the interpretation of foraminiferal •5180 values in that they add the constraint of vertical density stratification that helps to narrow down possible TS scenarios as inferred from foraminiferal õ•80 to physically plausible solutions [Zahn and Mix, 1991; Labeyrie et al., 1992; $arnthein et al., 1995; Weinelt et al., 1996]. The modem •5 c fractionation lines shown in Figure 6a have been computed using the paleotemperature equation of Shackleton [1974] and a •Sw:salinity slope of 0.5 (•5 w is the oxygen isotope composition of seawater) which results from a North Atlantic freshwater end-member around -18%o in fiw (SMOW). This slope represents the •Sw:salinity relationship for modem North Atlantic deep and middepth waters. We have fixed NADW at a benthic •5 c of +3.6%o PDB and TS values of 2øC/34.9 that are commonly measured at coring sites in the North Atlantic [e.g., Labeyrie'et al., 1992]. For the TS values of 11 øC/36.3 that have been measured in hydrocasts near the site of SO75-26KL [Zahn, 1993] we obtain a •5c value of +2.1%o PDB. This value is consistent with late Holocene •5•80 of C. wuellerstorfi of +1.48 %o PDB in core SO75-26KL (Figures 2 and 4) once the values have been corrected to the Uvigerina scale by adding 0.64%o. For a first evaluation of water mass distribution at the LGM, the õc fractionation lines have been computed using the same slope as for the modern õc lines (Figure 6b). Glacial maximum lower NADW (LNADWLGM) was fixed to TS values of 0øC/35.8 and a õc value of +5.3%o PDB (Figure 6b) [Labe•yrie et al., 1992; Sarnthein et al., 1995; Jung, 1996]. From glacial-interglacial variations of planktonic foraminiferal õ•80 and from planktonic foraminiferal census counts along sediment cores from the Mediterranean it was concluded that salinities of glacial maximum Mediterranean waters were higher by 1.2-2.7 and temperatures were lower by 4ø-6øC [Thiede, 1978; Thunell, 1979; Thunell and Williams, 1989]. Using mean values of 5øC cooling and a salinity increase of 2 at the LGM, we have set MOWLG M as it enters the North Atlantic to TS values of 8ø/40.3 (Figure 6b). To determine the potential contribution of MOWLG M to ambient water masses at core SO75-26KL we need to determine TS values for middepth waters in the glacial maximum North Atlantic. One end-member water mass is Glacial North Atlantic Intermediate Water (GNAIW) [Duplessy et al., 1988] or Upper North Atlantic Deep Water (UNADWLGM) [Sarnthein et al., 1994; Jung, 1996]. The reconstructions by Duplessy et al. [1992], Labeyrie et al. [1992], and Sarnthein et al. [1994, 1995] infer that salinities of UNADWLG M were similar to those of LNADWLGM, 35.8. Using a benthic õ•80 of +5.0 PDB that is documented in sediment cores from the northern North Atlantic at water depths of 1100-1500 m [Jung, 1996] and a salinity of 35.8 yields a paleotemperature of +IøC for UNADWLG M (FigUre 6b). The TS values for UNADWLGM thus are 1 ø/35.8. Few paleodata are available to trace the glacial-imerglacial evolution of NACW as the second end-member for mixing with MOWLG M. Slowey and Curry [ 1995] infer a cooling of 2øC at middepths (1-2 km) around the Bahamas, pointing to similar cooling of surface waters in the North Atlantic subtropical gyre, the potential source region for NACW. Summer sea surface temperature (SST) in the glacial maximum North Atlantic ranges from 5 ø to >13øC at the northern margin of the subtropical gyre [Climate Long Range Investigation, Mapping, Prediction (CLIMAP), 1981; Sarnthein et al., 1995; Weinelt et al., 1996]. Winter temperatures were lower, approaching 0øC in the northern North Atlantic [Sarnthein et al., 1995]. Thus glacial NACW temperatures could have been anywhere between Present-day values (13øC) and close to upper glacial deep water temperatures. Temperatures warmer than todays appear unlikely given that upper North Atlantic water masses were colder during the last glacial [Slowey and Curry, 1995]. We discuss the following two scenarios (Figure 6b) using the temperature and salinity estimates given by Duplessy et al. [1991] and Sarnthein et al. [1995]. (1) a cold subpolar NACWLG M that formed in the area of the positive salinity anomaly at 51ø-54øN inferred by Duplessy et al. [ 1991 ], with TS values of 6ø/36, and (2) a warm NACWLGM that formed along the northern margin of the subtropical gyre (around 40øN), with TS values of 13ø/37.5. In the first scenario, NACW would have been considerably colder than today (AT = -7øC) and salinity increased by 0.4, whereas in the second 704 ZAHN ET AL.: THERMOHALINE INSTABILITY DURING MELTWATER EVENTS A: modern •5w:S=1:2 •w(fm,hw, w)=-18%oSMOW NA•CW •o.6 MOW 13 .i t • . t•k I...t . .•. ,•,,• .I.. • ' , I I I, I I lO' 965431o- -1 - • MOW:NACW:LSW = 0.34: 0.50: 0.16 -2 i i i i i i i i Ji i ! ! ! i ! 33.5 3,1 3,1.5 35 35.5 3• 3•.5 3? 3?.5 38 38.5 39 39.5 ,10 ,10.5 ,ll. ,11..5 42 B: last glacial maximum •w:S=1:2 •w(fre•hwatmr) = -18%o SMOW :].3 I I . I I I I •al'm • • • • • • : • , / / / 3 z OW:NACW = 0.10: 0.• 0 L•DW (2) MOW:•CW:UNADW = 0.10: 0.42: 0.• 33.5 34 34.5 35 35.5 36 36.5 37 37.5 3• 38.5 39 39.5 40 40.5 45 45.5 42 I._.• •5= (•. POe) density (oa) mixing triangles C: last glacial maximum •w:S=1:1.25 •w(fruhw,=) = -30%, SMOW • 7 = 6 NACW ' • • 0 Figure 6. (a) TS diagram showing water mass end-members that contribute today to the site of core SO75-26KL. Density lines are for the 200 dbar surface (approximately, 2000 m water depth). Isolines of •5 c equilibrium fractionation are computed using the paleotemperature equation of Shackleton [1974] and a •Sw:salinity slope of 0.5 for the modern North Atlantic; freshwater •5 w is -18%o (SMOW). NACW (triangle) is North Atlantic Central Water, MOW (solid dot) is Mediterranean Outflow Water, NADW (rectangle) is North Atlantic Deep Water, and LSW (diamond) is Labrador Sea Water. The cross shows TS values at the site of core SO75-26KL. Dotted lines give mixing triangle between water mass end-members. Today, a mixing ratio of 34% MOW, 50% NACW, and 16% LSW contributes to the hydrography at the site of core SO75-26KL. See text for discussion. (b) Same as in Figure 6a, except for the last glacial maximum (LGM). TS values for mixing end-members have been changed to estimated LGM values (see text). UNADW (diamond) is Upper North Atlantic Deep Water. Crosses indicate TS values at a benthic •5180 value of +4.0 Pee Dee belemnite (PDB) for highest possible MOW contribution if MOW mixes with cold (cross 1) or warm (cross 2) NACW. Mixing ratios for crosses 1 and 2 are indicated. The •5 c fractionation is computed using the same parameterization as in Figure 6a and correcting for a mean glacial •w increase of 1.2%o. (c) Same as in Figure 6b, except that the dw:salinity slope is set to 0.8, and freshwater •5 w is -30%0 (SMOW). This configuration accommodates glacially lowered precipitation temperatures and the contribution of meltwaters. See text for discussion.