Quaternary Science Reviews 27 (2008) 2041–2047 Contents lists available at ScienceDirect Quaternary Science Reviews journal homepage: www.elsevier.com/locate/quascirev A 4500-year reconstruction of sea surface temperature variability at decadal time-scales off North Iceland Marie-Alexandrine Sicre a, *, Pascal Yiou a, Jón Eirı́ksson b, Ullah Ezat a, Elwrick Guimbaut a, Imane Dahhaoui a, Karen-Luise Knudsen c, Eystein Jansen d, Jean-Louis Turon e a Laboratoire des Sciences du Climat et de l’Environnement, IPSL, CNRS/CEA/UVSQ, Domaine du CNRS, Ave de la Terrasse, 91198 Gif-sur-Yvette, France Earth Science Institute, University of Iceland, Askja, IS-101 Reykjavik, Iceland Department of Earth Sciences, University of Aarhus, DK-8000 Aarhus C, Denmark d Bjerknes Centre for Climate Research, University of Bergen, Bergen, Norway e Département de Géologie et Océanographie, Université de Bordeaux I, Talence, France b c a r t i c l e i n f o a b s t r a c t Article history: Received 30 January 2008 Received in revised form 13 August 2008 Accepted 19 August 2008 Marine paleo-records acquired at high temporal resolution provide critical data for testing numerical climate models and help to understand processes underlying ocean variability. This study presents a unique 4500-year reconstruction of sea surface temperature (SST) obtained from alkenones in the North Atlantic Polar Front area off North Iceland, at an average temporal resolution of 4–5 years. Spectral analysis of this signal shows dominant multidecadal oscillations which occurred with a stronger amplitude between 2500 and 4200 years BP, hand in hand with fluctuations of bottom currents indicated by paleomagnetic proxies. Contemporaneous large excursions of the Inter-tropical Convergence Zone (ITCZ) are also recorded by the distant Cariaco titanium time series, suggesting a link with low latitude Atlantic climate. We speculate that the oscillations reflect changes of the Meridional Overturning Circulation (MOC) induced by increased ENSO activity. Ó 2008 Elsevier Ltd. All rights reserved. 1. Introduction Climate in Europe is modulated by the northward transport of heat and moisture by the North Atlantic Current (NAC). Any change of this transport will impact on the winter temperature and precipitation patterns of northwestern European countries (Visbeck, 2002). Consequently, it is important to improve the understanding of the role of the ocean in broad-scale climate changes and the physical mechanisms underlying them, if we want to improve predictions. Advances in model performance and increased observational capacity have concurred to foster our comprehension of ocean variability over the past few decades to centuries. Yet, reliable quantification over longer time-scales remains a major challenge for climate research. Knowledge can be gained from palaeo-data but high-resolution and high-quality multi-proxy time series of the surface and deep ocean circulation need to be developed further. In the past few years, major efforts have been made to better describe marked changes of the past millennium, like the Medieval Warm Period (MWP) and the Little Ice Age (LIA) (Keigwin, 1996; Keigwin and Pickart, 1999; deMenocal et al., 2000; Eirı́ksson et al., 2006; Lund et al., 2006), or ocean * Corresponding author. Tel.: þ33 1 69 82 43 34; fax: þ33 1 69 82 35 68. E-mail address: sicre@lsce.ipsl.fr (M.-A. Sicre). 0277-3791/$ – see front matter Ó 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.quascirev.2008.08.009 circulation perturbed states such as the 8200 year BP melt water event (Ellison et al., 2006). Other studies have produced Holocene records at increasing temporal resolution to document the surface ocean variability, but few have achieved characteristic time-scales of the atmosphere/ocean coupling and the Meridional Overturning Circulation (MOC), i.e. decadal to centennial (Risebrobakken et al., 2003; Cronin et al., 2003, 2005; Black et al., 2007; Sicre et al., 2008). The MOC is affected by the North Atlantic Oscillation (NAO) (Latif et al., 2006), the dominant pattern of atmospheric variability in midlatitudes North Atlantic (Hurrell, 1995). The NAO and associated wind fields alter the surface ocean buoyancy which in turn affect the convective activity of major regions of deep-water formation, i.e. the Labrador, Irminger and Greenland seas, as well as sea ice dynamics (Dickson et al., 1996). The northern North Atlantic is thus a key region for investigating MOC variability and its links to NAO. Shelf sediments off North Iceland, characterized by high sedimentation rates, provide exceptional archives to capture surface ocean variability at decadal time-scales, and thus represent ideal sedimentary settings to undertake such studies. Furthermore, the presence of well-known tephra layers from volcanic eruptions in Iceland has allowed developing accurate tephro-chronological age models on marine cores (cf. Larsen et al., 2002; Eirı́ksson et al., 2004), thus reducing uncertainties associated with radiocarbon dating of marine calcite. 2042 M.-A. Sicre et al. / Quaternary Science Reviews 27 (2008) 2041–2047 The present study extends the 0–2000 year alkenone SST record published by Sicre et al. (2008) in order to investigate mid-Holocene ocean variability, when large amplitude shifts of the ITCZ are seen in the Cariaco basin (Haug et al., 2001). An SST time series of the last decades obtained from box-core sediments is also presented and compared to instrumental data to evaluate the reliability of our proxy reconstruction. 2. Materials and methods 2.1. Core location and oceanographic setting The MD99-2275 core (66 33N; 1742W, 470 m water depth) was retrieved on the North Icelandic shelf (Fig. 1) during the 1999 North Atlantic IMAGES cruise on the R/V Marion Dufresne. Because the top of the core is lost during normal coring operation due to overpenetration of the Calypso corer, a box-core (BO5-2006-GBC03C; 66 33.18N; 1742.04W) was retrieved in 2006 (Millennium project, R/V Bjarni Sæmundson B05-2006 cruise) to recover recent sediments in the same site. The mean sedimentation rate over the Holocene is on the order of 250 cm/1000 years. To resolve temperature changes at subdecadal temporal resolution, the MD99-2275 core was continuously sampled at 1-cm sampling step, corresponding to a temporal resolution of 2–5 years. As can been seen from Fig. 1, the coring site is located close to the marine polar front, in a climatically sensitive area where two important components of the North Atlantic circulation mix, i.e. the warm and salty waters of the Irminger Current (IC), a branch of the NAC, and the cold and low-salinity southward flowing waters of the East Greenland Current (EGC) (Østerhus et al., 2005). The surface hydrology is also affected by sea ice and drifting ice exported from the Arctic Ocean and East Greenland. Table 1 Depth in centimeters (cm), ages in year cal. BP and in year AD of the tephra layers identified in core MD99-2275 used to build the age model Depth (cm) Age (cal. BP) Age (AD/BC) Marker horizons 101 179 209 239 275 321 460 687 941 1552 230 470 540 650 850 1080 1818 2980 4200 7125 1720 1480 1410 1300 1100 870 132 1030 2250 5175 Veidivötn AD 1717 Veidivötn AD 1477 Veidivötn AD 1410 Hekla AD 1300 Hekla AD 1104 Settlement layer Snæfellsjökull I Hekla 3 Hekla 4 Hekla 5 2.2. Age model The tephro-chronological age model of the MD99-2275 core is described in detail by Eirı́ksson et al. (2004). The list of tephra layers used to build the age model for the past 4500 years is given in Table 1. The age of the core-top has been estimated to be 1950 AD. The age control for box-core B05-2006-GBC03C (abbreviated as GBC03C) is based on 210Pb and 137Cs measurements of a multicore (B05-2006-MC04C) from the same site. This core has been dated using a modified CRS-model (Appleby, 2001). The results of this dating are given in Table 2. The lowest level with measurable content of 137Cs (22.5 cm) is dated to 1953 which is in excellent agreement with the first releases of this isotope in nature (1954), adding confidence to the dating result. There are irregularities in the topmost 10 cm of the unsupported 210Pb profile, which may be due to changes in sedimentation rate or to physical disturbances. In that case the dating of the uppermost part of the core is associated with some unknown error. Ages are expressed in Fig. 1. Map showing the location and core sites used in this study: MD99-2275 (66 33N; 1742W) and BO5-2006-GBC03C (66 33.18N; 1742.04W). The main surface currents are indicated by the arrows (modified after Hurdle, 1986). M.-A. Sicre et al. / Quaternary Science Reviews 27 (2008) 2041–2047 2043 the methodology and interpretation of the magnetic results can be found in Rousse et al. (2006). Table 2 Dating model used for the B05-2006-GBC03C box-core Depth (cm) Age (cal. BP) Age (AD) 0.5 1.5 2.5 3.5 4.5 5.5 6.5 7.5 8.5 9.5 10.5 11.5 12.5 13.5 14.5 15.5 16.5 17.5 18.5 19.5 20.5 21.5 22.5 55.2 53.6 52 50.7 49.3 48 46 44 42 39 35.5 32 27.25 22.5 17.75 13 9.7 6.3 3 0 3.3 6.7 10 2005 2004 2002 2001 1999 1998 1996 1994 1992 1989 1986 1982 1977 1973 1968 1963 1960 1956 1953 1950 1947 1943 1940 2.4. Alkenone determination 0 years Before Present (BP; with 0 year BP ¼ 1950 years AD) for the MD99-2275 core, and years AD for the GBC03C core. 2.3. Magnetic parameters Anysteretic remanent magnetization (ARM), isothermal remanent magnetization (IRM) and volumic low field susceptibility (k) were measured along the MD99-2275 core using a pass-through high-resolution cryogenic magnetometer with DC-SQUIDs that is housed in a shielded room at LSCE. As earlier discussed by Rousse et al. (2006), the magnetic mineralogy is uniformly made of magnetites, thus allowing to use ARM/k or ARM/IRM ratios to deduce changes in the magnetite grain size, and ARM, the concentration of fine grain magnetites. These parameters were thus used in the MD99-2275 core as proxies of bottom current changes to identify time span of major ocean circulation changes. Details on SSTs were calculated from the alkenone unsaturation index, UK 37 , which is now a well-established tool in paleoceanography (Conte et al., 2006), using the widely applied calibration of Prahl et al. (1988). The C37 alkenone distribution was dominated by the C37:3, with C37:4 representing less than 1% of the total C37 ketones, including for the coldest SST values of the record (w6  C), justifying 0 K further the choice of UK 37 rather than U37 to derive SSTs (Sicre et al., 2002; Bendle and Rosell-Melé, 2004). Chemical analyses were performed following the procedure described by Ternois et al. (1996). Briefly, lipids were extracted from 1.5 g of freeze-dried sediments by a mixture of CH3OH:CH2Cl2 (1:2 v/v) in an ultra-sonic bath for 15 min, then centrifuged at 1000 rpm for 15 min and transferred in a pear-shaped flask. Alkenones were isolated by silica-gel chromatography using solvent mixtures of increasing polarity. They were then analyzed by gas chromatography (GC) on a Varian 3400 CX Series gas chromatograph. The oven was temperature programmed from 100  C to 300  C at a rate of 20  C/ min. The temperature injector was programmed from 250  C to 260  C and of the Flame Ionization Detector (FID) set to 320  C. We used a 50 m long capillary fused silica column CPSil-5CB, with a 0.32 mm internal diameter and 0.25 mm film thickness. A known amount of 5a-cholestane was added to the alkenone fraction prior to GC injection for quantitation. Concentrations of the (C37:2 þ C37:3) alkenones varied from 0.02 to 7 mg/g, which fall within the range of coastal sediment levels, including upwellings (Martinez et al., 1996). 3. Results Water column studies, based on sediment traps and hydrocasts, have shown that alkenones in polar oceans are mainly produced during the summer season (Sikes et al., 1997; Ternois et al., 1998; Sicre et al., 2002). This implies that the SST reconstruction off North Iceland likely reflects summer conditions. Possible bias by advection of CwarmD detrital alkenones by surface currents was found to be negligible (Sicre et al., 2008). As shown in Fig. 2, the North 12 11 Alkenone SSTs in °C 10 9 8 7 6 tephras 5 4 0 1000 2000 3000 4000 Age in years BP Fig. 2. Alkenone derived Sea surface temperature (SST) estimates over the past 4500 years in the MD99-2275 core. The calibration established by Prahl et al. (1988) was used to 0 convert UK37 into SSTs. Black diamonds indicate tephra layers identified and used to build the age model. The red curve represents the 10-point running mean of the data. 2044 M.-A. Sicre et al. / Quaternary Science Reviews 27 (2008) 2041–2047 Icelandic SSTs over the last 4500 years vary from w11  C to w6  C with strong high frequency variability. These values are comparable to the summer SSTs reconstructed by Jiang et al. (2005) using diatom distributions over the last 2000 years in the same core. The 10-point running mean (red curve in Fig. 2) shows a warming tendency between w4200 year BP and w2500 year BP, followed by a cooling towards present, except for a 3–4 century duration warmer interval which includes the MWP. This climatic anomaly is distinguished by a stepwise increase of 1–1.5  C around 1000 year BP and an abrupt decline at w600 year BP, marking the onset of a cold period which encompasses the LIA. The temporal characteristics of the SST signal were quantified by a continuous wavelet analysis (using Morlet wavelet) (Yiou et al., 1996). Spectral power was also computed with a multi-taper method (e.g. Ghil et al., 2002) to estimate the statistical significance of frequency peaks. Results of these calculations indicate significant variability at multidecadal (centered w150 years) and to a lesser extent at bidecadal (20–25 years) time-scales, both of which are discussed in the next sections (Fig. 3). SSTs reconstructed over the last 70 years from the GBC03C boxcore vary from 7.5  C to 10.5  C and also depict short-term oscillations of similar amplitude (Fig. 4b). These estimates are consistent with the recent compilation of in situ data produced by Hanna et al. (2006), reporting that since 1874, July and August SSTs measured from the nearby Grimsey island have varied between 6.7  C and 9  C (see Table 3 in Hanna et al., 2006). However, the latter values are summer month averages over 20–25 year time periods, but higher frequency measurements indicate summer values occasionally reaching 11  C. From this comparison we can conclude that alkenones reliably reproduce the temperature range of instrumental data for summer. It is worthy to note that, on average, these values are w2  C above the MD99-2275 core-top values and comparable to the MWP, suggesting a warming of the surface waters over the last decades to century. The SST offset between the top of MD99-2275 and bottom of GBC03C may indicate that the two cores do not overlap, and that the top of the MD99-2275 is older than 1950 AD. Additional coring and dating of recent sediments will be necessary for the construction of a box-core MD99-2275 composite. 4. Discussion 4.1. Bidecadal variability Bidecadal variability in the North Icelandic SST signal has been reported as the dominant oscillation mode over the last 2000 years, though expressing intermittently, while multidecadal oscillations were poorly characterized (Sicre et al., 2008). In this extended record, multidecadal excursions are more significant, in particular between 4200 and 2500 year BP. Yet, the 20–25 year period remains important. Interestingly, SSTs reconstructed over the last 70 years from the GBC03C core show four major oscillations, each one of roughly 20-year duration (Fig. 4b). The first two cycles depict a slight cooling ending by a temperature minimum of 7.7  C, while Fig. 3. Results of spectral analyses of alkenone derived sea surface temperature times series over the past 4500 years, in the MD99-2275 core (a). (b) Continuous wavelet analysis of the data was performed using a Morlet wavelet (Yiou et al., 1996). (c) Spectral power was computed with a multi-taper method (Ghil et al., 2002) to estimate significance of peaks. Red noise tests were performed in order to assess the significance of the particular frequencies/periods present in the SST time series. The colored lines indicate the confidence interval for red noise tests. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) M.-A. Sicre et al. / Quaternary Science Reviews 27 (2008) 2041–2047 NAO index a 1925 3.0 1935 1945 1955 1965 1975 1985 1995 2005 2.0 Alkenone SSTs in˚C convection, which lead us to conjecture that bidecadal SST variability off North Iceland could be MOC driven. 4.2. Multidecadal variability 1.0 0.0 -1.0 -2.0 b 2045 11 10 9 8 GSA 7 1925 1935 1945 1955 1965 1975 1985 1995 2005 Age in years Fig. 4. Comparison of (a) the winter NAO index values smooth with a 10 year running mean and (b) alkenone sea surface temperature (SSTs) from the box-core BCO5-2006-GBC03C. the next two tend to warmer values, w10.6  C, in 1996. The severe cooling of the late 1960s coincides with the Great Salinity Anomaly (GSA), a massive intrusion of Arctic polar waters and sea ice. At that time, a high pressure anomaly had increased to a maximum over Greenland producing strong northerly winds over the Greenland and Iceland seas, thus intensifying the EGC and spreading drifting ice southward into the sub-polar Atlantic (Dickson et al., 1996). The concomitant northeast shift of the Icelandic low and storm activity northwest of Iceland were concurrent factors enhancing the inflow of Arctic waters. Conversely, in the following years the Icelandic low was centered over Iceland and South Greenland, and southerly winds were more frequent. According to the recent study of Logemann and Harms (2006), reduced northerly winds/enhanced southerly winds favor higher transport rates of Irminger Current across the Denmark Strait leading to positive SST anomalies of the North Iceland shelf waters, that could account for rising SSTs after the GSA. Over the last 70 years, the NAO index experienced a major decadal shift: values decrease from the 1930s to the late 1960s and then increase until the mid-1990s (Fig. 4a). The comparison between Fig. 4a and b suggests a link between decadal SST variations and decadal NAO variations. We hypothesize that the SST cooling from 1925 to 1968, and subsequent warming up to 1996 are induced by the low frequency NAO forcing. Intermittent 20–25 year period oscillations along the MD99-2275 record could subsequently reflect the ocean response to periods of sustained NAO forcing, either in positive or negative phases. However, the accuracy of the sedimentary age model does not allow discussing at the year level the lead/lad relationship between the two records. Bidecadal variability has also been reported in the Mg/Ca and oxygen isotope records of ostracode shells of the Late Holocene estuarine sediments of Chesapeake Bay and attributed to NAO by Cronin et al. (2005). In a recent study, Latif et al. (2006) have shown, using hydrographic data and model results, that low frequency variability of NAO can induce MOC changes through the Labrador Sea Significant variance in the 50–150 year band dominates in the w2500–4400 year BP time interval. Multidecadal variability has been identified in modelling studies (Delworth et al., 1993; Delworth and Greatbach, 2000) and also seems to be a robust feature in globally distributed instrumental and proxy records over several centuries (Schlesinger and Ramankutty, 1994; Kushnir, 1994; Kaplan et al., 1998). This mode of variability has been linked to the MOC internal variability, although the dominant period differs between models, i.e. centered on 50 years in Delworth et al. (1997) and of about 35 years in Timmermann et al. (1998). Recently, wavelet analysis of the natural variability of the MOC simulated over 1600 years by the HadCM3 model identified maximum variance at time-scale of 10–30 years and predominantly of 70–200 years (Vellinga and Wu, 2004), a result that comes close to our data analysis. Few marine Holocene records from the Nordic Seas have documented the occurrence of SST oscillations at centennial timescale (Jiang et al., 2005; Bendle and Rosell-Melé, 2007). Yet, the lower temporal resolution of these time series, which additionally are based on 14C dating in a region of important reservoir age variability, prevents from decadal to sub-centennial scale correlations and investigation of forcing mechanisms of the ocean dynamics. In the Norwegian sea, where reservoir age variations are less significant, no major SST variations are recorded at the scale of the last 4000 years (Calvo et al., 2002; Risebrobakken et al., 2003) emphasizing the extreme sensibility of the northern Iceland oceanic domain to climatological changes. The high-resolution titanium (Ti) record of the ODP 1002 core from the anoxic Cariaco Basin of the Southern Caribbean (10 42N, 6510W, 893 m) is of interest for comparison to our record. This low latitude climatic signal has been interpreted as a proxy of rainfall/ fluvial inputs to the basin, driven by the N–S movements of the inter-tropical Convergence Zone (ITCZ) (Haug et al., 2001). Lower %Ti values indicate drier conditions and a more southern position of the ITCZ, while higher %Ti occur during wet conditions, when the ITCZ is more northern. Comparison with our data shows that major shifts of the ITCZ are contemporaneous to the large variations of the magnetic parameters (ARM and ARM/k) and multidecadal SST oscillations (Fig. 5). Episodes of coarser sediments, indicative of more vigorous bottom currents, coincide with a northerly position of the ITCZ, and a stronger phase of MOC. Enhanced transport by MOC creates a cross-equatorial SST gradient which causes the displacement of the ITCZ to the North (Vellinga et al., 2001). This shift in turn generates a negative freshwater anomaly that propagates to high latitude sinking regions thus starting to reverse the process, i.e. to slowdown the MOC, and completing a multidecadal oscillation. The large amplitude multidecadal SST variations, between 4200 and 2500 year BP, could thus be MOC driven and reflect alteration of the hydrological cycle in the tropical Atlantic. The MOC is sensitive to perturbations of the freshwater balance in the tropics because it influences the surface ocean density of the tropical Atlantic, and can then trigger climate variations at high latitudes (Vellinga and Wu, 2004). Today, there is a net export of freshwater from the tropical Atlantic to the Pacific Ocean through the atmosphere. According to Schmittner et al. (2000), this export flux can be modified by ENSO (El Niño Southern Oscillation) and induce changes in the Atlantic MOC. During El Niño years, more freshwater would be exported from the Atlantic to the Pacific, while during La Niña years this water export is decreased. The sensitivity experiments performed by Schmittner et al. (2000) using a coupled ocean–atmosphere model indicate that MOC is increased for larger 2046 M.-A. Sicre et al. / Quaternary Science Reviews 27 (2008) 2041–2047 12 0 1 2 3 4 a 10 0.55 8 0.45 0.35 6 b 0.25 North 4 0.15 ITCZ 0.05 Cariaco basin, 10˚42N; 65W ARM/k South 0 c 400 150 800 200 1000 d 1200 ARM (10-3 A/m) 600 ARM / K Less fine grains 200 100 250 Fine grains ARM (10-3 A/m) 0 50 Coarse grains % Titamium Alkenone SSTs in °C MD99-2275, 66˚33N; 17˚42W More fine grains 1400 300 1600 350 0 1 2 3 4 Age in kyears BP Fig. 5. Comparison of (a) sea surface temperatures derived from alkenone over the last 4500 year from MD99-2275 core, (b) bulk titanium content of the Cariaco basin sediments from the ODP site 1002 (Haug et al., 2001), (c) the pink curve shows the anysteretic remanent magnetization (ARM) values, and (d) the dark blue curve plots show the anysteretic remanent magnetization over and volumic low field susceptibility (k), (ARM/k) ratio values, from Rousse et al. (2006). Blue shades areas indicate time span of large fluctuations in the proxy records. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) freshwater export out of the Atlantic, i.e. for stronger El Niño and/or longer persistence of the perturbation (i.e. in a mean climate state shifted to more frequent El Niño events). To investigate further the possible role of ENSO on the MOC variability during the mid-Holocene, our data were compared to the high-resolution record of storm-derived deposits from Laguna Pallcacocha (Ecuador), which provides a unique high-resolution description of El Niño events over the last 15,000 years (Rodbell et al., 1999). This record establishes that the periodicity of El Niño was >15 years in Late glacial to early Holocene. Modern El Niño periodicity of 2–8.5 years became most apparent after 5000 years, in particular between 5000 and 2500 year BP when both strength and frequency increased (see Fig. 5 in Rodbell et al., 1999). As discussed by Haug et al. (2001), the large oscillations seen in the %Ti record at Cariaco at the time of this marked increase of ENSO activity could reveal a dynamical link between the two regions. Warm ENSO phases, by increasing freshwater export from the Atlantic would have generated salinity anomalies in the tropical Atlantic, thus affecting high latitudes sinking regions and subsequently the MOC and ITCZ dynamics. We speculate that the strong multidecadal SST variability off North Iceland between 4200 and 2500 years results from the hydroclimatic links between ENSO, the North Atlantic freshwater budget and the MOC. 5. Conclusions In this study, we generated a unique 4500-year SSTs record at 2–5-year temporal resolution for the high-latitude North Atlantic, in the oceanographic Polar Front area off North of Iceland, to explore ocean variability at decadal time-scale from mid- to Late Holocene. Spectral analysis of this signal reveals two modes of strong variance, at bi- and multidecadal time-scales, with strong and weak phases along the record. Our results suggest that low frequency NAO forcing could be responsible for the bidecadal variability of SSTs in our record. However, the occurrence of large amplitude multidecadal oscillations of SSTs between 4200 and 2500 year BP, coeval to major shifts of the paleomagnetic parameters requires a different explanation. The distant high-resolution %Ti record from Cariaco Basin and the gray-scale of Pallcacocha sediments suggest a dynamic link between the hydrological cycle of the tropical Atlantic/Pacific oceans and the high-latitude North Atlantic variability. Enhanced frequency and strength of El Niño warm phases, by increasing the freshwater export flux from the Atlantic, could have triggered MOC variability during this period of the mid-Holocene. Acknowledgements We are grateful to the IPEV for logistical assistance during the IMAGES cruise in 1999 and to the crew of the research vessel Marion Dufresne, with a special mention to Yvon Balut for the development and operation of the Calypso corer during all these years. We also thank CNRS (Centre National de la Recherche Scientifique) for salary support. This paper is a contribution of the PACTHOL project funded by the French LEFE program supported by INSU (Institut National de l’Univers), the PACLIVA project funded by the European Union 5th Framework Programme (Contract EVK22002-00143), and the Millennium project funded by the European M.-A. Sicre et al. / Quaternary Science Reviews 27 (2008) 2041–2047 Union 6th Framework Programme (Contract No. EVK-CT-2006017008). This is LSCE contribution number 3124. References Appleby, P.G., 2001. Chronostratigraphic techniques in recent sediments. In: Last, W.M., Smol, J.P. (Eds.), Tracking Environmental Change using Lake Sediments. Basin Analysis, Coring and Chronological Techniques, vol. 1. Kluwer Academic Publishers, The Netherlands. 0 K Bendle, J., Rosell-Melé, A., 2004. Distributions of UK 37 and U37 in the surface waters and sediments of the Nordic Seas: implications for paleoceanography. Geochem. Geophys. Geosyst. 5. doi:10.1029/2004GC000741 Q11013. Bendle, J., Rosell-Melé, A., 2007. High-resolution alkenone sea surface temperature variability on the North Icelandic Shelf: implications for Nordic Seas palaeoclimatic development during the Holocene. The Holocene 17, 9–24. Black, D.E., Abahazi, M.A., Thunnel, R.C., Kaplan, A., Tappa, E.J., Peterson, L.C., 2007. An 8-century tropical Atlantic SST record from the Cariaco Basin: baseline variability, twentieth-century warming, and Atlantic hurricane frequency. Paleoceanography 22. doi:10.1029/2007001427 PA4204. Calvo, E., Grimalt, J., Jansen, E., 2002. High-resolution UK 37 sea surface temperature reconstruction in the Norwegian Sea during the Holocene. Quat. Sci. Rev. 21, 1385–1394. Conte, M.H., Sicre, M.-A., Rühlemann, C., Weber, J.C., Shultz-Bull, D., Blanz, T., 2006. 0 Global temperature calibration of the alkenone unsaturation index (UK 37 ) in surface waters and comparison with surface sediments. Geochem. Geophys. Geosyst. 7 (2). doi:10.1029/2005GC001054. Cronin, M.T., Dwyer, G.S., Kamiya, T., Schwede, S., Willard, D.A., 2003. Medieval Warm Period, Little Ice Age and 20th century temperature variability from Chesapeake Bay. Global Planet. Change 36, 17–29. Cronin, M.T., Thunell, R., Dwyer, G.S., Saenger, C., Mann, M.E., Vann, C., Seal II, R.R., 2005. Multi-proxy evidence of Holocene climate variability from estuarine sediments, eastern North America. Paleoceanography 20. doi:10.1029/ 2005PA001145. Delworth, T.L., Greatbach, R.J., 2000. Multidecadal thermohaline circulation variability driven by atmospheric surface flux forcing. J. Climate 13, 1481–1495. Delworth, T.L., Manabe, S., Stouffer, R.J., 1993. Interdecadal variations of the thermohaline circulation in a coupled ocean–atmosphere model. J. Climate 6, 1993–2011. Delworth, T.L., Manabe, S., Stouffer, R.J., 1997. Multidecadal climate variability in the Greenland Sea and the surrounding regions: a coupled model simulation. Geophys. Res. Lett. 24, 257–260. Dickson, R., Lazier, J., Meincke, J., Rhines, P., Swift, J., 1996. Long-term coordinated changes in the convective activity of the North Atlantic. Prog. Oceanogr. 38, 241–295. Eirı́ksson, J., Larsen, G., Knudsen, K.L., Heinemeier, J., Sı́monarson, L., 2004. Marine reservoir age variability and water mass distribution in the Iceland Sea. Quat. Sci. Rev. 23, 2247–2268. Eirı́ksson, J., Bartels-Jónsdóttir, H.B., Cage, A.G., Gudmundsdóttir, E.R., KlitgaardKristensen, D., Marret, F., Rodrigues, T., Abrantes, F., Austin, W.E.N., Jiang, H., Knudsen, K.-L., Sejrup, H.-P., 2006. Variability of the North Atlantic Current during the last 2000 years based on shelf bottom water and sea surface temperatures along an open ocean/shallow marine transect in Western Europe. The Holocene 16 (7), 1017–1029. Ellison, R.C.W., Chapman, M.R., Hall, I.R., 2006. Surface and deep ocean interactions during the cold climate event 8200 years ago. Science 312, 1929–1932. Ghil, M., Allen, M.R., Dettinger, M.D., Ide, K., Kondrashov, D., Mann, M.E., Robertson, A.W., Saunders, A., Tian, Y., Varadi, F., Yiou, P., 2002. Advanced spectral methods for climatic time series. Rev. Geophys. 40 (1), 1–41. Hanna, E., Jonsson, T., Olafsson, J., Vladimarsson, H., 2006. Icelandic coastal sea surface temperature records constructed: putting the pulse on Air–Sea–Climate interactions in the Northern North Atlantic. Part I: comparison with HadISST1 open-ocean surface temperatures and preliminary analysis of long-term patterns and anomalies of SSTs around Iceland. J. Climate 19, 5652–5666. Haug, G.H., Hughen, K.A., Sigman, D.M., Peterson, L.C., Röhl, U., 2001. Southward migration of the intertropical convergence zone through the Holocene. Science 293, 1304–1308. Hurdle, B.G., 1986. The Nordic Seas. Springer Verlag, New York, 777 pp. Hurrell, J.W., 1995. Decadal trends in the North Atlantic Oscillation: regional temperature and precipitation. Science 269, 676–679. Jiang, H., Eirı́ksson, J., Schulz, M., Knudsen, K.-L., Seidenkrantz, M.-S., 2005. Evidence for solar forcing of sea-surface temperature on the North Icelandic Shelf during the late Holocene. Geology 33, 73–76. 2047 Kaplan, A., Cane, M.A., Kushnir, Y., Clement, A.C., Blumenthal, M.B., Rajagopalan, B., 1998. Analyses of global sea surface temperature 1856–1991. J. Geophys. Res. 103, 18,567–18,589. Keigwin, L.D., 1996. The Little Ice Age and Medieval warm period in the Sargasso Sea. Science 274, 1504–1507. Keigwin, L.D., Pickart, R.S., 1999. Slope water current over the Laurentian Fan on interannual to millennial time scale. Science 286, 520–523. Kushnir, Y., 1994. Interdecadal variations in North Atlantic Sea surface temperature and associated atmospheric conditions. J. Climate 7, 141–157. Larsen, G., Eirı́ksson, J., Knudsen, K.L., Heinemeier, J., 2002. Correlation of late Holocene terrestrial and marine tephra markers, north Iceland: implications for reservoir age changes. Polar Res. 21, 283–290. Latif, M., Böning, C., Willebrand, J., Biastoch, A., Dengg, J., Keenlyside, N., Schweckendiek, U., Madec, G., 2006. Is the thermohaline circulation changing? J. Climate 19, 4631–4637. Logemann, K., Harms, I.H., 2006. High resolution modelling of the North Icelandic Irminger Current (NIIC). Ocean Sci. Discuss. 3, 1149–1189. Lund, D.C., Lynch-Stieglitz, J., Curry, W.B., 2006. Gulf Stream density structure and transport during the past millennium. Nature 444, 601–604. Martinez, P., Bertrand, P., Bouloubassi, I., Bareille, G., Shimmield, G., Vautravers, B., Grousset, F., Guichard, S., Ternois, Y., Sicre, M.-A., 1996. An integrated view of inorganic and organic biogeochemical indicators of palaeoproductivity changes in a coastal upwelling area. Org. Geochem. 24, 411–420. deMenocal, D., Ortiz, J., Guilderson, T., Sarthnein, M., 2000. Coherent high and low-latitude variability during the Holocene warm period. Science 288, 2198–2202. Østerhus, S., Turrell, W.R., Jonsson, S., Hansen, B., 2005. Measured volume, heat, and salt fluxes from the Atlantic to the Arctic Mediterranean. Geophys. Res. Lett. 32, doi:10.1029/2004GL022188. Prahl, F.G., Muehlhausen, L.A., Zahnle, D.L., 1988. Further evaluation of long-chain alkenones as indicators of paleoceanographic conditions. Geochim. Cosmochim. Acta 52, 2303–2310. Risebrobakken, B., Jansen, E., Andersson, C., Mjelde, E., Hevrøy, K., 2003. A highresolution study of Holocene paleoclimatic and paleoceanographic changes in the Nordic Seas. Paleoceanography 18 (1), 1017. doi:10.1029/2002PA000764. Rodbell, D.T., Seltzer, G.O., Anderson, D.M., Abbott, M.B., Enfield, D.B., Newman, J.H., 1999. An w15,000-year record of El Niño-driven alluviation in Southwestern Ecuador. Science 283, 516–520. Rousse, S., Kissel, C., Laj, C., Eirı́ksson, J., Knudsen, K.L., 2006. Holocene Centennial to Millennial-scale climatic variability: evidence from high-resolution magnetic analysis of the last 10 cal kyr off North Iceland. Earth Planet. Sci. Lett. 242, 390–405. Schlesinger, M., Ramankutty, N., 1994. An oscillation in the global climate system of period 65–70 years. Nature 367, 723–726. Schmittner, A., Appenzeller, C., Stocker, T.F., 2000. Enhanced Atlantic freshwater export during El Niño. Geophys. Res. Lett. 27, 1163–1166. Sicre, M.-A., Bard, E., Ezat, U., Rostek, F., 2002. Alkenone distributions in the North Atlantic and Nordic sea surface waters. Geochem. Geophys. Geosyst. 3 (2), 1–13. 2001GC000159. Sicre, M.-A., Jacob, J., Ezat, U., Rousse, S., Kissel, C., Yiou, P., Eirı́ksson, J., Knudsen, K.L., Jansen, E., Turon, J.-L., 2008. Decadal variability of sea surface temperatures off North Iceland over the last 2000 yrs. Earth Planet. Sci. Lett. doi:10.1016/j.epsl.2008.01.011. Sikes, E.L., Volkman, J.K., Robertson, L.G., Pichon, J.-J., 1997. Alkenones and alkenes in surface waters and sediments of the Southern Ocean: implications for paleotemperature estimation in polar regions. Geochim. Cosmochim. Acta 61, 1495–1505. Ternois, Y., Sicre, M.-A., Boireau, A., Marty, J.-C., Miquel, J.C., 1996. Production pattern of alkenones in the Mediterranean Sea. Geophys. Res. Lett. 23, 3171–3174. Ternois, Y., Sicre, M.-A., Boireau, A., Beaufort, L., Miquel, J.-C., Jeandel, C., 1998. Hydrocarbons, sterols and alkenones in sinking particles in the Indian sector of the Southern Ocean. Org. Geochem. 28, 489–501. Timmermann, A., Latif, M., Voss, R., Grötzner, A., 1998. Northern hemispheric interdecadal variability: a coupled air–sea mode. J. Climate 11, 1906–1931. Vellinga, M., Wood, R.A., Gregory, J.M., 2001. Processes governing the recovery of a perturbed thermohaline circulation in HadCM3. J. Climate 15, 764–780. Vellinga, M., Wu, P., 2004. Low-latitude freshwater influence on centennial variability of the Atlantic Thermohaline circulation. J. Climate 17, 4498–4511. Visbeck, M., 2002. The ocean’s role in Atlantic climate variability. Science 297, 2223–2224. Yiou, P., Baert, E., Loutre, M.F., 1996. Spectral analysis of climate data. Surv. Geophys. 17, 619–663.