Patterns of climate variability in the north atlantic

Deliverables

Re-analysis data covering the period 1958 2001 were used to determine the relationship between regional, inter-annual snow accumulation variability over the Greenland Ice Sheet (GrIS) and large scale circulation patterns, cyclone frequency, and strength. Four regions of the GrIS have been identified that are highly independent with respect to accumulation variability. This result allows precipitation to be reconstructed over a longer period. Accumulation indices of three of these regions are associated with distinct large-scale circulation patterns: Central-western GrIS reveals an inverse relationship with a NAO-like pattern, the south-west a positive correlation with a high pressure bridge from central North Atlantic to Scandinavia, and the south-eastern GrIS a positive correlation with a high-pressure anomaly over the Greenland Sea. These large scale patterns also impact European climate in different ways. Accumulation variability in north-eastern GrIS, however, is dominated by cyclones originating from the Greenland Sea. Thus, Greenland ice core accumulation records offer the potential to reconstruct various large-scale circulation patterns and regional storm activity. The annually resolved ice core records from different regions over the GrIS were used to investigate spatial and temporal variability of calcium (Ca++, mainly from aeolian dust) and sodium (Na+, mainly from sea salt) deposition. Cores of high common inter-annual variability are grouped with an EOF analysis, resulting in regionally representative Ca++ and Na+ records for northeastern and central Greenland. Utilizing a regression and validation method including the ERA40 reanalysis data set, these common records are associated with distinct regional atmospheric circulation patterns over the North American Arctic, Greenland and Central to Northern Europe. In the north-eastern part of the GrIS a large fraction of the Ca++ variability is connected to a circulation pattern suggesting transport from the west and probably dry deposition. This pattern is consistent with the current understanding of a predominantly Asian source of the dust deposited over the GrIS. However, our results also indicate that a significant and dominant fraction of the inter-annual dust variability in NE and Central Greenland, respectively, are determined by the complex effect of frequency and intensity of wet deposition (including a potentially dominating effect of fog) during the spring season with high atmospheric dust loading, rather than representing the variability of the Asian dust source and/or long-range transport to Greenland. The variances in the regional proxy records explained by the streamfunction patterns are high enough to attempt reconstructions of the corresponding regional deposition regimes and the associated circulation patterns. Further, UNIBE collaborated with US colleagues on a study about the modes of temporal variability recorded in nitrate and accumulation records from six Greenland ice cores (Burkhart et al., revised manuscript submitted). Nitrate records from six Greenland ice cores covering the period 1794 to 1995 show a significant correlation in concentration for averaging periods greater than 10 years, as well as an approximately 60 percent increase in average concentration during the last 75 years. Annual nitrate fluxes contain low-frequency trends driven primarily by changes in concentration, while higher-frequency variability is driven by changes in snow accumulation. During the period of anthropogenic influence, nitrate is positively correlated with the North Atlantic Oscillation, while prior to that the correlation is negative, but less significant.

For the time period 1500-2000 a large number of high-quality climate proxy information and early instrumental station data exist for the European continent. This allows the reconstruction of time series of past climate variability and gridded climate information. Casty et al. (2005) reconstructed independent monthly European temperatures, precipitation and 500 hPa geopotential height fields back to 1766. Pauling et al. reconstructed seasonal European precipitation back to 1500. The approach is based on a Principal Component Regression, where the predictand is the gridded climate field information over Europe and the predictors are instrumental measurements or climate proxy information for the pre-instrumental period. They showed that it is possible to statistically reconstruct gridded climate parameter fields over the North Atlantic and European region for at least the last 500 years. Precipitation variability over the last half millennium reveals both large interannual and decadal fluctuations. Applying running correlations, major non-stationarities in the relation between large-scale circulation and regional precipitation are found. For several periods during the last 500 years, key atmospheric modes for southern Spain/northern Morocco and central Europe as representations of two precipitation regimes were identified. Using scaled composite analysis, we show that precipitation extremes over central Europe and southern Spain are linked to distinct pressure patterns. Due to its high spatial and temporal resolution, this dataset allows detailed studies of regional precipitation variability for all seasons, impact studies on different time and space scales, comparisons with high-resolution climate models as well as analysis of connections with regional temperature reconstructions. Casty et al. (2005) presented so-called combined winter climate regimes for the 1766 2000 period over the North Atlantic/European sector. They expand previous studies on recurrent regimes by combining spatially high-resolution independent reconstructions of the 500 hPa geopotential height, land surface air temperature, and precipitation fields. Nonlinear Principal Component Analysis was applied to the data in order to account for the underlying nonlinear dynamics of climate regimes. Three recurrent winter climate regimes are detected. One regime resembles in its pressure, temperature, and precipitation pattern the positive phase of the North Atlantic Oscillation (NAO), whereas the other two regimes are European blocking situations. Raible et al. (2005) investigated the decadal trend behaviour of the Northern Hemisphere atmospheric circulation using long-term simulations with different state-of-the-art coupled general circulation models (GCMs) for present-day climate conditions (1990), reconstructions of the past 500 years, and instrumental station observations. The multi-model simulations show that strong positive winter NAO trends are connected with the underlying sea surface temperature (SST) and exhibit a SST tripole trend pattern and a northward shift of the storm-track tail. Strong negative winter trends of the Aleutian low are associated with SST changes in the El Niño Southern Oscillation (ENSO) region and a westward shift of the storm track in the North Pacific. The observed simultaneous appearance of strong positive NAO and negative Aleutian low trends is very unlikely to occur by chance in the unforced simulations and reconstructions. The positive winter NAO trend of the last 50 years is not statistically different from the level of internal atmosphere ocean variability. The unforced simulations also show a strong link between positive SST trends in the ENSO region and negative Aleutian low trends. With much larger observed SST trends in the ENSO region, this suggests that the observed negative Aleutian low trend is possibly influenced by external forcing, for example, global warming, volcanism, and/or solar activity change. The above studies all aim at improving the understanding of the past climate variability of the North Atlantic realm. However, the gridded climate field reconstructions for this region are, until now, restricted to cover the last half millennium. Using the data collected in the project PACLIVA, it should be possible to extend such climate field reconstructions further back in time. It is therefore very important to obtain synthesised and homogenised marine sediment records and ice cores for the North Atlantic region as input variables for the mentioned Principal Component Regression approach to estimate gridded SST of the North Atlantic during the Holocene. Such methods, among others, could pave the way towards new spatial climate reconstructions on much longer time scales than possible today. This is a very challenging task due to difficulties in the dating, temporal resolution, or signal detection of each climate proxy record collected in PACLIVA.

Statistical analyses were used to characterise rainfall patterns in the Antilles, and their relationship to remote climate patterns. The Antilles archipelago stretches between Florida and eastern Venezuela. Rainfall climate in the Caribbean can be classified as dry-winter tropical with significant subregional variation in total annual rainfall, length of the rainy season, and timing of rainfall maxima. The North Atlantic Oscillation (NAO) and El Niño-Southern Oscillation (ENSO) are known to influence precipitation patterns in the North Atlantic Basin. Many meteorological stations in the Caribbean have limited temporal scope, and are not complete. The analysis of 35 stations from the time interval 1951-1981 provided an optimum balance between geographic and temporal scope. Time-series from these stations were subject to multivariate-statistical analysis in order to unravel regional precipitation climate, and to investigate the response of each subregion to NAO and ENSO fluctuations. The two-stage least squares (TSLS) factor analysis (FA) method was applied to identify distinct subregions that are each marked by common patterns of fluctuations through this time-interval, the geographical coherency of these regions, and differences in the seasonal precipitation variability among them. An advantage of TSLS FA is the possibility to compute standard errors for the factor loadings that are useful to assess the significance of the contribution of a variable to a particular factor. For analyses of the associations between the precipitation patterns and NAO the mean December-March NAO index (NAOI) for the time period 1826-2002 was divided into years with a low NAOI (lowest tercile of the NAOI) and high NAOI (highest tercile). For analysis of the effect of ENSO on precipitation the Niño 3.4 region, which has been shown to essentially characterize the ENSO behaviour, was used. A positive SST anomaly in of >0.4ºC in N3.4 represents a warm ENSO event, and a negative anomaly of >0.4ºC represents a cold ENSO event. The statistical significance of the differences between low and high NAO events, and warm and cool ENSO events, for different months was tested using t-tests. The problem inherent in multiple testing of the H0 and H1 hypotheses in climatology was considered in evaluations of the results. Application of TSLS FA indicated that four distinct patterns of precipitation fluctuations existed in the Caribbean through 1951-1981: Regime 1 - Westernmost Caribbean (western Cuba and western Bahamas), southernmost Dominican Republic, westernmost Puerto Rico. This is the most geographically disjunct region. Regime 2 - Eastern Cuba and the eastern parts of the Bahamas. Regime 3 - Northern Dominican Republic, northwesternmost part of Puerto Rico. Regime 4 - The rest of Puerto Rico, Lesser Antilles (U.S. Virgin Islands, St. Martin, and Barbados). These regimes have distinctly different seasonal precipitation patterns. Regime 1 is characterized by high rainfall in the 6-month period from May through October. Regime 4 is marked by low precipitation amounts through the winter and early spring (January-April) and a maximum peak in September, October, or November. Regimes 2 and 3 differ from Regime 1 by relatively lesser May-October and greater November-December precipitation and from Regime 4 by lesser May-December precipitation. Regime 3 is marked by more even distribution of rainfall throughout the year in comparison with Regime 2. Regime 2 stands out from the other regimes by lower total annual precipitation. The analyses of differences in precipitation between low and high NAO years for each of the four precipitation regimes indicated no effect of the NAO on precipitation in Regimes 1 and 2. In contrast, the analysis indicates significantly enhanced May precipitation during low NAO years in Regimes 3 and 4. Similarly, July rainfall is increased during low NAO years in Regime 3. For example, nearly 700 mm/yr more rain fell over all stations in Regime 3 in low NAO years than in high NAO years. A further analysis of stations with more than 30 years of continuous observations indicates that the pattern observed is not just a feature characteristic of the period 1951-1981. ENSO control of precipitation in the Antilles is also documented by the study. In Regime 1 most of the first five months of the year (all except February) are controlled by ENSO phases during parts of the previous year or the same calendar year. Regime 2 is relatively little affected by ENSO. Similarly to Regime 1, increased January precipitation in Regime 3 is related to warm ENSO phases the previous year. May is wetter during February-May warm ENSO phases. In Regime 4 increased July precipitation is associated with warm ENSO phases during the fall of the previous year and the spring of the same year. July-August cold ENSO phases cause enhanced rainfall in October.

Alkenones concentration have been measured on several sediment cores from the north Atlantic and the western Mediterranean Sea, and used to reconstruct Holocene sea surface temperature changes. The main cores used were ENAM9606 in Feni Drift, MD932343 in the Norwegian Sea and MD952343 in the western Mediterranean Sea, close to Menorca. Their composition has been compared with that obtained in other cores by other partners. Core ENAM9606 exhibits a long-term cooling during the last 10 kyr, ranging form 14ºC to 12ºC. This trend is consistent with insolation with maximal values between 11 kyr and 5.5 kyr BP and is consistent with those of the other cores included in Fig. 1 and with several continental records from the circum North Atlantic providing evidence of a long-term cooling during the Holocene. ENAM9606 can be divided in three intervals: early-Holocene (from 10 kyr to 7.7 kyr BP), mid-Holocene (form 7.7 kyr to 4 kyr BP) and late-Holocene (from 4 kyr BP to present days). The former exhibits the highest values from 14ºC to 12.7ºC. From 10 kyr to 9 kyr the SST amplitude changes about one degree every 150 years approximately, punctuated by some cooling events at 9.7 kyr, 9.4 kyr and 9.2 kyr BP. Then, SST stabilizes at around 13.5ºC between 9.2 kyr and 8.4 kyr BP punctuated by a cooling at 8.3 kyr BP. During the mid-Holocene the SST decreases from 13.7ºC to present-day values (12ºC). Finally the late-Holocene is the period that exhibits the major variability compared with the early-Holocene, although the amplitude of the SST changes is higher for the first part of the Holocene. From 1900 yr BP to present-days, the SST decreases until 12.2ºC exhibiting great variability. The most pronounced involves a drop of about 1ºC at 160 yr BP in the LIA. This event occurs nearly synchronously with temperature changes in Chesapeake Bay and humidity variations in Cariaco Basin. In the Norwegian Sea SST changed between 10.6 and 12.7 ºC. The highest values were observed in the interval between 6000-8000 years and after this period they decreased slowly to 12ºC as average with some episodes of lower values 2700 years BP. The temperature differences between 600 years and 2000 years BP ranged between 10.6ºC and 11.4ºC. In the Menorca core a significant SST drop from 17.6ºC at 10000 years BP to 13.74ºC at 62 years BP is observed. This decrease is consistent with the above mentioned SST decreases in the Holocene but this is the one involving highest difference among the cores included in the present project. The period between 6000 and 8000 years exhibit SST ranging between 15.8ºC and 17.6ºC with two marked drops at 7680 and 7205 years BP. Another marked drop occurred in 4753 years BP involving a change between 17ºC to 15.3ºC. The SST profile of this core is characterized by a strong well marked SST decrease between 1885 and 2530 years BP (13.9ºC) involving a decrease of 1.4ºC. Temperature changes in the order of 0.9-1.5ºC are observed in the last 2000 year period. The SST estimates in all cores studied shows a decrease from the beginning of the Holocene to present days. These high resolution studies confirm previous findings and reinforce the general trend in the marine environment. This trend is also observed in the continental masses, e.g. Mann et al., Nature, 392, 779-787 (1998), and likely constitutes one of the most robust trends of the climate evolution of the Holocene. In contrast, in the last 2000 years, the expected SST changes for the MWP and the LIA are observed in some of the cores studied but not in all. In some cases, e.g. ENAM9606 there are features that can be related to these climate episodes but they do not expand over equivalent time periods as in the continent. The marine records in which SST changes have been observed to exhibit a better defined pattern in agreement with these two episodes are located nearby the coast of Portugal (namely MWP) and nearby Iceland (core MD992270). Climate in the marine environments of these two records SST was maybe under significant influence of the nearby continental masses.

Alkenone analyses were performed along the core MD99-2203 (34°58 38N, 75°12 06 W, 620 meters depth), to determine changes in sea surface temperatures off Cape Hatteras on the Gulf Stream path. This time series covers the last 2000 years and complement the Mg/Ca ratios measured in the planktonic foraminifera G. ruber white.

There is coastal upwelling along the NW coast of Africa, and the Iberian Margin. Satellite SST data show a strip of cold coastal waters, 3-5°C colder than open ocean conditions, in the Cape Ghir upwelling centre, with an upwelling filament extending offshore. The Holocene record of SSTs in this region was reconstructed at high-resolution from three cores using alkenone paleothermometry Core GeoB6008-1 and short core GeoB6008-2 were recovered from 355m depth off the Moroccan coast (30°50.7 N, 10°05.9 W) and were used to estimate SST variability for the last 2000 years. Age control for GeoB6008-2 was established by 210Pb dating, and for GeoB6008-1 by both 210Pb and accelerator mass spectrometer (AMS) 14C dating. The upper part of SST reconstructions from GeoB6008-1 and GeoB6008-2 overlap with the instrumental period for the last 90 years of the 20th century, and, taken together, they show a steady cooling trend of approximately 1.2°C over this time. In addition, the pattern of variability for the 50 years where the two cores overlap (1912-1971AD) are remarkably similar, attesting to the consistency of the age model in this part of the core, and of the SST signal in each record. When viewed in the context of the gravity core record for the last 2.5 millennia, the last 100 years of the GeoB6008 alkenone records shows that the strong decrease in SST observed for the last century is larger and more rapid than any other change seen in the record and is 0.3°C colder than the 19th century and 0.6°C colder than the next coolest period from 1050-1150AD. The 1050-1150AD period corresponds to the middle of the so called Medieval Warm Period (MWP; 850-1300AD). Relative to the MWP the alkenone SST records show a warm Little Ice Age (LIA; 1400-1850). Prior to the 20th century, the GeoB6008 record shows pronounced millennial variability. Core GeoB5901-2 was recovered from 574m depth on the Iberian margin (36.380°N 7.071°W) and was used for high-resolution SST reconstructions during the 6000-8000 year period. Age control is based on AMS 14C dates. The alkenone SST reconstruction from shows a cooling trend of 1.3°C from 10000 to 1000 cal yrs BP. The warmest SSTs of 20.3°C occur in the early Holocene. Superimposed on the long-term trend is a multi-millennial scale oscillation, with relative peaks at 9000-9500, 6000 and 2500 cal yr BP. The GeoB6008 cores come from the heart of the Cape Ghir upwelling system, where the upwelling is driven by alongshore winds forcing offshore Ekman transport, compensated by the upwelling of cooler, nutrient-rich water. The alkenone SST proxy record from the GeoB6008 cores are interpreted as upwelling histories for Cape Ghir, and the decrease in SSTs for the 20th Century suggests an intensification of upwelling. The Cape Ghir SST record, including the rapid temperature changes of the last 100 years and the millennial scale variability, shows similar characteristics, though with opposite sign, to the Northern Hemisphere temperature Anomaly reconstructions. For example, from 1450-1850AD, the LIA, relative warmth is observed in the GeoB6008 SST record compared with relative cooling in the NHTA records. Both the high decadal-scale resolution and the overlap with the 20th century instrumental/anthropogenic period make the SST reconstructions for the GeoB6008 cores a unique dataset. Through this core we are able to learn about the global warming response of ecologically and socially important marine upwelling systems and further our understanding of recent climate variability. A high-resolution multi-proxy study reconstructed SST, river discharges and biological productivity in the Tagus deposition centre off Lisbon (Portugal) for the last 2,000 years. Century-scale SST variability of 2ºC allowed the identification of MWP and the LIA. High Fe and fine-sediment deposition with high n-alkane concentrations and freshwater diatoms during the LIA suggested increases of river discharge, whereas higher alkenone concentrations indicate increased river-induced productivity. During the MWP larger mean-grain size and low magnetic susceptibility, Fe, n-alkanes, and n-alcohols reflect decreased runoff. Increased benthic and planktonic foraminifera and upwelling related diatoms pointed to increased oceanic productivity. Based on the correlation between negative NAO phases and intensified Tagus River discharge over the last century, it was hypothesized that the increased influx of terrigenous material during the LIA reflected a negative NAO-like state or the occurrence of frequent extreme NAO minima. During the MWP, stronger upwelling was attributed to a persistent positive NAO-like state or the frequent occurrence of extreme NAO maxima. A peak in magnetic susceptibility was interpreted the 1755 AD Lisbon earthquake. This and the accompanying tsunami were estimated to have caused the loss of 39 cm of sediment and the instantaneous deposition of a 19 cm.

High-resolution magnetic analysis has been performed on the top 25 m of core MD99 2275 (66°33.06N, 17°41.59W; 440 m water depth) from the mid-outer shelf off North Iceland. This is a key boundary region for climate changes related to atmospheric and oceanographic variations influenced by the moving limit between the cold and warm water masses at the Polar Front and the low depression track with its associated westerlies situated across Iceland. An age model based on tephrochronology was previously reported for this core and has allowed precise timing of significant variations in the magnetic mineral concentration and grain size over the last 10 cal kyr. The Holocene climatic optimum identified in North Iceland between 10 and 6 cal kyr BP is characterized by minor variations in the magnetic record, while there is a clear increase in oceanographic instability from 6 cal kyr BP to the present. At large scale, a decreasing trend in the magnetic mineral content toward the present is presumably associated with a change in the circulation pattern and coincides in time with decreasing North Atlantic Deep Water formation. Short-term intervals of decreased magnetic mineral content are also identified at 5.2-4.9, 3.77-3.41, 3.01-2.7 and 0.93-0.62 cal kyr BP, the youngest one (AD 1020 to 1330) corresponding in time to the Medieval Warm Period. These events are interpreted to reflect periods of increased activity of the relatively warm, high salinity surface Irminger Current related to stronger input of North Atlantic waters into the Nordic Seas. In addition, spectral analysis of selected magnetic parameters demonstrates centennial-scale periodicities of 715, 240, 170 and 100 yr, which are more clearly expressed in the last 6 cal kyr BP than in the previous period. These periodicities might be attributed to intervals of persistence in the North Atlantic Oscillation (NAO) mode at centennial time-scales. This record, therefore, indicates that the Holocene climate was more unstable than is often assumed, and it demonstrates the importance of high-resolution climatic studies for this recent time period to improve climate model predictions.

Deep water hydrological characteristics have been determined for three cores in the North Atlantic, HM03-133-25 from the Faroe-Shetland Channel, MD95 2024 from Orphan Knoll in the southern Labrador Sea, and ENAM9606 from the Feni Drift in the Rockall Trough. Gravity core HM03-133-25 was collected from the southern end of Faroe-Shetland Channel (60° 07, 06° 04'W, 1156 m depth). The site is swept by relatively strong currents, yielding a high percentage of sand (20-60%), with a high of over 70% during the Bølling-Allerød (B-A). This 4.7 m long core has been dated by AMS C-14 and by correlation to core MD99-2284 from the northern entrance to Faroe Shetland Channel. The core spans the last 14-20 ka (there is uncertainty in the age of the base) with sedimentation rates relatively low in the late glacial (10 cm/ka) and high in the Holocene (early Holocene 65 cm/ka decreasing to 12cm/ka in the late Holocene). Flow speed has been monitored by current meters and ADCP for several years near the core site. Speed, temperature and salinity data reveal fast flows in both warm saline Atlantic inflow and cold less saline outflow. The magsus signature in cores HM03 and MD99 cores correlate during the B-A, Younger Dryas (YD) and 8.2 ka event. The magnetic mineralogy is presently being investigated. The 8.2 ka event does not appear in the isotopic records. The cores were scanned for magnetic susceptibility and for spectral colour reflectance. The fine fraction <63 µm was analysed for the mean grain size of the non-biogenic sortable silt fraction (abbreviated as SS) (mean of 10-63 µm range), with carbonate and biogenic opal removed. The grain size distribution was measured with a Coulter Counter. The sand percentage is regarded here as an indicator of flow speed as the faster flows through the channel are capable of transporting sand and preventing deposition of mud. The drop in sand to 14% around the 8.2 event time indicates relatively sluggish flow, while the peak over 80% in the B-A records fast flows under warm conditions. Holocene speeds increase after the 8.2 event up to the present. The YD does not show slow flow. A peak in flow speed is recorded between 7 and 6 ka and a plateau at 8 to 7 ka rising from the 8.2ka low. The period 0-2000 years BP is marked by high flow speeds. The sortable silt data are very noisy, and difficult to interpret as the silt is trapped as an interstitial component in sand. SS is lower over the last 2000 years than at other times. NEAP 15K/16B located at 2850 m on Gardar Drift is influenced by Iceland-Scotland Overflow Water (ISOW); while MD95-2024 at 3539 m from the Labrador Sea (Orphan Knoll), is affected by Denmark Strait Overflow Water (DSOW). Spectral analyses reveal similar pacings in the millennial band of ~1.5 ka, but deep water flow trends at centennial to millennial time-scales are opposing. This may be explained by variations in either the flux or/and the density structure of the lower water column. The latter is unlikely, since bottom water temperatures remained fairly uniform during the Holocene, suggesting a persistent dominance of DSOW at Orphan Knoll. Rockall Trough is a distal site along the flow path of the North Atlantic thermohaline circulation (THC), bottom waters are dominantly composed of recirculated North Atlantic Deep Water entering from the southeast. Sustained bottom current activity resulted in the formation of Feni Drift, one of the major drift deposits in the North Atlantic. This area will be capable of recording even relatively minor fluctuations of bottom current activity precisely because it constitutes a distal setting. The mean size of the sortable silt fraction is a proxy for relative bottom current speed. We apply this proxy on Feni Drift core ENAM9606 (55º39 N 14ºW, 2543m depth). We investigated the Holocene, at bicentennial or higher resolution. The record can be divided into four main intervals. The early Holocene displays a continuous decrease in SS mean size and inferred bottom current speed, culminating in a sustained minimum between 9.5 and 8ka BP. The mid-Holocene interval (7.7-5.6ka) is characterized by strong and stable bottom currents, punctuated by a single minimum around 7ka. This is followed by slightly weaker and more variable flow speeds until 3.4ka. The late Holocene (<3.4ka) shows drastically enhanced variability including several rapid drops in flow speed. The interval from 1.5- 0.9ka shows particularly marked fluctuations; here abrupt SS mean size minima alternate with maximum values comparable to the rather stable early mid-Holocene. Finally, over the following two centuries a continuous decrease in SS mean size finishes with moderately low values in the remainder of the record.

Detailed magnetic data have been obtained from core HM03-133-25 collected during a Norwegian-British cruise in the Faeroe Shetland region. The magnetic mineralogy is dominated by magnetites in the pseudosingle domain range. The size of the magnetic grains is a little larger during early Holocene than after illustrating either a slightly more active bottom current or more simply, a re-suspension of non-sorted material just after deglaciation occurred. Both long term and short term trends are observed both in the magnetic grain size and in the concentration of magnetic grains within the sediment. On the long-term, the data can be fitted by a fourth order polynomial in which more magnetic grains are synchronous with coarser grains. Both illustrate more active bottom current and this happen in the early Holocene and around 5-4 kyr B.P. When detrended, the grain size and concentration data also exhibit the same pattern with fine magnetic grains always associated with less magnetic grains (i.e. weak bottom current). Several events of weak bottom current activity are recognized and they are (within the accuracy of the preliminary age model) around 7.9; 7.5; 5.7; 4.8; 4.2; 3.6; 2.2 kyr. A spectral analysis will be conducted when the final reliable age model will be available.

Palaeoenvironmental conditions can be reconstructed from microfossil assemblages with a range of numerical techniques, known as transfer functions, which use information on the modern relationship between species distributions and the environment. Many taxonomic groups have been used to reconstruct a wide variety of environmental variables. The selection of the best transfer function method to use, and the optimal values of metaparameters, is usually guided by the prediction error of the training set under cross-validation. Work done in PACLIVA has shown that all of the numerical methods used assume that the observations in the training set are independent, but that oceanic environments are highly autocorrelated. This lack of independence between observations can severely bias the estimation of the true prediction error, and hence misinform model choice. The methods most sensitive to autocorrelation are those that seek local relationships between species and the environment (such as the modern analogue technique, and artificial neural networks). Because of their sensitivity, they give much lower uncertainty estimates than more robust methods (such as weighted averaging based methods), and are the most commonly used methods in marine palaeoecology. This work suggests that some of the most widely used transfer function methods should only be used with caution, and some popular environmental proxies (such as dinoflagellate cysts) are much less valuable that previously thought.

The potential of higher plant n-alkanes and n-alkan-1-ols as paleoceanographic/paleoclimatic proxies in marine sediments has been explored by PACLIVA. These biomarkers are major lipids in these samples. These compounds can be used to monitor the inputs of vegetation to the marine system. In addition, they have been used to obtain a marker of the intensity of the bottom currents which is based on the higher chemical labiality of the molecules containing hydroxyl groups to oxidation. A new tool for assessment of the relative intensity of the deep currents has been developed in PACLIVA [AI = (conc of higher plant n-alcohols) / (conc of higher plant n-alcohols + conc of higher plant n-alkanes). This ratio between higher plant n-alcohols and n-alkanes can be used for assessment of the changes in the intensity of deep water currents since n-alcohols are oxidized more rapid than the n-alkanes. A comparison was made of the index obtained from cores in the N Atlantic and the western Mediterranean Sea (ENAM9606, MD952011 and MD992343) to study of the intensity of deep water currents. The differences show significantly lower indices for core MD992343. This result is expected since this core is situated in an area under the direct influence of the Deep Mediterranean waters formed in the Gulf of Lions. This core exhibits a major index drop in the period between 7300 and 3000 years BP which must represent a significant increase in the rate of formation of these waters. Then, a strong recovery of index values must involve a slow down in the formation of this deep water mass and the most recent period (1640 BP to present) involves another increase in formation of Deep Mediterranean waters. In comparison to these values, the indices in the north Atlantic cores exhibit consistent higher ratios which much less variability. In principle, this lower variation suggests more constant deep water flows than in the Western Mediterranean, involving no significant changes throughout the Holocene. Although further studies will be necessary for a better assessment of the strengths and weaknesses of this proxy, their application to cores ENAM9606, MD952011 and MD992343 has given very consistent results showing that whereas major changes in intensity of deep water currents occurred in the northern western Mediterranean during the Holocene the changes in the north Atlantic (Feni drift and Norwegian Sea) in the same time period were much lower.

Reconstructions of surface layer temperatures and sea ice variability have been produced for two sites in the Nordic Seas over the past 1200 years, one site MD2011 from the Vøring Plateau along the Eastern inflowing branch, and one Site MD2275 off the north coast of Iceland. Sea ice variability was reconstructed from historical observations at Iceland, and from sediment mineralogy at both sites. Sea surface temperatures were reconstructed using transfer function based on diatom and foraminifera assemblages, and the foraminifera-based geochemical proxies, Ca/Mg ratios and stable isotopes. At each site, temperature-proxies correspond well with sea ice reconstructions on the centennial time scales. The sea ice proxies also document W-E movements of the Arctic and Polar fronts. There is a clear warming in the MWP, and earlier onset of the LIA North of Iceland compared with in the East, and a mutual variability within the LIA.

Sea surface conditions in the time interval 0-2ka and 6-8ka have been reconstructed from two North Iceland cores (MD99-2275 and MD99-2271. Due to the present position of the Polar Front on the North Icelandic shelf, this is a very sensitive area for reconstruction of climate change through the Holocene. In this region, the relatively warm Irminger Current, bringing Atlantic Water masses to the shelf, meets the cold, low-salinity East Icelandic Current (Arctic Water), which partly derives from the East Greenland Current (Polar Water) and partly from westerly eddies of the Norwegian Atlantic Current (Atlantic Water). Dating of the North Icelandic shelf records has been achieved by a combination of AMS radiocarbon dates and tephrochronology. Tephra samples from marine sediment cores and from soil sections in Iceland have been geochemically analysed for major elements at microprobe facilities at the University of Bergen, Norway. This has enabled the construction of age models that are independent of marine reservoir age uncertainties hampering reliable radiocarbon dating of marine samples. Mineralogical analyses have been employed to investigate deposition of sand-sized sediments released from melting sea-ice and icebergs. These records have been compared with the sea-ice index record from Iceland back to AD 1600. Methods have been developed to differentiate primary and secondary deposits of volcanic tephra, a key issue in tephrochronological work in settings proximal to volcanoes. Domains for primary versus secondary volcanics have been established employing morphometric techniques. Several different proxies are used to reconstruct high resolution sea surface temperature (SST): 1) diatom-based transfer function, 2) transfer functions based on planktonic foraminiferal assemblages (Maximum Likelihood, ML) and 3) planktonic oxygen isotopes (sinistrally coiled Neogloboquadrina pachyderma, NPS). The diatom-based transfer function uses a modern diatom-environmental variable dataset from around Iceland covering areas with sufficient environmental gradients for the deduction of quantitative relationship between diatom species and environmental variables. A Monte Carlo permutation test with forward selection shows that diatom distribution in the area is primarily controlled by summer and winter sea surface temperatures (SSTs, SSTw). WAPLS transfer functions for SSTw and SSTs have uncertainties of less than 1°C. The reconstructed temperature for the time interval 2-0 cal kyr BP on the North Icelandic shelf shows an overall decreasing trend through the last two millennia, with generally higher temperatures during the Roman Warm Period than during the Medieval Warm Period. There is a sharp temperature drop between AD 1250 and 1350, at the initiation of the Little Ice Age. There is a high correlation between the reconstructed marine temperature and historical records. The absolute level of reconstructed temperatures is almost similar for the diatom-based reconstruction and for the isotopic-based one, while the ML reconstruction results in generally lower values (a difference of about 4°C). The temperature reconstruction for the time interval 8-6 cal kyr BP show a clear climatic optimum between 8.0 and 7.3 cal kyr BP off North Iceland. After that, there is a generally decreasing trend in sea surface temperature. The 8.2 cooling event is clearly expressed in all the MD99-2275 data. This core site was located close to the Polar Front area at that time, an area which was therefore very sensitive to minor changes in the distribution of the surface water masses. The 8.2 event has, however, not been recorded at core site MD99-2271, which is located in the central part of the Atlantic Water masses of Irminger Current. This site must have been located to the south of the frontal area at that time. As seen for the 2-0 cal kyr interval, the ML-based temperature reconstruction gives temperature values, which are a few degrees lower than those based on isotopes and diatoms. A comparison between the two time intervals shows that the values are close to 2°C higher in the 8-6 cal kyr interval than in the 2-0 cal kyr time interval.

Alkenones are produced by algae and, extracted from marine sediments, are used as a proxy for sea surface temperature. This work improves the characterisation of the chemical composition of alkenones. A method for the location of the double bonds in these compounds was developed. Cyclopropylamine was used as derivatisation reagent for the first time. It was found to give high reaction yields, high GC retention time differences between derivatives and precursor compounds, and virtual absence of by-products. Examination of a large group of alkenones from cultures of Emiliania huxleyi, recent and ancient sediments with cyclopropylamine derivatization showed that in all cases the double bonds were located at the same carbon atom distance from the carbonyl group and spaced in intervals of five methylene groups either from the carbonyl or between them, e.g. at sites 7, 14, 21 and 28. This result corrects previous assumptions in which double bond positions were situated by reference to the methyl end. 4,4-Dimethyloxazoline derivatization of hexatriacontenoates present in the same samples showed that these compounds have also their unsaturations with seven carbon atom spacing and counting by reference to the carboxyl group. The concurrence of both series of isomers in compounds of different oxygen functionalities indicates that the precursor hapthophycean algal species have a major biosynthetic pathway leading to the formation of these lipids. The data presented in this work unify the structures of the known alkenones in the present and the recent past under a common metabolic pathway. For the first time the scientific community has a robust method for the determination of double bond positions in alkenone mixtures of interest for paleoclimatic studies. The application of this method to all types of alkenone distributions know to date has shown that all isomers have the double bond locations in the same position relative to the carbonyl group. This finding is important for paleoceanographic studies because it reveals a strong uniformity in the alkenone mixtures used for paleotemperature estimations in all sorts of marine environments and time periods. The relationship between alkenone distributions and sea surface temperature estimates is therefore reinforced.

Diatom-based sea surface temperatures (SST) have been reconstructed for 3 Nordic Sea cores (MD95-2011, LO0914, MD99-2269) using different transfer function methods: Imbrie & Kipp regression (I&K); Maximum Likelihood (ML); and weighted averaging partial least square (WA-PLS). The surface calibration set is based on 52 species from 139 surface samples. The reconstructed SST time series have been analysed with SiZer (Significance of Zero Crossings of the Derivative) to find significant features, and with Wavelet power spectrum. SST reconstructions from the Nordic Seas describe the general Holocene climatic development as a period of cooling in step with decreasing insolation at the northern hemisphere since 11.000 years BP. In contrast, the SSTs at the Reykjanes Ridge site (core LO09-14) exhibit highly variable temperature fluctuations throughout the Holocene, with a general trend showing a relatively cool and variable early Holocene period (11-7 kyr), a mid Holocene Climate Optimum (7-5 kyr), and a generally stable late Holocene period (5 kyr to present). Multi-centennial variability on the time-scales from 640, 950, 1000, 1500 and 2500 years is prominent in core LO09-14. Superimposed on the general Holocene climate development there is a high-amplitude millennial to century scale variability in the surface waters of the subpolar North Atlantic Ocean. Even though climate oscillations with 2500, 1500, 1000, 950 and 640 yrs periodicities are evident in the LO09-14 SST records, there is also a time dependent dominance of different periodicities through the Holocene indicating a clear change of mode from about 5 kyr BP. Before 5 kyr BP a strong 2500 and 1500 yrs periodicity is present in the records. After 5 kyr BP the LO09-14 record shows lower amplitude variations with periods of 900 and 1000-1500 yrs. Century scale variability in the LO09-14 SST record document several Holocene surface ocean cooling events (HCE). Three cooling events (HCE 6, 9 and 13) recorded in core LO09-14 show clear positive correlation to IRD events in core V29-191. The other HCEs show either negative correlation or none-conclusive relation. Whereas in core V28-14, HCE 1, 2, 3, 11 and 13 show positive correlation to the IRD events. We conclude that in cases where there is a positive correlation between the IRD events and HCE the advances of icebergs into the North Atlantic were accompanied by 2-3˚C SST cooling. SST results from cores MD95-2011 and MD99-2269 have contrasting temperature anomalies on a multidecadal time scale and indicate phase-locked intervals with a North Atlantic Oscillation signature. The results also show a close match to solar irradiance. We therefore suggest that oceanic climate variability during the last 800 years is a result of atmosphere-ocean interactions, i.e. the North Atlantic Oscillation and the Atlantic Multidecadal Oscillation, responding to solar forcing.

Sea surface temperatures have been generated at unprecedented high resolution using alkenone paleothermometry along the MD99-2275 core (66°33N; 17°42W, 470m) located off North Icelandic shelf. The achieved temporal resolution is 2-5 years for the 0-2000 year intervall and 5 years for the 6000-8000 years. Owing to its location, the core is influenced by the cold and fresh waters of the East Greeland Current to the North and the warmer and saltier waters of the North Icelandic Irminger Current a limb of the Gulf Stream. Therefore, these data allow reconstruction of the temporal evolution of polar front movement in a region where atmospheric forcing North Atlantic Oscillation has a dominant control on the interannual variability. This time series valuable for further comparison with model simulation of climatic forcings (solar, volcanic etc).

UoB developed a series of syntheses with respect to the northern paleoclimate time series with inputs from the other partners, in order to test the consistency of the results. The most robust methodology is the Threshold counting method employed by Osborn and Briffa (Science 2006). The data results from analyses of the last 1200 years displays a common, cross-basin signature of both the MWP and the LIA. The cold phases are common with the Northern Hemisphere reconstructions, and indicate strong volcanic forcing with cold events at about 1600 AD and 1800 AD. The results so far indicate that the NAO-pattern is not a consistent pattern on multidecadal and longer time scales, but that it is probable that a more NAO-like sustained pattern was apparent during the Early Holocene time period based on the difference between the eastern and western basins. Analyses of the results strongly indicate that the Early Holocene warming was very different from the modern in that the modern is apparent both at surface and subsurface implying advective dynamics are important, and the early Holocene reflects the impacts of the insolation anomaly only in the surface layer. Data sets are stored in project data base at BCCR. These will now be ported to the PANGEA data base and will be available for the new generation of transient model experiments under execution as part of several European initiatives (e.g. the Millennium initiative of the MPI-Hamburg (led by Johann Jungclaus). Further data-model analyses are planned with the transient model experiments for the. last millennium performed as part IPCC AR4, and with model experiments performed as part of the PMIP-2 program. These studies have been funded both in Norway, Switzerland and Germany, and secures that the PACLIVA data sets will become an integrated source for validation of millennium model experiments and for attribution studies relating climate forcing to responses in the ocean.

Due to the highly advective nature of winter-precipitation in Europe, precipitation-reconstructions are potentially a sensitive indicator of past atmospheric synoptic variations, including the North Atlantic Oscillation and other modes. In Scandinavia, a number of published glacier records are analyzed, and winter-accumulation signals have been extracted from these records by using the relationship between ablation-season temperature and winter-accumulation at Norwegian glaciers (Liestøl in Sissons, 1979). This has allowed the amount and relative precipitation to be temporally and spatially described in western Scandinavia. At the start of PACLIVA, NAO was regarded as the most important atmospheric mode. Recent reanalysis data have provided insight into a more complex non-stationary behaviour of high and low-frequent atmospheric variability (Jacobeit et al., 2003; Cassou et al., 2003; Hurell et al., 2004). Analysis of measured glacier and pressure-field data over Scandinavia has shown how different geostrophic vectors and sub-systems affect different regions during winter (Nordli et al., 2003; 2005). Taking these findings into account and considering the spatial patterns in the reconstructed winter-precipitation over western Scandinavia it has become evident that on low frequencies spatial non-stationarity may be more significant than variations in the NAO-associated meridional pressure-gradient in the North Atlantic (Lie et al., 2003, 2004, 2005). As spatial variations include several independent time-series, the temporal precision of these analyses is below the original PACLIVA-objective. Therefore we have calculated indices for a continuous time-series of non-stationarity over the last 6000 years. The Continental Index is the standardised glacier-size gradient between two glaciers in southern Norway, 90km apart along a strong precipitation gradient; we interpret differences between the sites as a winter-precipitation signal. The Coastal Index is the standardised winter-precipitation gradient between SW Norway and NW Norway (Bakke et al., submitted). The two indices capture the same low-frequent variability over the last 6000 years. These spatial variations in relative precipitation indicate that the atmosphere varies spatially as well as in its strength-domain during the Holocene, and suggest that the circulation-patterns vary on millennial time-scales. To test if similar low-frequent variations occur in instrumental time-series, 4 meteorological stations, with precipitation data available for the 20th century, proximal to the 4 glaciers were selected. The normalized instrumental data and the southerly geostrophic pressure-field centred over southern Norway (60°N, 5°E), both smoothed by a 31 yr running mean, are highly correlated and display multi-decadal variability. The Atlantic Multidecadal Oscillation (AMO) leads variations in the pressure-field and relative precipitation. As our precipitation-data does not include variations in total precipitation, known to be associated with the NAO, it is spatial variations of atmospheric circulation changes that dominate this picture. Recently, the AMO has been explained by variations in the overturning circulation, and shifts in atmospheric circulation patterns were suggested as a response (Sutton et al., 2005). To test if this assumption holds on millennial time-scales, we compared our data with a SST-reconstruction from the Vøring-plateau (MD95-2011; Risebrobakken et al., 2003), a record of North Atlantic deep-water ventilation (ODP-980; Oppo et al., 2003) and the PCI of geochemical analyses from GISP2 (Mayewski, 1997). These records behave similarly to the instrumental time-period on millennial time-scales. To examine variations near the second NAO centre of action, we collected 6 cores were retrieved from three lakes in the Grandes Rousses massif in the south-eastern French Alps. A core from Lac Blanc (2470 m asl) in the catchment of glacier St. Sorlin was chosen for analyses and subjected to a range of sedimentological analyses, including classical bulk-parameters, GEOTEC analyses and grain-size analyses. The top 100 cm has been analysed by micro-XRF. The data show evidence for a climatic optimum lasting from 325-375 cm, when little or no clastic glacier flour was deposited in the lake. There is a period of minor glaciation from 325-300 cm. Well-defined maxima in carbonate input are evidence for at least 8 distinct glacier maxima. The source of this carbonate is a carbonate-rich outcrop currently outside the glacier margin, but within the Little Ice Age maximum extent. This carbonate may affect bulk radiocarbon-dating through the hard-water effect. Wet-sieving the entire core recovered no terrestrial macrofossils for dating. We have postponed the dating of the core, and hope to date pollen extracted from the sediments. This work presents a continual record in which the Alps episodic moraine-derived glacier chronologies may be evaluated.

Sea surface temperatures have been reconstructed for the time intervals 0-2000 and 6000-8000 years BP in the Caribbean from three cores. Core JPA-02P, collected off the Virgin Islands, covers the period 800-1400BP. Core PRP07 (273 cm long) from south of Puerto Rico (17º52.82 N; 66°35.90W; water depth 273 m) covers most of the last 2000 years. The chronology is based on eight AMS 14C dates on the planktonic foraminifer species Globigerinoides sacculifer. Accumulation rates range between 0.2 and 3.6 mm/yr. The sediment core analyzed for the 6000-8000 BP interval, MD02-2582 (12.7 m long), was retrieved in 2002 by R/V Marion Dufresne from south of Puerto Rico (17º51.20N; 66º36.00W; water depth 380 m). Based on eight AMS 14C dates on G. sacculifer, the 6000-8000 BP interval was found between 37 and 44 cm. Accumulation rates range between 1.6 and 8.1 cm/kyr. Planktonic foraminifer census data were generated from counts of more than 300 specimens from the >150 µm fraction at 1 cm intervals in both cores from Puerto Rico, 29 taxa were identified. Planktonic foraminifer relative abundances were converted to warm-season (WS; August-October) and cold-season (CS; February-March) paleo-SST estimates using artificial neural networks (ANN) trained on the GLAMAP data-set. The ANNs used for reconstructions of paleo-SSTs were selected based on their ability to optimally reproduce measured summer and winter SSTs 1951-1993 (COADS data) using a high-resolution box-core from south of Puerto Rico for which 15 samples of planktonic foraminifer census data were available from this time interval. Paleo-SST reconstructions are represented by averages obtained from 10 separate ANNs, with different training and test sets. The average root-mean-square-error of prediction (RMSEP) relative to the COADS SSTs is 0.31ºC for summer and 0.51ºC for winter. A comparison of WS paleo-SST estimates using different quantitative approaches indicates that the ANN technique generates estimates that are by far closer to the COADS 1951-1993 estimates than any of the other techniques, some of which yield unreasonably low or high estimates. Present-day WS and CS SSTs in the NE Caribbean vary between 28.2 and 29.2°C, and 25.4 and 27.4°C, respectively (COADS 1951-1993 data). In the pre-1400 interval WS SSTs ranged between 28.2 and 28.6°C with a distinct low peak at about 1220 (27.8°C). CS SSTs ranged between 26.3 and 27.2°C, and a low value similar to that for the WS occurred at about 1220 (26.0°C). In the post-1400 part of the record distinctly low sea WS and CS SST anomalies occurred during the late 15th century and 1650 (minima of 28.2 and 26.2°C, respectively), associated with the Spörer and Maunder sunspot minima, respectively, and about 1920 (28.2 and 26.3°C, respectively). In addition, an extreme low CS peak is recorded in about 1975 (25.9°C). Highest WS SSTs are recorded in the 16th century (28.7-29.0 °C). Overall maximum CS SSTs are recorded in the late 15th and most of the 16th century (27.0-27.3°C). A comparison with a climate reconstruction for western Europe (Guiot, 2005) through the last millennium indicates that the Little Ice Age (LIA) seems to have been considerably less pronounced in the Caribbean. Estimated WS paleo-SSTs for the 6000-8000 BP time period (28.0-28.3°C) are cooler than most of the values for the 0-2000 BP period. This tendency is even stronger for the CS (paleo-SSTs in the range of 25.3-26.3°C). Hence, although the warmest climate of the Holocene would have been expected during the Holocene thermal maximum 6000-8000 years BP due to stronger solar insolation, relatively cool conditions appear to have prevailed in the Caribbean during this time interval. Stable isotope analyses were carried out on core PRP07. The oxygen isotope record does not show any clear relationship to the paleo-SST, and are most likely influenced by local salinity variations, possibly induced by hurricanes.

Analyses of oxygen and carbon isotopic compositions and the Magnesium/Calcium ratios of the planktonic foraminifera Globigerinoides ruber white were made on core MD99-2203 (34°58 38N, 75°12 06 W, 620 meters depth), located off Cape Hatteras on the Gulf Stream path. The Mg/Ca ratio is directly connected to the calcification temperature and the oxygen isotopic composition is a function of the temperature and the salinity of the water. This allows us to reconstruct surface hydrological variation of the Gulf Stream. The time resolution of the analysis are very high (8-10 years) for the interval 0-2000 years B.P. and ( 30-35 years ) for the interval 6000-8000 years B.P. that correspond to two different periods of solar forcing.

Sea surface hydrological characteristics have been determined for three cores in the North Atlantic, HM03-133-25 from the Faroe-Shetland Channel, MD95 2024 from Orphan Knoll in the southern Labrador Sea, and ENAM9606 from the Feni Drift in the Rockall Trough, using Mg/Ca (SST) and oxygen isotope (18Oforam which includes SST and 18Osw components) measurements on the planktonic foraminiferal species Globigerinoides bulloides. This surface dwelling species was abundant at all sites during both periods of interest (8000-6000 and 2000-0 years BP). Reconstructions salinity at Femi Drift are qualitatively consistent with multidecadal variability in the instrumental record, notably lower salinity in the 1920s and 1970s, but the amplitude of the variability is overestimated by a factor of 5. For the 20th century, both Mg/Ca- and 18Oc-derived calcification temperatures (the latter based on assumed 18Ow = 0.50 vs. V-SMOW) are consistent with late spring to early summer (May-June) instrumental temperatures. However, the two proxies yield divergent results, implying temporally variable18Ow, presumably due to changing surface salinities. Multidecadal to century-scale variability is common throughout for T(Mg/Ca), ?18Oc and derived ?18Ow records at Femi Drift implying recurrent hydrographic shifts and no long-lasting intervals with sustained warm/saline or cold/fresh surface waters. On longer timescales, the 18Ow record shows relatively lighter values at its base (prior to 300BC) and from 150-400 or 500AD, 750-1150AD and 1600-1800AD, suggesting higher salinities. The Mg/Ca SST record generally implies slightly higher temperatures during the same periods, but its structure differs in detail. In particular, climatic anomalies from 150-400/500AD and 750-1150AD are not readily apparent if one solely considers the Mg/Ca record. Subsequently, a pronounced SST drop occurs only after 1400AD, though the onset of an earlier (slight and highly irregular) SST decrease seems to coincide with the initial 18Ow decrease. These two climate anomalies mentioned above may correspond to the local expression of the Roman (150-400AD) and Medieval (750-1150AD) warm periods, respectively. This would imply an early onset of the subsequent Little Ice Age (after 1150AD in the 18Ow record), but pronounced cooling occurs only after 1400AD and then until 1600AD. The following 1600-1800AD interval seems to be at least as warm and saline as any preceding interval in the record. A second SST and 18Ow decrease comprises also the entire 20th century; accordingly our records seem to show no indication for a post-Little Ice Age recovery. The general correspondence between the Mg/Ca SST and 18Ow records at Femi Drift is consistent with variable northward advection of warm and saline surface waters during the last 2300 years. On the other hand, divergent patterns (most obvious from 1150-1400AD) may indicate additional local controls on sea surface temperatures such as variable air temperatures. The Feni Drift oxygen isotope record seems to show an intriguing antiphase relationship with a high-resolution record from Vøring Plateau. Due to the combined errors of both calibrated radiocarbon age models, this antiphase relationship cannot be rigorously assessed on multidecadal timescales. On the other hand, it is also apparent for the entire interval 400-800AD and during the 20th century, when age control in both cores is based on radionuclide records. The Vøring Plateau site is located upstream along the flow path of the same warm and saline surface waters, and its isotope record is dominated by temperature fluctuations. Accordingly, this data comparison strongly suggests significant salinity influences on our Feni Drift stable isotope record. Average foraminiferal weight varied only slightly between 2000 and 0 and between 8000 and 6000 years BP in the Faroe-Shetland Channel and at Orphan Knoll. Between 2000 and 1000 d18O at MD95 2024 was more variable that at HM03-133-25. Mg/Ca varied between 1 and 2.5 mmol/mol in core MD95 2024 and between 1.8 and 3mmol/mol in core HM03-133-25. In both cores, between 1800 and 1600 years BP a broad warm period can be detected, that may possibly be associated with the Roman Warm Period. During the Little Ice Age ( 900 years BP) SST at the Faroe-Shetland channel were cooler. However, cooler intervals can be also be observed 600 years BP and more recently. Between 2000 and 0 years BP, SST overall seems slightly warmer in the Faroe Shetland channel compared with Orphan Knoll (Labrador Sea), and between 7000 and 6000 years BP a similar trend can be observed. Between 8000 and 7100 years BP (Holocene warm period?) the two sites seem to have similar SST.

Searching for OpenAIRE data...

Deliverables

Share this page

Download