A multiproxy reconstruction of the evolution of deep and surface waters in the subarctic Nordic seas over the last 30,000 yr

Abstract

On the basis of various lithological, mircopaleontological and isotopic proxy records covering the last 30,000 calendar years (cal kyr) the paleoenvironmental evolution of the deep and surface water circulation in the subarctic Nordic seas was reconstructed for a climate interval characterized by intensive ice-sheet growth and subsequent decay on the surrounding land masses. The data reveal considerable temporal changes in the type of thermohaline circulation. Open-water convection prevailed in the early record, providing moisture for the Fennoscandian-Barents ice sheets to grow until they reached the shelf break at ∼26 cal. kyr and started to deliver high amounts of ice-rafted debris (IRD) into the ocean via melting icebergs. Low epibenthic δ¹⁸O values and small-sized subpolar foraminifera observed after 26 cal. kyr may implicate that advection of Atlantic water into the Nordic seas occurred at the subsurface until 15 cal. kyr. Although modern-like surface and deep-water conditions first developed at ∼13.5 cal. kyr, thermohaline circulation remained unstable, switching between a subsurface and surface advection of Atlantic water until 10 cal. kyr when IRD deposition and major input of meltwater ceased. During this time, two depletions in epibenthic δ¹³C are recognized just before and after the Younger Dryas indicating a notable reduction in convectional processes. Despite an intermittent cooling at ∼8 cal. kyr, warmest surface conditions existed in the central Nordic seas between 10 and 6 cal. kyr. However, already after 7 cal. kyr the present day situation gradually evolved, verified by a strong water mass exchange with the Arctic Ocean and an intensifying deep convection as well as surface temperature decrease in the central Nordic seas. This process led to the development of the modern distribution of water masses and associated oceanographic fronts after 5 cal. kyr and, eventually, to today's steep east–west surface temperature gradient. The time discrepancy between intensive vertical convection after 5 cal. kyr but warmest surface temperatures already between 10 and 6 cal. kyr strongly implicates that widespread postglacial surface warming in the Nordic seas was not directly linked to the rates in deep-water formation.

Introduction

Paleoceanographic studies have demonstrated that the oceanic circulation in the North Atlantic region was quite variable between the Lateglacial and Holocene periods (e.g., Broecker et al., 1988; Lehman and Keigwin, 1992; Veum et al., 1992). Salinity changes in regions of thermohaline circulation and deep-water formation have been proposed for being partly responsible for these rapid climatic shifts (Broecker et al., 1990; Rahmstorf, 1995). Sediment cores from the North Atlantic document that Last Glacial ice-sheets fluctuated in size on timescales not recognized previously (Bond and Lotti, 1995), and that such fluctuations were probably also causing some of the salinity changes at these mid- to high-northern latitudes due to enhanced iceberg thawing (e.g., Duplessy et al., 1991; Bond et al., 1993; Maslin et al., 1995).

The Norwegian–Greenland–Iceland seas (Nordic seas) is a region well suited to study glacial to interglacial climatic changes. This is because intense deep-water formation occurs here as a result of the poleward flow of warm and saline Atlantic surface water and consequent cooling of these waters upon heat release. This process is recognized as an integral part of the modern, interglacial climate system. It is believed that the present day situation, with barely any ice left in Scandinavia, is the result of the global climate change after the Last Glacial Maximum (LGM) triggered by an increasing solar radiance and a subsequent northward expansion of warm Atlantic surface waters into subarctic latitudes (Ruddiman and McIntyre, 1981; Imbrie et al., 1993). During the LGM, the landmasses surrounding the Nordic seas were largely covered by ice sheets. At this time and during the ensuing deglacial phase planktic foraminiferal isotopes and sedimentologic proxies show that the marginal regions of the Nordic seas have repeatedly received high but also variable amounts of sediments from icebergs/ice-sheet magins (e.g., Elverhøi et al., 1995; Laberg and Vorren, 1995), and that strongly varying surface salinity conditions due to meltwater input probably had a pronounced effect on the surface circulation and intensity of deep-water renewal (e.g., Veum et al., 1992; Sarnthein et al., 1995). The water-mass variability at the surface, particularly during the last deglaciation, apparently paralleled rapid atmospheric temperature changes on the nearby ice-covered landmasses (e.g., Lehman and Keigwin, 1992; Taylor et al., 1993).

Supported by planktic stable isotope studies, micropaleontological records have been used in the Nordic seas to elucidate past water mass changes in the Nordic seas (Kellogg, 1980; Jansen and Bjørklund, 1985; Baumann and Matthiessen, 1992; Koç-Karpuz and Jansen, 1992; Sarnthein et al., 1995; Hald et al., 1996). Despite considerable progress in the understanding of the glacial-to-interglacial climate system has been made on the basis of these studies, they could only partially unveal the specific conditions of the LGM. Previous interpretations saw the Nordic seas as an ice-covered region during the LGM. But more recent investigations suggest the presence of Atlantic surface water as well as open-water conditions during the main glacial phase, oxygen isotope stage (OIS) 2 (Bauch, 1992; Hebbeln et al., 1994; Dokken and Hald, 1996; Weinelt et al., 1996), implying that there must have been a notable water-mass exchange between the Nordic seas and the North Atlantic.

Understanding past environments is a rather challenging task if the region under study has no modern analogues to compare with. The interpretation of the Last Glacial paleoenvironment of the Nordic seas is, in this respect, particularly challenging because it is often hampered by an inconsistency in the downcore occurrence of important proxy tools. For instance, paleoceanographically crucial pelagic microfossil groups are either lacking from the fossil record or are extremely impoverished in species numbers. The bathyal conditions are usually best looked at using benthic forminifera. However, the glacial and interglacial species assemblages of the Nordic seas differ from each other in showing during the glacial interval just a few, mainly infaunal species (e.g., Struck, 1995). This common lack of epibenthic foraminifera in glacial core sections makes it difficult to establish dependable epibenthic δ¹³C records, and is the main reason why no profound knowledge exists on the glacial-to-Holocene deep-water mass evolution of the Nordic seas. So far, the only existing benthic δ¹³C records which also covers the entire LGM comes from the southeastern Norwegian Sea and has been used to interpret the water-mass development between the North Atlantic and the Norwegian Sea (Veum et al., 1992). But this core, as well as others most often used for paleoceanographic studies in the Nordic Seas, originates from the western Norwegian-Barents Sea continental margin, a region with hemipelagic sedimentation that has undergone strong depositional changes since the LGM making it more difficult to rule out local depositional effects on the various core records (e.g., Laberg and Vorren, 1995; Bondevik et al., 1997).

In order to overcome, at least partially, the various obstacles associated with sediment records from the Nordic seas and to obtain a constructive paleoceanographic interpretation with a more than regional implication it seems imperative to integrate a multitude of different proxy tools from a `pelagic’ core site with reasonably high sedimentation rates but less strong local overprint. By incorporating various benthic and planktic stable isotope and faunal assemblage data as well as lithological records giving evidence of the temporal variations in the input of specific types of iceberg-rafted sediments, an attempt is made in this study to interpret the complex water mass history of the Nordic seas since the final phase of the last glaciation. Besides interpreting the various proxy records from the main investigation area in the central Nordic seas which today is influenced by inflowing Atlantic surface-water and deep-water formation, we will also compare these data with records from a site in the western Fram Strait directly underlying the outflowing polar waters from the Arctic Ocean. This comparison will allow us to interpret not only the glacial-to-interglacial water-mass evolution between the Nordic seas and the North Atlantic but also to relate the observations made in the central Nordic seas with the crucial oceanic changes occurring near the Arctic Ocean gateway.