Holocene Variability in North Atlantic Climate and Deep Flow South of Iceland

G G Bianchi (Dept. of Earth Sciences, University of Cambridge, Downing Street, Cambridge, CB2 3EQ, UK; Internet: ggb20@esc.cam.ac.uk); I N McCave (Dept. of Earth Sciences, University of Cambridge, Downing Street, Cambridge, CB2 3EQ, UK; Internet: mccave@esc.cam.ac.uk)

Holocene climatic variability is well known in recent historical times from the Mediaeval Warm Period to the Little Ice Age, but poorly known from earlier periods. Millennial-scale fluctuations in climate have been suggested from records of ice-rafted debris from the deep ocean, sea salt ice, 14C production rates and instabilities in climate models.

Here we present the first Holocene data showing changes in the intensity of deep water flow related to climatic fluctuations in the northern North Atlantic. A high resolution Holocene record (~ 45 cm/ka) has been obtained from NEAPACC core 15K recovered in the South Iceland Basin, adjacent to the main flow of Iceland Scotland Overflow Water (ISOW), a principal component of North Atlantic Deep Water (NADW). Stratigraphic control was achieved through a series of 19 radiocarbon dates. The grain size parameter we employ as a proxy for near-bottom current intensity shows clear fluctuations in ISOW flow coinciding with the Little Ice Age, Mediaeval Warm Period, Dark Ages Cold Period, and Roman Warm Period. Furthermore, we recognise a series of similar quasi-periodic fluctuations throughout the Holocene, which are best defined in the last ~ 7,500 years. These have a spectral peak period of about 1,500 years, which coincides with Holocene events of apparent increased iceberg flux from the Greenland Sea to lower latitudes in the North Atlantic (Bond et al., 1997). The timing of these events coincides with periods of reduced deep water convection we infer from our grain size proxy. We therefore hypothesise that the freshwater supplied by increased ice rafting may suppress densification and sinking of water in the Norwegian-Greenland Sea. Prior to ~ 7,500 years BP there is a trend to slower speeds matching warmer climate, probably reflecting deglacial meltwater influence on convection.

We also note a striking positive correlation between our record and changes in SST in the Sargasso Sea (Keigwin, 1996), which often seem to lag fluctuations in the vigour of deep-water flow speed in the Iceland Basin. These observations importantly suggest that Holocene climate modulation is likely to be driven from the polar regions of the north-east Atlantic with repercussions on at least a wide regional scale.

Our results imply that deep water convection in the Greenland- Norwegian Sea is very sensitive to potentially small forcing mechanisms and that it can provide a means of amplifying local/regional climatic events to a hemispheric/global scale.