Meltwater injections and their impact on Atlantic meridionaloverturning circulation and climate during the time period ofHeinrich Event 1 and the last deglaciation

摘要

The temporal relationship between meltwater pulse 1a (mwp-1a) and the climatehistory of the last deglaciation remains a subject of debate. By combining theGRIP ?18O ice core record on the new Greenland Ice Core Chronology 2005(GICC05) timescale with the U/Th-dated Barbados coral record, it isconclusively derived that mwp-1a did not coincide with the sharp Bøllingwarming, but with the abrupt cooling of the Older Dryas. To evaluate whetherthere is a relationship between meltwater injections, North Atlantic Deep Water(NADW) formation and climate change (i.e., the long term change in the averageweather), a high-resolution magnetic (?ARM/?) proxy record of NADW flowintensity from Eirik Drift, south of Greenland, is presented. A record of meansortable silt grain sizes (an established proxy for near bottom current flowspeed), obtained from the same samples on which the ?ARM/? was measured,shows remarkable similarity to the magnetic record and validates ?ARM/? as aproxy for NADW flow intensity. The record of ?ARM/? indicates only a relativelyminor 200-yr weakening of NADW flow, coincident with mwp-1a. Thiscompilation of records also indicates that during Heinrich event 1 (H1) and theYounger Dryas there were no discernible sea-level rises, and yet these periodswere characterised by intense NADW slowdowns. Records of planktonicforaminiferal ?18O, as well as lithic and foraminiferal counts from Eirik Drift arecombined with previous studies from the Nordic seas and the ‘Ice Rafted Debris(IRD) belt’, and portray a sequence of events through the interval of H1. Theseevents progressed from an onset of meltwater release around 19 ka BP, throughthe ‘conventional’ H1 phase from ~17.5 ka BP, to a final phase between 16.5 and14.6 ka BP, characterised by a pooling of fresh waters in the Nordic Seas, whichwere injected hyperpycnally. This build up of fresh waters was purged from theNordic Seas, preconditioning the Nordic Seas for convective deep-waterformation. This allowed the abrupt re-start of NADW formation in the NordicSeas at the Bølling warming. In contrast to previous estimates for the duration ofH1 (i.e., 1000 years to only a century or two), the total, combined compositesignal of H1 presented here had a duration of nearly 4000 yrs (~19–14.6 ka BP),now spanning the established period of NADW shutdown. Clearly, deep-waterformation and climate are not simply controlled by the magnitude or rate ofmeltwater addition. Instead, the results presented here emphasise that the locationof meltwater pulses may be more important, with NADW formation beingparticularly sensitive to surface freshening in the Arctic/Nordic Seas.