Noen av de brå klimaendringene som fant sted under den siste istiden skjedde i forbindelse med større endringer i tilførselen av ferskvann til Nord-Atlanteren. Store mengder isfjell løsnet fra det nord Amerikanske isdekket i kalde perioder og ble transportert inn mot sentrale deler av Nord-Atlanteren via havstrømmene. Dette ferskvannet påvirket havsirkulasjonen, og dermed de klimatiske forholdene over store områder. Man har trodd at de nordiske havene var fullstendig dekket av sjøis under disse kalde periodene. Våre foreløpige resultater indikerer at der var hull, en såkalte polynya, i sjøisdekket i mindre områder helt nord i de nordiske hav, i Fram Stredet, i deler av en av de kalde periodene, selv om det var sjøis lenger sør. Da sjøisdekket i sør begynte å brekke opp var det antageligvis i en kort periode fritt for sjøis i ett avgrenset område i den østlige delen av de nordiske havene, nord til Fram Stredet. Men mens det forble sjøisfritt i sørøst, ser det ut til at Fram Stredet var dekket med sjøis i nesten hele den varme perioden.
The last glacial was characterized by a series of abrupt climate shifts between cold stadials and warm interstadials (DO-events). A link is documented between sea ice variability in the southern Nordic Seas and air temperature change over Greenland, with sea ice identified as a key determination factor for setting up these large, abrupt climate shifts. However, the spatial extent of the sea ice cover is still unknown and several unanswered questions remain, e.g.: When the sea ice cover broke up in the southern Nordic Seas during the stadial-to-interstadial transitions, how far north did it break up? How different were the timing and duration of the events in the northernmost Nordic Seas relative to the southern Nordic Seas? What were the magnitude and timing of changes in the vertical stability of the ocean? What triggered the abrupt, full basin scale changes? In ABRUPT we will reconstruct the sea ice conditions, hydrography and climate of the Fram Strait, with an unprecedented resolution for this area, over two targeted DO-events. These reconstructions and a regional data synthesis will be used in combination with multi-model output from four state-of-the art GCM glacial simulations, for an integrated analysis of the dynamics of abrupt Arctic climate change during the last glacial. Furthermore, the stadial Nordic Seas bear strong resemblance with the present Arctic Ocean, e.g. the existence of a strong halocline stabilizing the sea ice cover and a subsurface layer of Atlantic Water that bring warm and salty water masses into an interior basin, where it recirculates under the halocline. The rate of change seen for the warming around the Fram Strait over the recent decades, where Atlantic water meets the Arctic sea ice cover, is comparable to the rapidity of the DO-events. ABRUPT will evaluate the relevance of the DO-events for ongoing Arctic climate change to unravel mechanisms important for understanding the risk of similar changes in the future Arctic.