The northern North Atlantic sediment drifts have a much greater areal extent than has previously been indicated. The northern North Atlantic extends from the East Greenland to European continental shelves and from the Charlie-Gibbs Fracture Zone to the Greenland-Scotland Ridge. Within this region are seven major sediment drifts containing some of the best clues for pre-Quaternary bottom water circulation in the North Atlantic. Feni Drift is the oldest in the region, originating near the Eocene-Oligocene boundary. It was followed by accumulation of Bjorn, Gardar, Hatton, and Snorri Drifts from the late Early to Middle Miocene. Eirik Drift may have started to accumulate in the Late Miocene, and Gloria Drift in the early Pliocene. Through analysis of the Cenozoic sediment mass/age distribution and the change in area of seafloor for the study area through time, it can be concluded that there is a deficit of Oligocene to Miocene sediment in the region. This could be explained by sediment cycling; the erosion of older sediment to produce young sediment. The mass of Pliocene to Quaternary sediment is much larger than that of the older sediment and may have been derived primarily through the erosion of Eocene and Oligocene sediment by deepwater currents.

The sediment drift portion of the total regional sediment mass, increased significantly three times during the Cenozoic. Each of these drift growth phases, lasted between 3 and 4 m.y. The first two growth phases, from the Late Eocene to Early Oligocene and from the Early to Middle Miocene, were accompanied by northward shifts of depocenters to the newly forming sediment drifts. The third and largest growth phase occurred between 7 and 3 Ma, and may have been preceded by the initiation of accumulation on Eirik Drift, 8--7 Ma. This most recent growth phase was accompanied by high apparent accumulation rates over southern Gardar Drift as it expanded to the east. Erosion rates over Reykjanes Ridge may have been higher during the mid-Pliocene than in the Quaternary. There are two main sites where dense water could have formed and flowed south through the Rockall Trough to start the accumulation of Feni Drift.

If the atmosphere were cool enough, dense water could have formed in the Norwegian Sea or on the Faeroe Shelf. Dense water may also have formed on the Rockall Plateau due to salinity increase through evaporation in an arid climate. As an alternative to the hypothesis of Arctic Ocean water overflowing the Iceland-Faeroe Ridge in the early-middle Miocene, it is suggested that dense water formed along shallow segments of the Iceland-Faeroe Ridge and then flowed into the South Iceland Basin to begin accumulation of Bjorn, Gardar and Snorri Drifts. The initiation of sediment drift formation in the South Iceland Basin may have been caused by a decreased rate of deep water production combined witha higher rate of detrital input due to a Middle Miocene uplift event along the Iceland-Faeroe Ridge. A second uplift event along the Greenland-Scotland Ridge and a decreased production of Northern Component Water in the early Pliocene may be responsible for increased sediment accumulation in Bjorn, Garder and Eirik Drifts. ¿American Geophysical Union 1994