The Impact of the Final Lake Agassiz Flood Recorded in Northeast Newfoundland and Northern Scotian Shelves Based on Century-Scale Palynological Data - Abrupt Climate Change: Mechanisms, Patterns, and Impacts - page 139

1997]. Barber et al. [1999] proposed that the final meltwater

drainage from combined glacial lakes Agassiz and Ojibway
(hereafter Lake Agassiz),

1.1. Evidence for the Lake Agassiz Outburst Flood

Lajeunesse and St-Onge [2008] mapped large sand waves
and iceberg scours on the Hudson Bay seafloor, which they
attributed to a catastrophic subglacial release. Kerwin [1996]
recognized a marker red bed in sediments from Hudson
Strait, believed to be the result of settling of suspended
glacial sediment transported from NW Hudson Bay by melt-
water

flowed directly into the Labrador
Sea (Figure 1), perturbed the North Atlantic Ocean over-
turning circulation, and suppressed the northward transport
of heat, inducing the cold event. A large volume of meltwater
impounded by the Laurentide Ice Sheet was released in a
flood lasting about 0.5 year at peak transport rate of approx-
imately 5 Sv (1 Sv = 10 6 m 3 s 1 ) as the ice dam failed in
Hudson Bay [Teller et al., 2002; Clarke et al., 2004]. Some
computer models have shown that this volume of fresh water
would have been suf

floods [Barber et al., 1999]. The red bed was subse-
quently found to be a double set of graded sediments
[Lajeunesse and St-Onge, 2008]. Paleohydraulic modeling
by Clarke et al. [2004] suggested there were two episodes of
drainage, caused by the ice dam resealing to the seafloor after
the

cient to affect the overturning circula-
tion in the North Atlantic [Wiersma and Renssen, 2006].

first drainage, allowing Lake Agassiz to re ll and undergo

a second 48 years later. Radiocarbon dates in this

chapter are expressed in the form of 8.2 ka, for example,
meaning 8200 calibrated years before present (A.D. 1950).
Many North Atlantic records, which show evidence for
slowing of the AMOC at the time of the 8.2 ka event [Oppo
et al., 2003; Hall et al., 2004], were deemed inconclusive by
some [Alley and Agustsdottir,2005;Rohling and Pälike,
2005]. However, Ellison et al. [2006] present evidence of a
cooling trend in sea surface temperature (SST) and a reduc-
tion in deep water

ood 13 -

flow speed in the eastern North Atlantic
that are coincident with the Lake Agassiz discharge, all of
which culminated in a cold climate event 200 years later at
8.2 ka. Kleiven et al. [2007] have recently presented evi-
dence of reduced North Atlantic Deep Water flow in the
northwest Atlantic for a period of about 100 years beginning
at about 8.38 ka, coeval with the catastrophic drainage of
Lake Agassiz (dated 8.47 ka with an error range of 8.16 - 8.74 ka
by Barber et al. [1999]). Based on high-resolution chemistry
and isotope records from Greenland ice cores, Thomas et al.
[2007] estimated the duration of the event to be about 160
years, between 8.21 and 8.14 ka. Because of the relatively
short duration of the 8.2 ka event, high-resolution records are
needed if its relationship to oceanic change is to be detected
and understood.
Though Lake Agassiz

Figure 1. Location maps and surface currents along the eastern
Canadian margin. (a) Surface currents in the northwest North
Atlantic, Labrador Sea, and Baffin Bay, based on the works of
Greenberg and Petrie [1988], Han et al. [1999], and Shore et al.
[2000]. The acronyms refer to the following surface currents:
BLC, Baffin Land Current; EGC, East Greenland Current; WGC,
West Greenland Current; and LC, Labrador Current. Other acro-
nyms refer to the following locations: NDC, Notre Dame Chan-
nel; NFLD, Newfoundland; and NS, Nova Scotia. The areas with
a water depth between 0 and 200 m are shaded light gray; the 200
and 1000 m depth contours are shown. (b) Location of core sites
and other important sites mentioned in the text. Abbreviations are
LS, Labrador Shelf; BB, Baffin Bay; and GB, Grand Banks. The
locations of the two cores published in this chapter are shown by
bold numbers: 1, Notre Dame Channel core 87033-19 and 2,
St. Anne ' s Basin core 84011-12. Other marine sites are 3, Canso
Basin; 4, Emerald Basin; 5, La Have Basin; 6, Cabot Strait; 7, Gulf
of St. Lawrence; 8, Bay of Islands; 12, Cartwright Saddle; 13,
Orphan Knoll; 14, Flemish Pass; and 15, Laurentian Fan. Terres-
trial sites are 9, Everitt Lake; 10, Silver Lake; 11, Compass Pond;
and 16, Labrador region.

s drainage was large enough to
suggest transport volumes of approximately 5 Sv through
Hudson Strait [Clarke et al., 2004], Hillaire-Marcel et al.
[1994] and MacLean et al. [2001] found no evidence in the
delta 18 O records indicating a freshening of surface waters in
Hudson Strait, the Labrador Sea, or on the Labrador Shelf.
However, Andrews et al. [1999] observed a weak signal at
Cartwright Saddle, and one has been found spanning the
early Holocene at Orphan Knoll [de Vernal and Hillaire-
Marcel, 2000, 2006]. In fact, the lack of a widespread shift
in isotopic values has been noted [Andrews et al.,1999;
Keigwin et al., 2005; Hillaire-Marcel et al., 2007]. Hillaire-
Marcel et al. [2007, 2008] suggested that freshwater mixing

'