Episodic Sediment Failure in Northern Flemish Pass, Eastern Canadian Margin: Interplay of Seismicity, Contour Current Winnowing, and Excess Pore Pressures

Abstract:

Episodic sediment failures are recognised on continental slopes around Flemish Pass and Orphan Basin from multibeam bathymetry, seismic reflection profiles and piston cores. Seismic stratigraphy is tied to published long cores with O-isotope data back to before MIS 6 and carbonate rich Heinrich layers in places produce marker reflections in high-resolution sparker profiles. Heinrich layers, radiocarbon dates and peaks in diatom abundance provide core chronology.

Slope sedimentation was strongly influenced by the Labrador Current and the silty muds show architecture characteristic of contourites. Variation in Labrador Current strength is known from the sortable silt proxy over the past 125 ka. Large slope failures were mapped from seismic reflection profiles and their age estimated from seismic stratigraphy (3–5 ka resolution) and in some cases refined from cores (1–3 ka resolution). Large slope failures occurred apparently synchronously over margin lengths of 50–350 km. Such failures were earthquake triggered: other mechanisms for producing laterally extensive synchronous failure do not apply. Triaxial shear measurements show a S_u/σ' ratio of typical slope sediment of 0.48, implying considerable stability. However, some silty muds have Atterberg limits that suggest susceptibility to liquefaction under cyclic loading, particularly in Holocene deposits and by analogy those of past full interglacials. Basal failure planes of some large failures correspond with either the last interglacial or the MIS 6 glacial maximum. Comparison with seismological models suggests that the observed slope failures represent earthquakes ranging from Mw ~5.6 to ~7.6. Mean recurrence interval of M = 7 earthquakes at any point on the margin is estimated at 30 ka from seismological models and 40 ka from the sediment failure record. In northern Flemish Pass, a spatial cluster of several failures over 30 ka preceded by a long interval with no failures suggests that some other mechanism has preconditioned the slope for earthquake triggering. There is circumstantial evidence that such preconditioning is related to excess pore pressures, released by fluid drainage at head scarps.