Petrology, rock magnetism and stratigraphy of late quaternary Ice-rafted detritus at the southeast Grand Banks slope of Newfoundland, Atlantic Canada

Abstract

The Quaternary sediments of the Atlantic Canada margin usually contain numerous evidences of paleoclimate oscillations. The short-lived, abrupt and episodic surges of sediment-laden armadas of icebergs calved from Laurentide Ice Sheet (LIS) into the North Atlantic Ocean via ice stream conduits has deposited ice-rafted detritus (IRD) all across the northern North Atlantic oceanic basin. Such events have had considerable impact on ocean water circulation system (i.e. Atlantic meridional overturning circulation – AMOC) and deep water formation owing to the huge influx of meltwater delivered by subglacial discharge and iceberg melting. This catastrophic instability of the LIS has resulted in a stratigraphic sequence of detrital-rich Heinrich Layers (HL) during the cold phases of the Late Quaternary period. As IRD fragments are the direct evidence of the interplay between ice sheets and the North Atlantic Ocean, we still lack understanding of their hidden information on ice sheet dynamics, location of ice margins, the provenance and spatial dispersion of distinct terrigenous materials and the petromagnetic properties and contributions of different IRD species to magnetic HL records. In this PhD thesis, I investigated a sediment core (GeoB 18530-1) acquired at the foot of Southeast Grand Banks Slope of Newfoundland, that represents a continuous Late Pleistocene stratigraphic sequence of Heinrich Event layers 1 - 6 (H1 – H6). Discrete sample sets were taken at a very high resolution of 2 cm for various analyses including digital light microscopy, thin section petrography, X-ray fluorescence, rock magnetics, and grain-size sieving. In a first step of the thesis project, a total of 8243 coarse-grained (>1 mm) IRD particles were petrologically identified and their lithologies petrographically classified into 22 microscopically distinguishable categories. The abundances of these IRD lithologies were determined by counting individual particles at a high resolution of 2 cm steps throughout the continuous stratigraphic sequence of HLs (H1 – H6). This revealed the temporal and spatial change in the deposition of each IRD lithology. It was found that dolomitic IRD predominates the mid-part of all HLs whilst muscovite-biotite granite was found in greater abundance at the top and bottom of the HLs. Additionally, the statistical investigation of the count records of the 22 IRD lithologies manifested three well-defined groups based on whether their deposition is higher within HLs or interlayers or similar in both. Compositionally, it was found that H1, H2, H4, and H5 are kindred to each other while H3 and H6 are individually distinct. The second step of the thesis was an advancement of the finding from the initial part where the petrological information that was gleaned is used for magnetic analysis of particularly HLs and individual IRD lithologies. The pronounced magnetic susceptibility signal of HLs is usually intriguing compared to the interlayer units thus, it is momentous to discern the IRD lithology that influences the signal. Though, the >1 mm IRD underrepresents the total amount of IRD where finer grains are much more abundant, a lithology-specific identification can only be conducted using the coarser IRD fractions. From the magnetic properties of the coarse IRD it was found that out of the 22 different lithologies, as many as 16 possess a relatively lower magnetic susceptibility than the background sediment. Only 6 IRD lithologies are typified by a higher susceptibility among which the muscovite-biotite granite is four-fold higher magnetic than the background. With the complementary XRF data is was revealed clearly that the viii susceptibility enhancement by the (K/Fe denoted) granite is partly compensated by the (Ca/Sr represented) dolomitic IRD. Therefore, with the muscovite-biotite granite being second in abundance to dolomitic IRD, it mainly controls the pronounced magnetic susceptibility signals of HLs. In modeling magnetic susceptibility based on coarse-grained IRD count data a misfit of the model was observed proposing that IRD sedimentation does not only occur through detritus rain-out from icebergs but rather through other transport processes like advection or gravitational flow, at least for the terminal part of the sediment’s trajectory. Lastly, traditional magnetic susceptibility measurements are obtained in ways, where the inhomogeneity of sediment records is mostly played down. This is because the sensor coils of whole-core magnetic susceptometers average signals over a large volume, thereby integrating the microscale susceptibility distribution over a considerable volume. Split-core susceptibility logs produced with a spot sensor at high resolution average the magnetic signal over a much smaller volume of a material. Using a novel protocol, this spot sensor technique provides the freedom to scan across an entire split core. By logging multiple parallel tracts at a high resolution of 1 cm along and between each tract, the derived contour plot reveals coarse IRD distribution, where the color spectrum denotes lateral variability in magnetic susceptibility.