Magnetic Fabrics and their Application to Basal Crevasse Fills, Flàajökull, Iceland

Wednesday, 17 December 2014
William Russell Jacobson Jr, University of Wisconsin Milwaukee, Milwaukee, WI, United States and Thomas Hooyer, Geosciences Department, Milwaukee, WI, United States
Long, linear features consisting of sediment and ice, approximately 100 m long and 20 cm wide, run transverse to the margin of Flàajökull, an outlet glacier of Vatnajökull Ice Cap. These features, interpreted as basal crevasse fills, are thought to have formed by debris and water injected into a void under pure tension or a combination of tension and shear in response to an ice pressure drop at the bed. The debris content of these basal crevasse fills are between 5 to 10% by volume. The formation of these basal crevasse fills is uncertain, because direct observation is difficult.

To study these basal crevasse fills, we used the orientation of magnetic grains using anisotropy of magnetic susceptibility (AMS) to guide us in understanding their kinematics. The AMS technique is superior over other fabric methods because a three-dimensional susceptibility ellipsoid is used to determine strain. We sampled two basal crevasse fills and obtained 86 ice core samples for AMS analyses. We also cut nine blocks of ice to determine the magnetic mineralogy, grain size of the magnetic carrier and c-axis orientation of the ice. The AMS results demonstrate that at one fill, the fabric was most likely formed by a combination of pure shear and simple shear. At the second site the AMS results were not well clustered possibly due to insufficient strain. Hysteresis and high temperature susceptibility tests indicate a magnetite carrier that was silt-sized or smaller. Thin sections used to evaluate c-axis fabrics display several multi-maximums that suggested that the fabric developed through recrystallization during deformation. It is inferred that grain scale processes reveal deformation by grain-boundary migration and grain nucleation. Magnetic particles appear to have behaved as passive markers following the March model (1932). Given this data set, we argue that the crevasse fills were formed by multiple processes including injection of turbid waters followed by in situ-freezing and deformation.