Microstructure, CPO and rheology of ice: scaling from the laboratory to ice sheets and ice shelves

Abstract

Deformed polycrystalline ice has very strong crystallographic preferred orientations (CPO) with associated elastic and viscous anisotropy. When the loading that drives ice sheet flow changes, for example following an ice shelf collapse, the rate of response will depend upon the inherited CPO and its subsequent evolution in the new stress configuration. Lateral and vertical contrasts in mechanical properties occur where bodies of ice with different CPOs are juxtaposed. Ice shelves (where the ice sheet has moved offshore and is floating) such as the Ross Ice Shelf (RIS) provide good examples. Feeder glaciers with different CPO converge to create an ice shelf with lateral heterogeneity that controls subsequent mechanical behaviour, including break up. The mechanical response of the Earth's ice sheet systems to climate forcing depends, in part, on the evolution of the CPO. Thus, there is a need to improve our quantitative understanding of CPO evolution in ice and its impact on rheology. Cryo-EBSD data from new axial compression experiments demonstrate that ice deformation microstructures change as a function of both temperature and differential stress/strain rate. At high T and low stress, strain induced grain boundary migration (SIGBM) is fast. The CPO is dominated by grains with high resolved stresses on the basal (easy slip) plane. In axial compression, poles to the basal plane are arranged in a cone around the compression axis. At lower T and higher stress, deformation is dominated by lattice rotation, polygonization and grain rotation. Low T axial compression experiments have the poles to the basal plane clustered parallel to compression. Polar terrestrial ice sheets deform at rates at least two orders of magnitude slower than any lab experiment, so CPO prediction requires scaling relationships. We will present a preliminary scaling relationship for CPOs in axial compression. Most natural deformation however has significant simple shear or constriction and we will show results from some new experiments with these kinematic components. To validate scaling relationships to natural strain rate conditions, we measure CPOs from naturally deformed ice. We will present cryo-EBSD data from Antarctic glacier samples: some of these natural CPOs are explicable within the framework of existing experiments, others are not and require new experimental effort. We will also show ultrasonic measurments of P- and S-wave velocity anisotropy from natural samples and associated with some laboratory deformation experiments. The coincidence of measured velocity patterns with those calculated from the CPO gives us confidence to use seismic investigations to provide proxy CPO data on the scale of the thickness of ice sheets. We will present new data from an active source survey that used two seismometers frozen into the 220 m thick McMurdo ice shelf at depths of 39 m and 189 m. P- and S-wave velocities and S-wave splitting along different ray-paths enable us to constrain the bulk CPO in a way that would be much more difficult using only surface receivers. Future investigations using this approach should improve our ability to assess CPO and relate this to rheology at natural strain rates in ice sheets.