Abstract
One of the most substantial challenges posed to us today is global warming induced by human activity. Some recent studies have shown ice loss from polar ice sheets such as the West Antarctic Ice Sheet (WAIS) is accelerating towards tipping point and will result in ice sheet collapse and destructive sea level rise felt globally. Current ice sheet models are hindered by our lack of knowledge on the mechanical behaviour and response of flowing ice at a continental scale. Ice flow comprises a combination of sliding at the base of the ice sheet and internal deformation. The latter is controlled primarily by englacial temperature, the arrangement of ice grains into a crystallographic preferred orientation (CPO), and the stresses acting on the flowing mass. While many experimental studies in the past have contributed to our understanding of the mechanical behaviour of ice under stress, these are based largely on pure, isotropic samples deformed to low (e <20-30%) strains. Flow laws used in ice sheet modelling are defined by these studies, based specifically on the behaviour of ice before weakening is initiated. Natural ice contains soluble and insoluble impurities, mechanical heterogeneities like layering, fracturing and bubbles, and can flow and deform under different kinematic regimes resulting in changes to CPO and mechanical anisotropy throughout its long strain history.
This study combines experimental data and observations from a shallow ice core to contribute to the understanding of the mechanical behaviour of natural ice. Ices with a natural chemical composition and initial isotropic fabric have been deformed under uniaxial compression at varying strain rates (10-4, 10-5, 5x10-6 s-1) and temperatures (-10 and -30°C). Mechanical data demonstrates an increase in ionic concentrations strengthens ice only at warm temperatures, becoming more prominent at slow strain rates. Microstructural analysis shows impure ices with microstructures typically indicative of weakening. It is proposed that solid solution strengthening and grain boundary strengthening, processes typically seen in metals, may be strengthening the impure ices. This could have significant implications for marginal and basal regions of ice sheets, ice streams and glaciers, suggesting models could be overestimating weakening during ductile flow in these high temperature, relatively slow flowing regions.
A microstructural record of the near surface (<33 m) Priestley Glacier shear margin has been compiled and the seismic and dielectric anisotropy of the core has been constrained. The area is dominated by simple shear kinematics and is south of the glacier’s grounding line, so the glacier is floating, and basal traction does not need to be accounted for. The core samples have a very strong c-axis cluster CPO, which rotates generally from sub-perpendicular to 50° antithetic to the shear plane from the surface to 33 m depth. Microstructural and misorientation analysis indicates the shear margin is deforming by a combination of basal and nonbasal slip, with basal slip becoming more prominent below 5-10 m. Subgrain rotation is an active recovery mechanism, and grain boundary migration recrystallisation is active most effectively in shallow, seasonally perturbed (air temperature-controlled) ice and below 25 m, likely due to an increase in strain energy driving migration. Seismic and dielectric anisotropy strongly correlates with CPO strength and orientation with depth. The P-wave velocity patterns in core samples and observed horizontal velocities are geometrically similar, but the CPO rotation seen in the core is not present at depth. It is proposed that rigid body rotation is causing these rotations, which do not extend beyond 100 m. This highlights the usefulness of seismic datasets in attaining deep ice fabrics, but the difference in mechanical behaviour of ice at the near surface and ice at depth should be acknowledged in future ice sheet models.