Evidence of Ice Flow Switching from Carlson Inlet to Rutford Ice Stream Based on Polarimetric Radar

Abstract

The flow of polar ice is controlled by its viscosity that is spatially variable and depends, among other factors, on the orientations of the anisotropic crystals of ice, often referred as crystal orientation fabric. Ice crystalizes in planes represented by the c-axis, a direction perpendicular to the main plane of the crystals, and it is highly anisotropic: the viscosity along the c-axis is two orders of magnitude greater than across, and hence it can be a key factor for ice flow modelling. Interestingly, the ice crystals rotate to accommodate ice flow, similarly to how dominoes tend to align under strain, and ice c-axis orientation evolves to be perpendicular to the direction of the maximum strain rate. Thus, ice flow and crystal orientation fabric are related. However, critically for our work, c-axis evolution is not instantaneous and, particularly in currently slow deforming ice, crystal orientation fabric contains traces or past ice flow conditions. Here, we use data from the British Antarctic Survey (BAS) airborne radar PASIN2 for deep ice sounding in Rutford Ice Stream, collected during the 2019-2020 season, to derive crystal orientation fabric. Because electromagnetic waves propagate at different speeds depending on the wave polarisation being parallel or perpendicular to the c-axis, an optical phenomenon called birefringence, we compare signals from different antenna orientations in our array to derive englacial crystal orientation fabric. We then compare our radar-derived crystal orientation fabric with strain rate derived from satellite ice flow observations. To aid the interpretation, we use a numerical model that bounds the prediction of ice fabric from ice flow under different assumptions. We find that Carlson Inlet, now stagnant, show traces of past fast flow on its crystal orientation fabric. This agrees with previous studies that suggest flow-switching and water-piracy between neighbouring Carlson Inlet and Rutford Ice Stream (Vaughan et al., 2008). Our method provides a framework to investigate the timing and the causes of the flow-switching event. More in general, we demonstrate the use of existing and future airborne polarimetric data to investigate recent changes in the cryosphere.