Streaming flow on polythermal mountain glaciers: In-situ observations on Jarvis Glacier, Alaska

Accelerated melting of glaciers and ice caps has raised serious concerns about sea level rise. As we work towards a solution to address these concerns, it has become a chief priority to rapidly improve predictions of future changes in global ice mass balance. Numerical simulations projecting ice loss have uncovered a strong sensitivity to mechanical and/or rheological weakening of the shear margins of streaming ice. To accurately project sea level rise, future models will require careful treatment of shear margins. This necessitates a deeper understanding of the flow dynamics at shear margins and how streaming flow relates to the constitutive flow law for ice.

 

We developed an open source inexpensive tilt sensor (∼20% the cost of commercial sensors) for studying ice deformation and installed our tilt sensor systems in two boreholes drilled close to the shear margin of Jarvis Glacier, Alaska to obtain kinematic measurements of streaming ice. We used the collected tilt data to calculate borehole deformation by tracking the orientation of the sensors over time. The sensors' tilts generally trended down-glacier, with an element of cross-glacier flow in the borehole closer to the shear margin. We also evaluated our results against flow dynamic parameters derived from Glen's exponential flow law and explored the parameter space of the stress exponent n and enhancement factor E. Comparison with values from ice deformation experiments shows that the ice on Jarvis is characterized by higher n values than that is expected in regions of low stress, particularly at the shear margin (~3.4). The higher n values could be attributed to the observed high total strains coupled with potential dynamic recrystallization, causing anisotropic development and consequently sped up ice flow. Jarvis' n values place the creep regime of the ice between basal slip and dislocation creep. Tuning E towards a theoretical upper limit of 10 for anisotropic ice with single-maximum fabric reduces the n values by 0.2.