Search

The mechanical properties of glacier beds play a fundamental role in regulating the sensitivity of glaciers to environmental forcing across a wide range of timescales. Glaciers are commonly underlain by deformable till whose mechanical properties and influence on ice flow are not well understood but are critical for reliable projections of future glacier states. Using synoptic-scale observations of glacier motion in different seasons to constrain numerical ice flow models, we study the mechanics of the bed beneath Hofsjökull, a land-terminating ice cap in central Iceland. Our results indicate that the bed deforms plastically and weakens following incipient summertime surface melt. Combining the inferred basal shear traction fields with a Coulomb-plastic bed model, we estimate the spatially distributed effective basal water pressure and show that changes in basal water pressure and glacier accelerations are non-local and non-linear. These results motivate an idealized physical model relating mean basal water pressure and basal slip rate wherein the sensitivity of glacier flow to changes in basal water pressure is inversely related to the ice surface slope. [ABSTRACT FROM PUBLISHER]

The article focuses on a study conducted by the authors related to models of plan-view marine ice-sheets. It states that they conducted perturbation experiments which specified spatial variation in basal sliding and helped in evolution of grounding lines. It discusses intercomparison of the ice2sea MISMIP3d model. It highlights that examination of normal stresses perpendicular to the grounding lines reveals extension across the grounding lines.

The article presents the Tiling method to couple ice flow models of different orders of complexity. The goal of the method is to determine the spatial extent of where higher-order models are used to lessen the computational cost while retaining the model precision. A geometry for which all models yield similar results is used to examine the consistency of the method and a geometry where a full-Stokes model is needed in the vicinity of the ice front is used. The results reveal that the hybrid models show important improvements over mono-model approaches and lessen computational times.