Modeling experiments on the deceleration and reactivation of Kangerlussuup Sermusa, West Greenland

Abstract:

Seasonal variations in outlet glacier velocity due to basal sliding are well-documented and typically involve acceleration early in the melt season due to enhanced sliding as a result of inefficient drainage of surface water reaching the bed. However, velocity observations from Kangerlussuup Sermusa (KS) in West Greenland contradict this pattern. Instead, ice velocity at KS shows no significant change in early spring compared with the previous winter. This sluggish response of the glacier to spring melt is often followed by an extreme, and short-lived, deceleration. For example, in August 2010, the lower 20 km of the trunk decelerated from about 1600 m a^-1 to less than 250 m a^-1; this event was followed by a rapid reactivation back to the previous velocity in less than 60 days. Available records since 2006 show that the sequence of steady spring velocity, followed by summer deceleration, and rapid fall reactivation occurs annually; however, the magnitudes of deceleration vary. In this regard, the response of KS to regional environmental forcings is unique compared to its neighboring glaciers.

In this study, we investigate whether the unique behavior of KS can be explained by the interaction between changes in basal conditions and the local geometry of the glacier. We model the glacier flow by solving full-Stokes equations using the finite element method in the open-source FEniCS framework. Assuming isothermal ice within the lower trunk, we run experiments on the mechanical properties and boundary conditions of the glacier. These experiments include spatio-temporal changes in basal slipperiness, periodic melt-water influx to the bed, and ice viscosity variations due to changes in melt-water supply to the bed. We also conduct sensitivity analyses on the glacier flow with different ice geometries (e.g. thickness and surface slope) to investigate conditions under which we can produce the unique seasonal behavior of KS. Finally, we assess the impact of the combination of these parameters on the ice flow and evaluate the credibility of our hypothetical scenarios based on available measurements.