Quantifying Feedbacks Between Ice Flow, Grain Size, and Basal Meltwater on Annual and Decadal Time-Scales Using a 2-D Ice Sheet Model

Abstract

Ice sheet flow is strongly controlled by the conditions at the ice-bed interface. While these processes are hard to observe directly, comparisons between numerical modeling and ice surface observations can be used to indirectly infer subglacial processes. Specifically, seasonal summer speed up near the margin of the Greenland Ice Sheet (GIS) has been linked to the presence of subglacial water. For decades, the Glen flow law has been the most widely-accepted constitutive relation for modeling ice flow. However, while the Glen law captures the temperature-dependent, nonlinear viscosity of ice, it does not explicitly incorporate ice grain size, which has been shown in laboratory experiments to influence ice rheology. To compensate for the lack of explicit grain size dependence, ice sheet models often utilize an “enhancement factor” that modifies the flow law to better match observations, but does not provide insight into the physical processes at play. Using a grain size sensitive rheology that incorporates grain size evolution due to dynamic recrystallization and grain growth, I model the effects of seasonal variations of subglacial hydrology in a 2-D vertical cross-section of ice flow on both annual and inter-annual timescales. The presence of subglacial water reduces the frictional coupling between the ice and the bed. Here I simulate the presence of water at the ice-bed interface during the melt season using patches of free-slip and explore a range of patch sizes and geometries to investigate their role in modulating ice surface velocities and grain size within the ice. I compare modeled winter and summer surface velocities to observations taken on the western margin of the GIS and find that realistic surface velocities are achievable using agrain size sensitive flow law without the introduction of an enhancement factor. Further, the grain size of the internal ice responds on an inter-annual timescale to these seasonal forcings at the bed, potentially leading to long-term changes in surface velocities.