Flow Is Plastic, It's Fantastic

Abstract:

Capturing the spatial variability in the flow is essential to any credible modeling effort of the Greenland ice sheet, yet it remains challenging for several reasons. First ice flow results from gradients in gravitational driving stress, making accurate ice thickness measurements indispensable. Second the way the driving stress is balanced between vertical shearing, and basal and lateral drag must be either parametrized or estimated from inversion of surface data. Third solving the Stokes equations is computationally expensive, prohibiting long integration times and/or large scale parameter sensitivity studies. In this work we address the latter two. We demonstrate that much of the observed flow variability can be explained by combing a computationally-efficient hybrid stress balance model, a sliding parametrization that includes the effect of basal effective pressure, and high-resolution ice thickness. We show how uncertainties in ice thickness and basal sliding influence ice discharge through the grounding line. In addition to commonly-used metrics of success such as root mean square error, we validate our modeling efforts by comparing simulated to observed profiles of mass flux through flux gates around the periphery of Greenland. We conclude that the cumulative mass flux across flux gates is an insufficient metric to assess the model's skill to simulate the spatial variability in ice flow.