An investigation into the stochastic physics of iceberg calving and universal calving laws

Regents Of The University Of Michigan - Ann Arbor, Ann Arbor MI

Investigators

Abstract

Given the sensitivity of glacier dynamics to the position of the glacier front, without a parameterization of calving, to determine the position of the glacier front, which can be implemented into continental scale ice sheet models, our ability to accurately forecast changes in sea level from the ice sheets in response to climate change will be seriously compromised. However, despite the important role iceberg calving plays in contributing to observed changes in Greenland and Antarctica, the glaciology community still does not have a parameterization of iceberg calving that can explain the diversity of observed calving behavior. The challenge that must be overcome is that the parameterization of calving must not only be capable of explaining the wide variety of observed behaviors, it must also be sufficiently simple that it can be incorporated into continental-scale ice sheet models without introducing a prohibitive computational tax on the models and modelers. Funds are provided to develop a parameterization of calving that (i) can be applied to grounded and floating glacier termini and (ii) can be implemented into regional and/or continental scale ice-sheet-shelf models. The primary goal is not to predict the specific behavior of a particular outlet glacier, but instead to attack the problem from a broader perspective and try to understand the qualitative envelope of behavior and stability of ice tongues, outlet and tidewater glaciers as a group. The calving law will be formulated as a stochastic process using techniques borrowed from the statistical-mechanical theory of fracture of disordered material. The newly derived calving law will be incorporated into a flowline ice-sheet-shelf model to simulate a range of different environmental conditions, basal topographies, and sliding laws. The core of this proposal focuses on determining the behavior predicted by the calving law throughout the entire range of parameter space.