BREAK UP OF ICE SHEETS: AN INTEGRATED NUMERICAL AND EXPERIMENTAL APPROACH

In the coming decades the need to understand how the world environment is evolving due to climate change will be great. To face these problems policy makers as well as the general public need information with ever-increasing levels of certainty. While a large suite of models exists to predict the oceanic and atmospheric responses to increased greenhouse gas emissions, very few models include the effects of ice that is not assumed to be at steady state under the climate warming scenarios tested. This is largely due to the numerical challenges associated with the simulation of ice: the material exhibits features typifying both a solid and non-Newtonian liquid. In particular, the break down and calving of ice shelves is, however, an important aspect of the mechanisms leading to mean sea level rise and we need models that can reliably simulate these dynamics.

Observations have shown that the bending that occurs as ice transitions from resting on land to floating in water promotes the failure of ice from the bottom up, called basal crevasses. These features appear with characteristic regularity in spacing, often persisting within the ice for long distances and promoting the eventual calving of ice. In order to simulate an idealized version of this scenario, we test the effect that a forced bend has on a parallel-sided slab. We use a visco-elastic finite element model called DynEarthSol to perform two series of experiments where we systematically change the ice thickness and the slope angle. One of the limitations of the numerical approach is that there is no real breaking of the ice and therefore no physical separation of the icebergs. To address this limitation we performed experiments on a two-phase model material using the same initial conditions as in the numerical experiments. The numerical and experimental results show that with an increase of the ice thickness and slope the crevasse spacing is increasing and becomes more regular.