Better forecasts of melting ice

Associate Professor Josefin Ahlkrona will receive funding from the Knut and Alice Wallenberg Foundation to recruit an international researcher for a postdoctoral position at the Department of Mathematics, Stockholm University.

The aim of this project is to enhance the efficiency, accuracy and theoretical understanding of computer models for melting ice sheets. Antarctica’s and Greenland’s melting ice sheets along with glaciers are the primary contributors to rising sea levels as Earth’s climate warms, and this may have serious consequences for coastal infrastructure and ecosystems. To prepare for this – by building dams, urban planning and societal adaption – we need reliable ice melt forecasts.

Current computer models that simulate movements and changes to the ice have considerable uncertainty. This limited accuracy is partly because physically rigorous models are computationally prohibitive at the high spatial resolution (less than 1 km) and long temporal spans (100–100,000 years) required. Researchers therefore tend to rely on simplifying assumptions, that lead to significant uncertainties in their forecasts.

The solution may come from a new method for time-stepping schemes in ice-sheet models. The method allows for taking long time-steps in some parts of the simulation. The aim is to develop an algorithm that automatically selects the largest possible time step at any given moment without compromising accuracy. At the same time, rigorous error analysis should clarify how the accuracy of the simulation depends on the selected time-step. Preliminary estimates indicate that these calculations could thus be 10 to 100 times faster than traditional methods.

Photo: Stefano Papazian