Skip to main content

Improving the Physics in Ice Sheet Models

Final Report Summary - IMPISM (Improving the Physics in Ice Sheet Models)

This project addresses the numerical modelling of ice sheets and their contribution to sea level rise. Melting of ice masses is one of the most obvious and dramatic consequences of climate change, with the potential to change sea levels by over 60m in the long term. Changes in the melting behaviour of glaciers also affects freshwater availability and flood hazards, with potetential impact on millions of people globally. There is a growing need to base strategic decisions on computer simulations that forecast the expected and possible range of future behaviour. But existing models fail to account for much of the physics that is known to drive rapid change in ice sheets. This project aimed to improve the representation of this important physics in computer models. It has provided improved parametersations of some processes, and yielded fundamental new insights into the mechanistic behaviour of ice sheets. Some of the newly developed models are currently being implemented in numerical ice sheet models and will help to facilitate improved predictions of future behaviour.

The specific objective of this project was to target aspects of ice-sheet physics that are poorly represented in current models. These include basal sliding, basal hydrology, temperate ice (ice that is at the melting temperature and contains a small amount of water), melting at the ice-ocean interface, and the compaction and thermodynamics of firn (snow transitioning into ice at the surface). The work has involved developing and solving fluid mechanical models of these processes, comparing their predictions against observations, and using these detailed models to derive improved parameterisations for use in large scale ice-sheet simulations.

Specific scientific highlights include the following:

Glacial hydrology: a new method has been developed for modelling water flow under ice sheets. This accounts for water that resides in sediments, in larger cavities and lakes, and water that flows through subglacial 'tunnels', melted into the ice. The new approach has been adopted as a component of some ice-sheet models, where it is important in routing meltwater to the ocean. A specific prediction of this model is the way in which channels are expected to enlarge where they emerge into the ocean, and this has been used to help explain some recent Antarctic observations of a growing 'esker' (sediment ridge under the ice).

Basal sliding: Coupling the models of subglacial water with a 'friction law' for ice flow has enabled better representation of the changes in lubrication that occur in response to seasonal melting of the ice sheet. The model has lead to the discovery that reduced lubrication, which causes the ice to move more slowly, can in some cases lead to faster ice loss. This runs counter to standard thinking, and was previously unrecognised.

Temperate ice: A new model has been develoed to describe the dynamics of temperate ice (ice at the melting temperature that contains interstitial water and which is throught to be much more deformable than cold ice). This model predicts the temperature and water content of such ice, and allows for comparatively easy inclusion in ice-sheet models compared to existing codes. The model is now being used to understand the formation of ice streams (particularly fast moving regions of ice).

Snow and firn dynamics: A new model has been developed to describe the heat and mass transfer of meltwater through snow on the surface of ice sheets. This model gives specific predictions of the proportion of meltwater that we expect to be stored, refreeze, and run off; and will consquently improve projections of sea level rise from the Greenland ice sheet where surface melting is the dominant mass loss.

More information can be found at