Arbeitsbericht NAB 10-34

Glacial Erosion ModellingResults of a workshop held in Unterägeri, Switzerland, 29 April – 1 May 2010

Deeply incised troughs and overdeepened valleys are common features in the Alps and in the Alpine foreland. These troughs and valleys as well as their sediments are of large practical and scientific interest, influencing the dynamics and hydrology of Pleistocene ice masses, presenting opportunities for glacial and climate reconstruction, and raising management issues related to aggregate, groundwater and hydrocarbon resources and radioactive waste disposal in deep geological repositories. Although the mechanisms and principles of glacial erosion are generally known, the formation of deeply incised troughs and overdeepened valleys beneath the frontal reaches of glaciers in the distal foreland of the Alps, referred to as deep glacial erosion, remains incompletely understood. In particular, the question as to when, where and how often future glaciations can lead to deep glacial erosion is of great importance for the siting and long-term safety of radioactive waste repositories in northern Switzerland.

In collaboration with the University of Zürich, the National Cooperative for the Disposal of Radioactive Waste (Nagra) organized and held a workshop in April 2010 aimed at evaluating the state-of-the-art of modelling glacial erosion as a means for developing a better understanding of the subglacial processes governing landscape evolution in the Alpine foreland of northern Switzerland in past as well as future cold environments. During the workshop a draft assessment was developed by an international group of leading experts on what quantitative information glacier erosion modelling can provide in view of safety aspects related to radioactive waste repositories in Switzerland.

During the last decade, three-dimensional models of ice flow have been developed that account for sediment erosion and transport beneath glaciers and ice sheets. The deformation and thermomechanics of glacier ice are reasonably well understood and ice cover extent and ice flow paths can be modelled with good confidence. In contrast, basal processes including glacier hydrology, ice-bed interaction, sliding, sediment transport and interactions with permafrost are comparatively poorly understood and require considerable additional model development to be usefully incorporated into comprehensive coupled models. In this respect, glacier hydrology was particularly acknowledged by the experts as being important for realistic simulations of ice dynamics and erosional processes.

There was general agreement by the experts that prognostic modelling is beyond our present capabilities and not recommended because of substantial uncertainties in the quantification of basal processes as well as the relevant climate forcing for future glacial cycles. Instead, it was suggested that diagnostic modelling has potential value for understanding processes and quantifying uncertainties and for testing sensitivities and process parameterizations.

According to the experts, modelling strategies that hold promise include ensemble modelling to explore the range of outcomes over the entire range of uncertainty in variables and extremal (end-member) modelling to bracket best and worst cases. These approaches can lead to the identification of potentially important processes and parameters and thus the component models needed (e.g. glacier hydrology, glacial sedimentary processes) for comprehensive coupled models. In deciding on an approach to the simulations, it was suggested that the relative value of ensemble runs with simpler models versus fewer runs with more complex models should be considered.