Marine Ice Sheet Model Intercomparison Project

page maintained by F. Pattyn @ ULB

Benchmark Experiments for Marine Ice Sheet Models and Grounding Line Migration

FULL TEXT (PDF): version 24/11/2009


Christian SCHOOF, Department of Earth and Ocean Sciences, University of British Columbia, Vancouver, Canada (email: cschoof@eos.ubc.ca)

Richard HINDMARSH, British Antarctic Survey, Cambridge, UK (email: rcah@bas.ac.uk)

Frank PATTYN, Laboratoire de Glaciologie, Départment des Sciences de la Terre et de l'Environnement, Université Libre de Bruxelles, Brussels, Belgium (email: fpattyn@ulb.ac.be)


The last few years have seen a great deal of effort invested by various research groups in developing numerical marine ice sheet models that are intended to represent accurately the motion of an ice sheet grounding line, and this effort has seen a variety of models emerge. It is at present unclear to what extent these models agree with one another, or indeed, how well they are able to model real marine ice sheets.

The purpose of the proposed experiments is to address the former question: how well do our models agree with one another? In addition, the purpose of the experiments is also to assess how well the numerical schemes used solve the partial differential equations on which they are based. Not all the models used in marine ice sheet simulations deal with the same equations, but many of them share basic characteristics. For instance, most include a representation of the longitudinal stresses required to couple ice shelves to the ice sheet. We therefore do not restrict this intercomparison to a single type of model, but invite a variety of models described below to participate.

The impetus for the intercomparison has been provided amongst others by the recent papers of Vieli and Payne (2005), Pattyn et al. (2006), Hindmarsh (2006) and Schoof (2007), which have demonstrated not only the strong possibility of numerical artifacts in marine ice sheet simulations, but also the importance of grid resolution and of accurate represention the sheet-shelf transition zone. These papers have also shed some light on the question of ice sheet stability. No stable steady states have been found on upward-sloping ice sheet beds, in line with an early hypothesis of Weertman (1974), but some of the numerical results in Vieli and Payne (2005) and Pattyn et al (2006) have left open the possibility of `neutral equilibrium': the possibility that a perturbation in grounding line position in a steady-state ice sheet could lead to a similar, but distinct, new equilibrium shape. This possibility has, however, been discounted by Schoof (in press), who also found that marine ice sheets may undergo dramatic and irreversible changes - hysteresis - under changes in physical properties (sliding, ice viscosity) or external forcing (accumulation rates, sea levels). The basic questions that arise from these papers and earlier work are the following:

What type of experiments?

To facilitate comparison, we will use the simplest physical set-up that can represent a marine ice sheet:

One recent development that will hopefully facilitate the intercomparison is the development of a boundary layer theory for sheet-shelf interactions (Schoof, 2007). This theory takes a complementary approach to numerical marine ice sheet models: it uses a systematic set of approximations to parameterize the sheet-shelf transition zone in a simpler, shallow ice model. The difference with other models is that the theory is not specific to a particular numerical method. Importantly, it allows steady states to be computed semi-analytically and at low computational cost, and we will use these steady states to guide the experimental design. We emphasize, however, that the boundary layer theory is itself only approximate, and part of the objective of the experiment is to test it against other models.


PDF with full description of the experiments

The following MISMIP_distribution .tar file contains SMsolver2, which is a MATLAB scripts that solves for steady state profiles based on Schoof's boundary layer profiles. Specific instructions are given in the header for the script. Just input user_grid at the top of the program and it plots the computed profiles and outputs ice thickness user_thickness at your grid points. The header explains how to change the bed profile and how to change parameters.

MISMIP_distribution.tar (updated 29/04/2008)

SMsolver2_function.m (all-in-one function - 24/11/2009)

Uploading your result files

Results files of the experiments can be uploaded on the ftp-site of the ULB at the following address:


When you have uploaded a file, please send an email to Frank Pattyn (fpattyn@ulb.ac.be) because these files will only reside here for two weeks.


The MISMIP intercomparison has been closed.


Results of the intercomparison are published in The Cryosphere. The full citation of the paper is: Pattyn, F., Schoof, C., Perichon, L., Hindmarsh, R. C. A., Bueler, E., de Fleurian, B., Durand, G., Gagliardini, O., Gladstone, R., Goldberg, D., Gudmundsson, G. H., Huybrechts, P., Lee, V., Nick, F. M., Payne, A. J., Pollard, D., Rybak, O., Saito, F., and Vieli, A.: Results of the Marine Ice Sheet Model Intercomparison Project, MISMIP, The Cryosphere, 6, 573-588, doi:10.5194/tc-6-573-2012, 2012.