Lessons from the short history of ice sheet model intercomparison

Abstract. Intercomparison should include measurement of differences, between model results and observations, among the model results themselves, or between model results and exact solutions. The processes of measuring differences and critically analyzing those differences are vital. Without such measurement as a component of intercomparison, the only expected benefits of an intercomparison project are participation, possibly the discovery of communal confusion, and the establishment of public, non-proprietary data sets.