Polar ice sheets and global sea level: how well can we predict the future?

Geophysical models are based on hypotheses that are often derived from theoretical arguments or from observations, or a combination of both. Owing to the open nature of geophysical systems, these models cannot be verified in the sense that it cannot be proven that the model is an accurate representation of the physical reality, At best these models can be confirmed by comparing model predictions against independent observations. The more such observations the model agrees with, the greater confidence can be placed in the model and its reliability as a basis for decision making, A review of mass balance models used to predict past and future contributions to global sea level change arising from changes in snow accumulation and surface ablation on the polar ice sheets is presented. Observations on which these models are based are ambiguous and there is evidence suggesting that these models do not capture all relevant physical processes, some of which-such as changes in atmospheric circulation patterns-may have an equally important effect on changes in surface mass balance as does the direct temperature forcing. In the context of greenhouse-warming-induced sea level change, uncertainties in model parameters are sufficiently great to yield a range of projected contributions from Greenland and Antarctica that encompass sea-level lowering and rise in 2100 A.D. for each of the warming scenarios considered. The uncertainty associated with ice sheet mass balance parameterizations is of similar magnitude as that associated with temperature projections. These results are broadly similar to those of the Third Assessment Report issued by the Intergovernmental Panel on Climate Change (IPCC). (C) 2002 Elsevier Science B.V. All rights reserved.