Reference
Eisenman, I., Untersteiner, N. and Wettlaufer, J.S. 2007. On the reliability of simulated Arctic sea ice in global climate models. Geophysical Research Letters 34: 10.1029/2007GL029914.
What was done
The authors used two standard thermodynamic models of sea ice to calculate equilibrium Arctic ice thickness based on simulated Arctic cloud cover derived from sixteen different global climate models (GCMs) that were evaluated for the IPCC's Fourth Assessment Report.
What was learned
Eisenman et al. report there was a 40 Wm-2 spread among the sixteen models in terms of their calculated downward longwave radiation, for which both sea ice models calculated an equilibrium ice thickness ranging from one to more than ten meters. However, they note that the mean 1980-1999 Arctic sea ice thickness simulated by the sixteen GCMs ranged from only 1.0 to 3.9 meters, which is a far smaller inter-model spread. Hence, they say that they were "forced to ask how the GCM simulations produce such similar present-day ice conditions in spite of the differences in simulated downward longwave radiative fluxes?"
Answering their own question, the three researchers state that "a frequently used approach" to resolving this problem "is to tune the parameters associated with the ice surface albedo" to get a more realistic answer. "In other words," as they continue, "errors in parameter values are being introduced to the GCM sea ice components to compensate simulation errors in the atmospheric components."
What it means
The three researchers conclude that "the thinning of Arctic sea ice over the past half-century can be explained by minuscule changes of the radiative forcing that cannot be detected by current observing systems and require only exceedingly small adjustments of the model-generated radiation fields," and, therefore, that "the results of current GCMs cannot be relied upon at face value for credible predictions of future Arctic sea ice."