Projections of greenhouse gas (GHG)-induced future climate change are based on climate models whose complexity ranges from the simple energy-balance, climate/upwelling-diffusion ocean model we used to make our Intergovernmental Panel on Climate Change (IPCC) projections [Houghton et al., 1990] and the revised projections of our Nature paper [Schlesinger and Jiang, 1991b], to the coupled atmosphere-ocean model which Cubasch et al.  have now used to repeat our earlier projections. It is likely that even the most comprehensive model may not include some physical processes, and such absent physical processes, along with those included by potentially inadequate parameterizations, will render uncertain the validity of the results.
Risbey et al. [1991b] are correct that the Cubasch et al.  model does not include melting of the polar ice caps or the potential release of methane from methane clathrates. However, based on our simulations of glacial onset from 115,000–105,000 years BP with our atmospheric general circulation/mixed-layer ocean-ice sheet/asthenosphere model (M. E. Schlesinger and M. Ya. Verbitsky, Simulation of glacial onset with an atmospheric general circulation/mixed-layer ocean/ice sheet/ asthenosphere model, in preparation, 1991), it appears that the time scale of changes in the polar ice sheets is too long to be important for the greenhouse issue during the next century. Furthermore, Cicerone and Oremland  conclude that: “In principle, there is a clear possibility of future atmospheric methane increases due to methane hydrate destabilization, but the size of the effect needs better quantification.” Consequently, the only way to assess the significance of not including these and other processes in our most comprehensive, contemporary climate models is to include them and then repeat the simulations, in the vein of those recently performed by Cubasch et al. .