Questions are not Answers
IMHO the current state of climate change models enables them to raise serious questions but not to answer them. The state of knowledge and quality and range of data available to test them is simply insufficient.
As evidence of that, one of the latest and supposedly best refined models adapted to current data I am told is DePreSys. Eg:
Improved Surface Temperature Prediction for the Coming Decade from a Global Climate Model
Science, Aug 10, 2007, Doug M. Smith, et al
Note the conclusion: "the year 2014 predicted to be 0.30° ± 0.21°C [5 to 95% confidence interval (CI)] warmer than the observed value for 2004."
In other words, the predicted ten year warming is in the range 0.09°C to 0.51°C with 90% confidence in the estimate. But of course this error range does not include errors due to fundamental deficiencies in the model. It is merely calculated on the scatter the model produces when initial conditions are randomly tweaked. So at the very best we are told the model has an uncertainty in its ten year predictions of a factor of five.
Extrapolate that to 100 years and the uncertainty can only increase.
Furthermore, most science I have seen uses 95% confidence levels. Why not here? The obvious answer is that if the authors gave the error range for the usual 95% standard it would not exclude the possibility that the ten year warming outcome would be zero.
As far as I can see the history of Viking farms in Greenland does in fact seriously question the doomsday scenarios regarding its melting icecap.
Eg, see: http://www.nytimes.com/2001/05/08/science/08VIKI.html
Greenland was occupied by settlements of up to 1500 people for four centuries until about 1350 during much of which they farmed cows and sheep. The claim that they were supported during that time only by supplies from Scandanavia seems untenable.
There are no reports that London was under water during that period.