Professor Michael Kelly, Prince Philip Professor of Technology, University of Cambridge, kindly sends us his comments on a letter this month to Nature Geoscience, Test of a decadal climate forecast, by Myles R. Allen, John F.B. Mitchell and Peter A. Stott. I previously commented on the letter in Climate forecasts fulfilled or what? Mike just returned to England after spending seven months as Visiting Professor at the prestigious MacDiarmid Institute, Victoria University of Wellington.
The recent paper of Allen et al. does a careful job of estimating errors in forward projections of global temperatures from earlier calculations on global circulation models of the atmosphere. Given the simple question — are the models doing a good job or not — the increasing level of sophistication needed to defend them is of concern. For many of us, a temperature stasis of 17 years is enough to suggest that the models are not as robust as some of their advocates maintain.
The Met Office is predicting five more years of statis. I am content to wait another decade to see if the divergence of the models and projections continues, and thus deepens and widens the level of concern. That said, I would greatly prefer to have models of greater discrimination that are more capable of earlier critical judgement.
I am not content with the current economic and engineering madness in rolling out renewable energy systems that are uncompetitive and immature; this should not last another day. The growing divergence between models and data (as shown in the leaked AR5 report of WGI) is equivalent to lengthening the odds of catastrophic climate change in the current century, and if climate mitigation is viewed as an insurance policy, the premiums we are prepared to pay now should come down sharply.
PS: The projections of Akasofu [S-Y Akasofu, 2010, Natural Science 2 1211-24 ‘On the recovery from the little ice age’] seem to be out-performing all the climate models for the last 17 years!