By Rud Istvan,
In several previous recent post comments, I have been cryptically alluding to an alternative ECS derivation that strongly supports the observational ECS estimates, and also the Willis Eschenbach emergent thermoregulation hypothesis. Since my comments were scattered about and are largely non-searchable via WUWT, I thought a guest post with lots of links and illustrations might be a useful summation avoiding more same-same comments. In the future I can just comment to relative newbies by referencing this summary guest post.
My arguments are somewhat technical, so please forgive the sometimes unfortunately lengthily explanations intended for those who newly visit WUWT. Its logical history goes back to 2011, as the below linked references prove.
AR4 said the most likely ECS (equilibrium climate sensitivity, what the resulting GAST temperature will be after a doubling of CO2 when everything else settles out after some hundreds of years hence (see essay Sensitive Uncertainty in ebook Blowing Smoke for the inherently uncertain ECS time frame issues) was ~3C. Their CMIP3 climate model ensemble actually had a mean of 3.2C, so IPCC could even claim they were being ‘conservative’.
AR5 offered no such most likely ECS estimate. The reason was that the CMIP5 models still predicted about 3.2C ECS, while the observational energy budget papers since 2013 estimated about 1.7C. So IPCC said no central estimate could be given because of the model/observation discrepancy (IPCC AR5 SPM.5.D2).
The questions arise, which is the more reliable ECS estimate, and can the AR5 differences be reconciled? This guest post attempts to answer both in a simple logical fashion, while omitting many distracting ‘climate science’ false nuances.
Basic Climate Science
There are two notions of ECS. The first is without feedbacks (mainly water vapor and clouds). It estimates a simple non-condensing gas CO2 doubling. AR4 implicitly had that at a 1.1C. Lindzen’s 2012 curve reproduced below used ~1.2C. Using Monckton’s equation and his parameters yields 1.16C, so one can legitimately argue Lindzen’s following curve merely rounds up given uncertainty.
To understand the following Lindzen graph, one also has to understand his underlying BODE f/(1-f) ‘total feedback factor’ equation reasoning derived from electronics amplifier circuit theory. It posits that the net system feedback gain is a linear sum of the individual system feedbacks. It works. And, if the no feedback result is just about 1.2C (Lindzen), then it looks mathematically like this:
Simple graphical analysis says that per Bode, if the AR4 ECS is ~3, then the Bode equivalent must be about 0.65. Unlike Monckton’s previous CE posts here on Bode math, it is still quite reasonably well behaved.
AR4 also said that water vapor feedback (wvf) by itself about doubles the no feedback ECS. So ~1.2C * 2 = ~2.4C, so from the above graph a WVF Bode of ~0.5. AR5 also said that cloud feedback must comprise most of the rest, since all else per AR5 nets to about zero.
So per AR5 cloud feedback must be about positive Bode (0.65-0.5) 0.15.
Dessler (paywalled, A Determination of the Cloud Feedbacks, Science 330:1523-1527 2010)) wrongly asserted his paper found a positive cloud feedback. Steve McIntyre took him severely to task. In fact, Dessler’s R^2 of 0.02 shows no correlation whatsoever, so no cloud feedback at all. So the real Bode is 0.65-0.15~=0.5.
Wentz (paywalled, Wentz et al, How much more rain will global warming bring? Science 317: 233-235, 2007)) found in 2007 that the CMIP3 models produced about half of ‘observed’ ocean precipitation. His paper can be criticized because the oceans were not then well ‘observed’. ARGO changed that, and one of its three main design goals (via salinity) was ocean precipitation. ARGO data now says that in fact CMIP5 produced about half the ARGO sensed precipitation.
That implicitly means that the water vapor feedback is actually about half of modeled because of rainfall humidity washout. So Bode 0.5/2 is about 0.25.
Willis Eschenbach thermoregulation theory rules.
Plug Bode 0.25 into Lindzen’s above base ~1.2C Bode feedback curve and out pops about 1.7C ECS, very close to the observational energy budget ECS estimates, and about half of the CMIP5 climate models. Model ECS is provably high by about 2x. And that unavoidable observation/math explains two things:
- Why there is no climate emergency at all.
- Why Schellnhuber’s admittedly arbitrary 2C ‘tipping point’ was later changed to 1.5C by alarmists. Otherwise, their new climate alarm would have to also be mathematically cancelled.