Arguments
Dessler Demolishes Three Crucial 'Skeptic' Myths
Posted on 8 September 2011 by dana1981, Rob Painting
Andrew Dessler's new paper, which we first examined in a post yesterday, has some very far-reaching implications in terms of refuting climate "skeptic" myths. In fact, its results are relevant to three seperate myths in the Skeptical Science database. 
As a result, we have incorporated the findings of Dessler (2011) into the three myth rebuttals as follows.
It's not internal variability
Roy Spencer is the driving force behind the "internal variability" hypothesis, which posits that some unknown and undefined mechanism is causing cloud cover to change, which, by changing the overall reflectivity of the Earth, is the driving force behind the current global warming.
In attempting to substantiate this internal variability hypothesis, Spencer & Braswell (2011) assumed that the change in top of the atmosphere (TOA) energy flux due to cloud cover changes from 2000 to 2010 was twice as large as the heating of the climate system through ocean circulation. Dessler (2011) used observational data (such as surface temperature measurements and ARGO ocean temperature) to estimate and corroborate these values, and found that the heating of the climate system through ocean heat transport was 20 times larger than TOA energy flux changes due to cloud cover over the period in question.
This empirical finding contradicts Spencer's hypothesis that cloud cover changes are driving global warming. However, it is consistent with our current understanding of the climate: ocean heat is exchanged with the atmosphere, which causes surface warming, which alters atmospheric circulation, which alters cloud cover, which impacts surface temperature. While Spencer hypothesizes that the changes in cloud cover are the main driver behind global warming, Dessler concludes that they're only responsible for a small percentage of the changes in surface temperature from 2000 to 2010. Spencer's internal variability hypothesis is contradicted by the observational data.
Spencer and Braswell (2011) is Contradicted by Observational Data
A highly-touted (and exaggerated in the media) claim in Spencer & Braswell (2011) was that their results suggested that climate sensitivity is low because climate scientists are misinterpreting climate feedbacks as climate forcings. In their paper, Spencer and Braswell analyzed 14 models, but they only plotted the 3 with highest and 3 with lowest equilibrium climate sensitivities. In the process, Spencer and Braswell excluded the three of the climate model runs which best matched the observational data, and also cherrypicked the data set furthest from the model runs (HadCRUT) (Figure 1).
Figure 1: Dessler (2011) reconstruction of Spencer & Braswell's Figure 3, showing relationship between top-of-atmosphere (TOA) net flux and surface temperature, as a function of lag between them. The blue line is the observational data chosen by Spencer and Braswell (HadCRUT). The red lines show other available observational data. The black lines show climate model results. The black lines with crosses show the climate model runs chosen by Spencer and Braswell in their paper.
Dessler found that these three model runs excluded by Spencer which best matched the data are also among those which best simulate El Niño and La Niña, which is not surprising, given that much of the temperature change over 2000-2010 was due to the El Niño Southern Oscillation (ENSO). Thus Dessler concludes that
"since most of the climate variations over this period were due to ENSO, this suggests that the ability to reproduce ENSO is what's being tested here, not anything directly related to equilibrium climate sensitivity."
Spencer's claim of low sensitivity and negative feedbacks is based on this test, which is actually a test of models' ability to reproduce ENSO, and based on his internal variability hypothesis, which as noted above, Dessler's paper has also put to rest. Thus Spencer's claim of low sensitivity and negative feedbacks is not supported by the empirical observational data.
Lindzen and Choi (2011) is Fundamentally Flawed
Lindzen and Choi (2009), slightly revised as Lindzen & Choi (2011), used measurements of sea surface temperature in the tropics and satellite measurements of outgoing radiation from 2000 to 2010 in an attempt to determine climate sensitivity, ultimately concluding that sensitivity is less than 1°C for doubled atmospheric CO2.
Lindzen and Choi plot a time regression of change in TOA energy flux due to cloud cover changes vs. sea surface temperature changes. They find larger negative slopes in their regression when cloud changes happen before surface temperature changes, vs. positive slopes when temperature changes happen first, and thus conclude that clouds must be causing global warming.
However, Dessler also plots climate model results and finds that they also simulate negative time regression slopes when cloud changes lead temperature changes. Crucially, sea surface temperatures are specified by the models. This means that in these models, clouds respond to sea surface temperature changes, but not vice-versa. This suggests that the lagged result first found by Lindzen and Choi is actually a result of variations in atmospheric circulation driven by changes in sea surface temperature, and contrary to Lindzen's claims, is not evidence that clouds are causing climate change, because in the models which successfully replicate the cloud-temperature lag, temperatures cannot be driven by cloud changes.
Major Myth Mashing
It's difficult to exaggerate the impact of Dessler's findings, because these are three of the most crucial arguments for climate "skeptics." In order for the man-made global warming theory to be incorrect, climate sensitivity must be low (see Climate Sensitivity: The Skeptic Endgame). Since all previous studies using many different lines of evidence point to the same answer, that climate sensitivity is not low, climate "skeptics" had to rely on Spencer & Braswell and Lindzen & Choi as the only game in town arguing otherwise. In one fell swoop, Dessler has demonstrated that the only two modern papers arguing for low climate sensitivity are both fundamentally flawed, and their assumptions are contradicted by observational data. In short, there's absolutely no reason to believe the IPCC's equilibrium climate sensitivity range of 2 to 4.5°C for doubled CO2 is incorrect.
Additionally, climate "skeptics" have yet to put forth a plausible, coherent, internally consistent alternative to challenge the robust man-made global warming theory. Despite its fundamental problems, Spencer's internal variability hypothesis was probably the best alternative presented to this point, and Dessler drove another nail into its coffin by demonstrating what a small effect clouds have had on global temperature changes over the past decade.
But isn't their a corollary to this question? Given that we're here to ask it, there must be some strong negative feedback that engages at some temperature, especially since CO2 lags temperature at every point in the ice core data. Why do expect this cycle to be the one that doesn't engage this negative feedback?
And what's your logical train that deduces a "strong negative feedback" from the observation that CO2 lags temperature changes in ice cores? I would have thought the ice core data is rather strong evidence of positive feedback.
My guess is that you are mixing up 'positive feedback' with 'runaway feedback' or assuming that feedbacks continue indefinitely and thus 'require' a "strong negative feedback" to kick in at some point and 'overwhelm' the constant positive feedback. That isn't how it works. Once a forcing stops the feedbacks associated with it perforce will do so as well unless they are so powerful as to be continually self-perpetuating (which no one has suggested is currently the case for AGW feedbacks).
Basically, when the orbital forcing behind the glacial cycle ended the CO2 and ice albedo feedbacks it was causing also ended (not immediately, but relatively soon thereafter). At that point there were no significant positive forcings or feedbacks and thus no need for this "strong negative feedback" which you hypothesized. Instead, the subsequent cooling came from the orbital cycle shifting the other way... causing a cooling forcing and the same CO2 and ice albedo feedbacks then working in reverse.
The negative feedback of "this cycle" is already under way... the orbital forcing has switched from warming to cooling. Nobody is 'expecting it not to engage'... it already has. If all else had remained unchanged that would result in slow cooling and another glaciation 10s of thousands of years down the line. However, human CO2 emissions have introduced a new, and much stronger, warming forcing.
I should have been more precise. If we hypothesize that CO2 causes regenerative amplification and this positive feedback is the reason small Milankovitch solar radiance variations can result in large global temperature variations, then we must also account for why there hasn't been run away heating or cooling in the past and why we see in the ice record, millennial periods of falling (rising) temps with rising (falling) CO2. Put another way, how can a forcing that's too small to account for the observed variance, overcome the the maximum positive feedback seen at the temperature maximum when CO2 concentrations are near their maximum unless their exists be some negative feedback that comes into play near the extremes?
One other possibility occurs to me. Non-linear, regenerative feedback systems whose complex poles are too dissipative to sustain oscillation can be subject to injection locking to a small period forcing. The output becomes quasi-periodic, synced to the input signal but with substantial phase noise. I presume this hypothesis has been examined and rejected.
As to the Milankovitch cycles - feedback operates on forcing changes both positive and negative. This means that when orbital mechanics decrease insolation, that negative change in forcing is amplified by feedbacks as the earlier positive change was. The forcing acts as a control knob.
Condescension is usually not conducive to dialog. We're running into cross-discipline semantics. Positive feedback is self-enforcing and will operate on any noise present to select and reenforce components near the system eigenvalues. The amplitude at this frequency will grow without bound until a limiting mechanism is encountered. This limiting mechanism is best viewed as a countervailing negative feedback which works to constrain the poles to the jw axis.
#57 - I was not trying to postulate a novel negative feedback mechanism. I was referring to whatever you want to call the thing that prevents thermal runaway. So rephrasing my original question, why won't this same mechanism mitigate the effects of AGW, limiting the temperature excursion to that which would of been encountered naturally?
Yes of course but if the open-loop gain <1 the closed loop gain is also <1. We are I thought, only considering a system capable of amplifying the small forcing induced by the Milankovitch cycles.
We could consider the dominant feedback to any atmospheric warming (whether greenhouse-forced or solar-forced); i.e. the water vapour feedback. Say the primary warming from enhanced [CO2] is 1 oC, and the water vapour feedback results in an additional primary warming of x oC, the total warming is something like 1 + x + x^2 + x^3 + x^4 ...
which is 1/(1-x).
So if the water vapour response to a 1 oC warming is 0.5 oC then the total warming when everything comes to equilibrium is 1/(1-0.5) = 2 oC.
The same argument applies for other feedbacks. In other words there doesn't have to be any "negative feedback"...under the effect of a forcing the system evolves to a new equilibrium.
If you're interested in what's actually happening during the transition from glacial to interglacial transition try these papers entitled Ice Age Terminations, and The Last Glacial Termination, respectively.
Does such a model exist?
Thanks