How’s this for “rapid response“? This rebuttal comes out at exactly the same time the press embargo lifts in Science. We were able to obtain advance copies of the Dessler paper, plus Dr. Spencer had seen it as a poster at the recent A-Train satellite symposium. – Anthony
Update: Dessler responds here at Real Climate and makes the full paper available here at his TAMU website.
2PM EST, December 9th, 2010 by Roy W. Spencer, Ph. D.
How clouds respond to warming – the ‘cloud feedback’ problem – will likely determine whether manmade global warming becomes either the defining environmental event of the 21st Century, or is merely lost in the noise of natural climate variability.
Unfortunately, diagnosing cloud feedback from our global satellite observations has been surprisingly difficult. The problem isn’t the quality of the data, though. The problem is figuring out what the cloud and temperature behaviors we observe in the data mean in terms of cause and effect.
So, Andy Dessler’s (a Texas A&M climate researcher) new paper appearing in Science this week is potentially significant, for it claims to have greatly closed the gap in our understanding of cloud feedback.
Dessler’s paper claims to show that cloud feedback is indeed positive, and generally supportive of the cloud feedbacks exhibited by the IPCC computerized climate models. This would in turn support the IPCC’s claim that anthropogenic global warming will become an increasingly serious problem in the future.
Unfortunately, the central evidence contained in the paper is weak at best, and seriously misleading at worst. It uses flawed logic to ignore recent advancements we have made in identifying cloud feedback.
In fact, the new paper is like going back to using only X-rays for medical imaging when we already have MRI technology available to us.
What the New Study Shows
So what is this new evidence of positive cloud feedback that Dessler has published? Well, actually it is not new. It’s basically the same evidence we published in the Journal of Geophysical Research.
Yet we came to a very different conclusion, which was that the only clear evidence of feedback we found in the data was of strongly negative cloud feedback.
But how can this be? How can two climate researchers, using the same dataset, come to opposite conclusions?
The answer lies in an issue that challenges researchers in most scientific disciplines – separating cause from effect.
Dessler’s claim (and the IPCC party line) is that cloud changes are caused by temperature changes, and not the other way around. Causation only occurs in one direction, not the other.
In their interpretation, if one observes a warmer year being accompanied by fewer clouds, then that is evidence of positive cloud feedback. Why? Because if warming causes fewer clouds, it lets in more sunlight, which then amplifies the warming. That is positive cloud feedback in a nutshell.
But what if the warming was caused by fewer clouds, rather than the fewer clouds being caused by warming? In other words, what if previous researchers have simply mixed up cause and effect when estimating cloud feedback?
A Step Backwards for Climate Science
What we demonstrated in our JGR paper earlier this year is that when cloud changes cause temperature changes, it gives the illusion of positive cloud feedback – even if strongly negative cloud feedback is really operating!
I can not overemphasize the importance of that last statement.
We used essentially the same satellite dataset Dessler uses, but we analyzed those data with something called ‘phase space analysis’. Phase space analysis allows us to “see” behaviors in the climate system that would not be apparent with traditional methods of data analysis. It is like using an MRI to see a type of tumor that X-rays cannot reveal.
What we showed was basically a new diagnostic capability that can, to some extent, separate cause from effect. This is a fundamental advancement – and one that the news media largely refused to report on.
The Dessler paper is like someone publishing a medical research paper that claims those tumors do not exist, because they still do not show up on our latest X-ray equipment…even though the new MRI technology shows they DO exist!
Sound strange? Welcome to my world.
We even replicated that behavior see in the satellite data analyzed with phase space analysis — our ‘MRI for the climate system’ – by using a simple forcing-feedback climate model containing negative cloud feedback. It showed that, indeed, when clouds cause temperature changes, the illusion of positive cloud feedback is created…even when strongly negative cloud feedback really exists.
Why Dessler Assumed We Are Wrong
To Dessler’s credit, he actually references our paper. But he then immediately discounts our interpretation of the satellite data.
Why?
Because, as he claims, (1) most of the climate variability during the satellite period of record (2000 to 2010) was due to El Nino and La Nina (which is largely true), and (2) no researcher has ever claimed that El Nino or La Nina are caused by clouds.
This simple, blanket claim was then intended to negate all of the evidence we published.
But this is not what we were claiming, nor is it a necessary condition for our interpretation to be correct. El Nino and La Nina represent a temporary change in the way the coupled atmospheric-ocean circulation system operates. And any change in the atmospheric circulation can cause a change in cloud cover, which can in turn cause a change in ocean temperatures. We even showed this behavior for the major La Nina cooling event of 2007-08 in our paper!
It doesn’t mean that “clouds cause El Nino”, as Dessler suggests we are claiming, which would be too simplistic and misleading of a statement. Clouds are complicated beasts, and climate researchers ignore that complexity at their peril.
Very Curious Timing
Dessler’s paper is being announced on probably THE best day for it to support the IPCC’s COP-16 meeting here in Cancun, and whatever agreement is announced tomorrow in the way of international climate policy.
I suspect – but have no proof of it – that Dessler was under pressure to get this paper published to blunt the negative impact our work has had on the IPCC’s efforts.
But if this is the best they can do, the scientists aligning themselves with the IPCC really are running out of ideas to help shore up their climate models, and their claims that our climate system is very sensitive to greenhouse gas emissions.
The weak reasoning the paper employs – and the evidence we published which it purposely ignores! – combined with the great deal of media attention it will garner at a time when the IPCC needs to regain scientific respectability (especially after Climategate), makes this new Science paper just one more reason why the public is increasingly distrustful of the scientific community when it comes to research having enormous policy implications.
===============================================================
Abstract:
On a global scale, clouds presently influence climate in a way that cools the planet. But, they will lose some of that cooling capacity as climate warms, according to a study that supports current ideas about how atmospheric carbon dioxide affects global temperature. Clouds can potentially have both positive and negative feedback effects on climate, and this is responsible for much of our uncertainty about the amount of warming that will be caused by increasing concentrations of carbon dioxide in the atmosphere. It’s generally agreed that overall this feedback is positive, with warming being exacerbated as clouds trap larger quantities of outgoing infrared radiation, but so far we have only a general idea of this effect. Andrew Dessler has estimated the actual magnitude of the feedback effect by analyzing ten years of satellite data on the flux of radiation through the top of the atmosphere. He concludes that the feedback effect is indeed positive and of a value that agrees with the canonical range of estimates of how much warming will occur for a doubling of atmospheric carbon dioxide.
Article #10: “A Determination of the Cloud Feedback from Climate Variations over the Past Decade,” by A.E. Dessler at Texas A&M University in College Station, TX.
Contact: A.E. Dessler at +1-979-862-1427 (office phone), +1-979-220-4513 (mobile phone), or adessler@tamu.edu (email).
I’m surprised that nobody has commented on the scatterplot in the main figure (2). Both visually and from the r^2=2% (ie 0.02), and a slope failing a signicance test, the proper conclusion is that there is no evidence either way. Roughly from the r^2, 98% of the variance is unexplained by the regression, which explains only 2%.
The scale in Figure 2 has negative values of T(K). Is it because of this breakthrough that it was published by Science? I think we should be told.
If Spencer’s work was dismissed as he used the period 2000 – 2010 which was full of natural variability and no obvious co2 induced warming – why than does Dessler not dismiss his own analysis???
Please let me cite Prof. Joni Mitchell, pioneering climatologist, who did valuable early work on clouds:
“It is error only, and not truth, that shrinks from inquiry.”
Thomas Paine
Dr. Dressler,
Congratulations on your prominent publication–pretty good for an Aggie!
I agree with your contention that authors have no control on the precise dates of publication of their articles. Dr. Spencer was quite ungracious to suggest that there was some funny business on your part about the issue of Science in which your paper appeared. However, I’m sure that you would agree that the staff at Science was keen to get it out in the timeframe of the Cancun meeting, and the fact that it was embargoed means that they expected a rather large splash.
My question for you: Did Science solicit your submission? In my area, it is common for Science and Nature to request submissions of work that their editors become aware of–often with the notion that they might be published to correspond with major meetings. It doesn’t mean that there’s some conspiracy afoot–it’s merely the practice of these “high impact” journals like to maximize visibility.
I would think this is a pretty easy problem to solve. If there is more moisture for clouds, there is more cloud cover. If evaporation increases from forcing, there is more moisture for clouds, and there is less forcing from the sun. When evaporation decreases from less forcing because of more clouds, there is less evaporation, and so there is less cloud cover. Which leads to more forcing from the sun. Which leads to increased evaporation and more cloud cover. And so on.
Of course, it would be nice to have some paleo-cloud data handy. Anyone know of a good dataset for paleonimbus?
From Realclimate:
***Short paraphrase:
Spencer: ENSO is caused by clouds. You cannot infer the response of clouds to surface temperature in such a situation.
Dessler: ENSO is not caused by clouds, but is driven by internal dynamics of the ocean-atmosphere system. Clouds may amplify the warming, and that’s the cloud feedback I’m trying to measure. ***
I do not see where Spencer states this explicitly.
Sorry to misspell your name, Dr. Dessler.
Negative feedback implies clouds will increase as the earth warms. If clouds are decreasing while the earth is warming, that would rule out negative feedback as the dominant process, even if it doesn’t prove positive feedback.
Jim D, “dominant” is problematic in that context. First, you ignore lag. Second, obviously if the earth is warming all immediate negative feedbacks are being dominated at least temporarily. That doesn’t tell us anything about individual feedbacks.
savethesharks says: “Your comments got snarked by one of the mods over there, Bob. I hope you will publish a retort.”
I replied.
&&&&&&&&&
Bob Tisdale says:
Your comment is awaiting moderation.
10 December 2010 at 5:07 AM
raypierre replied: “Sorry, not even close, and no cigar. You are ignoring the fact that to determine the effect of clouds on SST you can’t take just the downward shortwave effect and ignore the cloud greenhouse effect. The cloud greenhouse effect nearly compensates at the top-of-atmosphere, but it works its way into the surface budget through warming the atmosphere. But take heart — you are in good company in being confused about this. It was the basis of Ramanathan’s confusion in his famously wrong ‘thermostat hypothesis.’”
Interesting. In a reply to my same comment over at WUWT, Andrew Dessler wrote, “Bob Tisdale: You’re probably right about that. I put that into the e-mail without spending much time thinking about it.”
http://wattsupwiththat.com/2010/12/09/the-dessler-cloud-feedback-paper-in-science-a-step-backward-for-climate-research/#comment-547536
Additionally, if memory serves me well, Pavlakis et al addressed your concerns in the paper linked above and in their earlier companion paper:
http://www.atmos-chem-phys.org/7/2013/2007/acp-7-2013-2007.pdf
Andrew Dessler says:
“What I find is that while clouds cool, and will continue to cool, they will cool less as the climate warms.”
Maybe in a hypothetical model. In reality, clouds are constantly moving, any slight change in the balance you are modelling is far outweighed by movement of the clouds either towards, or away from the equator.
Please correct me if I am wrong.
Sir, thank you for taking the time to engage in the conversation in a venue which allows free speech, it really is appreciated.
So now, Spencer says he isn’t claiming that clouds are the cause of El Nino. His statement is,
” El Nino and La Nina represent a temporary change in the way the coupled atmospheric-ocean circulation system operates. And any change in the atmospheric circulation can cause a change in cloud cover, which can in turn cause a change in ocean temperatures. We even showed this behavior for the major La Nina cooling event of 2007-08 in our paper!”
This statement means that clouds are a feedback mechanism in the classic sense. If clouds are to be regarded as a feedback mechanism, then Dessler’s data , showing that when temperature increases, trapping of radiation by clouds increases, shows that behavior of clouds add to the increase in temperature. This is positive feedback.
It is true that clouds are complicated, as Spencer says, but that statement doesn’t negate the data and its interpretation by Dessler, if clouds are regarded as a feedback phenomenon. There is no logic in Spencer’s objection.
The objection to the publication date of Dessler’s paper seems beside the point.
It seems like whining on the part of those who don’t like the data and conclusions, and wish it would go away.
Each increment of increased Clouds in a warming world provides a feedback of -5, -4, -3, …. and so on.
In the unusual world of climate science trying to prove the theory right, this series of negative feedbacks turns into a positive feedback because -3 is not as negative as -4.
Nope, each increment of more clouds is still a negative feedback.
Clouds increase as it gets warmer. Clouds provide a net cooling. Therefore, clouds are a negative feedback even if they are less negative as the warming increases.
Dessler’s formulae need to be rewritten because there has been some mistake in the mathematical transforms. It is probably because he is mixing and matching deltas with absolute values which can lead one astray if you don’t keep in context what you are actually plotting and the signs of what you are plotting.
For example, Cloud forcing is say -40 watts/m2 but declines to -39 watts/m2 in the 2007-08 La Nina when temperatures fell to -0.3C. If you correlate the change in cloud forcing of +1 watts/m2 (a delta) against the -0.3C (which is now a simple absolute value) the correlation starts to looks like a positive feedback when it is in reality a negative one.
Wayne Richards says:
December 9, 2010 at 9:28 pm
Citing Dr. Mitchell; so it’s finally “settled”: “ice-cream castles”, “rain & snow”, yep clouds=negative feedback. Too bad she was never peer-reviewed, ah…the ‘team’ would have never allowed this landmark work through their ‘process’. 😉
On a more serious note:
Andy,
Thank you for posting your e-mail exchange with Roy at RC. I found it encouraging that the discourse was, well, amicable, intelligent, and professional. On reading this post (and not Spencer’s blog) I didn’t get the impression that he was accusing you directly (although you deny being on the receiving end of ‘being pressured’ on the RC thread), but rather alleges that the review and publishing process (or powers that be) wanted this in time for Cancun. If you discovered that somewhere the process had been “messaged” to expedite your work being published sooner than would normally be expected, for political ends, what would your response to that be?
I was disappointed at your “slapdowns” of Spencer on the RC thread: specifically the Cancun comment and the dismissal of his work. It gives the positive, professional tone of your email discourse a disingenuous “flavor” and denegrates your position. As a result (IMO) Spencer’s position of your method being likened to “…using only X-rays for medical imaging when we already have MRI technology available to us.” is bolstered. They also seemed as an intentionl mechanism to flagellate the regular RC crowd, and were successful.
After reading some of the research (from IPCC, you, Spencer, Cheng, et al) on clouds as feedbacks I come away with the impression that a definitive answer is still “out-there” but stacking up in the direction of general negativity/cooling depending on region, altitude and type of clouds involved (still many unanswered questions). The arguement for causation: chicken or the egg (temp/clouds) will only eventually be answered by continuing the work, through collaboration, cooperation, and amicable discourse between researchers.
Dressler states:
“It is important to note that while a slight negative feedback cannot be ruled out, the data do not support a negative feedback large enough to substantially cancel the well-established positive feedbacks, such as water vapor, as Lindzen and Spencer would argue. ”
Yes, water vapor is a well established positive feedback, but why does the cloud feedback have to be negative enough to cancel it out? Since Wentz showed that the models were only able to reproduce one-half to one-third of the precipitation increase (a negative feedback), it is not even clear that the water cycle viewed in total, not just the water vapor component, is a net positive or net negative feedback. Since even Dressler grants that his work doesn’t rule out a slight negative feedback, he is basically acknowledging that all the model based projections with correlated high positive cloud feedback and high total net positive feedback, might well be wrong.
A net negative or neutral feedback leaves the well mixed greenhouse gas hypothesis accounting for less than a third of the recent warming, and contributing only about one degree C of warming over the next century. One degree C is less than the natural variation, so a decade or even thirty year period 100 years from now might well be cooler.
Andrew Dessler says:
December 9, 2010 at 8:04 pm
latitude: If clouds cool less as the climate warms, then that is indeed a positive feedback.
==========================================================
Andrew, what is the magic number then?
Since CO2 has nothing to do with this, it’s temperature only.
After your research, you must have found a magic temperature where clouds
are no longer negative and they switch to positive.
If there is no magic temperature, that can only mean that as temps go up, clouds
become more positive, making it harder for temps to go back down.
That would be the run-away global warming.
Since we know that temps have always gone back down, then that would mean that
clouds have very little, if any, effect. Or it could mean that clouds were never positive in the first place.
What temperature would you think is that point?
At what temperature would you consider clouds no longer negative but positive?
Since anything I post at RC is removed (or never makes it) almost immediately I have to make my comment here.
I read this statement by Dr Dressler in his comment at RC, “A positive cloud feedback loop posits a scenario whereby an initial warming of the planet, caused, for example, by increases in greenhouse gases, causes clouds to trap more energy and lead to further warming. Such a process amplifies the direct heating by greenhouse gases. Models have been long predicted this, but testing the models has proved difficult. ”
I don’t believe he is correct in stating that “models have long predicted this (positive cloud feedback)”. I think the models are constructed with positive cloud feedback ‘built in’. The models can hardly be said to ‘predict’ something that is built into the equations as an assumption. That would be like saying the ‘model’ that contains the equation ‘2+2 = 4’ ‘predicts’ that 2+2 = 4 isn’t it?
Climatologists are amazing. They talk about positive feedback as if it was just an other measure of something. But positive feedback has consequences. Amplifiers saturate, servomechanisms speed up, reactors blow up, temperatures run away, things break.
Where are the consequences of positive feedback on the earths climate. The earths climate is stable. Even huricanes end in a whimper not a bang. There has never been a thermal runaway. Ice age yes but thermal runaway no. Untill climatologists can demonstrate the consequences of positive feedback, they should go back to school and study servomechanisms.
I’m sorry for my simplicity, but isn’t it true that the suns energy being centrally located in a area of earth causes a high radiation and convection system stacking up into the atmosphere to become a high pressure area. No clouds.
The lower pressure areas allow water vapor to form and condense into clouds. Lots of clouds.
So when the high pressure collapses and allows a lower pressure system to move in, the heat transfer to the atmosphere is slowed and trapped as water vapor(clouds). There is always water vapor in the air and when the sun is at it’s minimum in winter water vapor will form into ice crystal in the air with clear sky’s. A lowering of atmosphere allow nuclei to enter and form as a rain drop or a snowflake.
Clouds can temporarily trap earths radiant heat on its surface AND block out the heating energy reaching earths surface from the sun.
El Nino or El Nina is just a easy out to explain wrong predictions from a flawed CPU model. I`m sick of this BS (bad science). Man made/almost all natural CO2 is a flea on the suns back when influencing the warming on earth. Your explanation/modeling is just as wrong as the others saying the opposite. Water vapor is the 99.999% main GHG, what else influences earth more? Pfft.
Claiming you can model clouds when you have absolutely zero knowledge of how they are created is beyond belief. Modern science has become a GIS modeling program, how sad.
Bill Illis says:
December 10, 2010 at 4:55 am
Nope, each increment of more clouds is still a negative feedback.
Clouds increase as it gets warmer. Clouds provide a net cooling. Therefore, clouds are a negative feedback even if they are less negative as the warming increases.
=======================================================
Thanks Bill, that’s what I was working towards.
Andrew seems to be looking at each cloud. Where as temps increase, each cloud
has less negative feedback.
Warmer will create more clouds.
Where each cloud could possibly be slightly less negative, according to Andrew, but the accumulation of more clouds would have a greater net cooling.
It could be a possibility that the entire surface of the planet becomes covered in one big cloud.
Would that be positive, as Andrew likes to call it, or negative?
Dr. Dessler, Apologies for misspelling your name above. I hope you continue to participate here, since realclimate is highly censored. Regards.
http://virakkraft.com/Cloudcover-ENSO.gif
85-90: ENSO avg. 0 >> cloud cover unchanged.
90-00: ENSO positive >> cloud cover reduced
After 00: ENSO avg. 0 >> cloud cover unchanged
How can that be positive feedback?
Dr. Dessler,
“Of course clouds cool. That’s not the question. The question is whether clouds cool more or less as the climate warms. ”
In what way is this statement not self-contradictory? “Less cooling,” however awkward the language, means a net increase in temperature, inarguably. By positing a theory that clouds cool less as the climate warms, and then failing to explain why we would not inevitably have arrived at a cloudless hot planet very shortly after the first day warmer than the previous day, you have contradicted yourself. If these two phenomena were coupled with unchecked amplification, life could not exist. In other words, we all better hope and pray that clouds cool at least as much, if not more, when “climate” warms.
Do one of Einstein’s “thought experiments” about this. Automatic Control is well-understood, well-documented, and easy to study. Every engineer who reads this will say the same thing, at least to himself or herself.
Keep this up and I am going to start in with the Aggie jokes…
http://www.technologyreview.com/blog/arxiv/26105/?p1=Blogs
Interesting take on cloud formation.