Over on Master Resource Chip Knappenberger has redeauxed the Lindzen-PNAS papers and elicited a mess of Email, two of which Eli would like to discuss. Roy Spencer writes at #13 (how prescient)
-
Positive feedback for climate is not the same as for engineering…in the usual sense of the word, the climate system is stable, with net negative feedback.
But the MAIN climate stabilizing effect is NOT included in climate “feedback”: the increase in IR cooling to space as temperature rises (the Stefan-Boltzman(n) effect). It’s just semantics, and leads to much confusion.
For example, positive cloud feedback would reduce the rate of radiative loss to space with temperature below the Stefan-Boltzman(n) value…but it would still be a loss of energy with warming, and so negative feedback in the traditional sense.
Andy Dessler at #27 Choi's his Chou's or chews his toys, or whatever, but brings some interesting inside stuff
There’s one additional piece of information missing from this post: this paper was originally submitted to JGR, and it was rejected by that journal, too. When I talked to Lindzen last Oct., he railed about how unfair the reviews from that journal had been. At that point, I think Lindzen recognized that his paper was never going to make it through any kind of legitimate peer review, so he next submitted it to PNAS so he could select his own reviewers. Kudos to PNAS for not letting him select the entirely unqualified Happer or Lindzen’s wholly-owned subsidiary, Choi. But now Lindzen thinks PNAS is being unfair to him. Of course, after so many rejections by so many reviewers, there’s another possibility that Lindzen seems to not consider: his paper is not very good.Eli, Eli has been looking for the text of Happer and Chou's reviews. Don't seem to be everywhere.
Besides which, Roy is wrong about the Planck feedback not being included in usual discussions of feedback. That's in fact the term used in virtually all discussions of the subject, including papers by Bony et al and Soden and Held. Why should it be surprising that the net climate feedback be negative? If not, the Earth could not get rid of the absorbed solar energy no matter how hot we got, and we'd be a molten blob by now (as I pointed out at the start of my Phys. Today article). The other climate feedbacks (water vapor, ice albedo, and clouds) are positive, and thus make climate more sensitive to changes in radiative forcing, since they WEAKEN the net negative feedback. What could be simpler? You couldn't get confused about this sort of stuff unless you worked really, really hard at it.
ReplyDelete--raypierre
raypierre,
ReplyDeleteI understand you point but *at equilibrium* shouldn't the net feedback be 0. Energy in = Energy out and all that. I may have misunderstood that aspect and just what is meant, but in the end doesn't TOA flux out have to equal energy put into the system?
Like the captch: "logic"!
The incoming solar isn't a feedback, RN, it's a forcing, so the point is that the (positive) net forcing and (negative when you include the Planck feedback) net feedback balance. Just terminology.
ReplyDeleteSteve Bloom
Yeah, that's what I figured. The net of feedbacks over the whole system and the external forcing should be 0 (otherwise Venus of snowball earth) at equilibrium. The problem is that the terminology allows this sort of technically correct, but factually incorrect argument, such as made by Spencer can muddy the waters.
ReplyDeleteIs it just me or does it look like sometimes Spencer talks of the whole earth (including the atmosphere) and sometimes he talks just of the surface temperatures which (at least to me) would be the interesting part WRT gardening and the amount of snow I have to plow.
ReplyDeleteNo, Rattus, you're still confused. The net energy flux must balance to zero at equilibrium,not the net feedbacks. The net feedback is negative, meaning that the hotter the planet gets the more it radiates to space. This is what allows a planet to achieve a stable equilibrium at a finite temperature.
ReplyDeleteNote that there are circumstances when, locally considered, the net feedbacks can be positive (destabilizing). That happens when the system is sufficiently close to an unstable equilibrium. In that circumstance, the temperature wanders over into the the stable equilibrium in whose attractor basin it is located. The net feedback will be negative (stabilizing) there. I discussed this in my recent Snowbird dynamical systems talk on feedback and bifurcation, which ought to be up on the web soon.
--raypierre
> Spencer writes at #13 (how prescient)
ReplyDelete> Positive feedback for climate is
> not the same as for engineering
See http://school.maths.uwa.edu.au/~berwin/humour/invalid.proofs.html#17.2Interdisciplinaryproof
"In which a special form of an equation, definition, or technique used by one science is applied in a completely unrelated field. Best used by engineering undergraduates ...."
Hang on, hang on. Everyone is talking as though negative feedback is necessary for equilibrium. If so, that is a rather bigger difference to the concept of feedback in, say, electronic engineering.
ReplyDeleteIn electronics, positive feedback amplification is possible and doesn't lead to a run-away effect, as long as the feedback is small (for some value of small). This was exploited in early radios (super-regenerative receivers) to allow a high gain amplifier to be constructed from a single valve of transistor. The problem in electronics is that gain varies as the components warm up - that is the cause of instability in electronics, not the positive feedback itself.
Take the negative feedback amp on this page: http://en.wikipedia.org/wiki/Negative_feedback_amplifier
Now flip the sign of beta. The gain remains finite as long as beta*A_ol is less than 1. (Yes, really. Counterintuitive but true. But it's exactly the same result as an infinite series with a finite sum.)
Now, my understanding of climate science is more limited, so the rest of this may be wrong, but as far as I can see there is a direct parallel. The input voltage is the forcing. The amplification term A_ol is the climate sensitivity without feedback. The total amplification A_fb is the sensitivity with feedback. The output voltage is temperature. The system has positive feedback, but it still has an equilibrium for any level of forcing. The only difference between negative feedback and *small* positive feedback is whether the total sensitivity is less or greater than the sensitivity without feedback.
Kevin C
What Spencer says makes sens to me, given the context. In #4 it was claimed that you couldn't have positive feedbacks or there would be runaway warming, and it is this claim Spencer replies to. Pierrehumbert may be correct that in technical papers the blackbody radiation is included in the discussion, but in general discussions a statement about positive feedbacks usually only include stuff like clouds, water vapor etc but exclude the Planck response.
ReplyDeleteDr. Jay Cadbury, phd.
ReplyDelete@Ray Pierre
Can you please define what equilibrium means in the climate system? Since there was over 7,000ppm of co2 in the atmosphere at one time and the earth converted most of it into carbonate rock, my understanding of your argument is that the earth cannot do this again, even on a smaller scale. Furthermore, it is impossible to claim that man's influence is stronger on the climate than natural mechanisms. After all, the climate has risen to far higher temperatures naturally. Your telling us there is global warming even though the earth is below its GAT, so that doesn't make much sense. What could be simpler?
"The other climate feedbacks (water vapor, ice albedo, and clouds) are positive, and thus make climate more sensitive to changes in radiative forcing, since they WEAKEN the net negative feedback. What could be simpler? "
Ray, you make these feebacks sounds like they are living animals.
Additionally, I haven't ruled out that the feedbacks you mentioned are positive but a forcing value of 300% is way too high.
ReplyDeleteIt all seems straightforward enough to this philosopher of science:
ReplyDeletefeedback is a matter of how the system changes, given a forcing that drives it away from a (local) near-equilibrium. If the system responds in a way that produces a net increase to the forcing, the feedback is positive and the system will move further from the initial equilibrium than it would without the feedback. Because this notion of feedback concerns how the system responds to a given forcing when we begin at a particular near-equilibrium state, the value assigned to the feedback is not a universal 'truth' about earth's climate (sorry Dr. Jay).
I assume that we typically leave out the negative forcing of increased OLR because that's an obvious bit of background, and our aim is to determine how high the total forcing gets once (relatively quick) feedbacks are added in. That's what tells us how far (in terms of temperature, in this case) the system will move from its initial state before it reaches a new near-equilibrium. But the technically-minded (of course) recognize that the reason the system tends toward a new equilibrium state is that there is an underlying negative forcing that's strong enough to prevent the system from running away, i.e. from responding to the forcing by departing indefinitely from the initial state: what we usually call the 'forcings' here are acting against a background restoring force. (Think of moving the end of a spring anchored at the other end by pulling on it: a positive 'feedback' in this usage would be something that increases the net force acting to stretch the spring, while the spring constant corresponds to the background radiative response.)
Any corrections/ clarifications would be very welcome...
Bryson Brown
Eli, Eli has been looking for the text of Happer and Chou's reviews. Don't seem to be everywhere."
ReplyDeleteI very good question? What happened to openness and transparency being advocated by the 'skeptics', seems that they are happy to make allegations to that effect against climate scientists, but seem to think that those same rules do not apply to themselves.
And yes, Lindzen's paper is not good, not good at all-- the experts agree on that point. And Roy, well he is just spinning his wheels as per usual.
ML
Dr. Jay Cadbury, phd.
ReplyDelete@ML
I don't think Happer got to review the paper? Actually, I know him so if you want I can probably find out if he reviewed it or not. However, I wonder if that might violate a disclosure policy, I assume not since the critical reviews of the paper are available, although the names are kept anonymous.
The idea of climate feedback raypierre is talking about is straightforward, and easily digestible to someone like Roy Spencer.
ReplyDeleteSuppose the Earth is a Planck blackbody radiator at temperature T that radiates with a flux of σT^4. It follows that the climate sensitivity for this blackbody planet is 1/(4σT^3), in dimensions of Kelvin change per unit radiative forcing. For Earth, the radiating temperature is 255 K, so the sensitivity is about 0.25 K per W/m2 forcing. The radiative forcing for a doubling of CO2 is ~ 4 W/m2 at modern concentrations, so this leads to the result that a doubling of CO2 would cause a ~4 W/m2 * 0.25 K/(W/m2) = 1 K temperature change.
We can now define climate feedbacks to mean this. They are net positive if the combination of all feedbacks produces a sensitivity greater than 0.25 K/(W/m2) and net negative if it produces a response between 0 to 0.25 K/(W/m2). A response less than zero is unphysical because a positive radiative forcing cannot cause the globe to cool, by definition.
Physically, the positive feedbacks that operate on the thermal side of the Earth's radiative budget will reduce the slope of a graph of OLR vs. T, making the line more linear than T^4. This means that the planet can still equilibirate with the incoming solar radiation, but it takes a bit more global warming to do so. The same logic applies to a negative radiative forcing scenario; in this case, positive feedbacks will make the equilibrium response even colder than what you would anticipate with just a blackbody radiator alone.
The extreme cases of a snowball Earth or a runaway greenhouse can be a case where the net feedback is no longer negative, at least until the planet is ice covered or the oceans have boiled off, at which point it is possible to equilibriate again but in a regime far-removed from the previous regime. The "equilibrium" state is defined by a point where the absorbed solar radiation equals the outgoing terrestrial radiation.
ReplyDeleteCan you please define what equilibrium means in the climate system? Since there was over 7,000ppm of co2 in the atmosphere at one time and the earth converted most of it into carbonate rock, my understanding of your argument is that the earth cannot do this again, even on a smaller scale. Furthermore, it is impossible to claim that man's influence is stronger on the climate than natural mechanisms. After all, the climate has risen to far higher temperatures naturally. Your telling us there is global warming even though the earth is below its GAT, so that doesn't make much sense. What could be simpler?
Very little of this makes any logical sense. For one thing, Earth can still balance its outgoing radiation with the absorbed solar radiation (i.e., "equilibrium") with 7,000 ppm CO2. It just has to do it at a higher surface temperature for the same solar radiation & albedo. See Figure 2 in my SkS piece on the planetary greenhouse effect.
http://skepticalscience.com/Planetary_Greenhouse.html
Secondly, there's no reason CO2 needs to stay locked up in rocks forever, and no reason humanity can't release ancient carbon in the form of fossil fuels and compete with natural processes that have operated in the past. In fact, weathering processes are trying to bring CO2 to near zero over geologic time, and outgassing from the interior balancing this process is what keeps our planet in a relatively habitable range.
Chris Colose, in your short explanation you manage to use two different definitions of "net feedback". First you state that net positive feedback corresponds to a sensitivity larger than 0.25 K/(W/m2), but in the last paragraph you state that net positive feedback leads to instability. It's exactly this ambiguity in definitions that Spencer was trying to explain.
ReplyDeleteRoy Spencer may have some strange ideas regarding feedback, but he does understand the basic facts as discussed here, and should be lauded for being one of the "skeptics" who on a regular basis responds to the more crazy people on that side by trying to explain the basics, such as that there really is a greenhouse effect etc.
Thomas,
ReplyDeleteYou're right that I was ambiguous. By "net positive" (in the first paragraph) I should specify that I am excluding the Planck response, and thinking of the sum effect of things like water vapor, clouds, albedo, lapse rate, etc. This is what "positive feedback" means in the context of climatology. In other words, these things can all act in concert to amplify the "baseline" 0.25 K/(W/m2) response by a factor of two or three without any need to call it an unstable scenario. I used the word "net" because I was thinking that individual feedbacks acting alone (such as the lapse rate response in a moist adiabatic atmosphere) can act as a negative feedback. In that sense, the Planck radiator "feedback" is not a feedback at all, but a baseline by which we evaluate whether positive or negative "feedbacks" dominate. As raypierre explained though, all this means is that they reduce the efficiency at which increased emission actually brings you to a new equilibrium.
And yes, concerning my usage of the phrase when I talked about instability, I included the Planck response in this meaning. Sorry. This is all well understood by the feedback community, and is more a matter of semantics. The casual reader reading Roy Spencer's comment would get the impression that scientists were completely neglecting a big stabilizing feature in the climate system, which is absurd.
In the final paragraph, I am including the Planck response.
It's not just the technical literature that includes the Planck feedback to get a net negative (stabilizing) feedback; that is underlying all discussions of climate sensitivity. In the denominator of the expression for climate sensitivity, which is generally written as something
ReplyDeletelike beta0*(1-f), where f is a feedback factor (with sign convention that positive f means increasing climate sensitivity), that beta0 IS the Planck feedback. The reason that essentially all discussions center on cloud, ice and water vapor feedbacks is that the value of the Planck feedback is very well known, and there's really not that much to discuss. By now, the value of water vapor feedback is quite well known as well, but there's more to discuss there, since the reasons water vapor behaves as it does are fairly intricate. So Roy is addressing a point about which no doubt some people may be confused, but I don't see any widespread confusion myself in either the technical literature or in generally informed discussion on the subject.
--raypierre
Chris, did you read the rest of the thread? I had no problem understanding what Spencer meant given the earlier comments. It may be somewhat sloppily written, but it is a blog comment not a scientific paper, so I think he can be excused. I suspect that because Spencer has said some stupid things at other times some try to interpret this statement as unfavorably as possible.
ReplyDeleteThomas-- I think Roy was trying to clarify for some other confused readers on that thread, and I'm just trying to reiterate the point here since a few readers here seemed confused. Not trying to jab at him too hard.
ReplyDeleteThomas, you are welcome to the first and the last paragraph, but the middle of the sandwich Eli says is spinach
ReplyDelete"But the MAIN climate stabilizing effect is NOT included in climate “feedback”: the increase in IR cooling to space as temperature rises (the Stefan-Boltzman(n) effect). It’s just semantics, and leads to much confusion."
Eli-- I'm sure Roy Spencer meant that when we talk about "feedbacks" we think of the Planck response as a reference system rather than a feedback, and thus don't include it when we say things like "positive feedbacks are dominating the climate system." Of course, as raypierre mentioned, everyone knows it is the reference system. Really, there's more productive things to talk about.
ReplyDeleteChris, Eli is sure that you and Ray get such kind consideration when you make a statement that is only correct with an extremely charitable interpretation, and no, Eli is not so sure about what Roy Spencer said.
ReplyDeleteYou are making the mistake that everyone is playing science. Spencer and Lindzen and Singer and the like are into Calvinball and if you play science and they play Calvinball, you lose.
Actually, as stated above- I think Happer may not have reviewed. The new PNAS rules are that all reviews are handled through the main office, so you wouldn't necessarily have them in hand when you initiate the "Contribution" process.
ReplyDeleteEli: I'm very disturbed by your call to see the reviews. Lindzen has every right to keep the reviews confidential if he chooses to. And indeed, he has a responsibility to do so unless he has express permission from the reviewers and the journal to do anything else. Lindzen is of course being an ass in claiming 'unfair review', but that's actually somewhat beside the point.
ReplyDeleteSigh.. .whatever happened to respecting the confidentiality of the process, let alone the idea of taking a negative review as an opportunity to learn something, and submit a revised and improved manuscript?
Re the role of the Planck feedback, IANAS but it seems to me that the unconfusing way to explain it is to start with the black body situation, where it's solar forcing and Planck feedback with nothing else. Then even the dull will be able to grasp Raypierre's valuble description of the non-solar positive feedbacks as making the Planck feedback less negative.
ReplyDeleteRe Spencer, his terminology slipped all over the place in the last paragraph but is correct if he's given the benefit of the doubt on the definitions. Don't criticize him for misunderstanding it, as the evidence for that seems not to be there. Do criticize him for yet another example of the sloppiness that, combined with some basic misunderstandings of the science, has resulted in what has to be the worst track record in the field.
Spencer deserves no credit whatsoever, TP.
Steve Bloom
Eric, of course Lindzen remains free to exercise his right to keep them confidential. Eli is just asking. The point is that Lindzen, in the course of making public a bunch of other material that wouldn't normally have seen the light of day (is it at all clear that he got permission?), implied an extraordinary claim about the quality of those reviews (if they exist). That being so, let's see the extrordinary evidence.
ReplyDeleteSteve Bloom
Eli, note how Spencer uses quotation marks around "feedback" in the paragraph you quote. He is talking about how the *word* is used which is made clearer by how he says "it's just semantics". If he really thought, or wanted to imply, that climate scientists had done a serious error and not included the Planck radiation properly, do you think he would have called it semantics? Even someone who doesn't understand exactly what Spencer says should realize that he doesn't consider the issue serious.
ReplyDeleteSteve Bloom, I have to disagree with you when it comes to giving credit. Posts like these deserve credit no matter what the authors may write elsewhere. It's far too common with people who refuse to try to correct people on their own side no matter what they say.On his blog, Spencer for example described how he went as far as to make some simple experiments to prove to some of his readers that there exist backradiation and a greenhouse effect. Scientifically this has, of course, not been in doubt for over a century, but nevertheless this is a myth that is spread among the contrarians, and Spencer is in a better position than most to dispel it.
The problem, TP, is that this is in the larger context of trying to create myths. The net effect of his blogospheric writings is to confuse.
ReplyDeleteSteve Bloom
Eric said: Eli: I'm very disturbed by your call to see the reviews. Lindzen has every right to keep the reviews confidential if he chooses to.
ReplyDeleteWell yes, Eric it's unseemly and self-serving to dump confidential reviewers reports and editors letters to accompany these dreary acts of whining about unfair review. But Dr. Lindzen has decided to do so, so we're kinda stuck with it (unless we resolutely avert our eyes).
Having seen the (very thorough and rather fair it seems to me) reviews of Lindzen and Choi 2011, and heard the associated bleating, it would be interesting on the grounds of getting a properly rounded view of this little farce, to see the reviews that Drs. Happer and Chou wrote, if they did in fact write reviews. Since much of the complaining seems to be about the PNAS review process (particularly in relation to Track III "Contributed" manuscripts), it would certainly help us to assess whether Dr. Lindzen was trying to take advantage of the system.
It's Dr. Lindzen that has turned the confidential review process into a spectator sport. One can hardly then complain if the spectators feel they'd like a more complete view. There's nothing worse than settling down in ones's seat and finding that a stanchion is obstructing the view of the pitch...
I don't understand how Eli is compormising the process. Lindzen said who his invited reviewers are, and invited discussion of the process, and since the "Contribution" process is not blinded in respect to the author, this isn't really the same process as for most papers. That said, I think the reviews do not exist because the request for those reviewers was denied. I hypothesize that the reviewers were probably cherry picked and likely already indicated a positive disposition to the paper. This is speculation and it is moot at this point.
ReplyDeleteActually Steve, as one who started his wanderings in the climate blogsphere completely in disbelief of CO2 having any effect, Dr. Spencers' blog was one of the main reasons why now I try to read everything with an open but critical mind.
ReplyDeleteSpencer has set a few useful stakes in the sand -- stating facts about the physical world that can't be denied. When he draws a crowd averse to one of those and ready to copypaste the topic to death -- back away slowly, edge of Internet quicksand.
ReplyDeletehttp://www.google.com/search?q=spencer+climate+"Yes+Virginia"
Bloody hell,
ReplyDeletehttp://blogs.forbes.com/patrickmichaels/2011/06/16/peer-review-and-pal-review-in-climate-science/
Eli, can you and the lagomorphs please tackle this BS.
Bloody hell,
ReplyDeletehttp://blogs.forbes.com/patrickmichaels/2011/06/16/peer-review-and-pal-review-in-climate-science/
Eli, can you and the lagomorphs please tackle this BS.
Everything about this is ridiculous, but consider the source.
"Double blindness" may be common in some fields, but it is certainly not universal, and as far as I am aware not applied widely in the Earth sciences. There are a couple problems with all the complaining done about this whole PNAS episode:
1) The denialosphere is running wild with this notion of "The Team," as well as thinking that "climate" is in some way done differently than many other disciplines like astrophyiscs or geology. Creating this artificial distinction allows their rheotoric of claiming unfairness to get by more than it should. In individual cases bad reviews will come up, either in letting bad papers slip by or on the flipside, very bias reviewers not letting scientifically sound papers get through (this is the trouble we had with Halpern et al. in our G&T rebuttal to IJMPB, before we took it up with higher-ups). There are of course thousands of papers on climate science, and with countless people required to review work (some of this task may be carried over to graduate students for experience). The Michaels article makes it sound like a high church of 10 review priests that sit there and review papers all day, only accepting those who agree with pre-conceived notions, but this is of course absurd. Besides, what does a person gain from letting a paper by that they are not even authoring which turns out to be bad?
2) If the denialists had their way, all reviewers would either be biased or unqualified by definition. We don't live in a world with expert robot reviewers. Scientists of all fields are human, and being experts, naturally have their own thoughts on the subject-matter of their expertise.
Eli, can you and the lagomorphs please tackle this BS.
ReplyDeleteDouble blind reviewing. I must have reviewed a couple of hundred papers by now in the fields of Biochemistry/Biophysics/Mol Biol. I've participated in exactly one example of "double blind" peer review. Of around 80 papers published myself exactly zero were reviewed by "double blind" peer review.
Of course poor old Dr. Michaels (and all of the other hand-wringing peer review concern trolls) are at a disadvantage compared to pukka scientists. They lack some of the attributes that make science, and scientific publishing (by and large), work, namely a burning desire to find stuff out about the natural world, and the personal standards of integrity to write good papers and (by and large) review papers properly an honestly.
So maybe we can understand their angst a little, but we shouldn't feel too sorry for them, much as we have only a little sympathy for the little boy who is unhappy because his mother won't let him pull the wings of that beautiful butterfly...
For those who don't already know, Pat Michaels has for some years been Chip Knappenberger's employer.
ReplyDeleteSteve Bloom
Double blind has been proposed and indeed is used in some fields/journals although over a lot of years and reviews Eli has never had the pleasure. The usual response is that it is up to the editors to pick good reviewers and that often it is easy to figure out what lab something came from. It's really not an issue for most.
ReplyDeleteOf all of the proposals, the Eurojournal open reviews appear to Eli to be the best idea going.
James has some comments on open review
ReplyDelete