Here comes a border collie

BorderCollie Whew, that's a relief. The Arbiter has finally spoken, and I can stop trying to think for myself. Not that I was trying very hard. No one really likes thinking for themselves anyway - if you can its hard work, and if you can't its not pretty.

Yeah, I should probably have had a tl;dr version, which is that sensitivity is still about 3C.

Well, that's the only bit people care about really. Mialambre continues:

The discerning reader will already have noted that my previous posts on the matter actually point to a value more likely on the low side of this rather than higher, and were I pressed for a more precise value, 2.5 might have been a better choice even then. But I'd rather be a little conservative than risk being too Pollyanna-ish about it.

Of course that's not the bit the septics picked out (incidentally, notice that Young James has become a Hockey Team scientist; I'm not quite sure what he's done to be awarded that distinction; probably its a sort of attempt at a Lomborgian "I was once an env" or summat); they chose:

makes a high climate sensitivity increasingly untenable. A value (slightly) under 2 is certainly looking a whole lot more plausible than anything above 4.5.

(that from an email to Andy Revkin [ref]). That's a somewhat regrettable quote, because its easy to misinterpret it. The wackos alreay have, but I've already seen sane (but, it would seem, careless) people interpret it as "likely < 2", which isn't what it says. Since he's already said in the blog post "sensitivities above 4.5°C are extremely unlikely (less than 5%)", all that says is that "< 2 is more plausible that extremely unlikely" which isn't very parseable.

The point of James's post isn't to argue for a particular value, but to attack the "long tail" of high sensitivities stuff, and the IPCC reporting. Its a very readable post, so I see no point in re-writing it here.

This is all wrapped up around pre-release of the IPCC AR5 report, and the wildly exciting Lewis estimate of sensitivity. Like most people who've looked at this JA has doubts about Nic Lewis' analysis, as I think some of his choices are dubious and will have acted to underestimate the true sensitivity somewhat. I still think that if Lewis thinks his estimate has any value, he should write it up nicely and try to get it published. But he won't.

Meanwhile... remember how AW was going to sue the pants off Greg Laden, or something? What happened to that? Did it just fade away like a pile of ill-thought out ranting under the light of reason?

More like this

My hunch for a while has been somewhere between 2.5 and 3. This is partly based on Annan's earlier work, and it seems to still be where he would put the bulk of his pdf. Anything low is ruled out by physics and paleo, anything high is ruled out by observations and GCMs. Watching AW and his idiot followers seize on Annan's pretty clear remark (about 4.5) should be a lesson for us all.

By Nick Barnes (not verified) on 02 Feb 2013 #permalink

James has been pretty consistent both in numbers and methods. Eli and Willard will have something to say about the latter later in the week (this is a teaser:)

By Eli Rabett (not verified) on 02 Feb 2013 #permalink

I'm sure the value is 2.718281828459045

By David B. Benson (not verified) on 02 Feb 2013 #permalink

david b., now that would be cool. if sensitivity changes with conditions and time we might define it just that ;-). then there would just be some (other) forcings that twist the number this way and that.,

Named after Eli. Uncle E

Lovely David B:)

By Eli Rabett (not verified) on 02 Feb 2013 #permalink

It's all about me! Thanks for the cartoon. Better than anything that Josh person could do, I'm sure. Yes my point about 2 and 4.5 is that the IPCC "likely" range indicates an equivalence between these values which IMO is not sustainable.

i tried playing hockey once. I thought it was a rather crap game, though that might have been down to the hacked up rugby pitch, or more likely my incompetence. As for the abomination that is ice hockey, any fule no that the only proper use for an ice rink is curling.

By James Annan (not verified) on 03 Feb 2013 #permalink

Peanut gallery question.

I believe somebody said something like:

observations since 2000 necessarily lower sensitivity (estimates I guess).

I've read numerous explanations for observations since 2000. Most, not all, anticipate a reversal. If the reversal commences in 2014, does that mean observations post 2014 would necessarily raise sensitivity?

[Presumably not sensitivity itself, only estimates of it -W]

(I concur with praise for David B's precision.) :-)

There are real arguments among serious people (like JA & IPCC) about the shape of PDFs describing climate sensitivity.

Then there is a lot of blogosphere-beloved confusion about over-interpetiing words used to grossly-approximate PDFs or say something about them..

A real argument is: here are two PDFs A&B, superimposed on same graph, and with first 4 moments (mean, std, deviation, skew and kurtosis), and people can argue whether the data better supports A or B.

Talking about ranges without being clear about the assumed shapes of the PDFs is a recipe for confusion, given that no plausible PDF for sensitivity is uniform. (Whether the shape is normal, lognormal or something else is always open to much debate.) But I hate the unadorned use of ranges for such things.

Simplistic analogy:
One honest die: uniform distribution, range 1:6.
Two honest die: binomial,range 2:12 ... but just saying that its range is 2X bigger is accurate ... but mis-communciates because the two have very different PDFs.

At the risk of over-interpreting JA, it think his comment that stirred the blogosphere was ~ saying that a 0 dice roll was less unlikely than a 14. (Which might be correct, if the die rolled off a table and a border collie sitting under the table thought they were food and swallowed them, so no spots showed :-))

In any case, what really counts isn't sensitivity, but the actual temperature rise = f(sensitivity, CO2 rise). Some people seem to want to think "if it's only 2X, all is well" akin to the "if it's only 2degC by 2100, all is well.

By John Mashey (not verified) on 03 Feb 2013 #permalink

"i tried playing hockey once. I thought it was a rather crap game,"

I used to love playing hockey, though that may have been down to it being in my late teens and early twenties, and largely in mixed games.

I heartily agree with David B. Benson's estimate and wonder if the discussion should move from pdf's of S to pie charts?

Annan's post also comments on how a yet unnamed climate expert "lied" (his choice of words, not mine) about inflating the climate sensitivities " to help motivate political action". For some reason no mention thus far on that point here - funny that.

[As I said: "Its a very readable post, so I see no point in re-writing it here" -W]

I've coined the term midwarmer to describe those who ascribe to 2.7 to 3.2. Lukewarmer, midwarmer, assuredly damaging warmer, existential threat warmer - name the tribes.

Going back to the 4th climate belief that we should do something about it, what place does sensitivity have or should have in discussions about that something?

If it’s only 2degC by 2100, all will not be well, but it will be looking pretty good. It will mean that net emissions will likely be significantly reduced, and/or that concentrations have been stabilized.

[Don't forget that dT isn't sensitivity. But, other things being equal, dT will scale with sensitivity. So yes, it has the obvious effect: the lower it is, the less dT we'll see for a given level of emissions. I don't understand your "It will mean that net emissions will likely be significantly reduced" -W]

By Paul Kelly (not verified) on 04 Feb 2013 #permalink

I still think that if Lewis thinks his estimate has any value, he should write it up nicely and try to get it published. But he won’t.

It is difficult to see how Nic Lewis could publish his analysis, based as it is on analysis of a leaked AR5 report.

[That will do for an excuse, if he wants one, but it doesn't stand up. The stuff in AR5 will be based on published papers; all AR5 does it point to them -W]

I hope that the IPCC lead authors take Lewis's arguments into consideration before writing up the final version of the IPCC report. I suspect James' comments are intended in the same vein. If they don't, perhaps Nic Lewis will publish a paper that criticises the IPCC AR5 sensitivity analysis at that time.

By Alex Harvey (not verified) on 04 Feb 2013 #permalink

I am sure we would be equally excited if an actual peer reviewed estimate of the equilibrium climate sensitivity was published by Nic Lewis, whereas we may agree that a comment on a leaked IPCC draft would not be published.

While waiting there are of course lots of other interesting papers in the literature that come to similar conclusions as Nic Lewis, e.g.

Determination of Earth's transient and equilibrium climate sensitivities from observations over the twentieth century: Strong dependence on assumed forcing. Schwartz S. E. Surveys Geophys. 33, 745-777 (2012). DOI 10.1007/s10712-012-9180-4 .

[Isn't that the same Schwartz who made such a hash of it before? -W]

My hope is that the slew of papers appearing arguing for a climate sensitivity at the low end and beyond of the AR4 'likely' range have convinced enough scientists to be brave enough - as your friend James Annan has been - to at least discuss the possibility without fear of persecution.

And who knows, perhaps discussion of the Lindzen/Choi argument I've followed for so long will finally become a reality?

Choi, Y.-S., H. Cho, C.-H. Ho, R. S. Lindzen, S.-K. Park, and X. Yu (2013), Influence of non-feedback variations of radiation on the determination of the climate feedback, submitted.
https://sites.google.com/site/yongsangchoi/publications

By Alex Harvey (not verified) on 04 Feb 2013 #permalink

I saw somewhere the following statement;

The probability of a Climate Sensitivity greater than 4 is equal to the probability of killing yourself in a single-chamber-full game of Russian Roulette.

[http://tamino.wordpress.com/2012/01/04/what-is-epsilon/#comment-58420 perhaps? But its inverted. The other way round, you have to allow yourself at least 20 chambers -W]

Since we've come to a fairly dangerous place after .9C of warming, 2C of warming hardly seems like All Clear.

By Jeffrey Davis (not verified) on 04 Feb 2013 #permalink

Schwartz has a history of poking his nose into things he doesn't really understand - eg, he is one of many who think they understand the carbon cycle better than the carbon cycle experts, stating, "No evidence that if CO2 emissions were halted, CO2 would plateau out to a value substantially greater than preindustrial." http://www.ecd.bnl.gov/steve/pres/WellKnownAGU10vgphs.pdf

(this is one of my pet peeves. He isn't a "5 year lifetime" crazy, but he should still know better - fitting historical data is not the same as understanding actual mechanisms and physical limits)

Not sure why Alex highlighted the Schwartz paper in this context. As far as I can tell it's basically a review article documenting the relationship between forcing and sensitivity across various studies. He states an ECS range of 1.2 to 4.9ºC, which is very consensusy, but I'm not sure how independent that is from other estimates.

Paul Kelly:

"I’ve coined the term midwarmer to describe those who ascribe to 2.7 to 3.2."

So that includes the "about 3C" consensus view. It also includes the sensitivity that falls out of GISS Model E, which is 2.7-ish.

How does one differentiate a "midwarmer" from a "climate scientist"?

dhogaza,

Most climate scientists are midwarmers; at least the ones i pay attention to are. It is mostly activists and politicians who push the long tail, high sensitivity line.

By Paul Kelly (not verified) on 04 Feb 2013 #permalink

After reading Alex's post#13 I'm wondering if we should have a whip-round for James Annan as a small gesture to help him cope with all the persecution he's braving.

Send him a pizza

By Eli Rabett (not verified) on 04 Feb 2013 #permalink

Actually, seeing that the GISS CMIP5 model yields a sensitivity of about 2.5C (down from earlier versions which were at about 2.7), so I guess these modelers are "lukewarmers".

Only pointing this out because Kelley's attempting to position the mainstream view of sensitivity - the lukewarmer-to-midwarmer view - towards the realm of the self-proclaimed lukewarmer skeptics, and by association wants one to believe that the problems we face are being exaggerated by "politicians and activists".

The reality, though, is that with a sensitivity of 2.5-3C and continued growth in emissions, limiting warmth to 2C by 2100 is not an attainable goal.

"If it’s only 2degC by 2100, all will not be well, but it will be looking pretty good." - 0C would be even better, but neither is likely to happen ...

Paul S.,

You must be looking at the wrong study.

From the abstract,
Equilibrium sensitivities determined by
two methods that account for the rate of planetary heat uptake range from 0.24 to 0.75 K (W m-2)-1 (CO2 doubling
temperature 0.88 to 2.75 K), less than, to well less than, the IPCC central value and estimated uncertainty range, and
strongly anticorrelated with the forcing used to determine the sensitivities.

By Alex Harvey (not verified) on 04 Feb 2013 #permalink

William,

Dhogaza's "with a sensitivity of 2.5-3C and continued growth in emissions, limiting warmth to 2C by 2100 is not an attainable goal" is the basis for my saying a 2c limit by 2100 would mean that emissions will likely be significantly reduced.

dhogaza,

I don't want to position the mainstream view of sensitivity anywhere but the dead center middle. That is the most logical starting point for policy discussions. Yes, i do think there is a lot of exaggeration by activists and politicians which hinders rather than promotes action.

By Paul Kelly (not verified) on 04 Feb 2013 #permalink

Not to mention a lot of exaggeration by the malinformed and contrarians which hinders rather than promotes action.

By American Idiot (not verified) on 04 Feb 2013 #permalink

"I don’t want to position the mainstream view of sensitivity anywhere but the dead center middle. That is the most logical starting point for policy discussions."

And, of course, it is becoming increasingly clear that it's inconsistent with the 2C increase by 2100 that you'd describe as "looking pretty good".

Ignoring the fact that you have no basis for a 2C increase as "looking pretty good", it's simply inconsistent with your antecedant, which is that mainstream climate science has sensitivity more or less nailed.

You seem to be arguing that science is right about sensitivity, but wrong about the negative affects that come with it.

State so clearly, please, and don't dance around the 2C increase by 2100 which is not consistent with CO2 emissions growth and the mainstream view of sensitivity.

dhogaza, GISS model E would be closer to "denier", since it has ECS as 2.5 (according to Gavin) ;-)

Alex,

You must be looking at the wrong study.

Nope, here's the link: http://link.springer.com/article/10.1007%2Fs10712-012-9180-4

It matches the DOI, publication, volume and page numbers you quoted. Abstract states: 'Equilibrium sensitivities determined by two methods that account for the rate of planetary heat uptake range from 0.31 ± 0.02 to 1.32 ± 0.31 K (W m−2)−1 (CO2 doubling temperature 1.16 ± 0.09–4.9 ± 1.2 K), more than spanning the IPCC estimated “likely” uncertainty range, and strongly anticorrelated with the forcing used to determine the sensitivities.'

Large pepperoni please, with olives.

[Large? I thought you complained about the gross portions over here. Small for you -W]

By James Annan (not verified) on 05 Feb 2013 #permalink

Persecution pizza with pepperoni. Only at Antonio.

Paul S., that's bizarre - it's the same abstract except that my copy has completely different numbers. And it states clearly 'accepted for publication, Jan 2012'. I wonder what happened...

By Alex Harvey (not verified) on 05 Feb 2013 #permalink

Looks like James gave Judith Curry another "Wow!" moment, although it seems her jaw dropped so low it must have come out something like "Roohhwah"

[She doesn't seem to like me much, either. Ah well, I'll live -W]

This made me laugh

http://wingod.newsvine.com/_news/2013/02/01/16809711-james-empty-blog-a…

This is Dr. James Annan's blog. He is one of the core scientists of the climate consensus being the co-author of the "Annan and Wahl, et al." paper on tree ring climate reconstructions that the Mike Mann MBH 1998 paper is based on.

[Oh dear. It doesn't look like any of the commentators have noticed -W]

PeteB,

They didn't mention that he was a prominent figure at the UN until his recent resignation over the Syrian situation.

Could this be even worse than not being able to spell 'Connolley'?

;-)

Could someone clarify something for me: does this cutting off of the long upper tail account for the possibility of non-linear feedbacks? Is the possibility of significant additional feedbacks considered small enough to be discounted?

[Most of the feedbacks are technically non-linear: the sea-ice albedo one for example. In practice they are effectively linear, within a certain range of forcing, though probably not all the way up to 2*CO2 (summer sea ice will disappear and hence the linearity must disappear, at some point). I think your question is ill-posed. I think you're thinking of "tipping point" type stuff, but that is hard to work with -W]

dhogaza,

You're manufacturing a disagreement where none exists. We agree on the most likely sensitivity. We agree that, at that sensitivity, current emission trends mean warming will be greater than 2C by 2100. We should agree that warming by 2100 limited to 2C would mean that emissions had been significantly reduced in the interim.

By Paul Kelly (not verified) on 05 Feb 2013 #permalink

it seems to me that this latest molehill episode means bupkiss from a risk management perspective. any sane strategy would consider outcomes that can't be ruled out. you don't design your policies around the most likely outcome. saying that S is most likely less than 4.5C is not the same ruling it out altogether.

By Marlowe Johnson (not verified) on 05 Feb 2013 #permalink

PK: "I don’t want to position the mainstream view of sensitivity anywhere but the dead center middle. That is the most logical starting point for policy discussions. "

Risk management much? No, didn't think so.

By Steve Bloom (not verified) on 05 Feb 2013 #permalink

Agreed, Marlowe, but getting from even a correct value for ECS (given present boundary conditions) to what's actually going to happen under a given emissions scenario is difficult, to say the least. For one thing, and this point seems to be frequently glossed over, ECS can never be reached since the fast feedbacks won't be able to reach equilibrium before the slow ones start kicking in (as indeed they already are). Raypierre's advice to always bear in mind ESS should be paramount in these discussions, notwithstanding that our understanding of any equilibrium state has its limitations given that what we're dealing with is a fast transient with the potential to trigger unprecedented feedbacks (in scale if not in kind).

Opatrick, further to the foregoing, I think the key point is that even if we can substantially minimize the ECS long tail, a real world one remains.

By Steve Bloom (not verified) on 05 Feb 2013 #permalink

PK: "We should agree that warming by 2100 limited to 2C would mean that emissions had been significantly reduced in the interim."

Well, OK, if by "interim" you mean ~1995, we're in complete agreement. :)

By Steve Bloom (not verified) on 05 Feb 2013 #permalink

Has James actually complained about the portions in Blighty? I'd thought that criticism was limited to the colonies, reconfirmed annually by yet another SF gourmand-a-thon.

I like the persecution pizza concept, Neven, but I don't think you can get a decent pizza in Chico. Or is that the point?

Anyway, what do I know about pizza, but Chico's pretty close to SF, so maybe this December we could get an expert assessment? Sticking to those high-class joints must be so boring after a while.

Chico: Where a restaurant's Michelin rating is based on the number of tire tracks left by the last drunken Chico State student to have driven through the front window! :)

By Steve Bloom (not verified) on 05 Feb 2013 #permalink

"I think your question is ill-posed." Undoubtably.

[I meant it in a nice way, of course -W]

"I think you're thinking of "tipping point" type stuff" quite possibly. I was thinking of things like permafrost melt and changes in rainforests, as well as sea-ice albedo changes. not necessarily as dramatic as a tipping point, but feedbacks that have had little impact so far but may become much more significant.

[That kind of stuff comes under "carbon cycle feedbacks". They are in some models, but not all. I think that they are (in the great scheme of things) relatively small. Rainforest dieback adds extra carbon, but not a huge amount. I'd guess that, compared to the disaster of the rainforest dying, the extra carbon would be a smaller matter -W]

"but that is hard to work with". Yes, I think that was my question. Does the fact that they are hard to work with mean they are excluded from the calculations? And is it right that they should be if they are?

Oh, com'on Steve Bloom, I've rather enjoyed the meals I've had at Sierra Nevada Brewing :-).

By Rattus Norvegicus (not verified) on 05 Feb 2013 #permalink

Well I've never met Caspar Amman, he may be my friendly ghost. On the other hand, I'm frequently round at Great Uncle Kofi's place for drinks, now that the famous graffito Banksy Moon has taken over the UN. The stories Kofi tells about the IPCC are enough to turn my hair grey...or even make it fall out...

By James Annan (not verified) on 05 Feb 2013 #permalink

>>> Paul S
>> Alex,
http://link.springer.com/article/10.1007%2Fs10712-012-9180-4

Do either of you have the full text?
It would be interesting to know if the actual text released as "accepted for publication" is different from what was published. (And who released it and who wrote the abstract, of course).

One of those early lessons from grad school -- abstracts are not citable sources: you have to read, understand, and cite the actual paper.

I recall suspicion that abstracts often are drafted by the same people who write those awful university press releases tpumped out before journals appear.

By Hank Roberts (not verified) on 06 Feb 2013 #permalink

Hank,

Stephen Schwartz's home page contains a link to an abstract relating to the paper, dated 19 January 2012, and a preprint which indicates the paper was accepted on 24 March 2012. The pdf document itself is dated 25 March 2012.

The January abstract contains numbers quoted by Alex, whereas the March preprint contains the published numbers I quoted.

The top line of the abstract page indicates the paper was in press as of January, which presumably means it had been accepted in that form. The March preprint pdf indicates a revision was made on 20 March, which was then accepted on 24 March. The short timespan suggests the paper had at least been preliminarily accepted prior to that but Schwartz wanted to make a small revision due to further discussion or identification of an error/inconsistency.

Hank/Paul S.,

I am sure there is an innocent explanation.

I can't remember where I got my copy from but it would have been before the March revision. (I'll check the timestamp on my file next time I am on the right computer.)

I've sent an email to Prof. Schwartz to ask him - if he responds and gives me permission I'll let you know.

I have read my copy of the paper several times and the body of the article supports the abstract.

I think the most likely explanation is that an error was found - perhaps by Schwartz himself - and the paper was revised after it was accepted.

Maybe it was that damned fat tail problem. ;-)

By Alex Harvey (not verified) on 06 Feb 2013 #permalink

- Hello, Mr. Amman. We have here for you one persecution pizza with pepperoni!
- You must be mistaken. I am currently toeing the party line, and thus not persecuted.
- Would you like to be persecuted by us? We have Team Member discount. 10 for the price of 1.
- No, thanks.
- Are you sure?
- Yes.
- So, what do we do with this persecution pizza?
- Take it to Michael Mann.
- Where does he live?
- In Hollywood.
- Where exactly?
- Ask Roger Pielke.
- Which one?
- Whatever. Can I go back to the game now? Lewandowski just scored.

>"Does the fact that they are hard to work with mean they are excluded from the calculations? And is it right that they should be if they are?"

Some are included and should be included like water vapour.

Carbon cycle feedbacks are excluded by definition of sensitivity (to a doubling of CO2 level). Should they be excluded? Well that depends on whether you want a description of the system. For this you want them excluded.

If you want to know the sensitivity of the system to carbon emissions then you want the carbon feedback effects included in some manner. The easiest manner may well keep the sensitivity to CO2 separate from ascertaining carbon cycle feedbacks from a carbon cycle model.

In short, it may well be right to exclude carbon cycle feedbacks as long as you remember they have to be included when figuring out effects and permissible emission levels.

Sea ice albedo is a reason why sensitivity from now on when we are committed to loss of summer sea ice should be lower than sensitivity from pre-industrial state. We are interested in sensitivity from pre-industrial state. What is included is generally what we have seen occur. Models seem to make it hard for the ice to disappear in summer and therefore only predict a little more of the same.

There increasingly seems a significant risk the summer arctic ice cover will disappear rapidly. What occurs if that happens seems worth considering? Do we get a noticeably large temperature effect from the disappearance or not? The albedo effect is presumably lessened after that. But are we more interested in this long term effect or in the short term effect?

I think the increasing risk of this means more work should be directed at looking what climate effects we get from rapid collapse of arctic sea ice just in case it plays out as trends indicate rather than relying on model projections of a slowing rate of sea ice loss.

Russel, perhaps it would be more convergent if you wouldn't mix apples and oranges. What's that ~1963 spike, BTW?

By Steve Bloom (not verified) on 06 Feb 2013 #permalink

Paul S./Hank,

Dr. Schwartz responded and confirmed that he did indeed find an error himself after it was accepted, and was able to contact the editor and submit a revision before actual publication. All copies of the earlier paper were removed, but I got my copy before the error was found.

The actual conclusion w/r/t climate sensitivity is best captured in the following paragraph -

Values of the equilibrium sensitivities were determined for five of the six forcing data sets examined, all of which are within the ‘‘very likely’’ range2 given for this forcing by the IPCC (2007) Assessment. These sensitivities range from 0.31–1.32 K (W m-2)-1, corresponding to delta T_2x 1.2–4.9 K, and, with the exception of the MIROC forcing data set, are
less than the best estimate, 3 K given for this quantity by the 2007 IPCC Assessment, Fig. 12. Two of the five forcing data sets yield sensitivities within the IPCC ‘‘likely’’ range (see
Footnote 2), 2–4.5 K, with two below and one (MIROC) above this range, although the large uncertainty attached to the latter extends well into this range. For the PCM and GFDL
forcing data sets, the equilibrium sensitivities are below the lower bound of the ‘‘likely’’ range for this quantity, and indeed are nearly at, or below, the limit of the ‘‘very likely’’ range for this quantity, delta T_2x = 1.5 K.

By Alex Harvey (not verified) on 08 Feb 2013 #permalink

Alex,
Yet another indication that you should post a link when you refer to abstracts and claims and evidence.

Regarding Schwartz, it seems that he has changed his estimates of climate sensitivity quite drastically over time :

In 2007, he estimated "equilibrium temperature increase for doubled CO2 of 1.1 ± 0.5 K." :
http://www.ecd.bnl.gov/steve/pubs/HeatCapacity.pdf
which was hailed by contrarians, even though it was debunked scientifically.

Then 2011-07-20 he submitted a draft which suggests "Equilibrium sensitivities .... CO2 doubling temperature 0.88 to 2.75 K"
http://www.ecd.bnl.gov/pubs/BNL-96153-2011-JA.pdf
Which abstract matches the one you just posted (again without giving a link).
Is this the draft you refer too ?

The final paper, which was correctly linked to by Paul S
http://link.springer.com/article/10.1007%2Fs10712-012-9180-4
is available in full here :
http://www.ecd.bnl.gov/steve/pubs/ObsDetClimSensy.pdf
which was Revised 2012-03-20; Accepted 03-24, to the final version :
"Equilibrium sensitivities .... CO2 doubling temperature 1.16 ± 0.09 to 4.9 ± 1.2 K"

By Rob Dekker (not verified) on 10 Feb 2013 #permalink

Rob,

No, I didn't refer to a draft; I referred to a version that was accepted for publication in Jan, 2012. And yes my Jan, 2012, accepted version has the same abstract as your Jul, 2011, submitted version.

I've read both the accepted and published versions side by side now and it was very obviously an honest mistake and a fairly unlucky one at that given the large effect it had on the one thing that everyone cares about (climate sensitivity estimates), which wasn't the main point of the paper.

Making a mistake and then revising your opinion after you find the mistake seems to be doing good science to me. If 'contrarians' have blown Schwartz's conclusions out of proportion that's hardly his fault. We all seem to agree that scientists must be free to publish evidence that may even contradict the IPCC without fear of persecution - yes? :-)

I would say that the published version of Schwartz's is worth having a look at. If you just eyeball one figure, go to fig. 12 and note that the aerosol forcing as inferred from several GCMs (the exception being MIROC, which is an outlier, and moved the upper ECS bound way up to 4.9 K) is still suggestive of a lower to considerably lower climate sensitivity, when this is combined in an observational determination of climate sensitivity with the 20th century data. The PCM model aerosol forcing, for instance, implies a climate sensitivity of about 1.1 K/doubling of CO2.

I don't know if Schwartz's opinion has changed much over the years w/r/t climate sensitivity. But his work has been instrumental in setting off a thread of discussion in the literature on the question of why the GCMs all reproduced the 20th century temperature anomaly so well despite having completely different treatments of the aerosol forcings in them.

E.g. Kiehl, J. T. (2007), Twentieth century climate model response and climate sensitivity, Geophys. Res. Lett., 34, L22710, doi:10.1029/2007GL031383.

Knutti, R. (2008), Why are climate models reproducing the observed global surface warming so well?, Geophys. Res. Lett., 35, L18704, doi:10.1029/2008GL034932.

Huybers, P. (2010), Compensation between Model Feedbacks and Curtailment of Climate Sensitivity, J. Clim, DOI: 10.1175/2010JCLI3380.1.

Schwartz appears to have been one of the early voices calling for this investigation. So, if his research has led to better understanding of a very important question, that seems to be someone doing good science to me.

You'll note there are some interesting discussions of all this in the forthcoming IPCC AR5 report.

By Alex Harvey (not verified) on 10 Feb 2013 #permalink

> All copies of the earlier paper were removed,
> but I got my copy before the error was found

There's a reason for pointing to the original, not to a copy. "Rule One of Databases" as I was taught it.

By Hank Roberts (not verified) on 10 Feb 2013 #permalink