The following waste of paper journal article

A methodological note on the making of causal statements in the debate on anthropogenic global warming
http://www.springerlink.com.proxy.bib.uottawa.ca/content/n238rl53w2571262/

is bound to create quite a stir within the skeptic community. An idea that links to the peer-reviewed skeptical science article is that we respond with a comment on this paper (if able). I have put together some thoughts about this article that I think are an okay start (though rambling) to a dissection to the article either as a comment or SKS article. I would love some comments and references to put in certain places and some structure etc...

Kampen (2010), hereafter KP10, makes the argument that the basis for current understanding on anthropogenic global warming (AGW) is constructed almost exclusively from correlation research. KP10 furthermore makes the claim that there is significant motivation for the maintenance of the current view on AGW because of monetary investments associated with this theory.

KP10 initial application of Bunge's theory (et al.) is not correct in making the assumption that it applies to the earth's climatic system. According to KP10, Bunge's theory is only applicable where the relationship between two variables (implicitly CO2 and Temperature) is unique with one having a cause and the author an effect. KP10 also claim that Bunge's theory can be applied because the two variables are asymmetrical with one being able to force the other but not vice versa. The application of Bunge's theory to AGW is thereby unwarranted as it ignores that CO2 can act as both a forcing and a response to atmospheric temperatures (et al.). Well understood oceanic carbon absorption rates show that with climatic warming, CO2 is outgassed into the atmosphere because the carbon storage capacity of the oceans is lower with higher temperatures (Et al.). Furthermore, the best available scientific evidence also indicates that CO2 provides the amplifying effect on global temperatures during interglacials which show it as being a forcing also (Ruddiman 2006) not to mention the plethora of evidence from other geological periods indicate CO2 is an important driver of atmospheric temperatures (Ruddiman 1990). 

KP10 makes the claim that some skeptics believe that proponents of AGW point to natural climatic disasters as verification of a human impact on climate. Although this claim is not supported by the scientific literature, it is likely something that those in the sceptical community have indicated in the past. KP10 must also acknowledge therefore in making this statement, that there are instances where those sceptical of AGW point to extreme cold events as being evidence that AGW is in effect not operating. Neither instance by either the proponents of AGW or the skeptics of AGW is completely warranted however scientific evidence can in fact suggest that certain extreme events are more likely to occur with increasing atmospheric temperature (et al.) Furthermore as the recent heat wave in Russia suggests, some recent events are unprecedented over large time intervals (>= 1000 years) and therefore support conclusions that the climate system is perturbed by an external forcing (IPCC 2007). KP10 further claims that the theory behind AGW is not methodologically sound because it is claimed to “cherry-pick” pieces of evidence which confirm the theory. The author points to a Greenland Ice-Core Bore Record which does not show unusual warmth post WWII as being evidence that a piece of falsifying evidence has been ignored by proponents of AGW. What the author of KP10 does not acknowledge however is that although ( et al) is an important study, it is a single location, covering a particular timespan. In his criticism of proponents of AGW he himself “cherry-picks” which study to use as evidence to support his claim rather than using hemispheric or global reconstructions that do not support his claim. When the best available paleo evidence of climatic changes is compiled into comprehensive studies with multiple proxy sites, it is clear that the late 1990s to early 2000s warmth is unprecedented over the last 2000 years (Moberg et al. 2005, Mann et al. 2008, Ljunqvist et al. 2010).  

KP10 further makes a comparison to the IPCC consensus as being similar to the consensus’ which disagreed with propositions by Galileo, Copernicus and Darwin. What makes this criticism invalid is that there are virtually no publishing climate scientists that disagree with the key tenets that increased emissions of CO2 will cause some global warming. The disagreement often exists as to how much warming and this relates directly to feedback processes and climate sensitivity. Therefore although the author concludes that there’s no guarantee that the majority will reach the most sensible conclusion, the actual majority in this case is not simply the majority of scientists related to AGW but rather it is important to note that 97% of actively publishing climate scientists have identified AGW as being a real phenomenon (Doran and Zimmerman 2009). This is a majority amongst a very small group of scientists who are experts in climate sciences.

KP10 also makes the speculative argument that scientists will not “bite the hand that feeds them” implying that much funding is received through AGW and therefore scientists would not discuss their uncertainties or results if they counter AGW theory. This statement is completely without merit as the scientific literature has shown time and time again that authors who are proponents of AGW still publish studies that reduce the implications of AGW. An example of which is a study by Bamber et al. (2009) which concludes that a collapse of the West Antarctic Ice Sheet would have a significantly small sea level rise than previously known (3.2 m compared to previous estimates of 5 m). If scientists were not countering AGW such as KP10 contends, then this sort of reassessment would likely not be found in the literature because it seems to reduce potential impacts of AGW on coastal regions. Furthermore, if the anecdotal evidence presented by KP10 were correct, then articles questioning the scientific evidence behind AGW would not be published in the literature, however there have been many examples of articles published which do not support the scientific consensus (Mclean et al. 2007???, Douglas et al. 2008, Lindzen and Choi. 2009, Mckitrick and Mcintyre 2010). The claim made by KP10 is not substantiated by the scientific literature and can best be described as unsupported by scientific evidence.

KP10’s central claim is based on the implicit premise that there is a lack of experimental (or empirical) data which supports AGW and that therefore proponents of KP10 use causal relationships derived from correlational data to make assumptions. This premise is without merit as it is reliant upon a false claim that there is a lack of experimental data to support AGW. There is in fact a multitude of empirical data which supports AGW. This data has been collected from satellites, ground meteorological stations and spectrometry. In particular over the period 1970 to 2006 there has been an observed decrease in the amount of longwave (LW) radiation escaping from the Earth’s atmosphere into space (Harries et al. 2001, Griggs and Harries 2004, Chen et al. 2007). On the ground measurements using from 1973 to 2008 have measured an increase in downward LW radiation across the globe (Wang and Liang 2009). This is corroborated by Philipona et al. (2004) who analysed atmospheric longwave downward radiation over the central Alps and found statistically significant radiative forcing due to an enhanced greenhouse effect. These distinct empirical evidences demonstrate that the greenhouse effect is increasing, the foundational principle of AGW theory. Furthermore, using spectrometry Evans and Puckrin (2006) measured the wavelengths at which LW radiation is returning to the earth and found the forcing of the greenhouse effect from CO2 and other trace gases is very close to the forcing as predicted by climate models. Other model predictions specific to greenhouse forced warming include that nights would warm faster than days (Alexander et al. 2006, Fan et al. 2010), polar regions would warm faster than the rest of the planet, the tropopause should rise with greenhouse warming (Santer et al. 2003) and both the ionosphere and upper atmosphere would cool (Laštovička et al. 2006). These AGW signatures identified by climate models have all been confirmed through empirical measurements (et al et al et al et al).

Let me know what you think. Plus references are welcome.

neal - this is intended as a comment to be published by the same journal which published the paper in question.

A couple notes - you need to define AGW.  In the third paragraph I would take out the discussion of Fox News and the media in general.  I think it's sufficient to say this claim isn't supported by the scientific literature.  Maybe throw in a sentence that perhaps it refers to the mainstream media, in which case there is a preponderance of examples of cold weather cited as evidence against AGW.

Fourth paragraph there's a sentence fragment - "Therefore although the author concludes that there’s no guarantee that the majority will reach the most sensible conclusion."  For the 97% statement, reference Doran and Zimmerman (2009).  I don't think it's necessary to state that 97% is a majority, I think that goes without saying ;-)

In the fifth paragraph I'd add some examples of 'skeptics' getting published.  There are many recent examples to choose from - Lindzen and Choi '09, McLean et al., the troposphere temps paper by Spencer, Christy, Douglass, Singer, etc.

In the last paragraph the second sentence is basically a repeat of the first.  The third sentence is missing the word "exists", or something similar.

I can help come up with some more references later on.

This statement "There is in fact a multitude of empirical data which supports AGW. This data has been collected from satellites, ground meteorological stations, spectrometry and climate models." - should "and climate models" be there? Do you collect empirical data from a climate model?

Is enhanced polar warming a signature of greenhouse warming? My understanding is it should happen due to any warming - we've seen polar amplification in past periods of warming.

Some references on OLR:

In particular over the period 1970 to 2008 there has been an observed decrease in the amount of longwave radiation (LW) escaping from the Earth’s atmosphere into space (Harries et al. 2001, Griggs & Harries 2004, Chen et al. 2007).

Updated dates on Wang 2009:

On the ground measurements from 1973 to 2008 have measured an increase in downward LW radiation across the globe (Wang and Liang 2009).

More on DLR

Furthermore, using spectrometry Evans and Puckrin (2006) measured the wavelengths at which LW radiation is returning to the earth and found the forcing of the greenhouse effect from CO2 and other trace gases is very close to the forcing as predicted by climate models.This is corroborated by Philipona et al. (2004) who analysed atmospheric longwave downward radiation over the central Alps and found statistically significant radiative forcing due to an enhanced greenhouse effect.

Other signatures

Other model predictions specific to greenhouse forced warming include that nights would warm faster than days (Alexander et al. 2006, Fan et al. 2010)

...both the ionosphere and upper atmosphere would cool (Laštovička et al. 2006)

Do you want full references?

Alexander, L.V., X. Zhang, T.C. Peterson, J. Caesar, B. Gleason, A.M.G. Klein Tank, M. Haylock, D. Collins, B. Trewin, F. Rahimzadeh, A. Tagipour, P. Ambenje, K. Rupa Kumar, J. Revadekar and G. Griffiths, Global observed changes in daily climate extremes of temperature and precipitation J. Geophys. Res., 2006, 111

Chen, C., Harries, J., Brindley, H., & Ringer, M. (2007). Spectral signatures of climate change in the Earth’s infrared spectrum between 1970 and 2006. Retrieved October 13, 2009, from European Organisation for the Exploitation of Meteorological Satellites (EUMETSAT) Web site: http://www.eumetsat.eu/Home/Main/Publications/Conference_and_Workshop_Proceedings/groups/cps/documents/document/pdf_conf_p50_s9_01_harries_v.pdf Talk given to the 15th American Meteorological Society (AMS) Satellite Meteorology and Oceanography Conference, Amsterdam, Sept 2007

Evans W.F.J., Puckrin E., (2006). Measurements of the Radiative Surface Forcing of Climate. P1.7, AMS 18th Conference on Climate Variability and Change.

Fan Z, Brauning A, Thomas A, Lid J, Caoa K, (2010). Spatial and temporal temperature trends on the Yunnan Plateau (Southwest China) during 1961–2004. Int. J. Climatol.

Griggs, J.A., Harries, J. E., (2004). Comparison of spectrally resolved outgoing longwave data between 1970 and present. Proc. SPIE, Vol. 5543, 164.

Harries, J. E., Brindley, H. E., Sagoo, P. J. and Bantges, R. J. (2001). Increases in greenhouse forcing inferred from the outgoing longwave radiation spectra of the Earth in 1970 and 1997, Nature, 410, 355– 357.

Laštovička, J., Akmaev, R. A., Beig, G., Bremer, J., and Emmert, J. T. (2006). Atmosphere: Global change in the upper atmosphere. Science, 314(5803):1253-1254.

Philipona, R., B. Dürr, C. Marty, A. Ohmura, and M. Wild (2004). Radiative forcing -measured at Earth’s surface - corroborate the increasing greenhouse effect. Geophys. Res. Lett., 31, L03202.

Wang, K., Liang, S., (2009). Global atmospheric downward longwave radiation over land surface under all-sky conditions from 1973 to 2008. Journal of Geophysical Research, 114 (D19).

Folks, I think this article is being read the wrong way. And I think it's a mistake to attack this paper. This is not a paper on science, nor of disinformation: It is a paper on philosophy. And if we handle it right, we can use it to bolster the case that AGW is real.

I don't know about the journal in which this is being published, but Kampen's article itself is not really an evaluation of the evidence for AGW, but a discussion on the philosophical structure of the scientific case for AGW. His question is not, Is AGW happening?, but rather, How can we establish a framework in which the case can be made cleanly?

He brings up certain difficulties, which are well known, but he poses them as problems of scientific methodology:

- Since we don't have multiple identical Earths, we cannot directly explore the results of different CO2 production rates on climate. How then do we convince ourselves that we have a handle on the validity of the causal attributions that we make?

- How do we come to an understanding about the quality of our data when some scientists, apparently following accepted scientific practices, come to the conclusion that these data are no good?

I believe he understands that, as a practical matter, there is such a thing as a working consensus among scientists, and that all scientists, frankly, are NOT equal. However, his question is, Given that there are two sides to the dispute, how do you know that you are on the side of the angels?

His concluding paragraph is:

"As said, the challenge in corroborating any causal
hypothesis is to determine what kinds of evidence constitute

actual proof of the hypothesis. The field of verification
and falsification of recursive spatial–temporal causality [THAT MEANS COMPLICATED MULTI-CAUSAL MUTUALLY INTERDEPENDENT SYSTEMS WITH FEEDBACK, LIKE CLIMATE] is
underdeveloped and merits future research. Meanwhile,
consensus among scientists will remain to play a great role
in deciding when empirical evidence suffices as proof of a
causal hypothesis. However, in adopting Swanborn's (1996)
“regulative idea of striving after truth by consensus within
the scientific community over research results”, we are always
in danger of replacing the purpose of science (knowledge) with
a by-product of science (consensus in the form of “common
sense”). Karl Popper's requirement of a sound scientific
theory, that it should produce counter-examples that falsify its
validity, serves as a first line of defense against this danger.
Of course, failure to find falsifying evidence in empirical data
will render the AGW hypothesis much stronger."

What he's basically saying is that it's hard to pin down the causal power of CO2 when all of the evidence we have is based on palaeoclimate evidence that entangles all sorts of different physical aspects of the world at the same time: solar intensity variation, variation of orbit parameters, volcanic activity, continent shifts, and so on. So his suggestion is to look for the points that would prove the theory wrong: Identify specific predictions that, if proven wrong, would prove that the hypothesis of AGW is wrong.

Friends, the point is as follows:

- We know that there are a bunch of fakers out there that are just making stuff up. Some of them have Ph.Ds and professorships. Some of them are being paid off, and some of them are just scared. But pointing that out is not an argument that will convince the bulk of the population, because most people are just a paycheck away from being pretty scared themselves. We need an argument that cannot be pointed back at us. ["Scientists get paid too!"]

-  So Kampen's asking, How do you prove you're right? And his suggested solution is, Find clear-cut cases where you FAIL to FALSIFY the theory. This is according to Popperian lines: Karl Popper's dictum was, "You can never prove a scientific theory right; you can only prove a theory wrong. So if you've done your very best to disprove the theory, and you fail, then that's the best anyone can do to support the theory."

- Now, the opportunity in this situation is that there ARE some Popperian test-points: One is the stratospheric cold-spot associated with the greenhouse effect: It's a clear prediction of the theory, and if it were not there, there would be a problem. Fortunately, it IS there.  We need to identify more of these.

- The weak point is that there are also places where the measurements are NOT supporting the theory: I hear there is a tropospheric hot-spot that is missing. This is not devastating, given the known problems with atmospheric measurements. Remember we had problems with the radiosome temperature trends for more than ten years. But such missing evidence has to be placed in proper context. (For example, I had a brief note from Gavin Schmidt about that: He indicated that the uncertainties were large enough that it wasn't clear there was a problem.)

So my proposal is:

a) Congratulate Kampen's insight in identifying how difficult and subtle it is to validate our current scientific understanding of the climate and AGW. (This may take some effort: to interpret his comments on evidentiary problems as issues of scientific epistemology.)

b) Identify and fully describe all the Popperian "failures to falsify".

c) Identify and explain the apparent falsifications (the "failures to fail to falsify").

d) Point out that, although the current scientific understanding of AGW cannot get full marks at the present, essentially ALL the alternative explanations flunk Popper's test spectacularly. SkS is, in fact, devoted to demonstrating that.

In this way, instead of fighting Kampen's article, we use it to explain that, in fact, the current scientific understanding is the best game in town.

The general approach I take whenever I rebut skeptic arguments is to take this mind-set. We have truth and evidence on our side. Therefore, we should use the examination of any skeptic argument as an opportunity, not a problem. By rebutting the skeptic argument, we can actually flip their argument on its head and show that the evidence actually shows the opposite of what they're arguing. It's that Rovian style of jujitsu argumentation - use your opponent's strength against them.

The take-home from Kampen 2010 is that AGW is unfalsifiable. So the response should address that central message. Whether we critique the other flaws in his paper, I don't know, but the central message that Neal proposes is a strong and clear message that would also work for a more general audience. Eg - the simple question - what does AGW predict and do we see it happen?

BTW, the tropospheric hot spot is not a signature of AGW, it's a signature of any type of surface warming. So a failure to find the hot spot would cast doubt on our understanding of the change in lapse rate, not of greenhouse warming. As it is, the lapse rate is confirmed by short-term measurements so the most likely explanation is the lack of hot spot is due to measurement uncertainty.

I would not waste time with the majority of flaws in Kampen:

- from a philosophical point of view, they're not really flaws: He's just pointing out how, due to the complexity of the situation, someone could rationalize it.  He's not saying it's true.

- If he wants to go on about the philosophy of causality, a la Bunge, it doesn't matter: in fact, he already states that it's not very useful for anything much beyond elastic collisions, so why should we disagree with him? One of our points is that climate science is multifactorial and interdependent and recursive. He's supporting that view.

- Anyway, why make an "enemy" out of him when we're going to take his advice on how to build up a strong case for AGW?

What would be really useful would be to have an exhaustive list of the Popperian check-points: b) The strong predictions that would be show-stoppers if they came out wrong. And also, c) the missing points, that don't work out as fully as expected (but, as John points out with respect to the missing hot-spot, are not total killers). Those are the points that Kampen is missing: If we provide those, he will have to say, "Oops, I didn't know that. I guess the case for AGW is better than I had thought!"

Much more satisfying than giving him a "bloody nose."

And if he accepts our examples of Popperian check-points, who is going to have the bloody noses? WUWT and friends. Maybe we should bring kleenex.

The most convenient way to handle Kampen's remarks about the evidence is to slide it into the issue of evidentiary subtlety: It defuses the argument while avoiding a direct challenge. And then we hand him his head with the Popperian check-points. I think he will concede. If he doesn't, he will have to argue the evidence (which he has explicitly avoided doing up until now); and he has already admitted that he's no real judge of the evidence or the data. Then the net result would be merely "article vs. article": nothing special.

But we have a chance that he actually concedes the point: Checkmate. That would be sweet.

I'll be back soon to dig up some references, but I'd just like to mention that an official comment to a scientific article doesn't need to be a "response" to full paper. Comment's usually just point out some error(s) in the article in question. I'm not saying that full response shouldn't be done, I'm just saying that there's nothing wrong in the original Robert's response in that sense. Pointing out errors is not considered "attacking" in scientific media. Of course, it is best if the used phrasing is as polite as possible.

Also, I do think that an official comment is a way to go.

The answer is yes. Eg - I say we submit a comment and also do a blog post. The main detail to hash out is how to time it all. Do we wait to publish a blog post until the comment is accepted? Or do we get a blog post out quick now? I'd be tempted just to get out a quick blog post now as this paper does give us an ideal opportunity to talk evidence which is our strength, not our weakness. The blog post could also be used as the rebuttal to the "AGW is not falsifiable" argument.

As I type this, I lean more towards quick blog post just so we can have a rebuttal - then if this paper gains traction, we have a rebuttal we can link to in comments. If the concern is that a blog post/online rebuttal will undermine or steal the thunder of the comment, the blog post could always address the general argument which predates this paper.

Just thinking out aloud.

Comment publication will probably take quite long, the thunder of this paper will be long gone by then. :) Go public with blog post first if you want to make thunder. Official comment is just for setting the science straight officially.

Co-incidentally, just today this paper was published:

http://www.springerlink.com/content/n31866651q820822/fulltext.pdf

It's an editorial comment, so I'm not sure if it's peer-reviewed, but it might be a possible general reference:

"The essential findings of mainstream climate change science are firm. The world is warming. There are many kinds of evidence: air temperatures, ocean temperatures, melting ice, rising sea levels, and much more. Human activities are the main cause. The warming is not natural. It is not due to the sun, for example. We know this because we can measure the effect of man-made carbon dioxide and it is much stronger than that of changes in the sun, which we also measure."

"The greenhouse effect is well understood. It is as real as gravity. The foundations of the science are more than 150 years old. Carbon dioxide in the atmosphere traps heat. We know carbon dioxide is increasing because we measure it. We know the increase is due to human activities like burning fossil fuels because we can analyze the chemical evidence for that."

RC did a post about falsifiability back in 2006 which makes a similar argument to what we've been discussing:

http://www.realclimate.org/index.php/archives/2006/02/happy-birthday-charles-darwin/

There's nothing new under the sun :-)

Have just emailed Richard Somerville asking if he's ok with me using that quote on my climate scientists page along with his photo :-)

...UPDATE: Just heard back from Richard Somerville who is very happy for that quote to be used, it's been added to the Quote Page.

The problem I have with doing an evidence-specific take-down of Kampen's paper is that it focuses on the wrong thing: the doubts about AGW. His point, as I understand it, is that you can use Popperian check-points to establish your case even in the presence of doubts. Arguing the credibility of the specific doubts themselves is equivalent to telling people that they are not allowed to have doubts.

Analogy: You're on a mountain top surrounded by wolves, trying to attack you. But you have a helicopter. Using the Popperian check-points is like taking the helicopter out of there; arguing about credibility of the evidence is like getting a rifle and shooting the wolves.

In addition to the fuss & bother of shooting the wolves, you get heat from conservation groups. 

More Popperian checkpoints?.

- Changing radioisotope ratio of atmospheric CO2

- Accompanying decline in atmospheric oxygen matching estimates of fossil fuel combustion.  

I think it makes sense that since the article is centered around the statement that AGW isn't falsifiable, our reponse should be focused on giving examples to the contrary.  The human fingerprint in global warming rebuttal should be quite useful in that respect - plenty of examples and references we can use.  I think diurnal temperature range, cooling upper atmosphere (not just stratosphere), tropopause height, and changes in IR radiation are particularly good examples.  All signatures of anthropogenic warming which are empirically observed.

I also think we should try for both the comment and blog post.  If the journal will print the comment along with the article, then maybe hold off on the blog post.  But if there's a delay in the comment, just go with the blog post first.

I think the paper is essentially philosophical in intent, and so it has to be dealt with philosophically.

That doesn't mean that it will lack science: The Popperian check-points are all scientific points.

If we can get the comment published, that should really take precedence over the blog post: I think many journals will not publish something that has already been posted. Why should they?

dana1981, what would be terribly useful as a starting point: A table of check-points:

"Do or die" phenomenon;   Result ;   Reference

We have to think about what to include: dana's examples have to do with the phenomena and characteristics of heating, whereas Rob Painting's have to do with the build-up of CO2 from fossil fuels. These are different parts of the story: We may need a few sub-stories to build up the big picture, each with its own check-points.

I will take a stab at the philosophy...

I think the general consensus here is that we have a go at submitting the comment first before doing a blog post. If we can get the comment published with the article, that would be ideal. In that case, we release a blog post to coincide with publishing. If our comment isn't accepted, then we just adapt the comment into a blog post and probably just publish it straight away.

I'm not convinced we need to include the build-up of CO2 from fossil fuels - that is not disputed by anyone but extremists on blogs so I'd argue it's not required in a comment in a scientific journal. So I think the focus as far as Popperian check points go should be on evidence that the increased greenhouse effect is causing global warming.

Re a table of checkpoints, I came upon a fantastic resource last night by Barton Paul Levenson, Are the Models Untestable. The focus is on the success of model predictions but it equally applies to theoretical expectations of greenhouse theory. He has a table that lists a whole bunch of model predictions, papers where the prediction was first made and papers where the predictions were confirmed. He talks about general model results, not just greenhouse signatures. I've reproduced the table below including just the greenhouse signatures.

Hansen et al. (2005):

http://pubs.giss.nasa.gov/cgi-bin/abstract.cgi?id=ha00110y

"Our climate model, driven mainly by increasing human-made greenhouse gases and aerosols among other forcings, calculates that Earth is now absorbing 0.85±0.15 W/m² more energy from the Sun than it is emitting to space. This imbalance is confirmed by precise measurements of increasing ocean heat content over the past 10 years."

The PhD thing is a non-issue. Papers and comments are evaluated by their content, not by the credentials of the authors. There perhaps might be some odd journals where you only can submit papers if you are a member of some recognized scientific organisation but those kind of journals are minority I think. But if the submission can be done, after that it's just about the content, nothing else.

Edited to add: it's ok to concentrate on the falsifiability thing, but if there are clear errors in the paper, those should also be mentioned (like the straw-man about Greenland ice core).

One more point:

Kampen said: “Most of the observed increase in global average temperatures since the mid-twentieth century is very likely due to the observed increase in anthropogenic green house gas concentrations, and it is likely that there has been significant anthropogenic warming over the past 50 years averaged over each continent except Antarctica” (IPCC 2007: p. 5). In this proposition, which was part of the summary for policy makers of the famous IPCC report on climate change which has influenced politicians throughout the world, the phrase “very likely due to” needs clarification.

It has been clarified in the IPCC report:

http://www.ipcc.ch/publications_and_data/ar4/wg1/en/tssts-2.html#box-ts-1

"The standard terms used in this report to define the likelihood of an outcome or result where this can be estimated probabilistically are: ... very likely > 90 % probability ...".

My summary of this paper: All hat and no cattle.

I have gone through this whole thing, and re-present so it can be appropriately responded to, in a philosophical context. Here goes:

/////////////////////////////////////////////////////////////////////////////////////////////////////////////

Introduction:

1)   The IPCC’s statement is a causal statement: “Increase in anthropogenic GHGs has probably caused increase in GAT.”

2)   Bunge’s framework for causality is clear-cut but rigid.

3)   I don’t quite get how he’s bridging.

4)   American pragmatism: An idea is true if it helps you navigate the world.

5)   Since AGW has been useful in navigating the phenomena of climate science, from a pragmatic view, it can be considered true: the causal connection between anthropogenic GHGs and global warming is established. (QED).

6)   But from the beginnings of science, the concept of causality has always been a bit elusive.

7)   The earliest criterion for the validity of a scientific theory was that it gives the right answers to empirical data (the results of experiments and observations). But Hume’s critique of causality (“A causes B means B always follows A”) gradually convinced scientists that something conceptually more satisfying was needed.

8)   A specific problem with the theory of causality of that time: You could only judge on the basis of what happened. But Popper’s insight was that you can also learn from what didn’t happen: If a theory says X should happen, but it doesn’t, the theory is wrong. So Popper’s advice was: Try to prove your theory (“A causes B”) wrong, in as many ways as possible. If you can’t do it, the theory is good! (Or anyway, as good as it gets.)

9)   Kampen recommends, “In the case of anthropogenic global warming (AGW), the theory should list one or more counter-examples that could (potentially) disconfirm the theory. This listing of potential disqualifiers appears to be missing in the present debate on AGW.” This is where Kampen has made a false step, and this is where we start to correct him.

10) To paraphrase Kampen: “Some skeptics point out that all natural climate disasters are asserted as evidence of AGW, whereas evidence for a lack of warming in the post-WWII period (as indicated in the Greenland Ice-Core Bore Record) is not taken onboard as a falsification.” He goes on to say that cherry-picking the evidence is not sound; a statement with which that no one can object. But his preceding claims have to be countered: A) If you read the newspapers, you will see that every time there is a hot spell or flood, a prominent climate specialist states that this cannot be taken for “proof” of AGW, because one swallow doesn’t make a summer. This has been pretty consistent. So that dog won’t hunt. B) What is the issue with the post-WWII record of the Greenland Ice-Core Bore Record? We need to clarify what the record is, and what has been said by the relevant scientists. [Robert says that this is just one location.]

11)  We need to remind Kampen that the skeptics are erring by attributing to climate scientists the overblown interpretations of people who are inconvenienced by extreme weather. (We could, of course, point out that every time there is a cold spell, Fox News claims that AGW is laughable – but I think it would drag down our argument: We are talking about how scientists should behave, not about how media organizations do behave.)

12) Kampen states: “Unfortunately, when a theoretical phenomenon such as AGW becomes a global political program, it becomes vulnerable to methodological fallacies… the IPCC report aimed at reaching a consensus.” Here he is making an error: He is confusing the social consensus which the IPCC was supposed to foster with the intellectual consensus upon which the IPCC reports are based. There is an essential distinction: In order for humanity to focus on an appropriate response to the threats of climate change, there must be a consensus in international society: This is not a scientific issue, but a social issue, and this consensus can be built by publishing reports and fostering discussion of them worldwide: This is a legitimate role for the IPCC. However, what is informing the activity of the IPCC is the recognition of the intellectual consensus among climate scientists as a community: This is not something the IPCC can do. The IPCC only reports on the previous four years of peer-reviewed scientific research. Among climate scientists, it has no consensus-building power, because the climate scientists are already familiar with the research that went into the report. The IPCC does not fund research, so it cannot even direct study along any particular lines. Therefore, among scientists contributing to the understanding of the climate, the IPCC has no consensus-building capability: Its output can only reflect what intellectual consensus already exists among them.

13)  (It is, by the way, obvious that a social consensus is needed: If the AGW issue pertained to a remote planet, it would be a matter of merely academic interest as to the question of causality, and no social consensus would be needed. Inasmuch as the planet in question is our own, and drastic changes to it will affect us and our descendants, it is necessary to arrive at a social consensus as to what to do, if anything. In this setting, a failure to arrive at a social consensus about action is, by default, a social consensus to do nothing. Does anyone think we know enough to know that doing nothing is the best strategy?)

14) He goes on to assert that attainment of an (intellectual) consensus is not a defining feature of scientific truth; with the usual invocations of such scientific heroes as Galileo and Darwin; the argument that the majority is not always right; as well as the usual chestnut that scientists “won’t bite the hand that feeds them.” Much could be said about why it makes more sense to go with the overwhelming majority rather than with outliers, but if someone takes this stance, there’s no way you can force him to be reasonable. This cynical comparison of the scientific community to a pet dog is the low point of Kampen’s discussion. (But see the concluding remarks!)

15) He then goes on to point out that the IPCC explicitly relies on evidence to support their conclusions about the causal statements on AGW. Hence the question of the quality of the evidence comes up. He cites various authors (you know who they are) who have cast doubt on the quality of the data; but he does not invest himself in any judgment, he is just noting that there is disagreement. He admits to being in no position to evaluate these data himself.

16) So the entire point of this article is to examine the question, How does the scientific community establish a causal connection between increasing anthropogenic GHGs and global warming, if there is objection to trusting the data or to relying on intellectual consensus?

On the establishing of causality in time series:

- The consequence of the lack of experiment:

1)   Normally, to establish a causal relationship, you need to do an experiment: You’d like to set up a few identical Earths and see how they respond to the presence or lack of added CO2. Since this isn’t possible, we can only do longitudinal studies: Look for correlations over time for the Earth that we have.

2)   These are common in social sciences: If factor X is thought to cause factor Y, then you would expect an increase in X to be correlated with an increase in Y. From the correlation alone, you can’t tell whether X is causing Y, or Y is causing X. And there can be correlations that have nothing to do with actual causality, as well: spurious correlations. So if there is a causal relationship, there must be a correlation; but there can be correlations without causal relationships.

- Granger causality:

1)   Granger causality applies to time series: the time series of variable X “Granger causes” the time series of variable Y if knowing the history of X and Y allows a better prediction of the time series Y than by knowing only the history of Y.

2)   Granger causality has been applied to the question of AGW, but you still run into the question of spurious correlations (variables that may be missing from the model).

3)   Using the Granger causality framework, you can conclude that X causes Y; Y causes X; neither causes the other; or they each cause the other (causal recursion: a feedback loop). Feedback loops cannot be sorted out from just the correlations.

- Qualitative model validation and falsification:

1)   To deal with feedback loops, Kampen suggests that the precursor (cause) time series and the result time series can be sorted out by looking at the time series for the time derivatives. If X causes Y, then the derivatives of X should Granger-cause the derivatives of Y; and maxima/minima of X should precede maxima/minima of Y.

2)   For climate study, if we want to relate anthropogenic CO2 to AGW, we need to separate out anthropogenic CO2 from CO2 from other sources: volcanoes, out-gassing from the oceans, etc. Kampen states: “An ‘aggregate’ causal variable that combines radiative forcing of greenhouse gasses, anthropogenic sulfur emissions, and radiative forcing of solar irradiance cannot provide ‘direct evidence that, since 1870, human activity is largely responsible for the increase in global surface temperature’.” I guess what he’s saying is that if you don’t have the individual time series for the different types of factors, it’s bogus to define an aggregate time series.

3)   Going back to the succession of maxima/minima of X and Y, Kampen says: “Of course, evidence that the simple non-recursive causal chain is valid (conceptual hypothesis) mounts with increasing numbers of maxima and minima that follow the regularity constraint (operational hypothesis). However, the precise point when evidence changes into proof remains to be a subjective decision based on consensus among scientists.” [emphasis added]. So after all these mathematical mumbo-jumbo, it boils down to “a subjective decision based on consensus (what?!?) among scientists.”

Concluding remarks:

1)   Since we have only one Planet Earth, we’re stuck doing correlational studies.

2)   To establish a causal connection between CO2 and global warming, we should focus on doing falsification studies (in the Popperian sense).

3)   Kampen suggests: Maybe we could look at maxima/minima of anthropogenic CO2 and see if maxima/minima of global temperature follow (with a time lag). He is out of his gourd: The effect of global warming depends on the cumulative emission of CO2, so to wait for a maximum/minimum cycle will take a thousand years.

4)   “As said, the challenge in corroborating any causal hypothesis is to determine what kinds of evidence constitute actual proof of the hypothesis.” Hey, no sh*t. His answer? “More research is needed.” For this I read your paper?

5)   “Meanwhile, consensus among scientists will remain to play a great role in deciding when empirical evidence suffices as proof of a causal hypothesis.“ Didn’t we decide earlier that consensus didn’t count?

6)   “However, in adopting Swanborn's (1996) ‘regulative idea of striving after truth by consensus within the scientific community over research results’, we are always in danger of replacing the purpose of science (knowledge) with a by-product of science (consensus in the form of ‘common sense’).” So what else is new?

7)   “Karl Popper's requirement of a sound scientific theory, that it should produce counter-examples that falsify its validity, serves as a first line of defense against this danger.” OK: We have a whole list of them.

8)   “Of course, failure to find falsifying evidence in empirical data will render the AGW hypothesis much stronger.”  You don’t say? That’s deep, deep. Really.

/////////////////////////////////////////////////////////////////////////////////////////////////////////////

Well, after all that reading, the result is underwhelming.  I think the most important points to make in comment/rebuttal are:

- We are always going to be relying on an intellectual consensus among participating scientists: That's how science works, as you admit in your conclusion.

- Among active scientists with knowledge in the field, the intellectual consensus is pretty good already. We have a problem with the (in)active (non)scientists with no knowledge in the field.

- Your proposal to look for correlation between maxima/minima of anthropic CO2 concentration and of global average temperature is very very interesting, and we will get around to it in 1000 years or so. 

- Popperian counter-examples are of course useful, and we are providing a list of them.

dana1981,

That's kind of what I was thinking about: Start off with a general response to the article, trying to find points we agree about. Then point out where he's missed the boat in some cases. Re-affirm the importance of the Popperian counter-example, and give a detailed list. (How much detail depends on how much space we have; and the explanation should not be telegraphic, but comprehensible: It should be clear why each point relates to proving AGW: How do we know the CO2 is ours, how do we know that it's working, etc.. But anyway, a table of entries; the actual references can be listed at the end.) I am not sure how the article should end: We can't be too nasty, it's got to be more or less civilized. I like to end on a high note, with some kind of zinger: Probably on the point that we always end up relying, ultimately, on the judgment and consensus of scientists who are closest to the matter at hand; and they are the people with the most personal stake in getting the answer right. No amount of procedures and mathematical formalism is ever going to replace that.

I think the watch-spring of this article is the proposal to look for a time-delayed correlation between the maxima/minima of CO2 and the maxima/minima of global temperature. I think that's the only thing the author really cared about: The rest of it is stuff he wrote to get the argument "up to speed".

Because of the timescales involved, it's a dumb idea, because it takes a thousand years for CO2 levels to reduce. 

We can still write the comment in the way that makes the points we want to make, and that make a "good story." But Kampen probably won't care once we pop his balloon.

On the Greenland thing, I gathered some references:

http://agwobserver.wordpress.com/2010/10/29/papers-on-greenland-temperatures/

Chylek & Lohmann (2005) provide interesting analysis:

"Using double correlations between the Greenland temperature records, North Atlantic Oscillation (NAO) index and global temperature change we find a region of Greenland that is not affected by the NAO. Using this region as an indicator of Greenland’s temperature change that is related to global warming, we find that the ratio of the Greenland to global temperature change due to global warming is 2.2 in broad agreement with GCM predictions."

Here are some studies on CO2-temperature correlation (did Kampen cite any of these studies?):

http://agwobserver.wordpress.com/2010/02/18/papers-on-co2-temperature-correlation/

Ari,

- Yes, Kampen cites Triacca (2001), (2005) on the Granger causality concept applied to CO2 & temperature. He basically seems to be proposing that Granger causality can be applied to time-derivatives of the CO2 and temperature time series. I think that the concept of a time-derivative  of a highly noisy signal is fraught with difficulties, and looking for correlations between two such could be a fool's errand. Have you seen anything like that in your collection of papers? He refers to de Boor (1978), A Practical Guide to Splines, as a way to calculate derivatives. I'm suspicious:

"Notably, if, in a given time span, maxima (minima) of
X are not consistently followed (let alone preceded) by
maxima (minima) of Y, then the hypothesis that X is a
precursor for Y may be considered falsified. This, in
turn, requires a time span of measurements where
changes of the gradients must be detected for both X
and Y. Furthermore, the time series of derivatives must
be computed independently for X and Y (artifacts are
produced if they are derived from differentiation of the
autoregressive models used for establishing G-causality!),
for instance, by the use of splines (see, e.g., de Boor
1978)."

 - He has also cited Dahl-Jansen et al. (1998) from your list of Greenland papers (also Feldman & Marks (2009), Global Warming and Other Bollocks. Gee, I wonder what their attitude is?). But his interpretation (or his interpretation of the skeptics' interpretation - take your pick) is that this Greenland record is somehow "ignored as falsifying evidence." If Robert's summary of the situation (the issue is about the non-warming at ONE location) is essentially correct, it's pretty easy to argue that a local temperature history cannot be extrapolated to a global temperature history, so this issue cannot be considered a Popperian counter-example; if Chylek & Lohmann (2005) shows warming nearby in the same time frame, that underscores the point.

If you want to see what other papers he references, Kampen's paper can be found at:
http://www.springerlink.com.proxy.bib.uottawa.ca/content/n238rl53w2571262/

14) He goes on to assert that attainment of an (intellectual) consensus is not a defining feature of scientific truth; with the usual invocations of such scientific heroes as Galileo and Darwin; the argument that the majority is not always right; as well as the usual chestnut that scientists “won’t bite the hand that feeds them.” Much could be said about why it makes more sense to go with the overwhelming majority rather than with outliers, but if someone takes this stance, there’s no way you can force him to be reasonable. This cynical comparison of the scientific community to a pet dog is the low point of Kampen’s discussion. (But see the concluding remarks!)

Neal, rather excellent review of the paper. Thanks.

The Galileo reference always gets under my skin, probably due to my interest in Astronomy as a kid, and thinking of Galileo as a somewhat heroic figure. The vehement opposition to the Copernican heliocentric model was chiefly due to the religious views of the church (who preferred mankind at the center of the universe). Much like today's opposition to AGW science in other words, ideology based.

No, when skeptics trot that one out they are assigning themselves the wrong role in that historical drama, they're the present day equivalents of the Catholic church, not Galileo. If not for consensus, how did the Copernican system eventually come to be accepted?. 

Seems to me that invoking Galileo infers support for the meritocratic way in which science works.  (Just venting, you no doubt already know all this)

My reading indicates that the Galileo affair was as much a matter of internal Vatican politics as ideology: One particular school felt it "had ownership" of this issue, and didn't want Galileo butting in.

Galileo's basic argument was that questions concerning the configuration of the planets were not really important for the Church, since they really have nothing to do with religious matters; so why get involved in them? Plus, they were susceptible of empirical test (by telescope), so there was some risk of choosing the wrong side of a question you didn't need to be involved in.

Subsequent events proved him right:

- After his trial and recantation, he was home-jailed (too eminent to be tortured, fortunately): Score 1 for the Church

- Astronomers throughout Europe did NOT NOT look through telescopes; and so

- Support for the heliocentric model grew. Score 1 for Galileo

On balance, the Church took a bigger hit than Galileo did. 

Also, when I visited the Museum of the History of Science in Florence, they had on display one relic of Galileo: his finger.

His middle finger. 

Score 2 for Galileo.

Ari,

Can you get one of those papers on CO2/temperature correlation studies? In particular, I am interested in: Triacca (2005), "Is Granger causality analysis appropriate to investigate the relationship between atmospheric concentration of carbon dioxide and global surface air temperature?" I plan to argue that Kampen's proposal is inappropriate for AGW studies, and I would like to see what arguments Triacca used.

Robert,

The link I used last time, 

http://www.springerlink.com.proxy.bib.uottawa.ca/content/n238rl53w2571262/

no longer works: They want $$.

Can anyone get a copy of:

Triacca (2005), "Is Granger causality analysis appropriate to investigate the relationship between atmospheric concentration of carbon dioxide and global surface air temperature?"

Neal, can't help with that but I note a very recent paper out, which may be of interest to you:

http://www.springerlink.com/content/m3571631l5173210/fulltext.pdf

Evan Kodra & Snigdhansu Chatterjee & Auroop R. Gangul

Abstract:

Copy of the Kampen 2010 steaming pile of ............ here:

http://www.springerlink.com/content/n238rl53w2571262/fulltext.html

Neal

give me your email and I'll send you Triacca 2005

Neal: "If you want to see what other papers he references, Kampen's paper can be found at:"

Yes, I did have the full paper but while I was writing my last post, I didn't have it open and was too lazy to start digging it up for that sidenote. :) As Rob mentioned, the paper can be accessed from the original Springer page, your link is to some sort of proxy server:

http://www.springerlink.com/content/n238rl53w2571262/

Neal: "Can you get one of those papers on CO2/temperature correlation studies? In particular, I am interested in: Triacca (2005),..."

I only have full texts to those in my list that already have the full text link. However, late yesterday evening I saw your post and quickly took a peek to Google Scholar using search words granger causality climate and quite a lot of papers surfaced. I try to see today if I have time to compile a meaningful list from them.

Riccardo:

It's nealjking[at]gmail.com

Triacca (2001) would probably be useful, too.

Thanks!

dana1981:

Yes, but for some reason, we decided that this was not the killer point in the argument. Since the full graph post was not posted, I never got a sense for what it would look like: I just got the general impression that the final graph "wasn't very conclusive" about dramatic warming. But now we're asking a different question: "Does this history show a Greenland location not-warming?" Could you follow up on this point? I can't even identify the thread anymore, although I was involved in the argument about not building up to a joke with no punch line.

Ari:

Thanks, that could be useful.

I searched the causality papers and I'm really glad I did - there's some real gems in there. Some papers in the list have already been mentioned here:

Studies on Granger causality as a tool for climate research:

On the use of Granger causality to investigate the human influence on climate - Triacca (2001)

http://www.springerlink.com/content/57r2j2h8pn5dvcwp/

Is Granger causality analysis appropriate to investigate the relationship between atmospheric concentration of carbon dioxide and global surface air temperature? - Triacca (2005)

http://www.springerlink.com/content/grfehq8hmd4jd7nl/

Exploring Granger causality between global average observed time series of carbon dioxide and temperature - Kodra et al. (2010)

http://www.springerlink.com/content/m3571631l5173210/

Assessing causality from multivariate time series - Verdes (2005)

http://pre.aps.org/abstract/PRE/v72/i2/e026222

From Granger causality to long-term causality: Application to climatic data - Smirnov & Mokhov (2009)

http://www.nonlinmod.sgu.ru/doc/doc2009_3.pdf

Other examples of studies using Granger causality in climate studies:

Global Warming and Global Dioxide Emission: An Empirical Study - Sun & Wong (1996)

http://dx.doi.org/10.1006/jema.1996.0025

Econometric analysis of global climate change - Stern & Kaufmann (1999)

http://dx.doi.org/10.1016/S1364-8152(98)00094-2

Granger Causality of Coupled Climate Processes: Ocean Feedback on the North Atlantic Oscillation - Mosedale et al. (2006) (full text freely available in abstract page)

http://journals.ametsoc.org/doi/abs/10.1175/JCLI3653.1

Granger causality and Atlantic hurricanes - Elsner (2007)

http://onlinelibrary.wiley.com/doi/10.1111/j.1600-0870.2007.00244.x/full

Ok, I found the thread where we were arguing about the meaning of the Greenland measurements:

http://www.skepticalscience.com/thread.php?t=320&r=23

Relevant question:

By themselves, do the Greenland measurements indicate a non-warming during the 20th-21st centuries? Even locally?

Dana: "By the way I keep forgetting to bring this up.  Re: Kampen's claim that climate scientists are ignoring post-WWII temps from Greenland ice core data, didn't we decide that the Greenland ice core data only extends to about 1950?"

We were discussing about GISP-2 and Alley's temperature reconstruction specifically which we eventually found out reaches only to 1855. However, now in question is Dahl-Jensen et al. (1998) who use GRIP ice core drilling hole and another similar hole to measure the temperatures from the holes in order to do so called borehole temperature reconstruction.

http://courses.washington.edu/pcc589/2009/readings/Dahl_Jensen.pdf

Well, yes, but you should check out the abstracts here:

http://agwobserver.wordpress.com/2010/10/29/papers-on-greenland-temperatures/

The Greenland not showing much warming except in recent times (Hall et al. 2008, Chylek et al. 2006) seems to be a real thing (Chylek et al. 2004, Hanna & Cappelen 2003, Box 2002). However, there are things that are disturbing the situation (like there usually is when we are dealing with local temperatures instead of global temperatures). Somewhat counter-intuitively, Greenland seems to be affected strongly by volcanic forcing (Box et al. 2009, Box 2002). Box et al. describe the situation further:

"In contrast to the 1920s warming, the 1994–2007 warming has not surpassed the Northern Hemisphere anomaly. An additional 1.0°–1.5°C of annual mean warming would be needed for Greenland to be in phase with the Northern Hemispheric pattern. Thus, it is expected that the ice sheet melt rates and mass deficit will continue to grow in the early twenty-first century as Greenland’s climate catches up with the Northern Hemisphere warming trend and the Arctic climate warms according to global climate model predictions."

Another strongly disturbing factor is NAO (Chylek et al. 2004, Hanna & Cappelen 2003, Box 2002), but when it's effect is minimized Greenland seems to warm according to expectations (Chylek & Lohmann, 2005):

"Using double correlations between the Greenland temperature records, North Atlantic Oscillation (NAO) index and global temperature change we find a region of Greenland that is not affected by the NAO. Using this region as an indicator of Greenland’s temperature change that is related to global warming, we find that the ratio of the Greenland to global temperature change due to global warming is 2.2 in broad agreement with GCM predictions."

Additionally, troposphere over Greenland seems to have been warming quite normally (Box & Cohen, 2006).

One thing is also obvious from these studies. Greenland temperature evolution has not been ignored as Kampen claims.

Ari, thanks for compiling the references; and Riccardo, thanks for providing the chance to look at some of them.

It seems that the Granger causality concept is a way to detect a causal influence/relationship between two time series without having to know anything about what is actually going on between them. It seems to work better in the negative direction than in the positive, because it's basically an extension of the concept of correlation: If there is no Granger-causal (GC) influence of the X series on the Y series, then there is (probably) no causal influence of X on Y. 

But you need the "probably" because there could be a correlation that involves variables like Z, that were not taken into account during the calculation, and which were also changing with time. And there could be time delays; and in principle, the influence could act through time derivatives, etc.

Having looked or glanced at the papers or available abstracts of these GC papers, I have the following impressions:

- Sun & Wong (1996): They report a GC link from CO2 emission to global surface temperature. I haven't read the full paper, but I find it odd that they link the emission rate to temperature: I would expect a link from the CO2 concentration (or the cumulative sum of the emissions) to rate of increase of temperature.

- Stern & Kaufman (1998): They report a GC linkage from greenhouse gases to northern and southern hemispheric surface temperatures; and also detect a difference between them, which they attribute to sulfur emissions. They infer a causal link.

- Triacca (2001): He argues that the conclusion of Stern & Kaufman (1998) is too strong: Due to the subtleties of the correlations, it is not possible to draw the firm conclusion from the established GC links that there is a causal link: Other mathematically conceivable possibilities have not been ruled out. (He points out this his paper was rejected by Nature. Slightly miffed?)

- Triacca (2005): He runs the GC test for CO2 concentration and global surface temperature for the period AD 1860 - 2000, and for a range of sub-periods (all of which however begin at 1860).  He fails to establish a GC link. He does not reject the physical causality (he accepts the physics), but concludes that for some reason GC is not the right tool for studying this issue, possibly because the annual AGW signal should be about 0.005-C and the year-to-year variation is about 0.25-C, a factor of 50 greater.

- Mosedale et al. (2006): They report a GC linkage from daily measurements of sea surface temperatures (SST) and the north Atlantic oscillation (NAO).They regard the linkage as a feedback.

- Elsner (2007): He shows that global mean near-surface air temperature has a GC link to SST, but not the reverse: So the action on the sea is driven by the air temperature, but not the reverse.

- Kodra et al. (2010): They extend the GC methodology by taking into account lags, and differences between the sequential values (in effect, time derivatives). They conclude that it IS possible to establish GC link between CO2 and temperature, as generalized; however, they conclude that something else in the environment changed in the 1970s, causing an alternation in the correlations. What I find odder is that the best relationship they find is between the RC-2 and the GT-2: essentially, the derivatives of the radiative forcing and  temperature. I would have expected a more direct connection between RC-1 and GT-2 (radiative forcing and rate of temperature increase).

- Smirnov & Mokhov (2009): They extend the concept of GC to incorporate long-term causation, which they distinguish from short-term causation. I have not really read this paper: It looks as though it would take awhile to digest. However, they apply this extended concept to the issue of climate change, and conclude that the increase in CO2 concentration has caused at least 75% of the global surface-temperature rise in the years 1985 - 2005, while solar activity and volcanic activity have not led to significant trends. I wonder if other people think it's possible to pull out short-term from long-term trending in this way?

So how has Kampen responded to all this literature? Basically, he incorporates the idea of lags and time derivatives of the time series. Specifically, he suggests looking for maxima & minima: if X has GC influence over Y, then max/min in Y should follow max/min in X. I think this is a pretty goofy idea, because:

a) If we're talking about AGW, we're not going to get another maximum until we stop all CO2 emissions; and we won't get a perceptible minimum for about a thousand years after that.

b) If we're talking about CO2 generally, we can consult the palaeohistory; but that is burdened with many unknowns, which, to sort out, will just throw us back into the question of, How do we deal with proxies?

So I don't see that Kampen's proposal is really useful.

Ari & dana1981:

Thanks for the discussion.

wrt the impact on Kampen's paper: I think the main point is simply that a history of temperature at one location cannot be expected to match the history of the global average temperature, even if it happens to be located in Greenland. Therefore, its failure to match can hardly be viewed as a Popperian counter-example ("do or die"). End of story.

Yes, I think there are 2 points in response to Kampen's use of the Greenland temps as an example of anti-AGW evidence being ignored.

1. It's only one location, thus not representative of global temperatures, so it's not anti-AGW.
2. As illustrated by Ari, it has not been ignored, but rather the lesser Greenland warming has been studied in great depth.

dana,

Keep in perspective the point of Kampen's paper: He's proposing a methodology. The remark about AGW not addressing Popperian counter-examples is just an excuse for him to get to HIS proposal. We should point out his error; but if we spend a lot of energy countering incidental points, the editor will likely make a judgment that we have missed the basic point of the paper, and either cut back on our comments (perhaps without our input), or look for someone else to make comments that are more "on topic."

I think we have to make clear that we think his proposal is a bit off (and thus worth a comment); and we can toss in at the end that we agree with the motivation of finding counter-examples, even though his example of a missed counter-example is also off-base. Then we give the list of known counter-examples that have failed to falsify the theory.

OK.

I'm going to mull over my "line of attack" and then put out a 0th draft. Need to get some stuff done.

John,

I would like to start over from scratch, because the article has to be re-oriented towards a critique of the central intent of Kampen's article (a methodological/philosophical issue), or else I believe the editor will toss it out.

I would like to sketch it in outline form, with well-defined sections that can be filled in. The section on Popperian falsifications is very important: Even though structurally, it would not be central; something like: "We agree with Kampen that Popperian counter-examples are the way to go. We don't quite agree that the Greenland ice-core history represents a failure to take the falsification program on-board, because [of what dana said]. In fact, there are a range of Popperian challenges that have been posed and met to test the AGW theory, including [... and here we go...]", In actual fact, from our point of view, it would be the pay-off, the vaccine delivered by the injection. But I believe it has to be delivered as a "by the way", not as the main point of the critique, or else the editor will be entitled to conclude that we're just missing the point of Kampen's article.

My strategic concept is that Kampen's methodological center-piece is weak enough that when we critique that, we will also be entitled to critique his incidental errors about the evidence for AGW - which is what we care about.

Give me a day or so to outline something (I really have some personally urgent things to complete!), and then I think there will be places for people to fill in, or ways that the structure can be strengthened. 

Neal: "We should point out his error; but if we spend a lot of energy countering incidental points, the editor will likely make a judgment that we have missed the basic point of the paper, and either cut back on our comments (perhaps without our input), or look for someone else to make comments that are more "on topic.""

No, this wouldn't happen. Comment to a research article is for pointing out errors. It makes no difference if we even wouldn't comment the main point of the paper at all. Finding an error in an article is big deal to editors regardless if it's important to the big picture or not. Actually, it could be argued that we could make a stronger case if we would leave the causality issue uncommented.

One point I would like to highlight is that it seems to me that Kampen hasn't done any literature searches on the issues discussed in the paper (for example on Greenland, falsification or Granger causality). I think this needs to be shown in the comment.

Ari, I can't believe that the journal editor will publish every comment that comes into the office: Otherwise, I would have to conclude that very few people ever comment on a paper. As a practical matter, the editor must make some decision that some comments are not important enough to take up valuable space in the journal.

From that perspective, if someone wrote an article on the mathematics of Fourier analysis and happened to misquote the dates of Fourier's travels in Egypt, I don't think the editor would be favorably impressed by a comment that was focused on extensive correction of the account of Fourier's personal history. If someone else had made a similar comment about the mathematics, but left out the historical discussion, I think he would favor that; if there were no other comment competing for inclusion, he should ask the commenter to focus on the mathematics.

In the same way, Kampen's article is not about the truth or falsity of AGW, it's about the application of a particular statistical technique to this issue. In the course of his discussion, Kampen has made a few remarks about the evidence, just to bolster the case for the usefulness of his idea; but actually the value of his proposal has nothing to do with these remarks at all, they are just "ornamental". Our concern is that these ornamental remarks can be seized upon by skeptics as proof that contributing scientists believe that the theory of AGW is not being handled in the usual scientific manner; so we want to comment to make sure these points do not go unchallenged. However, to be honest, we are, to an extent, hijacking the article. I believe the other workers in the field (Triacca, Kodra, Mosedale and so on) would likely comment directly on the technique (if they would bother; because I think the technique has problems), but I don't believe they would comment on the issues concerning the AGW evidence, or the Popperian issues, at all.

Implicitly, we are competing for comment-space against these folks.

Neal: "Ari, I can't believe that the journal editor will publish every comment that comes into the office:"

They don't. The comments are peer-reviewed and the ones that pass the peer-review are published. I have seen comments on papers that haven't had anything to do with the main theme of the paper but have just pointed out some minor errors.

Ari,

So my point is that we are, potentially, in competition with with other people who will not focus on the AGW issues as such, but on the technical aspect, for the good opinion of the reviewers. I want us to look reasonably good in this competition, if possible: I don't want us to be the only duck that's flying west for the winter.

I think we are quite ready to start writing this thing. This thread even has some almost ready writing that could be included. We can write it point by point and at the end look it as a whole.

However, there's one thing we haven't yet addressed - this one:

"The quality of all scientific research depends of course, on the quality of the data that are being processed. Regardless of the quality of the (statistical) model used for analysis, if bad data are fed to the model, then the result of the analysis will be bad. This principle is known as garbage in–garbage out. In other words, if the data that are fed into climate models are open to dispute, then so are the projections of these models. In the scientific (i.e., peer-reviewed) literature, several authors have expressed doubts about the quality of the analyzed data and the possibility to derive at valid inferences on human impact on global warming (e.g., Jaworowski 1994; Soon et al. 2004; Michaels 2008; Pielke et al. 2007). However, since the author of this article is no expert on climate science, the issue of whether or not data used in climate science are of enough quality will be left for others to decide."

Isn't there an error here? If I have understood correctly, the climate models don't use existing data to do projections. Does someone here know this for sure?

Neal

when organizing the layout of the comment, consider to start with a brief introduction on the point that Kampen rises, then underline where you agree and end with what you think is a incorrect application of those ideas. In this last section you can rebut specific points or examples. Try not to give the impression that you consider the paper awfully wrong throughout, chances are you'll get the same referee of the original paper.

Riccardo,

That makes sense.

Ari,

The whole point you raise is a closed loop that leads to no consequence: See his last sentence. Our goal is not to beat his paper into the ground, but to correct some specific points (especially, the idea that AGW has not been exposed to Popperian counter-examples). Riccardo's advice above is relevant.  We don't need to fix every aspect of Kampen's paper: After all, he's not posting in support of SkS.

Ari, it's like Rob Way pointed out earlier, the paper is an excuse to trot numerous zombie skeptic talking points. Seems to be a lot of "but what if.........." 

- Michaels 2008 (one of the "institute" papers) - claims the surface temp record is biased and then tags his version of surface temps onto IPCC climate model projections. Epic fail. No wonder Ben Santer wanted to thump him.

- Pielke 2007. About the surface temp record (Watts photos etc). Relevance to climate models?

- Soon 2004. Surface temp whining again. And WTF is data padding?. He needs a thumping too. Relevance to climate models?

- Jawarowski 1994. - Waffle about ice cores, saying the aren't suitable for reconstructions. Relevance to climate models?.

Looks like he lifted the references from some denier blog, without actually reading them. Not suggesting that the comment address that issue, just pointing it out.

Rob,

My interpretation of Kampen's thought process:

"To promote my great new technique, I have to explain why it's needed. The AGW is a high-profile problem, let's apply it there!

Why is it controversial? People don't believe the results. Why not? They don't believe the data. What do they say about it? I'll check google and find out. Ooh, look at all that stuff at WUWT!

So now my technique can save the day! It'll be a Popperian counter-example: heads I win, tails you lose.

Oh, but don't they have other tests? Ah, but somebody says they're scandalously ignoring a counter-example in Greenland. That'll do."

So I don't think this paper is an excuse to promote bogus zombie anti-AGW arguments. Rather, he's scavenged websites to find bogus zombie anti-AGW arguments as an excuse to promote his technique. He doesn't care about the AGW question: not the results, nor the data, nor the quality of the data. It's just an excuse to write a paper where he can propose his technique, which is a solution looking for a problem.

It's not an assassination attempt, it's a drive-by shooting.

Neal, you made me do a bit of digging, I think you're right about Kampen. Whatever the case though, he looks young enough to need a snorkel in the coming decades - assuming he remains in the Netherlands. 

http://www.rme.wur.nl/UK/Staff/Dr.+Jarl+Kampen/

I think - and I'm no climate model expert - that the GIGO claim hypothetically could have some validity.  After all, climate models are tested through hindcasting (using the supposed "garbage in" data), and I presume various parameters are adjusted to make the models better fit the hindcasted data.

Ultimately climate models are mainly based on fundamental physics, but hypothetically if the instrumental temperature record were garbage, for example, I think it could influence the accuracy of climate model results.

I'm not sure this is a sufficiently important or well thought-out point by Kampen to warrant a response.  We could briefly mention a few studies which have verified the accuracy of the surface temperature record to show that the data isn't "garbage".  I wouldn't put any more effort into it than that, but it might be worthwhile to do that much.  Again, just a sentence or two.

As pointed out above, Kampen sprays a whole bunch of denier talking points in his paper. It would dillute our message to cover all of them. I get this in online debates all the time - you make a good point, then the denier throws out half a dozen talking points - your standard Gish gallop. The worst thing you can do is try to hit them all. Instead, I pick their central message and focus on that, refusing to be taken off topic.

In this case, Kampen has two major points - Granger causality and Popperian falsification (noting that his conclusion focuses on falsification). These should be our focus. I recommend not falling into the trap of having to debunk every single wrong point in his paper. There needs to be a single key point, perhaps "there is ample evidence that humans are causing global warming" then all aspects of our comment drives home that narrative. The multitude of papers on granger causality provide empirical evidence of causation. The numerous predictions of global warming theory that have been vindicated rather than falsified all, as he puts it, "renders the AGW hypothesis much stronger". I think we should use this wording in our comment.

Is there a deadline to get that comment in? I'm champing at the bit to write about falsification!

dana1981,

I agree largely with John: We should not give the impression that we have appointed ourselves as the super-editor, who is going to correct every last error of logic, data, grammar, spelling and typos: this would not only be insulting to the author, but to the reviewer who passed the article, and to the journal. Our public motivation should be to point out errors that would have a detrimental effect on the literature, that would actually mislead people.

Ordinarily, people don't publish papers with so many errors: He's obviously writing in an area where he has no expertise related to the actual content of the data. But for us to treat his paper like a high-school essay basically holds his paper up to ridicule (and reflects that ridicule onto the reviewer): It's not polite and it's not necessary. We can make our point by spelling out the issues that are important; if these points are substantial and well-made, then the reader will understand that our concerns about the paper are reasonable; will come to have some doubt about the trustworthiness of the rest of the article; and will basically ignore the paper thereafter. Mission accomplished. 

In other words: When hunting moose, don't shoot at rabbits: It's a waste of ammunition.

Or are you just being metaphorical?

Sorry, just curious :-)

I agree with Dana. The commenting is supposed to be about pointing out errors. If we see a clear error in the paper, there's no reason for not mentioning it. In fact, I would argue it's our responsibility to point out errors if we see them.

Perhaps we should write two different comment papers? :)

Robert,

In that case, I would think you wouldn't get many moose, except maybe deaf ones.

Ari,

Writing up two different sets of comments would take care of the internal controversy. But I think it would look strange to a reviewer to have overlapping authorship of the comments, wouldn't it? I imagine that two sets of comments would be reviewed by the same reviewer: What would s/he think about getting Comments-1 from authors A, B, C & D, and Comments-2 from A, D, E & F ? Wouldn't it be considered odd?

Here's my contribution on the IPCC's very likely -thing:

Kampen quoted a paragraph from the IPCC fourth assessment report (IPCC AR4, Solomon et al. 2007) commenting that the "very likely due to" in the paragraph needs clarification. Although we do not know what kind of clarification Kampen is after, we would like to point out that IPCC AR4 has very specific meaning for the words "very likely". IPCC AR4 says: "The standard terms used in this report to define the likelihood of an outcome or result where this can be estimated probabilistically are: ... very likely > 90 % probability ..." (IPCC AR4, TS.2, Box TS.1).

[To reference section:]

Solomon, S., D. Qin, M. Manning, Z. Chen, M. Marquis, K.B. Averyt, M. Tignor and H.L. Miller (eds.), Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, 2007, Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA.

(Neal: I don't know if the editor would consider that odd but at any case my suggestion was somewhat tongue in cheek and I think we should continue to build one comment. I suggest that we write all our stuff point by point and then see what we got. Nothing written is wasted because at least every argument we build will find their place in the blog posts to come, if not in the official comment paper. These minor points I and others have suggested don't really distract from the main points of the paper, we'll just put these in the corner of the paper somewhere as few sentence statements per point. However, these minor points do make a larger point; they will show how sloppy work Kampen has made preparing for this paper.)

Here's suggestion for the Greenland claim:

In the introduction section, Kampen writes: "In fact, some skeptics in the debate on AGW point out that all natural climatic disasters are used as evidence (verification) for the human impact on climate, whereas evidence that a post WWII global warming is absent in, e.g., the Greenland Ice-Core Bore Record is ignored as falsifying evidence (see, e.g., Dahl-Jensen et al. 1998; Feldman and Marks 2009)." While it is true that some skeptics in the debate of AGW point out such things, it is elementary to notice that the anthropogenic global warming is a global phenomenon and falsification would generally require more than temperature measurements in a single location, such as Greenland given here as an example. Furthermore, there has been some studies on the temperature evolution of Greenland. It seems to be true that Greenland has shown little warming same time as the globe has been warming (Box 2002, Hanna & Cappelen 2003, Chylek et al. 2004). There are some signs of very recent warming there (Chylek et al. 2006, Hall et al. 2008) and the troposphere over Greenland seems to have warmed (Box & Cohen, 2006). More importantly, there are factors disturbing the global warming signal in Greenland, especially North Atlantic Oscillation (Box 2002, Hanna & Cappelen 2003, Chylek et al. 2004) and volcanic forcing (Box 2002, Box et al. 2009). Interestingly, when an effort is made to minimize the effect of the North Atlantic Oscillation, the Greenland seems to warm in broad agreement with GCM predictions (Chylek & Lohmann, 2005). It seems to us that at least in the scientific literature the question of Greenland temperatures has not been ignored as shown by these few examples presented here.

[To reference section:]

Box, J. E. (2002), Survey of Greenland instrumental temperature records: 1873–2001, Int. J. Climatol., 22(15), 1829–1847, doi:10.1002/joc.852.

Box, J. E., and A. E. Cohen (2006), Upper-air temperatures around Greenland: 1964–2005, Geophys. Res. Lett., 33, L12706, doi:10.1029/2006GL025723.

Box, J. E., L. Yang, D. H. Bromwich, L-S. Bai (2009), Greenland Ice Sheet Surface Air Temperature Variability: 1840–2007. J. Climate, 22, 4029–4049, doi: 10.1175/2009JCLI2816.1.

Chylek, P., J. E. Box and G. Lesins (2004), Global Warming and the Greenland Ice Sheet, Climatic Change, 63(1-2), 201-221, doi:10.1023/B:CLIM.0000018509.74228.03.

Chylek, P., and U. Lohmann (2005), Ratio of the Greenland to global temperature change: comparison of observations and climate modeling results, Geophys. Res. Lett., 32, L14705, doi:10.1029/2005GL023552.

Chylek, P., M. K. Dubey, and G. Lesins (2006), Greenland warming of 1920–1930 and 1995–2005, Geophys. Res. Lett., 33, L11707, doi:10.1029/2006GL026510.

Hall, D. K., R. S. Williams, S. B. Luthcke, N. E. Digirolamo (2008), Greenland ice sheet surface temperature, melt and mass loss: 2000-06, Journal of Glaciology, 54, 184, 81-93(13).

Hanna, E., and J. Cappelen (2003), Recent cooling in coastal southern Greenland and relation with the North Atlantic Oscillation, Geophys. Res. Lett., 30(3), 1132, doi:10.1029/2002GL015797.

Ari,

Yes, I'm on the case; but I am fighting some fires right now. Will get to it this weekend.