Wednesday, February 17, 2016

Making Sense of Bits of the Argument

Many parts of the argument over AGW are sufficiently complicated so that, unless you are a professional in the field, there is no good way of telling who to believe. The result is that most people simply believe whatever story supports what they see as their side. An example would be the question of whether AGW will increase or decrease food supplies.

Once in a while, however, someone offers a contribution to the debate that is obviously dishonest—obviously enough so that a careful reader can spot the trick. The fact that there are dishonest people on one side or the other of the argument does not tell you that that side is wrong. But how their argument is treated by other people on that side does give you some information on who you should or should not trust. Any source of information—newspaper, web page, scientific journal, blog—that takes seriously work which a careful reader can see is dishonest should not be trusted, since it is either dishonest or incompetent.

I have two and a half examples from the CAGW side of the argument, the claim that, unless we take strong measures to slow AGW, the net results will be very bad. One is an article that tries to obscure the positive effect of CO2 fertilization by claiming that a change which increases the yield of every nutrient threatens human nutrition—because it increases some by more than others. One is the source of the much quoted 97% figure, along with a later piece in which the lead author misrepresents the result of his own work. One is a piece in a high profile popular publication by an economist arguing that AGW is a threat that requires immediate action—and reporting results of his own research that imply the opposite, but reporting them in a form that hides the fact. I count that as a half because I am not sure that the deception was deliberate.

What I find depressing is how few people, faced with clear evidence that someone on their side of the argument is dishonest, are willing to accept it.

I invite people to offer similar examples on the other side. They should be publications by respectable figures that meet two conditions:

1. They are obviously dishonest--obvious enough so that the demonstration does not depend on trusting some other source with the opposite bias.

2. They are taken seriously by lots of reputable people on that side of the argument.

32 comments:

Darf Ferrara said...

Richard Tol is discussed by Andrew Gelman. I'm not sure if this counts as obvious, since it gets into the weeds a bit statistically.

Anonymous said...

I confess, I haven't read enough "respectable, reputable" work on either side to give examples off the top of my head. However, there's a logical problem with your challenge.

Suppose, hypothetically, that the pro-X faction had a variety of publications, some more scholarly or honest than others, some by more respectable figures than others, while the anti-X faction had NO publications by respectable figures, and/or NO reputable people to take them seriously. Then it would be impossible for a pro-X-er to meet your challenge, which would seem to strengthen the anti-X argument, not because of the strength of that argument but perversely because of its weakness.

I'm not going to claim that the situation for CAGW is that extreme, but it doesn't need to be. Suppose the pro-X side had 3 times as many "respectable" publications as the anti-X side, and both sides were (on average) equally honest. Then it would be 3 times easier to find dishonest and misleading publications on the pro-X side than on the anti-X side, which could be interpreted to support the weaker argument.

When I tried to find a CAGW example, the first thing I thought of was by Ted Cruz, who (I 'd like to think) doesn't qualify as "respectable" or "taken seriously by reputable people on that side of the argument". In a December, 2015 NPR interview, Cruz claimed that the data don't support the conclusion that global warming is happening at all, much less anthropogenically or catastrophically. To this end, he said "If you look at the last eighteen years of NASA satellite data, you see no statistically significant upward temperature trend at all."

Let's talk about "the last N years of data". If I were unscrupulously trying to persuade people that global warming was happening, I would choose N to start at an unusually cold year, making it more likely that from-then-to-the-present would show a warming trend. Likewise, if I were unscrupulously trying to persuade people that it wasn't happening, I would choose N to start at an unusually warm year. Eighteen years before 2015 was 1997, the warmest year on record up to that time. (1998 was even warmer, so Cruz can continue saying "eighteen years" in 2016 without his statement becoming false.) This doesn't necessarily prove that Cruz was lying with statistics to reach his desired conclusion, but he did exactly what such a person would have done.

The data: Cruz didn't say exactly what data set he used. I did a quick search on "nasa.gov global temperatures" and found this page, which includes a "Download Data" link. According to this data set, 1997 was the warmest year on record up to that time (although sixteen of the eighteen years since have been warmer). I did a linear regression from each of those years to 2014 (since 2015 data weren't available yet when Cruz made his statement): from 1997-2014 shows a slope of 0.0115C, a lower slope than if you start at any of the years 1996, 1995, 1994, ... 1944. In other words, if Cruz had picked 19 years, or 20, or any other number between there and 70, his statement would have been less true than with 18 years. (I really ought to do a significance test on the hypothesis that the slope is positive, but I would have to look up how to do that. Download the data and try it yourself.)

Glen said...

hudebnik, your Cruz quote specifically refers to "NASA satellite data", but the link you gave shows a chart of surface data, which is a very different thing. There are two standard satellite-based series, called RSS and UAH.

If we choose to be charitable, the satellite data set that is most favorable to Cruz's claim is RSS. Including a trendline, it looks like this for the last 18 years.

UAH doesn't work quite as well so Cruz probably meant RSS. I think the series is not literally a NASA data product, but comes from sensors in satellites that were launched by NASA, so: meh. The claim seems plausibly defensible to me.

You can find RSS data from a more official source here.

Anonymous said...

The most popular rejectionist blog, which is taken seriously by rejectionists, is chock-full of dishonesty. It's Tony Watts' site.

There are a number of blogs that regularly, and frequently, point out the many dishonesties there.

Many more than 2 1/2.

Eric Rasmusen said...

If Watt's site is so dishonest, how about pointing out one example instead of just saying it's dishonest?

Anonymous said...

Eric, just about any article at

http://blog.hotwhopper.com/

will suffice.

Art said...

My guess is that Anonymous who linked to blog.hotwhopper.com is actually a trolling rejectionist trying to present his side as more honest than it actually is.
Why else would he link to the source that accuses Watts of quoting a "shallow" article written "the time when all the research wasn't in" and present it here as an example of obvious dishonesty by Watts?

Glen said...

Anon, the blog "hotwhopper" makes no apparent attempt to distinguish between "this is dishonest" and "this is incorrect". Moreover when the claim being made is "this is incorrect", it often boils down to "when I consult MY sources based on MY existing priors, I reach a different conclusion than Watts reaches based on HIS sources and HIS existing priors".

So even if you think hotwhopper documents dishonesty rather than mere disagreement you're probably going to have to point to a specific article there.

Anonymous said...

Just about any post by Bob Tisdale on WUWT is dishonest. Ample documentation at hotwhopper.

Mark Bahner said...

"The (Cruz) claim seems plausibly defensible to me."

Even if Ted Cruz's claim was "plausibly defensible," in terms of 18 years showing no trend, it wouldn't address the criticism that he cherry-picked the starting year. Which he did.

But no politician is a reputable source for climate change information.

Glen said...

Mark: the RSS trend isn't just statistically negligible starting on that particular year, it's also approximately flat starting from nearly every year since 2000 - you can verify this using the link I gave above. So on what basis do you claim "cherry picking" is going on?

Also note that the claim being made back in 1998 was not merely that warming was happening but that the warming trend was *accellerating*. 18+ years of a mostly-flat trend is not really compatible with that view. So there's some mystery here and as far as I know it's still unresolved and still relevant.

Eric Rasmusen said...

THe Gelman-Tol dispute linked to by Mr. Ferrara above doesn't obviously qualify. Or does it? I didn't have time to untangle who was right in that. For others: Economist Tol did a meta-analysis in J. of Economic Perspectives of a dozen or so other papers to find whether a small amount of global warming (something like 1 degree centigrade) was beneficial or hurtful. It looks like the typical mess that meta-analyses are, and Tol got some numbers wrong and admitted it but said they didn't make much difference. It looks inconsequential to me since the Tol paper wasn't likely to get useful results anyway from combining mediocre studies asking a different question than the one he was asking. I could be quite wrong, though, and would welcome anyone correcting me here.
As Glen noted in his comment, Ted Cruz cherry-picking the starting year for hte global warming halt doesn't really matter. You have to pick some year as the start of the temperature plateau. The 1997 peak is okay, or 2000, but if you pick 2002 or something you get the same picture--- temperature levelled off. If you pick 1992, say, you'll get a bit of temperature increase, but you won't get much if you pick 1998 or 1999 instead of 1997.

Jonathan said...

It seems to me fairly obvious that the world's climate changes over time, and that the existence of humans must be having some effects on it. I doubt that anyone knows with any accuracy what effects humans are having on global climate, or can assess the impact of these changes on humans in future.

However, as there are billions of humans and I have no power or influence over them, my beliefs and actions are irrelevant and I don't need to concern myself about the subject. Whatever's going to happen will happen regardless of me. If I think the sea's going to rise, perhaps I should avoid buying a house at sea level. That's about it.

I don't know why some people get excited about things they have no control over.

alaska3636 said...

Jonathan,

It is typical tribalism mixed with political expedience mixed with awful academic incentives. Politicians divide the subject and campaign so as to appear useful; the constituency, who have no influence anyway over a bunch of careerist bull-doody artists, pick sides and engage rigorously in either confirmation bias or apathy while the academics supply whatever it is they are paid for: some of the academics even convincing themselves that their side is purely motivated by truth rather than putting bread on the table.

You are an outlier and the kind that I prefer - you are tentatively aware that forces outside your understanding are shaping things beyond your ability to influence and you act accordingly. The smart-people tribe is small and has large but very unpredictable influence. All of this means something, I just don't know what it is. I try very hard in life and then give it up to the gods to decide. Cheers!

Mark Bahner said...

Glen writes, "Mark: the RSS trend isn't just statistically negligible starting on that particular year, it's also approximately flat starting from nearly every year since 2000 - you can verify this using the link I gave above. So on what basis do you claim "cherry picking" is going on?"

I tried to use your link above, but had difficulties with it. I then found this link:

Temperature trends, Kevin Cowtan

I constructed the following decadal temperature trends data, for the periods ending in January 2015 (an end date of 2015 results in ending in January 2015). The first temperature trend is RSS (Remote Signal Systems) the second is UAH (University of Alabama-Huntsville). Only values that are negative have the sign in front (i.e., positive values do not have a "+" sign in front).

Beginning Year / RSS trend / UAH trend

1995 / 0.028 / 0.124
1996 / 0.021 / 0.122
1997 /-0.010 / 0.099
1998 /-0.043 / 0.077
1999 / 0.026 / 0.147
2000 /-0.003 / 0.117
2001 /-0.050 / 0.066
2002 /-0.062 / 0.050
2003 /-0.045 / 0.023
2004 /-0.015 / 0.089
2005 /-0.037 / 0.062
2006 / 0.031 / 0.130
2007 / 0.054 / 0.170
2008 / 0.134 / 0.265

So then I was going to acknowledge that the RSS values are essentially zero from 2000 onward...and even back to starting in 1995. But I was going to point out that the UAH values are consistently positive in that whole time period.

However THEN I found this post by Robert Tisdale, that indicated that UAH released a "Version 6.0" of their temperature calculation algorithm:

New UAH lower temperature data show no warming for more than 18 years

From that, I see that this UAH version 6.0 is not in the Kevin Cowtan temperature trends site, since that site was last updated in 2014. I further see that the UAH 6.0 version has lower temperatures than their previous version for post-2000:

UAH version 6.0 versus 5.6

Sooooooooo...Ted Cruz would have been essentially correct (essentially zero warming) if he'd chosen *either* RSS or UAH, and if he'd chosen any starting time from roughly 1995 to 2008. (And perhaps even earlier than 1995, but I wasn't ambitious enough to check.) So you're right, even if he'd started in 2000 or thereabouts, he would have been essentially correct.

P.S. I'll point out here that ~10 years ago, I predicted that there was a 50% chance of global warming (as measured by lower tropospheric satellite temperature measurements) of 1.2 degrees Celsius from 1990 to 2100, with less than a 5% chance of warming of less than 0.02 degrees Celsius, and less than 5% chance of warming more than 2.45 degrees Celsius. I still feel pretty good about that prediction:

<a href="http://markbahner.typepad.com/random_thoughts/2006/04/complete_set_of.html>GW predictions in April 2006</a>

Mark Bahner said...

Oops. David, if you slip a closing quotes after the .html in that last link, it will fix it. Sorry for the inconvenience.

Bravin Neff said...

George Mason University professor Edward Wegman and cohorts.

Wegman was the lead author of the so-called "Wegman Report," which was used in the 2006 congressional investigation of the famous MBH "hockey stick" AGW papers. It turns out that...

The Wegman report was largely plagiarized from prominant climate science and Network Analysis works, including a textbook written by the "Bradley" in "MBH". Read that again if your brain just did a double-take: The Wegman Report that was submitted to Congress, and used to criticize MBH, actually plagiarized a textbook written by the B in MBH (among other works).

LOL. You couldn't make that up.

It gets better. The Wegman Report was used as the basis of a paper published in the journal "Computational Statistics and Data Analysis." This paper was retracted and apologized for, following an investigation by a GMU committee.

And then there's GMU. While the one GMU committee reprimanded Wegman, parcel to the retraction of the journal artical, as far as I can tell the greater GMU administration body is still dragging its feet 6 years later on what to do about the original 2006 Wegman Report and the discoveries of plagiarism therein. Here's a write up from the person who discovered the original plagiarism, followed by some commentary.

https://deepclimate.org/2012/02/22/gmu-contradictory-decisions-on-wegman-plagiarism-in-csda-but-not-in-congressional-report/

Mark Bahner said...

"The Wegman report was largely plagiarized from prominant climate science and Network Analysis works, including a textbook written by the "Bradley" in "MBH". Read that again if your brain just did a double-take: The Wegman Report that was submitted to Congress, and used to criticize MBH, actually plagiarized a textbook written by the B in MBH (among other works)."

The blog post from which you quote discusses plagiarism/alleged plagiarism of introductory materials. Obviously, plagiarism a bad thing to do. But whether introductory materials are plagiarized or not says nothing about whether the technical analysis in the paper is correct or incorrect...let alone whether the technical analysis is honest or dishonest. (A technical analysis can easily be incorrect without being dishonest.)

Plagiarizing (quoting without attribution) introductory materials for a paper or report is nothing like what John Cook and his co-authors have done about their 2013 Environmental Research Letters paper, which is blatantly misrepresent (lie) about the results of their paper's analysis, even after it has been pointed out they are misrepresenting (lying about) their results.

Bravin Neff said...

Mark Bahner, you may want to dig deeper into the Wegman scandal than simply relying on the blog post I linked to. The person who did the major work on Wegman is computer scientist John Mashey. I assure you it goes far deeper than "quoting without attribution introductory materials," and by my lights is far worse than what John Cook is alleged to have done.

Follow this link for an attachment that is a cursory analysis by John Mashey (his full analysis is over 170 pages long). It demonstrates far more than simply missing attribution to quotes. Indeed, it goes even further than demonstrating that much of the analysis relied on plagiarism, selective omission, errors that skewed the narrative in the desired direction, bibliography padding, exaggeration, and more than I am willing to go into here.

Bravin Neff said...
This comment has been removed by the author.
Bravin Neff said...

Sorry, here's that link. Attachment by John Mashey at the bottom.

http://www.desmogblog.com/wegman-report-not-just-plagiarism-misrepresentation

Eric Rasmusen said...

The mistakes in the Mann hockey stick paper and in the criticism of it by McIntyre and McKitrick are pretty confusing, especially since the technique of principal components analysis that they're all doing is dubious to begin with. My take is that in the end the Wegman Report *is* a good example of an anti-warming deceptive analysis. See https://deepclimate.org/2010/11/16/replication-and-due-diligence-wegman-style/ .

But there is additional confusion because the "plagiarism" in the Wegman Report is not only irrelevant to the question of whether it is misleading, but doesn't really rise to the level of plagiarism either. The Wegman Report is presented not as original research, but as a report on existing work. Is it wrong to use somebody else's words in such a report? We would penalize a student for it, but that is only because the purpose of the student paper is to make them exert effort and to make sure they understand what they're writing about. In something you wrote up for yourself, or to inform someone else, you wouldn't bother with citation of cut-and-paste excerpts because it's not worth the effort. Nobody cares whether the words are original with you or not; it doesn't matter for the words' credibility or for giving anyone credit since the words are so mundane. So the criticism of the Wegman report for plagiarism seems misguided in two senses-- it is criticism for something not harmful, and it distracts attention from the way the report *is* dishonest.

Bravin Neff said...

Eric,

I agree with your claim the Wegman Report is deceptive analysis based on the work contained in the Deep Climate blog you linked to. And at this point I have read enough discussion and a dozen reconstructions of the “hockey stick” by other researchers using different data sets and techniques, to see that abandoning MBH’s PCA technique appears to keep the hockey stick firmly intact (which, while the MM critique and WR have been abandoned by anybody who cares about their reputation.

But you are downplaying the plagiarism in the WR. The forms of plagiarism that merely followed the form of “cut-and-paste-excerpts” were the smallest offenders. There were dozens of examples of plagiarized texts that included inversions of meaning, fabrications-via-biasing language introduced into the plagiarized texts designed to steer the narrative in directions not originally intended, and worse (my link above shows much of this). It is hard to see how this can be dismissed as cut and paste excerpts. If that weren’t true, one could simply expect the proper attribution of source material would right the wrong. How do you attribute for plagiarized text whose meaning you have inverted?

Worse, I think these offenses are not secondary to the shoddy analytical work, they are inherently linked. As the Deep Climate link you posted shows, the same offenses occurred within the techniques with which the WR dismissed credible authorities while attempting to disguise their recreation of the MM critique (that, it turned out, was really just a rerun of the MM algorithm). The simple fact is that “plagiarism” is remembered most because it is an attention-getting headline that requires no explanation, but the totality of offenses run far deeper because they are inherently linked to the analysis itself. The WR was a report submitted to Congress and federally funded. Quite a few people care about that.

Also, Wegman has denied culpability. If Wegman was motivated in the manner you excuse it, one would think it should be a trivial matter for him to take responsibility and laugh the thing off. On the contrary, the subsequent paper submitted to the journal Computational Statistics and Data Analysis, which was largely based on the Wegman report, has been retracted with an apology issued. A GMU committee reprimanded Wegman, and another GMU committee has been hung on the plagiarism issue for at least 4 years (as many commenters have said, it appears they are hoping the problem will fade away).

Eric Rasmusen said...

Yes, I agree that Wegman should have admitted that he (or his co-authors) cut and pasted or closely followed the summaries other people had given.

An interesting problem though, is whether the Wegman Report is to be treated as research or not. To me, it seems it is not--- it is just a summary of existing work. So if George Mason U. doesn't hold it to scholarly standards of citation, that's fine. This is not incompatible with Prof. Wegman having a duty to admit that it contains paraphrases of other people's work.

It should also be kept in mind that the hockey stick analysis was dishonest and flawed for reasons quite outside its use of principal components analysis. I'm not up on it, but as I understand it, the principal components errors were not a big deal, but the big problem was that Mann spliced together different data sets that were incompatible, and in, particular, he had to "hide the decline" in 20th century temperatures that his method would yield if consistently applied.

Bravin Neff said...

"Yes, I agree that Wegman should have admitted that he (or his co-authors) cut and pasted or closely followed the summaries other people had given."

You just aren't appreciating the rabbit hole of plagiarism Wegman's team spun. I'll ask it again: how would you expect Wegman to provide attributions to plagiarized text whose meaning he (or his team) altered substantially in numerous ways, including:

1. Inversions to the original meaning.
2. Alterations in tone.
3. Introductions of biases not originally included.
4. Raising of doubt not originally included
5. Etc.

? What would that look like?

In the link I provided (you can easily find the full version, however), the plagiarized text and the originals are juxtaposed side-by-side for anyone to see. Deep Climate also did much of this work. If the Wegman Report is "just a summary of existing work," that is the craziest piece of incompetent summarizing I have seen. It is hard for me to see how one could characterize it that way.

Lastly, I have no reason to believe the MBH 98 and 99 papers were dishonest. You mention the famous "hide the decline" quote-mine from the so-called "climategate" email hackings. That has been explained to death, that the "decline" referred to the tree ring proxies related to the long discussed divergence problem, not the global temperatures. Even if you doubt that (and frankly I doubt you have good reason to), the MBH reconstructions have been confirmed numerous times with independent data sets and alternate techniques, the latest I am aware of being the Marcott, Shakun, Clark and Mix paper ("A Reconstruction of Regional and Global Temperature for the Past 11,300 Years," Science, 2013).

Finally, note the following committees, panels and bodies:

1. National Academy of Sciences (2009 & 2010)
2. US EPA (2009)
3. Science Assessment Panel (2010)
4. Inspector General of the US Dep. of Commerce (2011)
5. Penn State University (2009)
6. University of East Anglia (2009)
7. National Science Foundation (2011)
8. Others

...all concluded the same thing: there is no evidence the Climate Research Unit members falsified data, committed research misconduct or some other practice outside of accepted norms. But you would never know it if you found climategate and Wattsupwiththat compelling.

Mark Bahner said...

"I assure you it goes far deeper than 'quoting without attribution introductory materials,' and by my lights is far worse than what John Cook is alleged to have done."

Why do you write, "...what John Cook is alleged to have done"? Don't you agree that John Cook and his co-authors are blatantly misrepresenting the findings of their 2013 Environmental Research Letters paper when they post all over the Internet:

"97% of climate papers stating a position on global warming agree global warming is happening and we are the cause."

?

97% of climate papers...

Do you think this abstract is agreeing that "global warming is happening and we are the cause"?

"Methane may have been released to the atmosphere during the Quaternary from Arctic shelf gas-hydrates as a result of thermal decomposition caused by climatic warming and rising sea-level; this release of methane (a greenhouse gas) may represent a positive feedback on global warming [Revelle, 1983; Kvenvolden, 1988a; Nisbet, 1990]. We consider the response to sea-level changes by the immense amount of gas-hydrate that exists in continental rise sediments, and suggest that the reverse situation may apply—that release of methane trapped in the deep-sea sediments as gas-hydrates may provide a negative feedback to advancing glaciation. Methane is likely to be released from deep-sea gas-hydrates as sea-level falls because methane gas-hydrates decompose with pressure decrease. Methane would be released to sediment pore space at shallow sub-bottom depths (100's of meters beneath the seafloor, commonly at water depths of 500 to 4,000 m) producing zones of markedly decreased sediment strength, leading to slumping [Carpenter, 1981; Kayen, 1988] and abrupt release of the gas. Methane is likely to be released to the atmosphere in spikes that become larger and more frequent as glaciation progresses. Because addition of methane to the atmosphere warms the planet, this process provides a negative feedback to glaciation, and could trigger deglaciation."

How about this one?

The magnitude and timing of a major rapid negative carbon-isotope excursion recorded in marine and terrestrial matter through the Early Toarcian (Early Jurassic) and Early Aptian (Early Cretaceous) oceanic anoxic events (OAEs) have been proposed to be the result of large methane gas-hydrate dissociation events. Here, we develop and evaluate a global carbon-isotope mass-balance approach for determining the responses of each component of the exogenic carbon cycle (terrestrial biosphere, atmosphere and ocean). The approach includes a dynamic response of the terrestrial carbon cycle to methane-related CO2 increases and climatic warming. Our analyses support the idea that both the Early Toarcian and Early Aptian isotopic curves were indicative of large episodic methane releases (∼5000 and ∼3000 Gt respectively) promoting warm ‘greenhouse’ conditions in the Mesozoic. These events are calculated to have increased the atmospheric CO2 concentration by ∼900 and ∼600 ppmv respectively and land surface temperatures by 2.5° to 3.0°C. However, we show that much of the methane released from oceanic sediments is rapidly sequestered by terrestrial and marine components in the global carbon cycle, and this effect strongly attenuated the potential for ancient methane gas-hydrate dissociation events to act as major amplifiers in global warming. An increase in oceanic carbon sequestration is consistent with the deposition of globally distributed black shales during these OAEs. Our analyses point to the urgent need for high-resolution marine and terrestrial carbon-isotope records to better characterize the nature of the Toarcian and Aptian events and improve our interpretation of their consequences for the global carbon cycle.

?

Bravin Neff said...

"Why do you write, '...what John Cook is alleged to have done"? Don't you agree that John Cook and his co-authors are blatantly misrepresenting the findings of their 2013 Environmental Research Letters paper...'"

Its not settled for me. My familiarity with the dispute is David Friedman's own posting in this blog space. Friedman took Cook to task over the use of the word "main," which frankly I found unconvincing. Friedman argued only level 1, from Cook's table 2 (2013), fit the meaning of the word "main." It was a poor basis from which to argue for 2 reasons.

1. "Main" doesn't just imply "more than 50%," it is also applicable to the highest contributor among a plurality of contributors. This means a contributor of less than 50% can still be the main contributor. Friedman's dispute seems to be with the latter, when he said "...which again would be consistent with holding CO2 was responsible for some but less than half of the warming," as if the CO2 couldn't thereby be the main contributor. That didn't make sense to me, and it got me second guessing my own understanding of the word "main," so I looked it up. The several online dictionaries I checked appeared to agree with me: "main" can simply be the chief contributor, or highest rank, or having the greatest influence, etc. If that is true, and if I'm getting Friedman's argument right, it was weakened.

2. Friedman's dispute with the "main" thing was based on the examples in the definitions section of Cook's table. At best what his argument shows is that _the examples_ don't fit the definition of "main," not that the contents of the tallied papers do not. Now its entirely possible the publications tallied inside the 97% don't belong there, because he or someone actually checked them. But that wasn't Friedman's argument: he argued level 2 and 3 don't belong based on the examples in the definitions table, and based on the word "main." He might be right. I find it unconvincing.

But if it makes you feel better, I'm willing to believe Cook lied. Its not out of the question. However, I certainly don't agree the Cook et al (2013) paper is the "source" of the oft-quoted 97% figure. It might now be considered the main source (LOL), but there are at least 4 other studies that seem to converge on the same conclusion. The one from the Proceedings of the National Academy of Sciences from 2009 (http://www.pnas.org/content/107/27/12107.full) seems a more credible source, and that's the main one I was accustomed to seeing cited before 2013 was even a year.

But then there's Wegman...

James Picone said...

Depends what you'd consider wrong on the face. For example, here's SkS pointing out several instances of Patrick Michaels removing data from charts in a way that seems pretty dishonest:
http://www.skepticalscience.com/patrick-michaels-serial-deleter-of-inconvenient-data.html

The Hansen 1988 one is particularly egregious. More about it on SkS here.

Here is Monckton taking the end-of-the-century CO2 concentrations from IPCC projections, multiplying them by an ECS number, and extrapolating the result back linearly to 2001 without error bars. Then he claims that's the "IPCC prediction".

Here is some criticism of a graph John Christy has presented to Congress, which is produced by baselining models and observations to a particular year, comparing model values for surface temperature anomaly to observed tropospheric anomaly, shows just median of the ensemble instead of the uncertainty envelope, etc.

Approximately everything to do with The Great Global Warming Swindle. Hard to find a link for that one, I'm afraid, but here's a section on Wikipedia with one of the scientists that was interviewed for it claiming that data for one of the charts they displayed was in part falsified.

I see somebody has already brought up DeepClimate's look at Wegman. The most relevant part there is probably the table purporting to show that Mann's analysis generates hockey sticks from red noise, where the graphs on the table had, in fact, been selected from a set that had already been sorted by the measure of hockey-stickness McIntyre was using. That is, he presented the most hockey-stick-like outputs he got from his monte-carlo analysis as the norm. Or something like that, I don't remember the full details, DeepClimate goes into it in more depth.

Part of the problem with this is that the denialists have reshaped the Overton window of what would be considered unsupportable bullshit to the extent that folks like Ridley and Anthony Watts appear to be moderates next to, say, Salby and Tim Ball. In any normal field, the bullshit around ice ("Sure, the Arctic is losing ice, but Antarctica is gaining ice!" or "The Arctic is in recovery!"), the repeated games with tropospheric trends, the crap over paleoclimate records, or just plain ignoring all the studies supporting climate sensitivities >2c would look ridiculous.

The other problem is that a lot of denialist bullshit relies on statistical stuff and/or complexities that most of the populace doesn't know about and doesn't understand. The 'pause', for example, is statistical nonsense, anyone pointing out that CO2 concentration has a log-relationship with forcing, people who say things like "But water has a larger greenhouse effect!", or the "Temperature changes a lot over a day, therefore an increase of 2c can't do anything bad".

James Picone said...

@Mark Bahner:
Both of those papers are discussing methane as a significant radiative forcing agent. One of them explicitly says that it "may represent a positive feedback", and the other specifically notes the effect on CO2 concentration and temperature, providing an implicit ECS. This highly reliable source (there's another one for you, David, I'd be very surprised if they weren't lying through their teeth) claims Jurassic CO2 content as 1800 ppm, Wiki says 1950. Let's take the lower number.

A 900 ppmv release in that environment leading to a 2.5c change (i.e. the extremes of the ranges given in the paper) is an increase of 50%, ln(1.5)/ln(2) = 0.584 so ~58% of the full effect at doubling, so an ECS of 4.2c is implied by the paper. Yes, that's very much in cat1, unless you're disputing that humans are putting more CO2 into the atmosphere.

Do you think that for a paper to be in cat1, it has to demonstrate the entirety of modern climate science?

Mark Bahner said...

"Both of those papers are discussing methane as a significant radiative forcing agent."

Yes, those two abstracts are discussing methane as a significant radiative forcing...long before the Industrial Revolution even began. So those two abstracts can not possibly be saying, "...global warming is happening, and we the cause."

James Picone said...

@Mark Bahner:
You could have just said 'yes' to my question "Do you think that for a paper to be in cat1, it has to demonstrate the entirety of modern climate science?"

Unless you dispute an anthropogenic rise in CO2 (Or maybe argue that modern CO2 behaves differently to Jurassic CO2?), the 4.2c ECS the second paper's figures imply means that >100% of the temperature rise since preindustrial is anthropogenic (reminder that the IPCC's ECS range is 1.5c-4c with best estimate 3c; 4.2c is /quite high/ for an ECS value). Cook's allowed to draw obvious and pretty-much-indisputable implications from an abstract. Under your rules, I'm not sure any complicated proposition has ever been supported by a single scientific paper. Maybe one paper per complicated proposition.

You could, of course, look at attribution studies specifically. Here's an SkS article that looks at five; you could probably find others. All of them find >50% anthropogenic influence on temperature increase (in fact, all but one of them find >100% if I'm reading the graph right, and the one that finds <100% isn't much less. This makes sense, given the IPCC's attribution pdf). I'd be impressed if you could find one that was a) published in a respected, peer-reviewed journal (i.e. not Energy&Environment or similar crap-peddlers) b) found <50% anthropogenic influence. I'd be very impressed if you found one that was also cited by someone and doesn't have glaring errors at first sight. Here's some suggestions.

Mark Bahner said...

"You could have just said 'yes' to my question "Do you think that for a paper to be in cat1, it has to demonstrate the entirety of modern climate science?"

Why would I say "yes," if the answer is "Of course not"?

Also, why should I have answered the question, since it is completely irrelevant?

Those two papers were *not* in Category 1. They were in Category 3. From that simple fact, and from simply reading the abstracts, anyone should be able to see that the abstracts are not saying, "Global warming is happening, and we (humans) are causing it."