19 October 2009

Sound and Fury at WUWT

From the question place, where a reader noted a high traffic item at Watt's Up With That  and asked for a science response.  Where to begin?  First, I guess I'll note that most of the post is bluster and personal attack.  Once you cross out those parts, it's a much shorter article.

Second, as always, go back to the original source.  In this case, it is a Mann et al. 2008 paper Proxy-based reconstructions of hemispheric and global surface temperature variations over the past two millennia, with supplementary material.

Then, consider exactly what the claims (in this case, at WUWT) are, and just what evidence is produced for it. 

The fundamental claim at WUWT is that the entire reconstruction is upside down.  (We're treated to pictures of other things that are upside down.)  Right off, we know WUWT is wrong. 

There are three major features of temperature over the past 1000 or so years -- the 'Medieval Warm Period', the 'Little Ice Age', and the warming of the past century.  We've known about the first two since at least the 1970s -- Hubert H. Lamb's Climate: Past, Present, and Future.  The Mann and others reconstruction shows a Medieval Warm Period and a Little Ice age.

If WUWT were correct about Mann et al. having the curve upside down, then they (WUWT) must be insisting that it was a Medieval Cold Period and Little Warm Period -- which we've known for decades it wasn't, irrespective of anything that Mann or coworkers have done.  WUWT is simply wrong from the get go.

There is then a lot of sound and fury regarding 'Tiljander'.  This turns out to mean a paper by Tiljander and others in 2003, cited in the supplementary material of Mann and others.  WUWT cites CA citing personal communication claiming that Mann et al. used this data set upside down.  This looks more like a game of 'whispers' or 'telephone' than a serious scientific claim.  If Tiljander (then CA, then WUWT) had serious evidence of error by Mann et al., the scientific literature is the place for it.  Or, at the very least, Tiljander et al. could place a short note on their own blog/web site/university press release/....  Neither CA nor WUWT seem to cite any such thing, so I will draw the inference that this is because it doesn't exist.

Still, there might be a question as to whether the data were used correctly.  A more significant question is whether the data and its usage materially affect the reconstruction.  If, for instance, the only reason for showing a warming in the 20th century is Mann and others' use of this data set, we might be more concerned about whether there's been such a warming.  Or at least it becomes a much more important question whether they did use the data set properly.  So let's go back to the original source and see what usage was made, how, and what effects it has.

Page 2 of the supplementary material, under Sensitivity Analysis (NH Temperatures)Potential data quality problems.  First, we'll note that the authors do indeed consider the possibility of data quality problems. I'll quote that paragraph here (all typos mine, see the original):
In addition to checking whether or not potential problems specific to tree-ring data have any significant impact on our reconstructions in earlier centuries (see Fig. S7), we also examined whether or not potential problems noted for several records (see Dataset S1 for details) might compromise the reconstructions.  These records include the four Tiljander et al. (12) series used (see Fig S() for which the original authors note that human effects over the past few centuries unrelated to climate might impact records (the original paper states "Natural variability in the sediment record was disrupted by increased human impact in the catchment area at A.D. 1720." and later, "In the case of Lake Korttajarvi it is a demanding task to calibrate the physical varve data we have collected against meteorological data, because human impacts have distorted the natural signal to varying extents.").  These issues are particularly significant because there are few proxy records, particularly in the temperature-screned dataset (see Fig. S9) available back through the 9th century.  The Tiljander et al. series constitute 4 of the 15 available Northern Hemisphere records before that point.
They also note 3 other data sets with problems.

So, do the authors proceed blindly, pretending that all data are good (and equally good, at that)?  No, there's the reconstructed figure in S7, using all data, and another using all data except for the tree rings.  And then in figure S8, they show what happens after removing the 7 problematic data sets (The Tiljander 4 plus 3 from elsewhere).  Same kinds of curves either way -- still a Medieval Warm Period, a Little Ice Age, and a warm recent century.  As Mann and others note that before the 9th century there are few data if one withdraws Tiljander, I'm ignoring that part of the reconstructions.

From just reading the paper, we don't know whether the Tiljander data were used correctly.  We do, however, know that the answers are quite similar whether they're used or not (S7 vs. S8).  If there's an error, in other words, it's an error with little effect.

Contrast that with WUWT's initial claim of the whole reconstruction being upside down -- thereby turning the Medieval Warm Period into an ice age, and the Little Ice Age into a Warm Period.  They give no evidence at all that this is the case.  Irrespective of whether the Tiljander data were used wrongly, WUWT is wrong in their main claim.


Scruffy Dan said...

The whole notion of the Mann curve being upside down is absurd even if you don't know anything about the MWP and the Little ice age.

If Mann's curve is upside down then over the past century we have experienced significant cooling (if I recall correctly Climate audit showed a curve claiming that not too long ago), but irrespective of any proxy reconstructions done by anyone, we know (thanks to thermometers)that the 20th century has warmed substantially. There is no sense looking at tree rings to determine temperature when one has a thermometer handy.

If you're proxy reconstruction shows cooling ovre the past century then there is clearly something is clearly wrong with it.

Phil M said...

The WUWT article isn't claiming that the 'whole' Manning reconstruction is upside-down

- what it is pointed out is much more specific

- it shows that the sediment data 'Tiljander' is used upside down

- the Mann response is more-or-less that it doesn't matter

- this cannot be an appropriate response!

- as far as I can see, the ClimateAudit article (http://www.climateaudit.org/?p=7411#more-7411)
has correctly described the situation
- the sediments are less dense in warm periods due to having more organic material
- and more dense in cold periods due to having more mineral material.

- the authors of the original data then go on to postulate that the samples in the 20th century were corrupted by various human activities, and so should be ignored
(they otherwise indicate a cooling in the 20th century, which is obviously incorrect)

However, Mann has used this data, including the corrupted 20th century data, and has therefore inverted the data in order to get this (corrupted) data to agree with his a priori assumptions!

- this is the accusation of the WUWT article

- obviously, if this accusation is correct, then any reconstructions based on this data should be discarded
- and new reconstructions should be made without this data.....

Robert Grumbine said...

The folks at WUWT, I believe, do not think that the 20th century shows a warming, so I didn't point to that. (Watts' station project being devoted in part to his claim that the station siting produces a false warming.)

Phil M:
At the top of the article, just below the pictures of upside down things, plus the reconstructions, it says:

Chances are, if you are not Dr. Michael Mann of Penn State University, you’d answer: “It’s a trick question, all of them are upside down”.

And you’d be right.

The Tiljander material is later window dressing that apparently succeeded in distracting you from this whopper. WUWT did not show ... anything, really. It asserts that Mann did various things, but it actually gives no evidence that he did them as WUWT claims. Evidence would be, say, to show the part of the paper where Mann and others note that they flipped the curves, or the part of the data analysis that would have done that automatically, or the like. What is given instead is unverifiable hearsay. That may be evidence to you, but it isn't to me.

Where you say 'this cannot be an appropriate response' (I'll note that you're misrepresenting Mann, even as quoted at WUWT), there's a point of science involved. It is a usual thing that we push data (models, theories, ...) hard. It is therefore also important to see whether our conclusion is dependant on those things we pushed the hardest to use. So Mann and others show what happens to their analysis if they don't use 7 data sets (4 being Tiljander) that they note there are concerns about. The answer is, using them and not using them give very much the same answers (over the period 900-present that I looked at).

As it stands, there's no evidence that Mann and others used these data incorrectly. And there is evidence that their conclusion doesn't depend on even whether that data set was used.

Let's stay away from mind-reading and personality guesses. Particularly after reading the WUWT article, which was mostly personal attack with side trips to mind-reading and personality guesses, my tolerance is low.

There is indeed a science point about McIntyre's approach, and I'll take that up sooner rather than later in a post.

Deech56 said...

I'll try again. My initial speculation was that the author and commentators of the WUWT post would bring up the dendroclimatology data (particularly the bristlecone pines), as each issue is addressed separately, but AFAIK not together, in the paper and in Figures S7 and S8. Turns out I wasn't reading the WUWT post carefully enough: "The SI to Mann et al 2008 conceded that there were problems with the recent portion of the Tiljander proxies (without mentioning that they were using them upside down from the interpretation of Tiljander and Finnish paleolimnologists), but argued that they could still “get ” a Stick without the Tiljander sediments. However, as I observed at the time, this case required the Graybill bristlecone chronology .... Thus their 'robustness' analysis used either upside down Tiljander sediments or Graybill bristlecones."

My non-specialist view is that Mann, et al. 2008 is a very important paper, as it represents the current state of the science and addresses the points brought up in the NRC review of their earlier papers, but I don't think these side issues regarding the use of the proxies will go away soon. I look forward to your further analysis.



p.s. This doesn't have to be published - I just thought I owed you a more complete explanation than I had in my earlier comment. Yes, it is tough to wade through the comments of the "skeptical" sites, but it's something we've all had to do from time to time. As a scientist, I get personally offended by the smears that are brought up time and time again. It's not always easy to take the high road. Please keep up the good work.

Anonymous said...

Dr. Grumbine,

You are being rather unfair to Steve McIntyre with your comments.

In particular: "If Tiljander (then CA, then WUWT) had serious evidence of error by Mann et al., the scientific literature is the place for it. Or, at the very least, Tiljander et al. could place a short note on their own blog/web site/university press release/.... Neither CA nor WUWT seem to cite any such thing, so I will draw the inference that this is because it doesn't exist."

You are quite wrong here. Steve actually did publish this finding in the peer reviewed literature - it was accepted as a comment in PNAS which is the journal in which Mann 2008 was published. Mann's response was that Steve's criticism is bizarre, saying multivariate regression methods are insensitive to the sign of predictors. I have yet to see anyone defend Mann on this point. Perhaps you will be the first.

I think you owe Steve an apology.

Anonymous said...

We don't know where Atte Korhola got the "upside down" claim from. He's not exactly a very well taken figure here in Finland, married to the famous anti-mitigation right wing pro-industry MEP, Eija-Riitta Korhola.

It could very well be that he got it from McIntyre or Watts or wherever.

He might not have spoken to Tiljander at all.

Korhola says in his blog that he actually participated in a PNAS comment writing with McIntyre in February 2009.

Of course, Watts, McIntyre et al can present it approximately that even a Finnish paleolimnologist says their colleague's series was inverted.

I can translate Korhola's comment here if you want it kosher.

Robert Grumbine said...

Thank you for restarting. You put some substance to the table, and that's to the good.

The assorted reconstructions (Mann et al. being only one among many at this point) are not my area of expertise, so I won't be going to much depth here. On the other hand, the WUWT claim about the whole curve being upside down takes no depth of knowledge to see the error. So, for what that's worth, I make mention.

If you're willing to take on the translation, and would like to post it here or at your place, I'll be happy to host it or link to it. The CA article references either anonymous people, or machine translation. None of those is persuasive to me. Please note your knowledge level of Finnish (native speaker, studied it for some years, etc.) as well. (Given your 'about' page and name, I'd guess native speaker?).

On the other hand, the reference to the blog article is really quite weak -- some blog somewhere says something -- kind of thing. Maybe it does, and to that degree it's worth having a more widely accessible language version. On the other hand, just because some blog somewhere says something isn't much reason (to my mind) to reject something from the scientific literature.

Digressing a second: The rest of you -- take a look at the Gravity Loss. Space exploration is the topic there, and is another of my interests.

McIntyre did indeed publish a note in the PNAS -- journal that Mann and others originally published in. But, according to the CA (climate audit) articles that WUWT linked to, the argument that the Tiljander and others data were used incorrectly was not in that comment. The argument for that error is in CA blog posts, and WUWT quoting CA quoting personal communication from Tiljander. That's a rather poor grade of hearsay. It might be that Mann and others didn't use the data correctly. But hearsay is not the way to show it.

It could well be that somewhere in the couple of dozen posts that McIntyre has made on this paper (counting those that he's tagged Mann 2008), or the several thousand comments, something better by way of evidence has been produced. I don't know. But it was the WUWT article I was asked to look at, and I assumed that the WUWT folks provided the strongest evidence they could. And that was to hearsay comments, not scientific publications.

Philip M: (not posted)
Repeating yourself is not the way to make progress here. Add evidence -- evidence about the science. I noted some sorts of evidence that would be meaningful. Instead, you stayed with third-hand sources and personalities -- not the science.

It's also a concern that your standards of evidence are so wildly different between WUWT and for Mann and others. For WUWT, you excuse that they open with a flagrant misrepresentation. But for Mann and others, you're satisfied to condemn the scientific publication (and its lead author) with third or fourth hand hearsay conveyed by WUWT.

An honest skeptic is skeptical of all sources.

Robert Grumbine said...

anon: (not posted)
If there's serious evidence that a scientific paper was incorrect, the scientific literature is the place for the correction, not just a blog. It's distinctly more work to publish in the scientific literature than a blog. But that's part of the point. The comment by McIntyre and McKitrick does not demonstrate that Mann and others used the Tiljander incorrectly -- it asserts it with hearsay support only. Mann and others respond that they did not use the data in the way McIntyre and McKitrick allege. c.f. Potential nonclimatic influences on the Tiljander and other proxies were discussed in the SI, which showed that none of our central conclusions relied on their use.

You could be lazy, and go with 'he said, she said'. Or you could take the issue seriously and start reading the scientific literature on the topic. If and when McIntyre produces his own analysis of paleoclimate, in full accord with what he thinks is the proper way to conduct such an analysis, that might be a contribution to our scientific understanding of the topic. He's explicitly said, however, that he is not interested in doing such work. So we have to rely on those who are willing to put in the work and take the risks (of being attacked, by, c.f. McIntyre) of trying to do such analyses.

If you have evidence beyond hearsay, and beyond 'he said, she said', please do show it.

Robert Grumbine said...

The original blog comment (cited by WUWT through CA) in Finnish, followed by a translation by gravityloss (a native speaker of Finnish) follows. I've (Bob Grumbine) corrected a couple of typos in the English (or at least I hope they were typos):

Olin itse mukana kirjoittamassa tuota kommenttia, mutta "kieltäydyin kunniasta" saada siihen nimeni mukaan (johtuen siitä, että katson omalla nimelläni esiintymisen tässä politisoituneessa tieteessä haittaavan varsinaista tieteelistä uraani). Valitettavasti PNAS sallii ainoastaan maksimissaan 150 sanaa ja viisi viitettä noihin kommentteihin. Paljon jouduimme jättämään kommentista pois. Itse asiassa Steve kysyi mielipidettäni siihen, että kannattaako (hän oli epäilevällä kannalla) koko kommenttia kirjoittaa, koska a) se jäisi väkisinkin rajoitusten takia torsoksi, b) kaikki asiat ovat joka tapauksessa jo dokumentoitu ClimateAuditissa ja c)kommetilla ei ennustettavasti ole mitään erityisempää merkitystä. Sain Steven ylipuhuttua tuon kirjoittamiseen. Ehkä minulla oli mielessäni kaltaisesi ihmiset.

Noita "väärennös" juttuja kommentoisin niin, että mielestäni näissä multiproxy rekonstruktioissa ei ole kyseessä "väärentäminen" vaan pikemminkin huono tieteenteko yhdistettynä "confirmation biakseen". Poikkeuksena Mannin alkuperäiset lätkämailapaperit (MBH98/MBH99). Olen yksi niistä kourallisista ihmisistä, jotka tuntevat Mannin lätkämailatyön niin tarkkaan, että pystyvät sen kokonaisuudessaan toistamaan. Mann on noissa töissä tehnyt useammassa kohdassa erinäisiä valintoja (*), jotka parhaan ymmärrykseni mukaan on selitettävissä ainoastaan sillä, että pyritään saamaan aikaiseksi tietty ennalta päätetty lopputulos. Minusta se täyttää tieteellisen väärentämisen tunnusmerkit.

(*) Parhaiten on julkisesti dokumentoitu Mannin kokeilut pääkomponenttianalyysilla (oikea/itse kehitelty PCA laskenta, kokeilut vihnemännyillä/ilman vihnemäntyjä "cencored"-hakemistossa...)."

I participated in writing that comment but I "didn't accept the honor" to get my name on it (because I see that showing my own name in this politicized science to be an impediment for my actual scientific career). Unfortunately PNAS allows only 150 words and five references in those comments. So we had to leave out a lot. In fact Steve asked my opinion (he was skeptical), is it worth to write such a comment at all because a) it would be left as a torso because of the limitations, b) all things are already documented and c)the comment would presumably not have any significance. I persuaded Steve to write it anyway. Perhaps I had people like you in mind.

Those "forgery" stories I would comment so, that in my mind in these multiproxy reconstructions what is not happening is "forgery" but rather bad science combined with "confirmation bias". An exception is Mann's original hockey stick papers (MBH98/MBH99). I am one of the handful of people, who know Mann's hockey stick work so closely that they can reproduce it in its entirety. Mann has in those works in many places done different choices (*), that in my best understanding can only be explained by trying to arrive a specific beforehand decided outcome. To me it fills the marks of scientific forgery.

(*) The best public documentation is available from Mann's experiments with principal component analysis (real/self developed PCA computation, experiments with bristlecones / without bristlecones in the "censored"-directory...).

MikeN said...

Here is some proof of the upside-down usage. Not in Mann, but in Kaufman 09, Arctic warming.

Here is the original paper. Figure 5 has the proxy in question.

Here is the data:

Here is a chart of Tiljander proxy data listed by Kaufman et al(10 year averages)

Here is the overall reconstruction, showing that higher numbers means warming.

So if you just compare the warm periods in the 1000s and 1200s, and see that they are cool periods in the reconstructions, you can see that the proxy is used upside down.

I mention Kaufman and not Mann because the source data is easy to track. Climateaudit.org has tracked down the data for Mann as well and Tiljander is used inverted there as well.

Also, Kaufman calls this an error and says he is issuing a correction. Bradley was a coauthor for both papers, so perhaps Mann will be corrected on this point as well.

Robert Grumbine said...

As I've already observed, repeating yourself is not a way to make progress. I've mentioned what sort of thing would be evidence. With your claimed expertise, you should be able to provide it. For instance, you make innuendo as to what happens if both bristlecones and Tiljander are removed. But merely innuendo. Do the work and publish it, that would be evidence. Please also read the comment policy.

Mike N:
Your evidence to defend WUWT saying that Mann and others' entire reconstruction is upside down is to say that somebody else used Tiljander and others data wrong?!

There's a minor point -- whether Mann and others used Tiljander and others' data correctly. Maybe they did, and maybe they didn't. The only evidence for error presented so far is hearsay (as you quote from the McIntyre and McKitrick article I've already linked to), or, now, you pointing to a different paper entirely. Evidence that Mann and others used a data set wrong has to be come from examination of what Mann and others did -- or by doing more work, as is the usual route in science, and constructing your (a general you, not necessarily you personally) own reconstructions, using the data correctly and showing that your answer looks more like Mann's if you use it incorrectly.

The major item is WUWT's claim that the Mann and others reconstruction is entirely upside down. That is a flagrantly false claim, and is the main topic here. Even before Mann entered college, it was already known that there was a Medieval Warm Period and a Little Ice Age. His reconstructions show that, as do everybody else's. They showed that in 1998, when he started this line of research, and still do today, regardless of whether he's the one doing the reconstruction, and regardless of whether the Tiljander data are used at all. And this was shown in the 2008 paper* that is the subject of WUWT's current egregious misrepresentation.

*(supplementary materials that is; and it's only by reading the supplementary material that you know that it was Tiljander et al data that were used).

If you read my blog routinely, you'll know that I focus on the ends of things that don't take great expertise. Getting in to whether Mann's methods do the right thing, or not, and did, or did not, handle Tiljander and others' data correctly is moderately technical. So not my focus.

Recognizing the falsity of WUWT's claim that the entire reconstruction is upside down takes no great expertise at all -- just to know that there was a medieval warm period, little ice age, and that the reconstruction shows those features. This post was tagged under 'weeding sources' because there are a lot of sources that say they are about science, in particular about climate science. But a number of them make egregiously false claims like WUWT did on this. I figure if we read, instead, sources that aren't making such enormous errors, we'll probably learn more, and faster.

Michael Smith said...

You say:

If there's serious evidence that a scientific paper was incorrect, the scientific literature is the place for the correction, not just a blog.


What if the "scientific literature" is edited by people sympathetic to the paper being criticized and they simply refuse to publish any critiques?

Why should a new technology like the Internet be arbitrarily ruled-out as a place for examining the scientific claims of a paper?

Why is it mandatory that one method of publishing scientific information -- the "peer reviewed journals" -- be frozen-in-place forever, regardless of other methods people might develop?

What makes those “peer reviewed journals” the eternal, unchallengeable, unalterable gold standard -- particularly when Steve McIntyre at ClimateAudit.org is finding error after error in papers that went through that very “peer review”?

A constant refrain from the pro-AGW advocates is that any criticism of AGW climate papers that appears outside "the mainstream" is automatically disqualified from consideration. However, I’ve yet to see any facts educed to justify this peculiar notion -- it is merely asserted as if it is axiomatically true. Well, it isn't.

It is true that any scientific claim that is made without supporting evidence may be ignored, no matter where it is published. But that is hardly what happens at ClimateAudit or WUWT. If you doubt that -- if you believe them to be making baseless claims -- then by all means go to the comments section of those blogs and blow them out of the water. (Unlike the site RealClimate, ClimateAudit does not censor comments.)

Science is the search for truth -- nothing justifies the notion that that search can only be conducted within the rigid confines of whatever the editors of today’s science journals are willing to publish. The process of science does not tolerate "gatekeepers" or censors.

Robert Grumbine said...

'what if', indeed.
If wishes were fishes, we'd all cast nets.

The case at hand alone shows that your what if is false -- McIntyre and McKitrick did publish a comment on the paper, and it's been cited here.

The reason for publishing things in the scientific literature is for scientists to read it. If a scientist got something wrong, and it got published anyhow, then the people who read the paper might do incorrect science if they rely on that paper. Best way to correct the error -- so that scientists who read the paper will see the correction -- is in the scientific literature.

There's nothing anti-technology about that. Best place for a correction of something I say on my blog here is on my blog here -- the same people who saw my error can then also see the correction. And there are a number of open source journals these days, so 'scientific literature' is indeed on the web. They engage in a different sort of review process. But there is one.

There is a merit to the non-blog world, which comments here and even more so at WUWT have illustrated. Namely, in the professional situations, there is far less haring off with hearsay and personal attacks, and far less flagrant misrepresentation of the sort WUWT did in the post I was responding to.

Nor would we see something both flagrant and silly like your implying that I think peer reviewed jounals are "eternal, unchallengeable, unalterable". Flagrant, since I never said any such thing.

Silly because the entire point of scientific journals is change and challenge. Come up with a new idea -- which therefore means change from the old -- and publish it. Read someone else's idea, think he's done something wrong, do the job right yourself, and publish it. The surest route to fame in science is to show that some commonly held idea is wrong.

But it does have to be the science you address, not people. However much you and folks at WUWT and CA detest Mann, which is apparently a very great deal -- the science at hand is global temperature reconstructions. Fame is not yours for saying that Mann made a mistake. That's rather minor stuff. Useful, if you then document it seriously, but minor.

Fame is yours for showing, as WUWT is asserting by saying Mann's reconstruction is upside down, that the Medieval Warm Period was actually a cold period, the Little Ice Age was a time of glacial retreat and warmth rather than a glacial advance and cold, and that climate has actually cooled over the past century. Put together serious evidence in favor of that reconstruction, and you're set for science fame.

Hank Roberts said...


Last nine-thousand years of temperature variability in Northern Europe

Hank Roberts said...

Ah, take a look at the comments on that paper -- the pattern of which season causes which layer in the annual pattern may reverse as a climate mode reverses. That's, grossly oversimplifying, modes of "cold and dry and dusty" versus "warm and wet" climate, I think.

So there's some basis for confusion for anyone who assumes the relationship of the light vs. dark material in the layers always means the same thing about the temperature. Plant growth and erosion will occur at different times of year in different climate modes.

I know nothing about any of this, just putting fragments together.

From the reply to the comments for that paper:

"We suggest that the most direct driver of the late-Holocene anomalies has been changes in the dominant atmospheric circulation type. This seems likely in an area, where the modern temperature and precipitation values are highly variable depending on the changing circulation patterns."

Horatio Algeranon said...

If blogs were science

Anonymous said...

I realize that this post is about the WUWT misrepresentation (which can be construed as a simple over statement to make a point) of the entire graph , but the claim that some data was used upside down suggests a question: What happens to the graph when the questioned data is reversed (used correctly according to Watts) and applied? Will it pull the tail of the graph down?

As you mentioned, Mann et. al. did consider the result without the questioned data but did they (or anyone else) consider the effect flipping the data would have if included?

If it can be shown that flipping the data has no effect on the graph then Watts' argument that Mann et. al. are being purposely selective disappears.

O.T.: Are you the R. Grumbine from talk.origins?


Deech56 said...

Hmmmm...in an earlier comment I wondered what would happen if both the "potentially problematic proxies" and tree ring data were removed. Guess I'm not the only one who wondered about that. Mike Mann has updated his supplemental information to answer this very question and came up with this (pdf). He shoots, he scores.