In March last year, I was intrigued by the following statement in the then recent IPCC Summary for Policymakers which stated:
“Studies since the TAR draw increased confidence from additional data showing coherent behaviour across multiple indicators in different parts of the world”
What exactly was the “additional data” since the TAR? At the time, AR4 had not been published. Even now, if you re-visit chapter 6, there is negligible support for this statement. Yes, AR4 lists many Team multiproxy studies, but, for the most part, these regurgitate pre-TAR data. Note that Mann et al 2008 will be of no help in this respect as it almost entirely relies on pre-TAR data as well.
Anyway, I was stimulated enough in March 2007 by this comment to re-canvass North American tree ring in the ITRDB for new additions. I noticed a recent addition of spruce chronologies from northern Alberta and did a post on Mar 19, 2007 showing a quick average of these chronologies to see if they supported the IPCC claim. I observed that, if these were temperature proxies (carefully expressing the matter in conditional terms), these did not evidence the SPM claim.
This post prompted a number of posts from angry/frustrated dendros and a discussion of what to do about CA at the dendro listserv.
Rob Wilson, a welcome and occasional visitor here, sharply criticized me for simply grabbing sites from ITRDB without determining that they had been sampled as temperature proxies, pointing out that Meko, the author of the majority of these particular chronologies, was known primarily for moisture studies:
For readers of this blog, PLEASE understand that one cannot randomly sample trees from any location and expect there to be a valid climate signal (temperature or precipitation)….
Dendrochronology is NOT just a study of climate. Tree-rings are used to study fire history, geomorphological processes and ecological aspects of tree growth (e.g. stand dynamics, insect attacks, pollution effects etc). Therefore, one cannot assume that ALL tree-ring data in the “North American tree ring data base at WDCP” was sampled with a climate related question in mind. Therefore, a simple average of 14 sites, without knowing why they were sampled in the first place, is arguably a meaningless exercise.
Mike Pisaric, another dendro, joined in the criticism, stating that my analysis was “flawed!”, that ITRDB chronologies had been collected for a multitude of purposes and that you could not assume that they contained a climate signal.
As Rob points out in 3 and 28, Steve’s analysis is flawed!… The ITRDB contains tree ring chronologies that have been used for a multitude of purposes and not just paleoclimatology. So no, you would not expect all tree ring chronologies from the ITRDB to contain a valid climate signal.
The dendro critique spilled over to the dendro listserv. (See CA post here) where the following question was asked:
So – should I (we) ignore this Blog? … Personally, I cannot do this. Although some of the criticisms and commentary are valid, some of it is simply wrong and misinformed, and in my mind, it is dangerous to let such things go. .. Overall, this is a matter of outreach. I believe that tree-rings are one of the most powerful palaeo proxies available. However, we cannot allow the discipline to be muddied by a few ‘loud’ individuals who’s motives may be suspect.
I responded to this criticism in a post The Dendrochronologists are Angry , stating that I had no interest in spreading misinformation and would undertake to promptly correct any such misinformation, opening up a thread for dendros to respond without any obligation to deal with CA readers. This did not result in much of a response.
I challenged the Dendro Truth Squad to root out the use of precipitation proxies, highlighting Dulan junipers as an example, a chronology used (either directly or indirectly in a number of multiproxy studies). One angry dendro replied via Rob Wilson (see here),
Those in the know, who really know the science, know not to use that chronology and know who still use that chronology. The work that uses that chronology for a temperature reconstruction is less-respected than others.
He accused me of ignoring “recent work that surpasses all others.“. I expressed my desire to visit this particular shrine and requested the identity of this work that “surpasses” all others , but no one was able to identify it for me. Perhaps the angry dendro had a draft version of Mann et al 2008.
There were about 10 posts at the time both on this dispute and on the more general question of North American treeline proxies, listed in this category.
As so often, my principal disappointment and frustration was on the one hand at the lack of substance in the dendro interventions and, on the other hand, at their inconsistency.
The inconsistency was particularly frustrating. While I didn’t think that I particularly deserved the criticism leveled at me for merely calculating an average of the 14 PCGL chronologies, I would have readily accepted the criticism if this led to the articulation of an industry-wide standard that dendros would apply to multiproxy standards. (In this light, I criticized dendros for feeling that it was important to speak out about an incidental post at CA, while remaining silent on MBH98-99 and similar studies.)
History repeats itself. If Peter Brown, Mike Pisaric, Rob Wilson and other dendros were outraged at my calculating the average of 14 Alberta PCGL chronologies, then how can they stand idly by while Mann et al 2008 essentially scavenges the ITRDB like a garbage picker? If they felt so strongly about my post last year, then they must be outraged about the total failure of Mann et al to assess “Location, location, location”.
Let’s go back to the questions asked at the dendro listserv and see which, if any, of the following applies to Mann et al:
So – should I (we) ignore this ? …
Although some of the criticisms and commentary are valid, some of it is simply wrong and misinformed, and in my mind, it is dangerous to let such things go…
I believe that tree-rings are one of the most powerful palaeo proxies available. However, we cannot allow the discipline to be muddied by a few ‘loud’ individuals who’s motives may be suspect.
What’s sauce for the goose is sauce for the gander.
83 Comments
Steve:
Your forceful tone seems warranted but I am less certain that it will have the desired effect. Has Dr. Wilson commented on Mann 2008 in any publicly accessible forums? Your commentary suggests that you do not know of any detailed analyses of Mann 2008 by other dendro scientists – positive or negative? Is it currently being cited in other paleo articles?
Steve, if the dendros won’t make comment on Mann’s work directly (for whatever reason), but do feel comfortable disparaging your work here at CA, then perhaps it would be a worthwhile exercise to do something similar to what Jeff Id has done. That is, rather than directly “attack” a study, simply use the data and methods of studies that they are silent on to present a conclusion that is “opposite” of the original study, then see which parts of it are “attacked” – I am reasonably confident you can get the math right, so any such “attack” would have to be based on the data or methods you select. Let them “settle” what’s wrong with it by themselves, then ask why such criticisms don’t also apply to the studies you got the “dodgy” data or methods from.
The “desired effect”? I don’t seriously expect Rob (or Mike Pisaric or Peter Brown) to wade into criticizing Mann or Briffa or anyone like that. They didn’t do it with MBH, so why now?
This post is about dendros as a group, not just about Rob Wilson. (And I don’t expect Rob to be fussed about this post. I emailed him that I was going to do a post on the Silence of the Lambs and I think that he was amused.) Quite a few dendros took exception to how CA dealt with their proxies – Peter Brown, Mike Pisaric and some unnamed angry dendros. I don’t expect Rob to carry all their burdens. They were quick to criticize a reference to Meko chronologies in an incidental blog posting here. Their silence in respect to Mann et al 2008 is deafening.
#2. I’ve done many such analyses.
I like to start with understanding the proxy data sets and am actually interested in seeing if there are some useful proxies in Mann et al 2008 that I wasn’t familiar with. The 1209 proxies of Mann et al take a little time to analyze.
Re: Steve McIntyre (#4),
Come on posters here at CA stop the piling on and listen carefully so that we can hear any break in that silence that Steve M describes above.
Re: Rob Wilson (#26),
There, there did you hear it? I did.
Maybe it’s just more fun beating up on conjectures made in the blog posts of ex mining executives.
It seems it would be easy to understand what exactly this
Or this
means in English. Perhaps it’s your Canadian accent Steve.
Or perhaps you’re correct and the logic is too subtle for those who deal with heavy equipment to out of the way places or to those who have difficulty getting measurement data stored on electronic equipment and available to those giving them grants, paying their government or academic research, publishing their papers, or other scientists and researchers.
Suitable responses would seem to be along the lines of “No, you’re mistaken, the PGCL white spruce are not held to be temperature proxies.” or “Since the widths don’t move along with temperature, they can’t be used as proxies for it. In fact, that brings up an interesting point; why would anyone use them in dendro reconstructions. I should go ask those doing so!”
You also said
If this is the case, then the inverted rings might be temperature proxies. However, you reported that Wilson and Pisaric observed that
which would lead one to believe there’s not enough of a signal one way or the other for temperature in the rings of those trees.
I’d think that correcting the idea that the 14 PCGL chronologies are related to temperature in any way would suffice, and moving along would be in order. Then, those using the chonologies in papers (or those gathering the data and not providing it) could be properly questioned.
Re: Sam Urbinto (#5),
If the PCGL trees have increased ring-width with increasing temperature, why don’t they grow better in more southern climes?
Re: Pat Keating (#13),
Well, Steve was in Mar 2007 under the impression that “white spruce (PCGL)” was “a species held to be a temperature proxy.” Further that IF the “hypothesis is” “PCGL ring widths increase with increasing temperatures”, the conclusion would be “these non-Jacoby chronologies are evidence against” such a link, using this ad-hoc impromptu experiment.
Now in actuality, if the ring widths were linked to temperature, higher=wider lower=narrower, then in a warmer climate the rings would be wider. Which means? Can trees have wide rings and be almost dead with hardly any (live) branches and puny leaf growth patterns? In that case (if the premise is true) even IF the temperature causes wide rings, that wouldn’t tell us how well the trees grow.
Obviously, if wide rings=growing well, that’s different than rings!=overall health. Which relates in certain ways to temperature and moisture and soil composition and chemistry and amount/strength of sunlight et al.
Wilson and Pisaric say Meko white spruce chronologies in northern Alberta are considered precipitation proxies. So what does that say about scientists using them as temperature proxies?
But there is another issue here; if a group of trees in question are precipitation proxies, what does that say or not say about the appropriateness of them in addition being in some way (tending to have some sort of inverse or direct relationship) also temperature proxies. Behavior and reactions aside.
But “These are precipitation proxies” isn’t the same as “These are not temperature proxies”.
As Steve said, the point of this post. He in his role is getting all this from simply wondering if one might “be inclined to say that the average of these 14 chronologies is inversely related to temperature”.
Based upon the reaction to Steve, if they are aren’t temperature proxies at all, what would be the appropriate response to those publishing studies as if they were?
Things seem a little backwards here.
Awsome post! It would be great if the dendros got together and published a paper that detailed just how one goes about determining whether a given tree-ring series is likely a temperature proxy.
Re: jae (#6)
Perfect. So despite of the lack of pamem we would at least be rewarded with some great circenses. Let’s move on.
Re: Wolfgang Flamme (#9),
Metaphore not appreciated. We, indeed, should move on from this 10 year impasse. Let the dendros show just ONE tree-ring series that they ARE CERTAIN are reliable temperature proxies. I dare them to provide a clear case! They went out on a limb (pun intended) when they got into this fray. Trees just ain’t that simple. Neither are you.
Re: jae (#18),
Addendum: with the millions of $ these guys have gotten over the last 20 years, they could at least have done some simple greenhouse studies to prove that temperature is a critical parameter affecting ring width or MXD. Come on.
Re: jae (#18),
jae, would you shortly explain what´s wrong with Grudd and his Torneträsk study about tree ring correlations with temperature:
Click to access Grudd%202008.pdf
Re: Andy (#23),
I like the Grudd study, because it gives results I like. 🙂 But seriously, maybe trees in northern Scandanavia ARE good temperature proxies. As I see it, the important confounding variable in tree ring studies is moisture stress; and it may be possible to show that that is not an issue there. But that (demonstrating that other variables are not overshadowing temperature) is what needs to be done for ALL series selected for a reconstruction. It looks to me that the selections are more likely to be done by eyeballing the results to see if they match the pre-conceived relationship. That is poor science, IMHO.
Also, I have reservations about solving the “divergence problem” by replacing data from old trees with that from younger trees, as is done in the Grudd study. Why is that reasonable? And a question for Rob or other dendros: There are good reasons for standardizing ring width measurements, because of the well-known and easily explainable decrease in ring width with age. But what is the rationale for standardizing MXD? Why would it change with age?
Re: jae (#18)
Looks like you got me wrong. IMO the main problem of dendro-/palaeoclimatology arises from the vast number of degrees of freedom. Inevitably to come to conclusions much is left to deliberate judgement – which has taken place.
Now giving the field a comprehensive set of authorative and tight rules and procedures much of the pasttime work would be in danger of having used ‘bad practices’. In addition, the field would be somehow plutoed from creative research to mere engineering. Thus my conclusion that if it will happen we will go without panem anytime soon but will at least be provided with lots of circenses.
I have to say I would dearly love to see you or Jeff Id or someone else with the talent and patience to submit a quality analysis of the Mann proxies.
Do it once with all the proxies Mann used, but without arbitrarily deleting post-1950 data, without infilling, without pasting fake data to bring the series to 2007, without bad PCA techniques, and without creating bogus hockey shapes at the end.
Then repeat it with a more suitable proxy criteria, dumping the Luterbacher and other temperature records and including updated dendros like the various Siberian series.
Send THAT to the same corresponding editor at PNAS and see where the chips fly.
Re: Clark (#7),
The seeming lack of untainted original data for replicating things might make such an effort a tad difficult to accomplish.
Maybe dendros feel it’s easier to pick on retired statisticians rather than well-funded, outspoken climatologists who get a lot of media time. Though the point of your post seems to be, if they’re attacking you directly, then they’re attacking Mann et al indirectly for doing the exact same thing.
This story has more twists and turns and sub-plots than even the Wire.
My favorite quote has to be
Fantastic. Now could you just point out the scientific rationale as to why this is so instead of telling us that those really in the know in the dendro community have this knowledge by seemingly supernatural means? And it appears that some work has been done in defining a relationship between proxy and precipitation that seems so hard to do when its temperature. Equations. Errors. Anyone?
I have yet to see an equation for how a tree grows past Growth = A B exp (C) and have never seen a paper make a stab at explicitly trying to empirically define a dendro proxy like a physicist would. Maybe that’s the point. Maybe that phrase I heard once that “Other science slips on a sliding slope from the height of physics” is more than just elitism. Maybe the physicists should tackle this instead. Its Harry Hill time..FIGHT!!!
I find it pretty rich that SMcI is getting criticized for picking the wrong trees. It isn’t like he put them in a hockey stick himself after all, the other guys do that. It’s also not like Mann group is doing such a great job picking out super tree, they have to chop the data off the end and paste on new data.
On one link above the Rob discuss one group of MXD trees that makes a somewhat similar shape to temp (for a very short time) and they act like they are holy grail thermometer trees not just a lucky spurious correlation.
I understand now that dendros imagined physical location makes some special trees temp limited but the results of their efforts have not IMO been particularly fruitful and has led to the reasoning behind the clearly false method of correlation selection. I think Craig Loehle has pretty well nailed this one down with his last paper, I hope everyone has a chance to read it, it’s simple and quite clear.
Trees are great but they make lousy thermometers.
Perhaps you might have a thread to answer the question – When it comes to Dendros, their Bark is worst than…..???
their pith?
Re: bernie (#1), Bernie says:
I.e. Pithing into the wind?
If tree rings are going to be used as temperature proxies, someone has to (and I repeat SOMEONE) has to show these trees actually respond to higher temperatures by increasing ring widths. I haven’t actually seen any such study.
And you can’t just core drill a tree and tell anything about ring width at all because almost all trees do not (and I repeat DO NOT) grow perfectly concentric rings. The centre of the tree is always weighted to one side or the other and ring widths vary from side to side over time. One drill core from one side of a tree will have a completely different signal than a drill core from any other side.
Drill a core from the East side of the tree and the ring-width will say it has been very hot here for 400 years. Drill a core from the other side and it will say it has been very cold here for 400 years.
Here is nice cross-section showing a (previously) 400 year-old Ponderosa Pine which has been subject to repeated fire burns. No drill core could tell you anything about the temperature history of this particular tree. You have to cut it down and do a cross-section to know anything at all. Even then you still can’t tell anything. Show me one drill core that provides a reasonable level of confidence about temperature conditions or even growing conditions after viewing this cross-section of this particular 400 year old Ponderosa Pine. Show me ONE cross section which provides a reasonable temperature signal. It totally depends on where you drill core the tree and obtain your sample from.
http://www.ncdc.noaa.gov/paleo/impd/firescar-photo.html
Steve: This is the sort of issue that dendros are aware of and can to a considerable extent be dealt with by replication. It’s a lower order issue IMHO but is the sort of thing of thing that a visiting dendro would select to debate while ignoring other issues.
Re: John Lang (#17),
But, now this is going too far. Burned trees and strip-barked trees, yes. But trees that have not have had such trauma still have the POTENTIAL to tell us something about temperature. Get away from the fixation on rate of growth on different radiaii. You want to look at the relative growth on one radius. BUT, you need to go to the pith to do that, and many of these folks have ignored that.
The point about tree rings has been made hundreds of times here. Enough piling on.
It’s tangential to the issue here – the failure of dendros such as Peter Brown, Pisaric and Wilson to hold Mann et al to the standards that they advocate.
Steve,
It’s all a matter of asking the right question:
http://www.liveleak.com/view?i=f2a_1223683039
Their bark is worse than their saps??
Andy #23,
The problem with Grudds’ Tornetrask series is obvious. The Warmers have assured us that the current warming is higher and faster than at any time within the last 1000 years. Ergo, the study is poorly founded!!! Proof of this comes from Mann et alls…
I guess Mann and other dendros are aware of very recent findings at the Uni. of Penn:
http://www.sciencecentric.com/news/article.php?q=08061131
“The research, published online in this week’s Nature, contradicts the longstanding assumption that temperature and relative humidity in an actively photosynthesising leaf are coupled to ambient air conditions. For decades, scientists studying climate change have measured the oxygen isotope ratio in tree-ring cellulose to determine the ambient temperature and relative humidity of past climates. The assumption in all of these studies was that tree leaf temperatures were equal to ambient temperatures.
Researchers at Penn, using measures of oxygen isotopes and current climate, determined a way to estimate leaf temperature in living trees and as a consequence showed this assumption to be incorrect.
This is an unfortunate finding for the potential to reconstruct climate through tree-ring isotope analysis but a boon to ecologists because it creates potential for the reconstruction of tree responses to both average climate and climate change over the last couple of centuries.”
Nature is full of surprises for the theoreticians. It pays to check those initial assumptions because what seems obvious isn’t necessarily correct.
Unfortunately, due to the “millions of $” I am paid, I must remain silent.
Baaaa!
Re: Rob Wilson (#26),
That’s actually pretty funny, but truly, we know it’s not the money.
=========================================
Re: Rob Wilson (#26), Rob, I’d love to hear more of your opinions and I’m sure that I’m not the only one. There are lots of us who choose to lurk and listen more than we comment.
Steve
Is this something for you?
Click to access Grudd%202008.pdf
Steve: I’ve reviewed this article and applied it in my Erice presentation, for example.
Just think, Rob, if you and your brethren were actually as even-handed as we would like, applying your criticism equally to all science you deem flawed, not just to that with conclusions you don’t like,
snip – that sort of comment doesn’t need saying
Steve,
IPCC Fig 6.10b has several post-TAR papers, RMO2005, DWJ2006, MSH2005, HCA2006… and the data for these is on Tim Osborn’s website as you recently pointed out. Are you saying that these are all rehashed old data? Is this discussed anywhere? It is sometimes difficult to see the ‘wood for the trees’ on your blog (sorry).
The “Studies since the TAR…” sentence is one of the very many grossly misleading statements in the SPM, one which coincidentally I have just written a page about, though my amusement is more with the “increased confidence” and “coherent behaviour” rather than the “additional data”.
Steve: I understand the wood and trees problem. But the re-hashing of old data is something that’s been a very regular theme. Rutherford, Mann et al 2005 (RMO) used the data sets from MBH98 and Briffa 2001 (both reported on in TAR). The Moberg proxy recon ended in 1979 and virtually all of the underlying data was pre-TAR in origin. The Hegerl recon ended in 1960 and again virtually all of the underlying data was pre-TAR in origin. The D’Arrigo, Wilson recon, which received far less attention than the others, is the most meritorious of the bunch and has some new data. Unfortunately it’s hard to come to grips with since there was no data archiving at the time and while D’Arrigo and associates have subsequently archived a considerable amount of data, it’s still spotty. For the medieval comparison, the proxies used in D’Arrigo et al are essentially identical to Briffa 2000. You see the same old Graybill bristlecone chronologies over and over, the Briffa 2000 Yamal study over and over.
We have one more example of a visiting dendro canvassing a thread for a weak comment and using that as an excuse not to respond to the substantive issue. I wish people would bear that in mind before making irrelevant piling-on comments.
Truthing dendro studies is a meritorious enterprise, but, as I noted above, it has nothing to do with the silence of the lambs – so why bring it up? Now Rob goes away thinking that he’s made a clever retort. It’s good spirited but has nothing to do with the substantive issue – that practising dendros like Rob refuse to muck it up with anyone on the Team, should they abuse a dendro canon. But he’s not going to admit here or anywhere else.
But he knows that my point is valid. And he knows that I know that he knows that it’s valid. But cone of silence will hold.
The below from Wilson’s dendro list epistle reiterates to me what Wegman, Steve and others have been saying for years: that these folks don’t understand statistics and are afraid to expose their “finished” work to examination by competent statisticians. This is incredibly shortsighted and completely unnecessary. Hire a competent statistician before you write the paper! Or does their fear also extend to their “unfinished” work?
” relevant basic literature and I was wondering if it would be worth while putting a simple web page together with links to relevant PDFs with regards to sampling strategies, data processing, calibration and verification methodologies etc. Some links to some case study examples might also be a good idea, although that may lead to an ‘audit’ of these studies on the CA Blog.”
Rob’s prepared to call me out on Meko spruces in an incidental post (even expressed conditionally), but not to call Mann out in a major study for matters where Rob would vehemently disagree with Mann’s proxy selection.
Rob’s real reason is simply that he doesn’t want to get into a dispute with Mann. Why would he? That’s why so many issues in the field never get resolved.
Re: Steve McIntyre (#36),
I agree. Rob knows that getting into a dispute with you carries few, if any, consequences, plus, I’m sure he understands you’re light-hearted enough to take things in stride. The same cannot be said for a dispute with Mann. He’d quickly become “the dendro fool that agrees with that Canadian amateur.”
Mark
In some fields of science, a good debate is welcomed. In other fields, and especially when it involves political correctness, to question orthodoxy is to risk career suicide. So it is not surprising that in the small field of dendros, where there are some big fish like Mann, that everyone sticks to their knitting and doesn’t start a fight. Can’t blame them. Often in such situations, it is outsiders who overturn the prevailing paradigm.
Re: Craig Loehle (#37),
But Craig this is the Internet age, in the past it was fine for scientists to sit on their hands and say nothing, the chance of I or anyone trawling through shelves of books, reading, noting, cross-referencing were frankly less than sero. But as the lies begin to unravel I or probably half the planet could google and find people who knew but kept quiet, the world has changed, I can well imagine Rob, realising the benefit, but not the downside – it’s the problem with always looking to the past. Even now a google on “rob wilson mann”, should have them quaking, they know about sloppy science abusing their craft and kept quiet.
Ian
#37, Craig Loehle,
Sadly, I expect the silent few will go down with the ship also, if it gets overturned. Perhaps that is rightfully so?
What happened to Stan? I will answer your question anyway. It is not a matter if the dendro science is incorrect or if it is the reasoning that is incorrect; they are not mutually exclusive. Apply proper, truthful and consistent scientific method and reason comes out of it. I would expect to see any science done this way.
So all the what ifs begin to fall into a simpler truth which may be something that can be unpalatable but if the assumptions, method of taking measurements and results/analysis are fair then this it. This is one of the main reasons to have this blog. We can’t deny an effect (AGW) if there is no consistent data as to what we are denying. So in relation to this post it seems quite odd that noone seems to have stood up and said ‘Hey Mann (sic) you can’t just go cherrypick a few new proxies from the vault, spin some new temperature anomaly without seeing what we think of it. We told the other guys that that wasn’t on so its not okay for you to do it – snip- ). Its just so inconsistent’.
Re: MC (#41),
MC,
Apparently my question got snipped out of existence. I know my reference to sloppy science and sloppy reasoning was not very respectful of the dendros, but there was a serious point underlying it. Perhaps Steve will allow me to make it without snipping.
snip – Obama and McCain are, for the present, prohibited words here.
We see this same type of sloppy reasoning in the dendros’ response to the legitimate questions raised about their work. Attacking the questioner does not rehabilitate sloppy science. If the emperor has no clothes, it matters not that the first to raise the point is a small child. Unfortunately, it often takes someone unconnected to the emperor’s power structure to be willing to raise the point. And note that the child’s questioning of the emperor would have raised a point every bit as valid had the child been blind.
We see the dendros attack Steve’s motives, his profession, etc. But they fail to realize that doing so does not resolve the deficiencies in their work. There is a defect in reasoning here. It is sloppy. And looking at their work, that same defect seems to affect their studies.
snip – enough piling on.
Theres a lot of squealing on the RealClimate site. Not the sort of thing one would expect from professional scientists.
I wanted to leave a post suggesting they re-read the Wegman et al report before they create any more graphs and to suggest maybe they should also employ a statistician as the reports recommendations. Unfortunately I could not see where to post.
Maybe you guys could recommend someone?
#22 — Andy, the problem with Grudd’s Tornetrask dendro study is the same fatal problem as for all dendro studies, which is that statistics is no substitute for physics.
Trees are only qualitatively judged to be temperature limited by reference to external indicators. There is no independent physical metric for temperature limited growth and no way to derive a quantitative temperature from a tree ring.
To justify a quantitative metric by reference to a qualitative judgment of temperature limitation is a scientific charade. Statistical renormalization of TR width or MXD to a measured temperature series cannot ever produce a true paleo-temperature metric or series.
The entire field of tree ring thermometry is a pathological abuse of science; in the vernacular, it’s a crock.
I for one am curious about how dendroclimatologists feel about the M08 proxy choices. I have been wrong plenty of times in my life and am quite willing to be wrong now if it improves my understanding.
I got snipped a bit. Yay! I can see why, its okay. Anyway the details may be off-topic but I want to tell a story about how there can be an inertia in changes to scientific thought and how this related to a field of work I was studying in. Its relevant because I think that some dendros may indeed start to question the Mann 08, but they may just feel a need to have more substantial evidence…
I grew thin film ceramic superlattices for my PhD and the main gist was to try and artificially enhance the dielectric properties of different ferroelectric films by layering them finely enough. Now, and here’s the crux, a paper by a guy called Hitoshi Tabata in Applied Physics Letters (you can go look it up, 1994 I believe) showed that there was enhancement over naturally occurring solid solutions (chemically obtained mixtures) when you layered the films. This kicked off a whole area of research around the world including research at my university (Queen’s University Belfast).
Now some colleagues of mine tried to repeat the results using the same materials; I used different materials to see if the method was repeatable. After much attempts it became clear to them that there was an apparent enhancement but that it also produced a very large dielectric loss. When you measure capacitance (dielectric constant in essence) the LCR meter does not distinguish high dielectric constant from an intrinsically good capacitor and essentially a bad conductor. You have to look at the loss. Which by the way you can never not read off the LCR meter. It always gives you capacitance and loss. You would think that people would always report the capacitance/dielectric constant and the loss. Well (you can see it in the paper) Tabata was not specific enough. It turned out it was very hard to get an intrinsic enhancement using those materials and then what was seen was an effect over 100 years old.
But now at this stage people are not coming out and saying look this result wasn’t correct. It just wasn’t said. So when I did my work (on similar materials called relaxors) and actually showed that you could get an enhancement of the dielectric constant and still have low loss, it showed that Tabata was on to something but he hadn’t done the science correctly. I don’t believe it was intentional but maybe his focus was on the trees of detail and not the wood of a more mundane emprical reality. In retrospect you can say this because of the simple fact: why aren’t you reporting the loss as well as the dielectric constant. This then became the norm. But it took about 10 years!
(My paper is M.H.Corbett et al, Applied Physics Letters, Vol 79, 815 (2001))
So I expect that at some stage someone will realise that maybe you have to be a bit more exact in your assumptions and methods with these proxies because their Spidey sense has learned how to tingle. But it takes time. What is frustrating is that it takes no time at all for most people who post here.
I just noticed on Real Climate that Tamino says he is a “professional statistician.” Does anyone here know that to be true?
It is really immaterial what he says he is, what he actually is, or what his background is. If he’s using that claim to add some validity to his comments… shame.
Mark
Of course, Ian, the massive hit count also lends an air of anonymity. This is particularly so when most people just don’t care about the details, at least not to the same level as this blog does.
Mark
Re: Mark T. (#50),
True, but as we’re talking to one of my areas of expertise, sadly for those relying on obscurity – I have some really, really bad news. Googling as I’ll refer to it is 1st generation, think about the progress from the pony and trap to your current mode of transport (true greenies please leave at this point the “internet” wastes more power than you want to think about). We know how to get rid of the garbage and how to make the real connections, the search engine of the future will not be the dumb index of today, it will progress faster than they can think of excuses for inaction.
Ian
Well, I re-read the link to Steve’s previous post on the matter, and I realized that the silence is deeper than I thought. I had asked Rob Wilson a question twice on that link that never got answered. (He had answered other questions of mine, and for me, is always a welcome participant here.) This was the second time I asked:
Unfortunately, it never got answered. Now, I can understand the dendros calling Steve a paleocriminal for not following the recommendations of the relevant authorities in the field regarding site selection … but just who are those authorities?
Yes, I know that site selection is one of the “fundamental principles” of dendroclimatology. On Grissino-Mayer’s excellent web page he says:
Thats fine as far as it goes … which, unfortunately, is nowhere near far enough.
So, dendros, here’s your chance to save a life in peril, an opportunity to save a youth from a lifetime of crime. Where can I find the relevant authorities whose rules will keep me from being a paleofelon like Steve?
w.
Re: Willis Eschenbach (#51),
Willis, I always had the impression that there are some general criteria for selecting sites, but in the end, I seemed (perhaps incorrectly) to get the impression from Rob Wilson’s comments that what was critical was selecting those trees that gave a reasonable correlation of MXD and/or TRW to maximum or average temperatures for the months of the year that gave the best correlation for the calibration (and validation?) period– and this was apparently not a priori, but after sampling the trees.
Without some recorded understanding of what is critical in the basic science for capturing a temperature signal from TRW and/or MXD and applying it to an a priori criteria, I would think that the potentional for data snooping would be considerable.
#52. Ex post picking has always been a bone of contention with Rob. At least he is aware of the issue, unlike most dendros who simply view this as a way of life and not as an odd and hard to justify practice.
Re: Steve McIntyre (#54),
If you say so Steve M, but then we have:
Re: Rob Wilson (#60),
The impression that Rob Wilson seems intent on imparting here is that he lacks time, interest and motivation to criticize Mann et al. and that might be so. I personally view his comments on cherry picking as similar to those of the developers and adherents to investment strategies that have had a difficult time seeing the statistical issues with developing their strategy with in-sample data that had great potential for being data snooped. Do the reference books Rob listed cover the dangers of data snooping? The replies are too oblique for me to interpret properly as evidence by Rob’s, “It just means there is room for improvement.” That criticism of Mann et al., if indeed it was, would I guess not ruffle any of the authors’ egos, but appears weak on specifics, albeit, a step up from baaa, however incrementally small.
Steve, you say:
I agree, and it’s a true oddity in climate science, a field filled with oddities. I know of no other field where people routinely claim that they can get significant results simply by throwing out data that they don’t like for some reason. In the dendro’s case, the reason is often “because it doesn’t contain a signal” … so what? If I sample 50 people to find out if a given medicine is any good, I can’t just throw out the patients whose vital statistics show no “signal” from the medicine … what’s different about trees?
There is also a terrible disconnect between throwing out proxies because they don’t pass some comparison to temperature, and at the same time allowing in proxies that have no possible connection to temperature. Yes, the Tiljander proxies rise in the last century, so they pass the “Pick Two” test … but a third grader would look at them and say “What’s that funny rise at the end?”, and a high school student would say “How can we use that, there’s no temperature record that looks anything like that”. (Of course, a scientist would say “These are stated by the original investigator to be contaminated post 1850 so we can’t use them” … but I digress.)
Finally, the “Pick Two” test will allow the inclusion of about 20% of false positives … how can that be justified in any study? Sometimes, I feel like I’ve wandered through the looking glass to some alternative universe ruled by different statistical rules …
.
w.
.
PS – Or to misquote “Alice in Wonderland” …
PPS – Am I getting it backwards, or do these folks have it backwards? Seems to me that temperature is the independent variable, and proxy response to temperature is the dependent variable. Aren’t they calculating it the other way? What am I missing here?
Some thoughts in this op-ed for supporting some posted thoughts.
and
John Lang, #16, provides the following photo of a fire-scarred ponderosa:
What could have caused the astonishing growth spurt after 1742? (Aside, that is, from an otherwise unknown Enlightenment Warm Period…) Did the fire of 1742 eliminate all competition for light and allow this survivor to explode? Or did the fire fertilize the ground with ash?
Re: Hu McCulloch (#58),
Yes, this is the type of metadata that should be included in the forthcoming “Guide to Selecting Tree Ring Series for Temperature Proxies” It is very clear in this case that there was a big loss of competition in 1742 here. And this is almost always a function of MOISTURE, not temperature or sunlight. That’s why competent foresters thin forests! I am convinced that most tree ring series are precipitation proxies, not temperature proxies. ESPECIALLY bristlecone pines.
Craig Loehle 37
“Often in such situations, it is outsiders who overturn the prevailing paradigm.”
Not quite correct. It should be:
“Often in such situations, it is COURAGEOUS, NOBLE and AUDACIOUS outsiders who overturn the prevailing paradigm.”
Regards,
Perry
Greetings CA,
Flippant comments aside, I feel a short serious comment is needed. I will not reply to any more comments, because frankly, I have better things to get along with.
Ok.
Fact 1. Earlier this year, I did criticise Steve for randomly averaging some tree-ring chronologies together without prior assessment of the time series. I stand by my criticism.
Fact 2. Steve has criticised my colleagues and I for “cherry picking” data. Fair enough. I would personally not use a precipitation proxy to try and reconstruct temperature even if there was an inverse correlation between the two parameters and/or some known spatial relationship teleconnecting such parameters.
Fact 3. Mike Mann, in his latest study has tried to collate as complete data-base as possible, and then screened the data for use in his reconstructions. This is the opposite to “cherry picking” in my mind so give him a break. Admittedly, there are some strange omissions in some of the proxy records he chose to consider prior to screening and of course I am particularly miffed that he did not use some recent records I have been a co-author upon in recent years, but that was his choice and I will not lose any sleep over it. It just means that there is room for improvement.
Fact 4. I am not going to publically criticise other peoples work. I am quite busy enough being critical of my own work. Please read the DWJ2006 paper. We were quite clear that we feel that replication was poor prior to ~1400 and that there are some calibration issues that need to be addressed. These are weaknesses that stimulate continuing research which we, as a community, are thoroughly engaged in rather than spending endless hours blogging.
Fact 5. For those with a continued poor understanding of the basics of dendroclimatology, please read the following two books.
Fritts, H.C. 1976. Tree Rings and Climate. London: Academic Press Ltd.
Cook, E. R. and Kairiukstis, L.A. (eds). 1990. Methods of Dendrochronology: Applications in the Environmental Sciences. Dordrecht: Kluwer Academic Publishers.
Rob
Re: Rob Wilson (#60),
Hello Rob —
I wanted to thank you for your response. I tend to agree with Steve’s criticisms here, but I sincerely appreciate your straightforward and honest intent. Thanks for taking the time to comment, and I hope you find time to keep contributing in the future.
Re: Rob Wilson (#60),
Fact 1: Criticizing Steve for “randomly averaging some tree-ring chronologies together without prior assessment of the time series” seems like helpful in getting Steve’s work less open to attack. One can imagine that means others are beyond attempting to help. If not, perhaps the others might benefit also from assistance such as this.
Fact 2: I take that as you saying you wouldn’t use precipitation proxies for temperature reconstructions even if there was an “inverse correlation between the two parameters and/or some known spatial relationship teleconnecting such parameters” means the White Spruce doesn’t have either of those going for it.
Fact 3: This seems to be “I wouldn’t have done it that way and it’s not perfect.” Okay. However, the independent observers here seem to have a consensus that ex-post screens for temperature correlations isn’t valid. Invalid is a bit different than imperfect.
Fact 4: I will assume you mean you won’t criticise other people’s published work, and only use constructive criticism on blogs. If so, seems reasonable. Perhaps the manner in which things were carried out could be improved in the future, however.
Fact 5: How about those that do understand the basics, can they read those books also?
Re: Rob Wilson (#60),
First, as others have already done, I offer you my thanks for coming here to say what you think. I hope you come back often and try to engage the people here. Some will not appreciate it and may give you a hard time, but others will appreciate your effort; in neither case will you get a free ride. That, however, is the real benefit to you, but only if you can manage to see it that way.
An example of this is Roger Pielke Jr.’s comment #61 above: what seemed so straightforward to you that you could write it out in just five facts has a basic inconsistency in it that you most obviously were not aware of. Had you not posted, you would not be aware of this. To my way of thinking, seeing it like that is valuable. The only reason that I write this is that I suspect you feel the same way. In short, if you can manage to not let the “sportsfans” get to you, I benefit from your presence in this forum and I do believe you will as well.
Rob-
The divergence problem in your facts is what has many observers in academia outside of climate science, but who pay attention, wondering what is going on in this community. Specifically:
Fact 1: “Earlier this year, I did criticise Steve . . .”
Fact 4: “I am not going to publically criticise other peoples work.”
Is Steve not a “person” who does “work” in this area?
What I believe is troubling in this situation is an understanding for the double standard. Such a double standard is of course not exclusive to this area, but it does suggest that extra-scientific factors play a (large) role in the process of evaluating knowledge claims. It would be a strange turn indeed if publication via peer review means freedom from criticism;-)
Steve should be publishing in the peer reviewed literature, to be sure. But the fact that he has chosen not to does not mean that his critiques or analyses do not exist. There is probably a middle way.
Re: Roger Pielke. Jr. (#61),
Steve is, of course, not of the fraternity so he’s fair game by the fraternity’s lights.
Re: Rob Wilson (#60),
And when did criticism stop being part of the process by which science is done? Afraid you might hurt someone’s
research grantsfeelings? It’s part and parcel to the the process.Re Jae #58, Thanks for the explanation of the post-1742 growth surge in John Lang’s ponderosa photo!
Might the characteristic BCP stripbark growth surge then be due to the problem that caused the trauma to the surviving tree (fire, infestation, or whatever) killing off neighbors that compete for moisture? I had previously been assuming that it must be due to the tree-specific factor of an increased ratio of root to foliage after the trauma.
If the trauma was a lightning strike, could it have been a direct and lethal strike to a neighboring, competing tree that merely damaged the tree in question?
Re: Hu McCulloch (#63),
Hu, the BCP strip-bark growth surge probably has nothing to do with competition. It is a whole different deal. I would surmise that when part of the bark is destroyed (by the processes Steve mentions), those cambium cells that are still alive get greater concentrations of photosynthesis products produced by the foliage, so growth rates are increased in those areas.
Incidently, it has been many years since I read Fritt’s book, but I will wager that he cautions against using tree ring series that show such abrupt changes in growth rates. They are always related to sudden events, like fire and disease. Need to read that book again if I can find it.
Nice point, Roger. I think, Rob, you laid bare your true level of objectivity in this area.
Mark
Rob Wilson, #60, writes,
One cherry-picking aspect of the new paper that Steve has validly called attention on this blog is its “pick-two” procedure, of selecting the better of two correlations, and then applying single-correlation critical values to claim “significance”.
I tried applying Mann’s pick-two procedure to my golf game last weekend, with sensational results: Instead of just playing my first tee shot as customary, I tried taking two tee shots on each hole, and then playing the better of the two balls. This was a lot more fun, and enormously improved my score!
As long as I’m just trying to have fun and get some exercise while supporting global-cooling greenspace, no harm is done by making up my own rules. However, it would be wrong for me to enter a tournament and claim that I was the winner unless I was playing by the standard rules.
It is similarly misleading for Mann et al 2008 to claim that there is a significant correlation between a proxy and local temperature if the better of two nearby series has been cherry-picked.
Steve has identified a number of other important problems with the new paper that I hope he will report in a letter to PNAS. (Reminder — letters commenting on PNAS articles must be submitted within 3 months of publication of the original article. Otherwise, all errors in PNAS are unassailable there. Letters must be excruciatingly brief, but can be backed up with a PNAS-archived SI.)
— Hu McMulligan
#63. Hu, Pete Holzmann observed some incipient stripping – there are some pictures somewhere: a limb of the tree collapsed tearing the bark off below the limb, but not killing the tree i.e. “when the bough breaks”.
He thought that heavy snow might cause some boughs to break, but there’s lightning on the mountains as well. The 1840s seem to have been particularly severe in Colorado – heavy snow and severe winters are blamed in some articles (Brunstein) for much of the bison disappearance. So one could have a situation where there was a disproportionate incidence of physical stripping in the 1840s, leading to a growth pulse in the surviving portion. Making a physical model for observed RWs is daunting.
ANyway Pete’s thought is that the trauma would be specific to a tree, but that the incidence of trauma might not be time-independent amont trees.
Cherry picking and ex post picking are related sins – we haven’t really defined either one.
Our usual concept of “cherry picking” is somewhat more manual – e.g. D’Arrigo picking the old Cook version of the Gaspe series with its big HS rather than an updated version without the HS. Or the multiproxy authors choosing Briffa’s Yamal version rather than Esper’s Polar Ural update. Or Juckes not using the Indigirka River series. Or using Graybill’s Sheep Mt rather than Ababneh’s. Or Briffa’s Tornetrask rather than Grudd’s. In small networks of 6-18 series, even a few such choices “matter”.
Mann 2008 is admittedly more of an example of “garbage can picking” rather than “cherry picking”, but that is hardly an improvement. Ex post “screening” for temperature correlation is not a valid procedure.
The only dendro procedure that seems to make sense to me is to what Schweingruber appears to have done but failed to properly report – select a large network of sites chosen ex ante as temperature proxies according to dendro criteria and report large scale averages. Of course, that way lies divergence. If the dendros learn in the study that their ex ante criteria need to be modified because they learned something about qualifying sites, then they need to collect fresh data seeing if the new new ex ante criteria work. You can’t just ex post pick series that go up.
Of course Mann excluded all the Schweingruber ring width sites from Mann et al 2008 – something that Rob hasn’t mentioned.
Re: Steve McIntyre (#70),
It seems a particularly fine liqueur is being distilled from the aged garbage in those cans.
===================================================
Rob,
Thanks for braving the storm, it takes a bit of guts. I will look for your book references.
Please reconsider your point 3. The math behind sorting distorts the actual temperature curve de-amplifying historic trends such that the true temp scale of the plot does not have a linear shape and is distorted both in time and amplitude. This type of screening, correlation should be rejected.
I wasn’t very clear.
I meant to say distorted in amplitude and offset in a manner which varies with proximity to the calibration period.
Sorry, it looked great the first time I read it.
#60 — Rob Wilson wrote, “Fact 5. For those with a continued poor understanding of the basics of dendroclimatology, please read the following two books.
“Fritts, H.C. 1976. Tree Rings and Climate. London: Academic Press Ltd.
“Cook, E. R. and Kairiukstis, L.A. (eds). 1990. Methods of Dendrochronology: Applications in the Environmental Sciences. Dordrecht: Kluwer Academic Publishers.”
Neither of those books, nor any other book, can provide a way of obtaining a physical metric from a purely statistical method, nor a way of deriving a quantitative measure from a qualitative judgment.
There is no holy writ in science, Rob. What’s being represented as proxy thermometry is an abuse of Fritt’s methodology.
Rob, thank you as always for your contribution here. I’m not sure if your “Fact 5” is in answer to my question on site selection for temperature proxies, but in any case, it is much appreciated.
You said above:
Having read Fritts 1976, I knew that he had very little on site selection for temperature reconstructions. His main interest was selection of sites with a clear moisture signal. I had great hopes for the Cook and Kairiukstis text. Unfortunately, they say little more than Fritts. (emphasis mine)
OK, climate signals will be stronger at the fringes, I agree with that. I also agree that there is climate information in plant growth variation. But the devil, as always, is in the details. Which climate information is in the variation of which tree rings? Soil moisture? Nighttime temperature? Timing of frosts?
Since water and temperature are usually the two biggest variables in plant growth, to find the temperature signal, we should look in places where there is plenty of water. That makes perfect sense.
Unfortunately, this advice runs directly contrary to the modern practice of using sites near elevational treelines. These treeline sites are often quite arid, both in the atmosphere and in the soil. This means the moisture signal will be quite large. Thus they are a very poor ex-ante choice for extracting a temperature signal.
In fact, all arid and semi-arid regions are not good places to be looking for temperature information. The moisture force is strong here, Luke …
I’m sorry, but I have to call “hand-waving” on that statement. You can’t say that there are statistical implications in the method, and follow that by saying it’s OK because of the “principle of limiting factors, the concept of ecological amplitude and replication”. There are interesting and important statistical consequences of the type of sampling that you are doing. You can’t vanish them by invoking the “principle of limiting factors”.
Next, the authors say:
Excellent advice. It doesn’t help us much regarding Mann 2008, except to warn us about catastrophic factors in mountainous regions, and to caution about the difficulty of establishing “dendroclimatic relationships” due to microclimates in the mountains. I note also that while the authors discuss establishing dendroclimatic relationships, authors like Mann often make no such attempt.
Excellent advice. But again, not much help with our current question.
This last advice (which completes their discussion of site selection) brings up an interesting question. Competition is in general a huge variable in the growth of trees in forest systems. How can we disentangle that?
Which leads back to the problem as I see it. Tree rings in many parts of the world contain a common signal. That’s the lure, that’s the hook, that’s the gold that drives scientists crazy. And it’s true, the signal is there.
Unfortunately, the temperature signal in tree rings is muddy both intrinisically and extrinsically. It is muddied extrinsically by the presence of a host of other confounding signals – frost timing, length of growing season, atmospheric moisture (humidity), amounts of mist and clouds, changes in soil nutrients, catastrophic excursions (fire, insects, landslide), CO2 levels, rainfall amount, timing of warm days, competition from other trees, wind strength and direction, date of last snowmelt, changes in insolation, rainfall timing, the list is very long.
In addition, it is muddied intrinsically by the inverted “U” shape of the trees’ temperature response curve, which means that both high and low temperatures make narrow rings.
To me, the steps involved in trying to extract the temperature signal from this morass should look like this:
1) Select sites based on some ex-ante criteria. Since according to the Cook reference the sites should be “where trees do not have a limited water supply”, this would in many cases rule out elevational treelines.
2) See if the sites contain a common signal.
3) If they do contain such a signal, see if it can be split into its component parts (e.g. temperature, moisture, competition, catastrophic events).
4) If the temperature signal can be disentangled from the other signals, adjust it (as far as possible) for the inverted “U” shape of the response.
5) Have the balls to put an honest error estimate, not just a statistical error estimate but an estimate of the overall error in the entire process, on the results.
.
Sadly, I fear that this is not what passes for dendroclimatology these days in many parts of the discipline …
.
w.
PS – Rob, I truly wish you would reconsider your stance, viz:
While that sounds all noble and all, and you present it as though it were a valid stance for a scientist to take, it is the exact opposite. We can show this scientifically by “reductio ad absurdum” as follows:
If every scientist took your position, and no scientist ever publicly said a critical word about another scientists’ work … please explain to me how this would do anything except bring science to a swift and grinding halt.
Q.E.D.
So I fear I don’t care how much you criticize your own work. Science doesn’t advance through people being critical of their own work. And criticizing your own work does not relieve you of a scientist’s obligation to advance science by pointing out the flaws in the reasoning and practices of others.
It is this reluctance to speak out that is slowly killing both dendroclimatology and climate science in general. You guys are all are sitting around pretending that the other guy’s … don’t stink, when everyone knows the room is reeking … and then you want us to believe your claims?
Sorry, Rob, but to be believable, first y’all have to believe in yourselves enough to stand up for what you believe in. I don’t mean just you. I mean that the silence of all of you climate lambs is what is making climate “science” the laughingstock of the scientific community.
(Latest joke making the rounds. A dendroclimatologist parks in a spot clearly marked “Handicapped Only”. When his friend asks him how he gets away with it without getting a ticket, he says “I just leave a copy of my latest peer-reviewed study hanging from the mirror.”)
Take a look at Linah Abeneh’s thesis, and the huge furor that it caused, and the uproar when it never even got published, as a quintessential example of the process I’m talking about.
What, there was no furor about her thesis? There was no uproar?
My point exactly. You guys are slitting your own throats, which is surprising and depressing in itself.
But what is truly sickening is to see you congratulating yourselves on how quiet the process is, declaiming how wonderful it is that you’re not publicly criticizing anyone, and remarking on how little noise there is as throats are slit and reputations die and science is hacked apart … as if the lack of noise, the lack of public criticism were some kind of shining, wondrous thing. The silence of you lambs is not noble, it is the silence of the lambs going to the slaughter without demur, without protest.
My quest for the elusive site selection criteria continues. In the book “Surface Temperature Reconstructions for the Last 2,000 Years (2006)” from the National Academies of Science Board on Atmospheric Sciences and Climate, the authors say:
So it seems like the question has been raised, answered … and forgotten. The NAS BASC knows that it is critical to ensure that sample selection (at the site, tree, and core level, please note) is “based on a priori rather than a posteriori criteria”. But individual practitioners are still defending a posteriori criteria … or no criteria at all.
And getting this information on site selection guidelines from the dendro community is like pulling teeth. Nobody to date has mentioned those two references … but I’ve dug them out, so I’m off to read some more.
w.
Re: Willis Eschenbach (#79),
Willis, I have been saying this for over 2 years. Right on. I’ve never seen a branch of science where it’s permissible to bias your sample selection by cherry-picking the way the dendros do. And when they actually admit in public that “you have to pick cherries to make cherry pie,” it is especially repugnant. The current dendro-science methodologies are truly pathetic, considering the most basic tenents of scientific pursuit. And thanks for your great posts on this matter.
OK, Steve, I suppose I’m “piling on,” but dammit I feel very strongly about this part of the “science.”
Steve: Take a deep breath.
Hughes at the NAS panel made an excellent distinction between dendro approaches – what he called the Schweingruber approach (as I recall) and the “Fritts approach” – now the MAnn approach.
The “Schweingruber approach” was to collect a large-scale network according to a priori criteria – the only sort of approach that makes sense to Willis and I. The other approach was to take a garbage can of tree ring chronologies and rely on software to sort out the mess – the Mann approach (the Hughes called it the Fritts approach.)
The problem with the Schweingruber approach was the divergence problem – which arose from the Schweingruber network.
One of the many ironies of Mann et al 2008 – which Rob and others have not acknowledged – is that Mann excluded every Schweingruber RW series, which are a very high proportion of the ITDRB network selected according to a priori criteria. This total exclusion of the very large Schweingruber RW collection is distinct from the deletion of post-1960 values of the Schweingruber MXD collection.
Willis, I agree with your comments above. Of course, whenever we engage someone on this, we always get a bait and switch. Sure, seemingly sensible criteria can be located in the dendro literature, but in the study at hand, we get a garbage can with ex post picking. And all the studies with small networks are completely unprotected against data snooping, as evidenced by the repetitive choice of proxies.
This is another thread that I think deserves more attention when any CA readers care to reflect. I found Steve’s reference to double (or conflicting) standards for proxy paleo studies to be of importance:
#78 — Willis wrote, “3) If they do contain such a signal, see if it can be split into its component parts (e.g. temperature, moisture, competition, catastrophic events).”
Thanks, Willis. That was a great post, as usual. IMO, item #3 is the kicker. By what biophysical theory can the effects you listed be quantitatively evaluated?
Metabolic theory applied to trees certainly predicts temperature limited growth, but not quantitatively. There are no known metrics derived from trees that can be cranked through a biophysical theory to produce a growth temperature.
This quantitative theory plain does not exist, and no statistical methodology can ever produce a physically meaningful metric where there is no physical theory.
7 Trackbacks
[…] Climate Audit […]
[…] However, surely it would have been far more relevant for them to speak up at the time of the publication of Mann et al 2008 and to have expressed this view as a comment on that publication. At the time, Climate Audit urged specialists to speak out against known misuse of proxies, but they refused to do so. (see Silence of the Lambs). […]
[…] However, surely it would have been far more relevant for them to speak up at the time of the publication of Mann et al 2008 and to have expressed this view as a comment on that publication. At the time, Climate Audit urged specialists to speak out against known misuse of proxies, but they refused to do so. (see Silence of the Lambs). […]
[…] However, surely it would have been far more relevant for them to speak up at the time of the publication of Mann et al 2008 and to have expressed this view as a comment on that publication. At the time, Climate Audit urged specialists to speak out against known misuse of proxies, but they refused to do so. (see Silence of the Lambs). […]
[…] However, surely it would have been far more relevant for them to speak up at the time of the publication of Mann et al 2008 and to have expressed this view as a comment on that publication. At the time, Climate Audit urged specialists to speak out against known misuse of proxies, but they refused to do so. (see Silence of the Lambs). […]
[…] Michael Tobis or Andrew Dessler speak out against UEA’s refusal? Or will they maintain the silence of the lambs that we have observed in the past. Like this:LikeBe the first to like this post. This […]
[…] more reasonable, or at least at-first-glance reasonable-sounding complaints from naysayers appear on that CA thread. This one from Eschenbach, for instance, one of the more coherent of the […]