David Appell has two trailers ( here and here) for his Sci American article [Oct 24 – url] on a “new” hockey stick article by Tingley and Huybers, not yet published, but said to have been submitted.
Tingley’s website contains two submissions discussing Bayesian methods, but only one submission (Tingley and Huybers 2009 url (h/t Jean S for pointing this out) describing a reconstruction with real data. (The same material is discussed in a 2006 AGU poster ) The network in the paper at Tingley and Huybers’ website is one that, within only 14 series, manages to include (1) (surprise, surprise) Yamal, (2) a strip bark foxtail series and (3) in a special feature appearance, Mann’s PC1 (though MBH98-99 are not cited). [Oct 24 – Also see followup post here; it appears that Appell was discussing another unpublished Tingley and Huybers not reported at their websites, about which some information is available from the PAGES 2009 conference. The criticisms in this post apply to the submission at the Tingley-Huybers website, but different criticisms apply to the network discussed in the PAGES 2009 presentation – see here. ]
Appell reported:
In any case, this new result ought to, I think, damp criticism that the PCA approach was somehow unsound or flawed, as some have implied…By the way, I asked Wegman for his thoughts on this new method, but he did not respond.
Somewhat smarting from Rob Wilson’s recent observation that I “had no idea what is being discussed w.r.t. methodology in many many meetings and workshops”, I asked the noted paleos, Clapton et al, for their thoughts on the Tingley and Huybers network and was very appreciative of Clapton’s prompt response linking to a workshop discussing selection methods.
UPDATE Oct 24.
In addition to the study linked above, Tingley and Huybers also have two pending articles comparing Bayesian analysis to RegEM – however, no reconstructions are presented in the two “Bayesian” articles. (Let me observe in passing that the Brown and Sundberg approach to multivariate calibration that we’ve explored here is strongly Bayesian in concept. So I have no objection whatever to taking a Bayesian approach to reconstructions.) The only article at Tingley’s website that presents an actual reconstruction (here) used a variant of CPS averaging on a small (9-14 series) set of “proxies” – NOT Bayesian methodology. I’ll discuss the network in this article in this post (comments on the network used in their PAGES presentation are in the accompanying post here).
Tingley and Huybers 2009 – the 1200 Year Study
While Clapton et al’s comment sum up the 1200-year reconstruction quite nicely, I’ll add some quick comments on some of the series in question. While the authors haven’t archived their data or methods, I’m familiar enough with the data to be pretty sure what they’ve used and I’ve been able to quickly develop code to sort of see what they’ve done up for at least part of the study. At a certain point, I lost interest in whether or not a composite including Yamal, Mann’s PC1 and strip bark foxtails was or was not invariant to a rotational null, deeming that issue of interest only to the Team.
Van Engeln: While Tingley and Huybers refer to IPCC AR4 (as Jansen et al 2007), they don’t appear to have consulted the Review Comments to IPCC AR4. Following Osborn and Briffa 2006, they use the Van Engeln record as a “proxy”. This record was also used in the AR4 Second Draft. In my Review Comments, I objected to its inclusion as “proxy” because its most recent portion was entirely instrumental, which I characterized as a “backdoor use of instrumental information, lending a false authority to the proxy records”. Unusually, this Review Comment was accepted and the Van Engeln series was removed from the AR4 proxy diagram. Given that even IPCC accepted this criticism, Tingley and Huybers should likewise have removed this record from their network. The IPCC exchange was as follows
6-1146 B 29:14 29:14 The van Engeln record only starts in 1251 and is a “shorter record” and does not meet the criteria of the caption. It should be excluded. It obviously wasn’t scaled over 800-1995. In addition, it uses instrumental information and contributes to a backdoor use of instrumental information, lending a false authority to the proxy records. [Stephen McIntyre (Reviewer’s comment ID #: 309-42)]
Accepted – the van Engelen record will be removed.
Mongolia: We’ve been following the history of the Mongolia series for some time (see CA post here which includes some very interesting comments by email from Gordon Jacoby. Tingley and Huybers say that their version comes from Osborn and Briffa 2006 (where a digital version of the annual data was archived after my request to Science). This version can be seen to be identical (up to rescaling) with the Mongolia version in Jones and Mann 2004, which proved to have been scanned from the original article – and not a very good scan. Jacoby commented as follows when I asked for a digital version of the data:
To clear the record; Mann and Jones obtained the data from unknown sources, published without any authorization, and Jones is distributing the data to colleagues. And, they published in GRL. Best wishes, Gordon Jacoby
Jacoby also warned me of a problem prevalent in paleoclimatology (and in this context, was not criticizing me, but warning me about some of the authors that I was studying):
You should also be aware another problem, the growing population of data parasites who produce nothing, do not understand data they use, do not present data accurately, and yet scream when all data are not served up to them. You have evidently been in communication with and about some of them.
Climate Audit 
125 Comments
GIGO, no matter what climo-matic you use, eh? The term for this kind of behavior is “maze-dumb.”
Yet another “new mathematical method”. Can’t wait to find out about it. Will it turn out to be Bistromathematics that is used to produce the hockey stick this time?
Steve: the issue is the network selection.
As the first rays of the new Sun broke over the horizon, I was filled with a sense of purpose and joy. Today would be the final day of my difficult climb up Mount Blind Faith to see the man.
snip – an attempt at irony, but too much on motives
No angry comments!!
I think the term you are looking for is “head bangers” 😉
Tingley and Huybers come to the “remarkable” conclusion:
Do we need to describe the millennium as a compound bow rather than a hockey stick?
Re: David L. Hagen (#4), What I find interesting about the “spatial variance” comment is that the recent warm period has tended towards less spatial variability.
Balling, R.C., Jr., 1998, Analysis of daily and monthly spatial variance components in historical temperature records. Physical Geography, 18, 544–552.
Wow with this kind of debunking, you should get a tenure position at Harvard – not those guys!!!
Re: bent-out-of-shape (#5),
‘scuse me while I kiss this guy
“Might as well face it, you’re addicted to [name that series].”
So who’s the opening act and who’s the headliner. If I had to guess, I’d say the Foxtails are the opener, with Yamal the headliner. Mann is providing the sound system.
I give Appell full marks for being consistent…and pathetically attached to the Hockey Stick. If anyone is guaranteed to demonstrate the fallacy of believing that PhDs are meant to know better about science, then its got to be Appell. He really doesn’t know any better.
snip – angry
Nope, guys; this is new stuff. It’s Bayesian Garbage.
H/t JeanS
=====
Jeez, Mc, that’s the best link to Eric’s stuff I’ve ever come across – completely amazing 🙂
And who said dendros were pale and boring 🙂
I’ve passed this link to my children, who went wild
Re: ianl8888 (#14),
Re: bender (#29),
I thought a bit about the link, in particular, whether things were “worse than we thought”. If the latter, then Clapton might be apt but maybe things have moved to a Velvet Underground stage. I also thought that Velvet Underground might be a little chronocentric to people growing up in the 1960s – I have what may now be a contemporary vinyl copy of their 1966 album (but also early Clapton things from the 1960s – Cream and even earlier.)
Velvet Underground H
Re: Steve McIntyre (#31),
🙂 🙂 🙂
We covered Bayseian methodologies almost three years ago in Road Map #3 through comment 550.
It’s interesting that, while the paper is about the MWP and indeed discusses the literature on the subject, it doesn’t cite Mann. This despite the fact it seems to use the Mann PC1.
He who shall not be named?
Re: Bishop Hill (#15),
It’s worth pointing out that in the 2005 SciAm hagiography by Appell on Mann, we get this statement:
So why is Appell telling us in 2009 that the Mann PC1 “proves” the MWP when Appell wrote in 2005 that the same disproves it?
OK here’s Tingley Figure 3:
Here’s the text for that graphic:
So (1) “Western USA” is our old friend the Mann PC1 aka “The Hockey Stick”. (8) is the newly debunked Yamal series. Doesn’t (3) also include the Mann PC1 in it as well?
Re: John A (#19), Hmm that’s interesting: The other russian series (other than Yamal) show an almost reverse hockey stick!
I’d call this ‘Zombie Science’ – the reanimated corpse of the hockey stick walks again.
Re: Paul Biggs (#20),
Voodoo Science Paul.
adjective, verb, -dooed, -doo⋅ing.
Use voodoo in a Sentence
See web results for voodoo
See images of voodoo
–noun 1. Also, vodun. a polytheistic religion practiced chiefly by West Indians, deriving principally from African cult worship and containing elements borrowed from the Catholic religion.
2. a person who practices this religion.
3. a fetish or other object of voodoo worship.
4. a group of magical and ecstatic rites associated with voodoo.
5. (not in technical use) black magic; sorcery.
–adjective 6. of, pertaining to, associated with, or practicing voodoo.
7. Informal: Usually Disparaging. characterized by deceptively simple, almost magical, solutions or ideas: voodoo politics.
–verb (used with object) 8. to affect by voodoo sorcery. (Dictionary.Com)
regards
I’m not sure I fully understand the reasons for different scientists to use the same data and then claim they have independently come up with the same results. Of course they will. It seems an awful waste of money to me.
Hold on. Even assuming that the data is flawed and/or not proxy worthy, why doesn’t this study seriously challenge the concern expressed here often that the PCA analysis was creating the hockey stick rather than the real world situation.
Re: Joe Hunkins (#22),
Joe, in which of the 14 resp. 16 lines of the diagram in #19 do you see a HS (just eyeballing)? If you have read the blog, do the descriptors of these ring a bell? And how about the rest – what reason can you give me that these do reflect the “real world situation” less exactly than the usual suspects?
Is there a kind soul that could send me (jean_sbls(at)yahoo.com) a copy of Appell’s article? I would not like to pay a cent for that man. Thanks. [edit: I got it. Thanks RM!]
Also I wish that people would not jump to any fast conclusions about the “new methodology” (which is NOT used in the paper linked by Steve) — liked Bayesian things or not — used by Tingley in his new “in revision” papers (here and here, Matlab code). Those seem worth considering in detail although I do not undertsand why these people are still neglecting the calibration literature. Also I do wish they are reviewed by someone actually understanding the methods.
Could it be that this nothing to do with funding, tenure, policy, politics or anything like that? Could it be that they just like thumbing their noses at Steve?
And then one wonders why dendros are not replying here…
Re: willard (#26),
You have GOT to be kidding. Did you LOOk at this paper? What’s in it? Clapton is *perfect*.
Re: bender (#29), it sure is perfect, but for what, a mocking contest in a masked ball?
Re: willard (#40),
It’s a joke, ok? Don’t worry. The analysis will come. This is just a teaser.
Re: willard (#26),
I do not know what you mean. Dendros do reply here. Not all of them, but this site is not inhospitable to facts and reason. And some dendros still attempt to use facts and reason.
Think of this site as a kind of intervention for addicted dendros. Of course, interventions do not always work. Much of the emotional driving force to “stop global warming” comes from the Mann hockey stick (and its successors with common errors). When you have a powerful graphic like that, it is hard to break the habit no matter how many times it has been debunked. Dendros may not like being the target of an intervention, but doing science right is the only way to persuade skeptics.
Tree hugging?
Tingley Proxy records ref 7
(7) Northern Sweden (Tornetrask; RCS tree-ring
Is obvious used an old proxy not the latest by Grudd 2008.
From this I quote:
The new data show generally higher temperature
estimates than previous reconstructions based on
Tornetra¨sk tree-ring data. The late-twentieth century, however,
is not exceptionally warm in the new record: On
decadal-to-centennial timescales, periods around AD 750 1000, 1400, and 1750 were equally warm, or warmer. The
200-year long warm period centered on AD 1000 was significantly
warmer than the late-twentieth century (p.05)
and is supported by other local and regional paleoclimate
data. The new tree-ring evidence from Tornetra¨sk suggests
that this ‘‘Medieval WarmPeriod’ in northern Fennoscandia
was much warmer than previously recognized”
How convenient to use old Proxys.
But there are no apparent Bayesian stats or definitions in the paper and Bayes or Bayesian isn’t mentioned. On the contrary it is all highly frequentist. Is it only David Appel who thinks it is Bayesian?
Steve: there is another Tingley and Huybers paper that refers to Bayesian stats http://www.people.fas.harvard.edu/~tingley/BARSAT_Part1.pdf .
It’s interesting that the abstract says “warm intervals are variable intervals”. Recall it was the MWP *variability* that Rob Wilson says caused him to reject the Polar Urals series, not the MWP warmth. Also note Ken Fritsch’s recent calculation that proxy variability is correlated with proxy mean. So is it the climate that is behaving this way, or the proxies’ response to it? If it’s climate, then Briffa’s substitution and Wilson’s defense are even less tenable.
Awesome video
Selecting the right proxies – “…It’s my wife, it’s my life”.
For tree rings, might a warmer period be a more variable one? Yes. For a tree, if you remove a limiting factor (like cold at Yamal) then the tree is free to grow faster, but other factors may still come into play. This means that at some times it can have a higher maximum, but the minimum can still be the same. For example, even in a temperate region like Ohio a very dry spring can cause very very narrow rings even though the growth potential is high.
Steve — It’s clear from the comments that many of my fellow readers don’t understand the point of your post. You shouldn’t allow data selection (they seem to have just selected the standard series for the sake of comparison) to take away from the fact that they’re doing exactly what you’ve been wanting the team to do for years: to stop using the same old tired Mannian methodology! Why not use this method without the Yamal garbage and see what you get? That would make your point pretty clear. Eric Clapton videos aren’t going to get your foot in the door with anyone who has interesting things to discuss with you.
Steve: There are two distinct sorts of issues in this field – there are weird Mannian methods in things like Mann et al 2008 where there are hundreds of proxies; and there are the little CPS networks with 5-18 picked proxies, where the issue is selection bias. I’ve spent much more time in the past 2 years on the latter sort of problem than on weird Mannian methods – though both are an issue, since one is said to support the other. I’ve been very forceful in objecting to the continued use of strip bark bristlecones/foxtails (which even the NAS panel said to “avoid” in temperature reconstructions) and the biased selection of Yamal over Polar Urals (this is not a new point.) I’ve commented on the “addiction” of paleoclimatologists to strip bark bristlecones/foxtails and Yamal – and have been condemned by critics for calling it an addiction. Here is a study that, within 14 proxies, uses strip bark foxtails, Yamal and Mann’s PC1. I think that it calls for a musical tribute. That’s not to say that I won’t show the effect of varying the network if and when I can figure out what they did (some I’ve figured out; I’ll post up code and resources and perhaps readers can figure out the balance.)
Re: Lahiri (#35),
Speak for yourself.
You say
Steve says
Wow…
I must be out of my league here, but can someone explain line 13 to me? Specifically, why does the “smoothed, then standardized” have the huge 20th century uptick when the “raw standardized” and “standardized, then smoothed” data doesn’t?
Scientific American wrote quote Reconstructing historical temperatures is difficult: investigators must combine information from tree rings, coral drilling, pinecones, ice cores and other natural records and then convert them to temperatures at specific times and places in the past. Such proxies for temperature can be sparse or incomplete, both geographically and through time. unquote
Pine cones?
JF
I’m sorry, but would someone explain something to me? I hate to be an idiot – although my wife assures me that I need to accept this condition as just “a part of [my] life” – but I completely do not get what Eric Clapton singing “Cocaine” has to do with the hockey stick, proxies, and so forth. Whoever among you is charitable enough to supply the explanation may need to use short, declarative sentences, composed of words fewer than four syllables, to ensure comprehension on my part.
(I feel like Elaine in the Seinfeld episode, asking the editor at the New Yorker to explain a cartoon.)
Re: Leighton (#42),
unbreakable addiction to active ingredient
Re: Leighton (#42),
We have some running jokes here that new readers may not get. I don’t try to make each blog entry self-contained. Regular and semi-regular readers will get the connection.
In the past, I’ve said that paleoclimatologists were “addicted” to bristlecones and that the introduction of the Yamal series in 2000 – without any proper technical article – was like crack cocaine for paleoclimatologists.
I’ve recently been reviled by critics for impugning the dignity of paleoclimatologists. Thus, there’s an irony in the entry of one more paleoclimate study using two strip bark series and Yamal in the midst of this – one that is featured in Scientific American no less as supposed “independent” confirmation.
Re: Steve McIntyre (#46),
Steve, you were very kind to clarify the matter for me. Thank you. The couple of times I wrote a comment at RC, the folks were less hospitable. (Has anyone else had that experience? Just kidding.)
… oh, wait, I get it now. We’re saying that they’re addicted to Yamal etc. That was a little subtle for those of us in the cheap seats, but I guess after some additional thought that I was able to (*ahem*) extract the signal from the noise, as it were.
Hi Steve,
I am finishing my doctorate now in an admittedly non-hard science field (education), but one of the warnings we regularly receive is that we can have our defense torpedoed by using statistics we do not understand and pay someone else to develop. I do not know, but suspect, that this is not a limited phenomena in any field that does not have statistics as a core (which, it seems to me, is most). I am not throwing any stones, but wonder if this may play a part here with egos getting in the way of accepting a dead end based upon inappropriate methodology and/or conclusions.
Since funding is tight, I now get to teach social studies in addition to math and my students have been studying technocracies, which I find highly apropos to this situation. I think the IPCC represents a truly unholy alliance of politics and science and I am reminded of Arthur C. Clarke’s 2nd Law – Any sufficiently advanced technology will appear to be magic. We are fortunate that folks like you are around to critcally examine pronouncements from above based upon techniques that may as well be magic to me.
Re: chip (#48),
The one thing I have learned from CA is that abuse of statistics is such a common phenomenon of “peerreviewedliterature” that it ought to be given its own academic department and journal to study. If I were you, I’d get good statistical training under your belt because then you’d never run out of things to study
Before we had the addiction, we had the Pushermann.
Re: Freezedried (#49),
Actually before the addiction, you had “I’m waiting for the Mann”
also by Velvet Underground
She’s alright, she’s alright, she’s alriiiight C3H8.
——
So it takes a lot of guts to put together all the horsecrap proxies in the same paper and declare victory.
My maf not good- my not climatologist but me think maybe –
Bad + bad + really bad + bad === robust bad.
Re: jeff id (#51), ROTFLMAO…this thread is why I am so “addicted” to CA.
Re: jeff id (#51),
With apologies to Clapton (and Steve), I offer my own version:
You wanna an uptick, you simply cherry-pick; Go Team.
Wanna show a flat past, selectin’ Yamal’s a blast; Go Team.
It’s robust, it’s robust, it’s robust; Go team.
If you wanna make news, foxtails you will use; Go Team.
Polar Urals are out, and Ababneh casts no doubt; Go Team.
It’s robust, it’s robust, it’s robust. Go Team.
Sample sizes too small, no worries at all; Go Team.
Cuz you can hide what you do, peer review won‘t question you; Go Team.
It’s robust, it’s robust, it’s robust; Go Team.
If you get caught, just recall what you were taught; Go Team.
“Doesn‘t matter“ if your wrong, ’cause you’ve now “moved on“; Go Team.
It’s robust, it’s robust, it’s robust. Go Team.
jeff id on line 51, how about:
“She’s alright, she’s alright, she’s alriiiightt, CH4”
Yep, as we know, it’s a strong greenhouse gas, so prevalent in nature, and there is certainly a lot of it being “released” by the likes of Mann, Appell, Tingley, . . . (as well as genuinely worthwhile creatures such as cattle, etc.)
Steve and company, please keep up the good work debunking the foolishness which will without doubt lead to some stiff tax increases around the world – err, unless some of the foolish politicians finally wake up to reality?!?!?
David Appell on Thursday:
David Appell today:
Yes, the new title is: Novel Analysis Confirms Climate “Hockey Stick” Graph
Manfred Mann The Mighty Quinn
We are repeatedly told, correctly, that the Hockey Stick is a small and perhaps even irrelevant bit of support for the theory of AGW by the folks such as Gavin, et al, when a stake is driven through the latest version.
But we are also treated to innumerable attempts, so far all of them fatally flawed by using the same defunct body parts, to reanimate the corpse of the Hockey Stick. And as soon as the new zombie crawls out of the grave, Team members immediately begin walking along side propping it up as it as important new evidence of AGW.
Is cognitive dissonance a Team requirement?
Re: Brian B (#56),
Brian,
I think they know exactly what they’re doing.
For all of those people calling for Steve to enter into the “peerreviewedliterature” – why would he want to enter a process as lacking in credibility as this?
In the case of these reconstructions and the progeny of reconstructions, we have the proxy addiction that David Appell somehow diverts away from by noting similar results from different reconstruction methods. But is it denial or simply ignorance of what the criticisms have pointed to: methods and cherry picking proxies.
.
Given cherry picking, two methods can give similar results, but what about cherry picking. Do the one out sensitivity test and if you have cherry picked sufficient proxies, you can show similar results.
Why are these defenders of the reconstructions not able to look at the entire bases of the criticisms? Could another addiction (not to proxies) have this affect on this thinking?
RE Dean P #36,
Good question — but there isn’t necessarily an error here.
The gray line is the raw data, standardized to have unit s.d. Then the red line is a smoothed version of it, which necessarily has a much smaller s.d. The blue line is equivalent to taking the red line and scaling it up to have unit s.d. So if the data turns up just a little more at the end than it has over most of its history, the blue line can jump outside the data. Curious, but not necessarily erroneous.
A deeper question is how the data was endpadded to enable the 19 point Hanning window filter to run clear to the end of the data. The smoothed versions should either end 9 points short of the end, where the fiter runs out of data, or else continue as dotted lines to show that the formula has been altered in order to continue the filter into the last 9 points.
RE Leighton #50, don’t feel bad — it took me a while to figure out Steve’s joke as well, even though I’m pretty regular here. Doh!
Re: Hu McCulloch (#58),
Hu,
Thanks! I think I understand what you’re saying. Still, it’s interesting! The raw data doesn’t seem to have any noticeable change (although it does seem to oscillate some at the very end of the 20th century). The red line has a slight uptick. But that blue line uptick is dramatic! It’s also an interesting proxy because it’s almost perfectly flat right up until the end.
But what can replace the addiction? Maybe this:
That was a funny start to my Saturday. Sort of made me wonder how many “new” reconstructions we will see in the new year. Similar to guessing how many Gorricanes we will have this year.
Steve —
Not that it makes much difference, but do they use Yamal 2000 or Yamal 2008? You’ve noted on the previous thread that the latter gives about 15% more HS.
Or should I say it gives 15% more high? 😉
Steve: Yep. The customers don’t seem to have famliarized themselves with the new brand yet. The downer from Taimyr has also been lessened.
Re: Hu McCulloch (#65),
15% is a lot when you’re on an artists budget!
RE Dean P #62,
Another factor making #13 (evidently the Van Engeln series mentioned by Steve in the headpost) so different from the others is that its final values are instrumental rather than a true proxy. This allows its final values to be predominantly positive, whereas the proxy portion is more like white noise. The result is that the red line is virtually flat up until the final years, when the run of mostly positives makes it hooks up a little. Then scaling it up to unit s.d. makes the final, instrumental years jump way out.
Fortunately one of the IPCC4 Expert Reviewers (our own Steve McI) caught this and got it banished from AR4, as noted in the headpost. But that evidently didn’t stop Tingley and Huybers from going back for another snort.
I’ve also noticed that Tingley and Huybers made a presentation at the PAGES July 2009 conference, in which they presented a 600-year reconstruction by poster and talk, applying their Bayesian methodology to a proxy network from 1400-present – but not back to AD1400.
The Appell article didn’t cite a specific article, but it looks like its about this presentation as opposed to the Tingley and Huybers 2009 article discussed in the post.
The network consists of 96 MXD gridcells – these look like they are obtained from the RegEM version of the Briffa-Schweingruber MXD network as calculated in Rutherford et al 2005 – also used in Mann et al 2008. Whether the post-1960 values are deleted and “infilled” a la Mann 2008 remains to be seen.
They also report the use of 7 O18 ice core proxy series (shown in a location map). Kaufman also used 7 O18 ice core series. The TH location map shows sites in Baffin Island and near Mt Logan (and two fewer Greenland sites). The Baffin Island data might be O18 from Penny ice cap (used in Mann 2008) and the other is presumably Holdsworth’s short Mt Logan O18 (but perhaps the recent Fisher version.)
They report the use of 13 sediment thickness series – versus 9 used by Kaufman plus 3 Finnish series. Their location map shows two in Alaska (Kauf -3); two in BAffin Island (Kauf -2 one in different location; one in Svalbard (Kauf – 1 in Iceland); 1 in Finland- (Kauf 3); 6 or 7 in the High Arctic islands (Kauf -2).
The poster presentation and talk refer to the use of 259 instrumental series as well. Different considerations will apply to this (still unavailable) paper than to the submission that is availale. The talk looks more along the lines of Mann 2008 than the small CPS studies.
Grace Slick also had a relatively early consideration of principal components and RegEM, noting not merely the differing effect of large and small coefficients, but also that many proxy series more or less played the role of white noise. We considered these topics in our articles on MBH (see for example the remarkable difference in size between Sheep Mountain and other coefficients.)
Thanks steve for proving that uber-geek math humor can be wicked-funny.
Re: PAGES reconstruction
Could some GIMP/PHOTOSHOP wizard overlay this image on a top of the IPCC spaghetti diagram?
Seems to me that this is not “confirming” anything, more like “contradicting”…
Re: Jean S (#71),
Here you go. [Nice distraction 🙂 ]
It’s not easy to see so I overlaid in front and in back. There’s a TON of vertical rescaling necessary…
snip – once again, please do not refer to policy
Was White Rabbit a thoroughbred?
(White Rabbit out of Black Hat by Arm Waving)
My fellow countryman is named Aryan van Engelen.
Not van Engeln.
Went to examine the Clapton et al. workgroup and noticed an alternative
presentation featuring Mark Knopfler collaborating on the presentation. I
say this expanded presentation is a better example of the dire straits of the Team’s
addiction.
Not forgetting Clapton’s advice in his alter-ego of Derek and the Dominoes: Tell the Truth
Can someone clarify what Steve means by the Mann PC1? i.e. does this refer to the decentred PCA that was denounced by decentred PCA by Wegman and Jolliffe or is it a variant that is correctly centered but provides results that have no statistical significance?
.
If the answer is the former can someone play devil’s advocate and explain how a well meaning but mistaken scientist could convince themselves that decentred PCA has any validity given the amount of ink/bits that have been written on this topic?
Re: Raven (#77),
It is the “decentered” PC1 from Jones & Mann (2004), see
http://www.climateaudit.org/?p=523
[Edit: ..and for further details, see http://www.climateaudit.org/?p=522 ]
Re: Jean S (#80),
After reading the long explanations in the links, I’m still having difficulty comprehending that TIngley’s “(1) Western USA (PC of tree-ring chronology” (comment 18) turns out to be PC1 from MBH99. As I understand, Mann calculated PC’s for nearby collections of tree ring data to avoid overweighing sites with lots of such data. (Since no calibration or test period was involved, these PC were presumably properly centered on the mean.) Then Mann apparently used these “local PC’s” and other proxies to extract from off-centered data the “hemispheric PC1 and PC2” used in the complete reconstruction. The name “Western USA” sure sounds like a “local PC”, not a “hemispheric PC1”. [If I understood the linked posts, “Western USA” is actual composite of two “hemispheric PC1’s” for different time periods.] Am I correct in concluding that Mann has mistakenly archived a composite of his “hemispheric PC1” data under the name Western USA?
Steve: you’ve got pretty much everything backwards. It’s actually the PC1 from Mann and Jones 2003 (See that category). You have to discount claims about the ostensible purpose of these methods “to avoid overweighting sites”, as the actual results are often opposite. This particular PC1 is hugely weighted by Graybill’s Sheep Mt series (the one not replicated by Ababneh.) You’ll have to fend for yourself a bit as I’m busy on current things.
Re: Frank (#107),
Let me try to explain in short terms how this series, labaled “Western USA”, orginating from Osborn&Briffa (2006), née Jones&Mann(2004)/Mann&Jones(2003) was constructed:
1) A “PC1” (with Mannian (MBH98 fame) off-centric methodology) was calculated from six tree ring chronologies listed here. This gives you a series spanning 200-1980.
2) [MBH99 step] First, a “PC1” (with Mannian off-centric methodology) was calculated from 27 tree ring chronologies. This gives a series spanning 1000-1980. This PC1 was “fixed” (“adjusted”) by “removing CO2 fertilization” from the end (post 1850). See here for details, and here for the effect in MBH99 (the effect in Mann&Jones should be similar). Recall how the hockey stick was plotted and presented in MBH99:
3) Sometype of recentering and rescaling of series in 1) and 2) were done, see here, and then the final “PC1” (“Western USA” in O&B) was obtained by splicing 1700-1980 part of series from 2) with 200-1700 part of series from 1).
Huh, a really reliable temperature proxy, isn’t it?
Pinecones?
Anyone?
JF
Is this an attempt to give a new name to the time around 1000AD?
Generally there is “Medieval Warm Period” (MWP), but there have also been “Medieval Warm Epoch” (MCE), “Medieval Climate Optimum” or “Medieval Climatic Optimum” (MCO), or “Medieval Climate Anomaly” or “Medieval Climatic Anomaly” (MCA), or “Little Climate Optimum” or “Little Climatic Optimum” (LCO). And then there are the older spellings such as “Mediaeval Warm Period“. [Links are to Google Scholar]
Now from Tingley and Huybers we will have “Medieval Variable Period” (MVP) or “Medieval Warm and Variable Period” (MWVP).
Thinking about this, Optimum should not be used because that implies best. And Warm does not sound bad enough to those of us approaching winter. But MVP also has positive connotations. Perhaps by giving it so many names, it will just go away.
nevket240: #66
Okay, voodoo and zombie are related. Who’s gonna post up Hendrix playing Voodoo Chile on YouTube?
Another possible musical link would be REM’s Fables of the Reconstruction.
Re: Clark (#82),
I’d personally go with Funkadelic’s A Whole Lot of BS m’self.
Here’s the two separated, with the TH graph rescaled and aligned. Perhaps this is easier to see…
The system doesn’t let me put a direct link to image.
[ed: sure it does. you need to make it into a tag using the buttons above or directly. <img src=’url’> ]
Re: lucklucky (#86),
You click “Img” and paste the location and clock OK twice.
“dignity of paleoclimatologists”: it’s the way you tell ’em.
I clicked in Img and put the link. Anyway thanks for the cleaning.
Oh “twice” probably it was that.
Tingley, reusing the Yamal uptick,
has just regenerated the Mann Hockey Stick.
Well, I done my duty and posted a comment to Appell’s SciAm article, at http://www.scientificamerican.com/article.cfm?id=still-hotter-than-ever, with a link to this discussion, and special mention of Grace Slick.
When (and if?) it clears moderation, it may be #3 of 3.
Steve: Per my earlier post, thanks for the additional commentary. One more question – you mention that they used 259 instrumental series as well as various O18 and trees… What are the other considerations that apply to these? Is there a problem inherent in using instrumental data, as long as it isn’t stuck onto the end of a proxy series? It seems like it would help to “calibrate” the proxy weighting, and if nothing else, ignore those which are mostly “white noise.”
Re: Jean S (#71), Who said it “confirmed” the hockey stick, apart from some random science blogger from a magazine? I think you’re barking up the wrong tree by paying any attention to (un)scientific american…
snip
Hope the Team are going to get ROBUSTED.
Re Julian Flood #79
Yes, Appell has been smoking pinecones!
RE #92
My comment on Sci Am went through.
Things good about the rconstruction method:
1. Bayesian approach different and interesting
2. Allows for varying sensitivity over time
Things fatally bad about the rcosntruction method:
1. Recycles same old crap proxies
2. Linear model assumed when responses are actually nonlinear
3. Uncertainties still ceiling to floor
4. No solution to the non-invertibility problem; you can’t reconstruct one variable when growth is co-limited by multipel factors
5. The method is completely opaque to audit
Conclusion: junk science that should never be published, yet gets published because of the “novel” methodology. Does not serve us well.
Steve: See my next post. The study discussed here (using the stale old proxies) also uses a CPS-type method. The “Bayesian” method is applied to a different network entirely and entirely different issues arise. I looked first at the raw materials and there are big questions on how you go from the raw materials in the “Bayesian” network to the recon.
Guitars are for incompetents and beginners – especially when they make noise driven by nuclear electricity, from a music score like the graph of John A (#18). Why not go back to the music of the MWP to look for allegorical material? Consider the insinuation and allusions in this simple piece of prose mentioning another, older type of music with exciting action surpassing strutting under spotlights in flares.
(Slightly modified from http://www.bbc.co.uk/dna/h2g2/A49254627 )
Has the feel of climatology action, no?
Re: Geoff Sherrington (#99),
I have a warped sense of humor so I found your post very funny. Thanks!
Appell’s Sci Am article, at http://www.scientificamerican.com/article.cfm?id=still-hotter-than-ever, is supposed to contain a graph, but all I can see is a picture of a tree section. How does one see the graph?
Re: Hu McCulloch (#100), Look at http://www.climateaudit.org/?p=7524 – it is probably related to the graphic from the PAGES conference.
Given Tingley isn’t a climate jockey: “Tingley, now a postdoctoral student at the Statistical and Applied Mathematical Sciences Institute in Research Triangle Park, N.C”, what is to be done?
Loved the music, sometimes there is a need to lighten up the discussion to bring “some” back down from the ceiling (hint wait for the end).
Re: schnoerkelman (#103),
I can’t get no perspective.
===============
Or for those with a liking for madrigals, this 16thC divergence problem is from Thomas Wheelkes:
Now ev’ry tree renews,
Why is your heart in winter’s raiment clad?
has anybody downloaded his matlab code yet and got it running?
RE Steve M #101,
I’ve now checked the print edition of the Nov. Sci Am, and indeed it contains an unidentified graph that is not included in the online version. It is not the graph you link (though it may be the same data, differently smoothed). I didn’t shell out $6 for the mag to take home and compare, but it looks for all the world like the MBH 98 or 99 HS (to 1000AD only), complete with gray “CI” and overlain red instrumental series that makes the end of the actual recon hard to see.
I’ve posted a request for clarification on Appell’s SciAm page.
Re: Hu McCulloch (#109),
Hu, does it look like an artistic version of this one (from the PPT)?
RE Steve #110,
The unattributed graph in the Sci Am print ed of Appell’s column does look for all the world like the graph on p. 13 of the PPT of T&H’s “A Bayesian Approach…” talk at http://www.pages-igbp.org/products/osmysmtalks09/YSM09_OralB_Tingley.pdf. This isn’t exactly sourced, but presumably comes either from MBH 99 or IPCC4, both mentioned on the same page.
Re: Hu McCulloch (#111),
P. 13 graph is the original MBH99 hockey stick from “Summary for Policy Makers” in IPCC TAR (here).
RE #110, 111,
Your graph appears to be the one on p. 11 of the same PPT I link in #111.
I seem to be confused, and maybe Steve or one of you readers can help me out. It appears to me that David Appell and Scientific American are the villians here, and not Martin Tingley and Peter Huybers. The Tingley-Huybers 1200-year temperature appears reasonable to me (complete with MWP and LIA, and devoid of a dramatic 20th century uptick). That reconstruction is Figure 4 in their report at
Click to access mean_variance.pdf
My take is that the Tingley-Huybers diagram in no way supports Mann’s hockey stick, and that the Scientific American article by David Appell at
http://www.scientificamerican.com/article.cfm?id=still-hotter-than-ever
is an ordinary blatant lie. Where am I going wrong in this?
Re: Jerry Lee Davis (#114),
The 1200-year reconstruction (800-2000, “T&H 2009”) has the most terrible proxy set one can have. So the T&H 2009 reconstruction is so unreasonable anything can be. However, the new methodology (BARSAT; not used in T&H 2009) they are proposing in two new papers seems reasonable, but that has been only applied to construct a 600 year (1400-2000) reconstruction based basicly on Briffa’s data with certain problems. Moreover, I don’t see how that supports Mann’s hockey stick (compare to this figure, where there is a somewhat comparable Mann’s reconstruction (MBH methodology) for latitudes 30-70N for the period 1600-1980). So the villians seem indeed be DA&SA, but for different reasons you are suggesting.
MATLAB weenies should get his code up and running
Re: steven mosher (#117),
The code is even more time consuming that RegEM so beware. Also we don’t have the actual data used in his 600-year reconstruction yet.
Also before people spend a lot of time on this method, it would be worthwhile working through the algebra both for this and RegEM. We noticed a year ago that Mannian RegEM recons yielded too-perfect reconstructions in the instrumental period i.e. the recon was in effect splicing the target instruments. This needs to be revisited with the additional knowledge of the method developed in the Steig analysis – it always takes a while to refresh this file and some clear time is needed – something that I won’t have until the Briffa thing is analyzed.
Re: Steve McIntyre (#118),
I agree completely. The EIV method has become pretty clear as to what it is but it’s taken time to figure out. I don’t think we’ll hear Mann saying -noone splices temp records on proxies anymore. It would be interesting to see the Xmis matrix behind the recon for the last iteration though. While I expect it correlates fine, there is a chance that it doesn’t do very well.
Re: jeff id (#119),
Re: Steve McIntyre (#118),
Is this “splicing effect” (for lack of a better phrase) documented/described somewhere?
Re: bender (#122),
In Steves original post the match wasn’t just too perfect, it was exact. The reason for this is that RegEM masks data which are available and infills the missing values. This makes sense for data which are sparsely missing or cases where the infilling data is of equal quality to the data being replaced. As in typical expectation maximization where only a few values are missing from data of the equal quality.
In the Steig recon, the sat data is lousy for trend and horribly noisy in comparison to the surface stations so Steve’s approach of replacing the entire dataset with the infilled values is the only way to go. I think in the case of fitting proxies to temperature the same kind of condition exists where there is a mismatch in fidelity of the proxies to the instrumental temperature.
If you look at the RegEM ttls code there is a line where the available data is masked and only the missing data is replaced.
Re: Steve McIntyre (#118), But you should have the data for his test of the system where he tested the method against Regem using thermometers.. I may be mis remembering but didnt he test the method by using CRU data for the US.. and them create psuedo proxies using thermometer data ( or did I dream that?)
Re: steven mosher (#120),
infilling gridded temperature data is a very different animal than reconstructions using proxy data. Since we don’t have actual data and runs, it would be more productive of my time (after analyzing the new Briffa data) to revisit RegEM. Plus it looks 99% certain to me that his BARSAT method has important elements in common somewhere with RegEM which is what causes the “too perfect” calibration period reconstructions. These sure look like hidden splicing is an effect of the algorithm (as Jeff Id also agrees.) In my emulation of RegEM, I tweaked the code to keep unspliced results (that’s the Xmis that Jeff mentioned) and maybe this can be used to perform statistical tests that the authors neglect.
Re: Steve McIntyre (#118),
Ahhh. So we now have de facto splicing. In all the Briffa/Kaufman stuff recently, someone (Rattus non Readus?) said (essentially): I don’t care about reconstructions or divergence after 1880 or so because we have the instrument record.
Prediction: Team paleos will “move on” to de jure splicing. We’ll see recon after recon showing the 1,000 years (2,000? 5,000? mill-yun?) prior to 1880 as flat as a pancake, followed by a CO2 induced blade from the instrument record thereafter.
Look on the bright side, it’s not as nutty as “teleconnections”, which was so bizarre it was hard to know where to start.
The previous workshop references for Clapton, et al, in this item and the resulting general trend of the piece and related posted observations put me in mind of an early commentary from Hunter, Garcia and Jerome made public around 40 years ago.
Although slightly repetitive, in the style of the times, the message is kept simple and is notable for predicting certain language traits that have become especially popular in recent times (“It was later than I thought ….”)
References to ships and boats clearly speak to issues about sea levels and unresolved issues about flotsam and jetsam or contentious papers are discussed. (“… all that could not sink or swim was just left there to float.”)
Matters concerning appeals to authority, the paucity of useful content as discovery over time erodes the initial hypothesis and the need to make ones concerns public are also subjects about which the authors wax lyrical.
I have found both text and the spoken word forms of the commentary and offer these for review by readers.
Note that the graphics used for the spoken word version linked here are not the best but it was the shortest version of the commentary I could find in the time available … readers are of course free to indulge their passion for search engines and seek alternative presentations – of which there are a few.
Looks like the links to Hunter, Garcia and Jerome were damaged in transmission, one going AWOL.
Here it is in a different form.
2 Trackbacks
[…] data revelations, which brings us back to the new paper by Tingley and Huybers. Steve McIntyre observes that “Their “new result” is based on a network that, within only 14 series, […]
[…] on” so quickly that it takes some care keeping track of their movements. The criticisms in my most recent post apply to the still unpublished Tingley and Huybers 1200-year reconstruction at their website (that […]