Friday, February 19, 2010

The Hockey Stick Illusion

As various readers will know, Bishop Hill alias Andrew Montford has just published a book called The Hockey Stick Illusion. I recommend that people read it and post their comments here, so we can have an informed debate about the merits of the claims made in the book and about its main protagonists.
I have not finished reading it and will not write a review here. Two things stand out for me. One is the peculiar genre of the book which is a mixture of hagiography and detective story which is not always working well. While the style of the book is remarkably accessible to non specialists there are, at times, too many technicalities which are described in too much detail so that the lay reader (ignorant of the importnace of specific statistical methods) easily gets lost.
The more important aspect is the book's reconstruction of the paleoclimate controversy as initiated by Steve McIntyre in 2003. He is the main hero in the book; HvS and EZ also make appearances but are not given comparable attention. The story I take away from this book is the resistance of Michael Mann to have his original papers (MBH98 and MBH99) scrutinized by McItyre and others. It poses the immediate question of how these papers could have passed peer review initially when no one could replicate the results based on the initial information contained in the published papers (or based on later additionally material that was later released reluctantly). Instead, there was a counter attack calling McIntyre incompetent. This line of rhetoric seems to reappear at every turn of events, we have seen it in the emails, and we have seen it on this blog many times. This rhetoric can no longer be taken at face value by anyone who reads the book.
The Hockey Stick Illusion will require a refutation (based on an alternative reading of events or new facts) if such claims were to be substantiated.

33 comments:

P Gosselin said...

I haven't read it yet, but I certainly will. It's a present for my upcoming birthday. I don't expect much new from it though, except for more background details on the greatest fraud ever perpetrated by man.
(As you can see, I'm going to wade into this book with an open mind!)

I didn't think it would happen this year, but it looks like we'll hit the 14 million sq km mark for Arctic sea ice.
http://www.ijis.iarc.uaf.edu/en/home/seaice_extent.htm

Anonymous said...
This comment has been removed by a blog administrator.
P Gosselin said...

http://www.youtube.com/watch?v=ODsN-uTP43Q&feature=player_embedded#
Ok! Let's allow Rahmstorf to run the planet. Let him write up his 10-point Lets-Save-Our-Asses Plan. Let's all do what he says.
No, the hockey stick is wrong...it actually should be more like a boomerang, or even better, like the trajectory of a rocket. That's it!

Is it any wonder that us sceptics are loud, and going to get a lot louder? Rahmstorf is a "scientist", along with equally shrill Schellnhuber, who has the Chancellor's ear. Think of the implications of what Rahmstorf is calling for. I've called him harsh names in the past, and I stick by them, because that's what he is.
I think it's a national disgrace that such a overheated head is allowed to drive policy. God help us.

I think Prof von Storch's thrust in the clip is right, but I think he needs to put a lot more psi's behind it. Otherwise his ship aint gonna reach orbit.

Rainer S said...

I am almost through - it sure is interesting reading.

Being an experimental physicist, I did not find the technical arguments hard to follow.

In fact, if the allegiations made regarding data quality/integrity, cherry picking, inappropriate use of mathematical/statistical techniques etc. are even only partly true, the palaeo people have a serious problem.

Likewise, the light the book sheds on the IPCC process is not flattering, either.

Since HvS and EZ are mentioned, they might at least add some valuable information on how they view the parts of the story they were involved in.

Rainer S said...

@Pierre

Although I share - at least in part - your opinion about some of the AGW protagonists:

Please try not to fill every post comment thread with OT statements. No offence meant, but it really is boring.

The contents of "The Hockey Stick Illusion" would supply enough material for discussions.

Anonymous said...

Rainer,
.
The allegations have more than enough supporting evidence to be presumed true (Two public inquiries in the US + CRU emails).
.
If you don't agree then how could we possibly acquire more evidence given the fact that the guilty parties deny it and the scientific institutions that are supposed to ensure the integrity of science (eg. journals, professional bodies) have decided that protecting the CAWG agenda is more important than ensuring the integrity of science?

grabstos said...

@6

It is not a question of me "not agreeing" with the book´s conclusions, it would just be interesting to hear the accounts of other parties involved like HvS and EZ.

And, mind you, I do have exp. physics training (German doctorate). Just from want I know about the dendro´s methods (I am not talking about the politics behind), I would recommend to rebuilt the whole edifice from scratch. And I would not like to have seen my name on any of these papers.
But that´s just a personal opinion.

Rainer S said...

@6

It is not a question of me "not agreeing" with the book´s conclusions, it would just be interesting to hear the accounts of other parties involved like HvS and EZ.

And, mind you, I do have exp. physics training (German doctorate). Just from want I know about the dendro´s methods (I am not talking about the politics behind), I would recommend to rebuilt the whole edifice from scratch. And I would not like to have seen my name on any of these papers.
But that´s just a personal opinion.

Anonymous said...

The statistics are important because this dispute is the only thing keeping the Mann hockey sticks alive. Defenders of the hockey stick point to Wahl & Ammann's paper, which can only defend it my making up new statistics, as it fails on the traditional ones.

Marco said...

@anonymous (and others):
One could ask oneself why there is such a desire to attack a 12-year old paper, when many others have come out with similar results using different methodology.

I think I know the answer to that question, but perhaps someone here can change my mind (and thus I won't say what my proposed answer is).

Anonymous said...

Statistical analysis, interpretative evaluation, depend wholly upon integrity of base-data sets. Absent full disclosure, including verifiable sources' time-series replicable in all cases to authors' and independent reviewers' satisfaction, no scientific report has any credibility whatever.

All anyone has to know is that underlying data are incomplete, manipulated, worked backward to show preconceived results. This requires no more ability than balancing a bank statement. What arrogant Warmists fail to realize is that blatant subterfuge blasts everything that follows at the root. Who cares what doltish theses they espouse when nothing they say means anything at all?

Over time, even covert rip-off artists like Al Gore must necessarily fold tents and slink away. Meanwhile, three months past Climategate, no major figure in this astounding scandal has dared spew forth more rubbish. Quite possibly, none of them ever will again.

mazibuko said...

I thought the tone of this blog and its followers was supposed to be a bit more constructive than the most recent offering by anonymous.

Such viteruperation only serves to get backs up, and take the debate back into finger pointing about motives, conspiracies, etc.

Anonymous makes pretty strong claims about the quality of the data that are not supported by the facts. Such a claim might match a few datasets, but there are many more out there that are robust.

Remember, when one points fingers, he or she will always find several pointing back.

TonyN said...

I think that the success of this book can be gauged by the fact that Real Climate hasn't dared attack it yet.

All good stories need a hero and this book is as much about one man's dogged determination to shine light into dark corners, and the problems he encounters where least expected, as it is about a particular scientific controversy.

Many of us who have followed this fascinating saga unfold over the last few years through a thousand blog posts will be grateful to Andrew Montford for bringing all the strands together in a form that is accessible to the general reader. With luck this will lead to informed discussion of the Hockey Stick and its problems outside the scientific community.

itisi69 said...

I've followed M+M's battle with the Hockey Stick over the years on both Climate Audit and Real Climate. I believe the disdain, arrogance and outright disgust Mann showed for Mcintyre (he left the room when McIntyre made his presentation during the Wegman hearing) and later on Unreal Climate gave McIntyre the best motivation to get to the bottom of this science debacle. Mann and co grossly underestimated "he who will not be mentioned" and Mann's arrogance (calling McIntyre a "fraud" and a "moron") ultimately led to his fall out of ivory tower. Never, ever underestimate your enemy, especially not a very meticulous and determined mathematical analyst like McIntyre.

Anonymous said...

Marco, there have been other papers, but they are not nearly as independent as the gang at RealClimate would have you believe.

They use the same proxies, and similar methods. For example, just looking at the NAS Panel report that supposedly vindicated Michael Mann, the report agreed with McIntyre that bristlecone pines should not be used as a proxy in temperature reconstructions. They then proceeded to support Mann's thesis of unprecedented warming by referring to studies that used bristlecone pines. Many papers use Mann's PC1 from the MBH 99 paper. Yamal is another proxy that gets used because of its extreme hockey stick shape.

Kooiti MASUDA said...

I do not read the book and I do not mean to discuss about it now.

But I have just arrived to another concept of "hockey stick illusion" by discussion at a "Talk" page of Wikipedia on "Hockey stick controversy" in the Japanese editon.

The graph of MBH99 had shades showing a range of uncertainty.

Later reconstructions often showed more undulations than MBH99 with smaller ranges of uncertainty.

Insiders of science see that the new curves are generally within the uncertainty range of MBH99, so MBH99 was basically still right though much less informative than more recent results.

Outsiders of science point out that new curves do not look like hockey sticks, so the famous or notorious hockey stick of IPCC TAR is now refuted. Since the figure came from MBH99, MBH99 must have been refuted.

The problem may be one of scientific communication. Most of the readers of popular media noticed the curve but not shades. It is no surprise as the figure was mostly used as a tool to say that the warming in the 20th century is unprecedented.

So, as a scientist, I want to say that it is a "hockey stick illusion" that is refuted, not MBH99. (Scientific evaluation of the merits of MBH99 is another matter.)

But people who feel what I call "illusion" as truth are not likely to agree with me.

Probably this is an instance of "perceived enviroment" discussed in the book of Stehr and von Storch (Climate and Society, Section 4.7).

Richard Tol said...

I have not read the book.

I have seen some of the work by McIntyre and McKitrick. They are sound statisticians.

itisi69 said...

McIntyres invitation to the World Dendro 2010 has been withdrawn. Apparantly "quite many planned speakers for the Roundtable discussion were not very willing to participate" the session in which McIntyre would make his presentation. Briffa still seems to be on the invitation list.

Ivory towers?

Leigh Jackson said...

RealClimate's take on the "Hockey Stick" I imagine looks rather different. Does Montford refute their analysis?
http://www.realclimate.org/index.php/archives/2005/02/dummies-guide-to-the-latest-hockey-stick-controversy/

Leigh Jackson said...

Richard Tol 17
There are a several versions of the HS written by several authors. Do you have any view on their professional acumen?

TonyN said...

Re: Leigh, #19

The best way to find out whether Montford's very recent book refutes Real Climate's five-year old article about the Hockey Stick would probably be to read it.

I wonder when Real Climate will get round to reviewing the "Hockey Stick Illusion"? As a serious science blog you'd expect them to be really interested in a book devoted the most famous achievement of one of their contributors. But perhaps no one there has heard about it yet.

Kooiti MASUDA said...

I have just encountered an "anti-hockey-stick illusion" as well.

A Japanese author, saying that the "hockey stick" was debunked, suggested that a figure in the IPCC's First Assessment Report (FAR, 1990) is more likely to be true. (That is the Fig. 7.1(c) of IPCC FAR WG1. It was not a scientific reconstruction but just a schematic drawing. See, for example, the article "MWP and LIA in IPCC Reports" of English Wikipedia).

The same author said in a book published in 2005 that Mann et al. (1998) was "corrected" by McIntyre and McKitrick (2003; MM03), and showed the graph of MM03 which had a peak in the 15th Century.

A curious thing is that the curve in IPCC FAR has a peak around 12-13th Centuries, and it has normal or a little below normal values around the 15th Century.

I am not sure whether the author is just a careless scientist who did not notice the inconsistency (a victim of the illusion) or is a propagandist who use anything that can give impression of an anti-hockey-stick to casual readers (a manipulator of the illusion).

Anna said...

@Kooiti MASUDA
"The Hockey Stick Illusion" actually is a book which tells the story about how McIntyre and McKitrick investigated Mann's hockey stick graph, and found that both the method and the data wasn't very good :-)

Try reading it. (I will, but I just got it yesterday, so unfortunately I can't contribute that much to the discussion yet).

Leigh Jackson said...

TonyN 21
I linked to that particular blog as Reiner says that the book centres on McIntyre's criticism of MBH and the blog attempts to explain the technicalities as simply as possible. Reiner suggests that the book fails in this respect. If the book does not effectively address such technical counter-criticisms I would prefer to save my money.

itisi69 said...

@Reiner:
"It poses the immediate question of how these papers could have passed peer review initially when no one could replicate the results based on the initial information contained in the published papers (or based on later additionally material that was later released reluctantly).

The Wegmnan Report (http://climateaudit.files.wordpress.com/2007/11/07142006_wegman_report.pdf) gives the answer to that, firstly it stated that:
"We have been to Michael Mann’s University of Virginia website and downloaded the materials there. Unfortunately, we did not find adequate material to reproduce the MBH98 materials. We have been able to reproduce the results of McIntyre and McKitrick"
.
with regard to the peer review process it stated:
"It is important to note the isolation of the paleoclimate community; even though they rely heavily on statistical methods they do not seem to be interacting with the statistical community. Additionally, we judge that the sharing of research materials, data and results was haphazardly and grudgingly done. In this case we judge that there was too much reliance on peer review, which was not necessarily independent."

TonyN said...

Leigh #24:

I can't quite see where the header post says that the book 'does not effectively address such counter-criticisms'. Perhaps you could direct me to the right place in the text.

In fact the post ends:

… there was a counter attack calling McIntyre incompetent. This line of rhetoric seems to reappear at every turn of events, we have seen it in the emails, and we have seen it on this blog many times. This rhetoric can no longer be taken at face value by anyone who reads the book. The Hockey Stick Illusion will require a refutation (based on an alternative reading of events or new facts) if such claims were to be substantiated.

Which seems to be saying the complete opposite, and makes it even more surprising that Real Climate has had absolutely nothing to say about Montford’s book.

Björn said...

Unfortunately, I haven't read the book, and probably I won't have the time (am building a house soon...). There is, however, one thing I really find amusing about the M. Mann discussion. I would expect everybody with a degree in physics could have seen it right away: That the MBH hockeystick can not represent sound physics. When reading the 2001 IPCC report (TAR), it was the shape of the MBH curve that drove me from the environmentalists into the 'sceptics' camp. (BTW, what a notion. I would every natural scientist expect to be a sceptics by profession and desire. It is the sceptics regarding established scientific theories that produces new questions, leads to new answers and better theories.)

First, everyone with a ruler could draw a straight line through the warming periods 1910-40 and 1970-99 and recognize that there was no significant difference in the slope. However, TAR stated something of "unprecedented" warming, which was obviously not backed by data. Jones confirmed this in his recent interview 10 years thereafter. But again, everybody with sound training should be able to see this even without a ruler!

Second, it is extremely unlikely that any dynamical system produces observables with very low variance for 850 years, only to hop to higher and higher levels by 1950 -- still 'pre-industrial'. This is exactly what the hockeystick is about. No medieval warming, no little ice age, only a relatively flat curve from 1000 - 1850. And then alternating phases of increasing and constant temperatures, each ~30 years long. If there were AGW, it would, according to this graph, have started in the year 1850. However, the regularity with which warming and constant phases followed each other since then hinted to a natural cycle.

Since realizing this, I knew that the influence of CO2 on the climate must have been exaggerated and I was quite desparate that most of my esteemed colleagues did not notice the discrepancy between some of the TAR claims and the underlying data not supporting it. Why it took a McKintyre to demolish the MBH paper is really puzzling for me, but probably only due to my lack of understanding of human psychology...

Leigh Jackson said...

itisis69 @25
The Wegman panel may not necessarily have been independent.

"We have been to Michael Mann’s University of Virginia website and downloaded the materials there. Unfortunately, we did not find adequate material to reproduce the MBH98 materials.

We have been able to reproduce the results of McIntyre and McKitrick (2005b). While at first the McIntyre code was specific to the file structure of his computer, with his assistance we were able to run the code on our own machines and reproduce and extend some of his results."

Wegman needed McKintyre's personal help to reproduce the results of a very recent paper. Perhaps we should not be unduly surprised to find that Wegman was unable to reproduce a paper published 7 years before MM05b from web materials alone. It is worth noting, however, that Wahl and Ammann reproduced MBH98 with materials which were publicly available at the time after asking Mann a few clarifying questions.

After the Wegman Report was published Mann stated that Wegman made no attempt to contact him or his colleagues.

Deep Climate has some interesting background and comment on Wegman as reported by Richard Littlemore.
http://www.desmogblog.com/wegmans-report-highly-politicized-and-fatally-flawed

http://www.ucar.edu/news/releases/2005/ammann.shtml

eduardo said...

In #10 Marco wrote
'One could ask oneself why there is such a desire to attack a 12-year old paper, when many others have come out with similar results using different methodology'

Marco, could you be more specific and explain what 'similar results' mean for you ?
I am just curious, thank you

Marco said...

@Eduardo:
as Kooiti also already indicated: with very very few exceptions (Loehle and Moberg), all subsequent reconstructions fit quite well in the uncertainty range of MBH98.

eduardo said...

I havent read the book completely yet, just pieces of it here and there.
As far as I can judge the parts of the story in which I am involved are correctly described.

My impression is that the book is mostly about the 'political' hockey-stick the and Mann-McIntyre confrontation, with some other secondary characters, and less about the 'climatic' hockey-stick.

Some aspects I would not agree with. For instance Gabi Hegerl, Tom Crowley or Rosanne D'Arrigo would not belong to the team, although this depends of what the definition of 'the team' is. If this definition is the closer circle of scientist working with Michael Mann, I do not think they are.

eduardo said...

@ 30
Marco, thank you.

well, we have to be a bit careful in interpreting the uncertainty range. Let us focus on the range for a certain year t1. If the ranges are correctly estimated, the true temperature for that year will lie within that range with a certain probability. The same can be said for another year t2, etc. So if another reconstruction oscillates at random around the MBH reconstruction within the uncertainty ranges, both would be compatible. The critical word here is at random: this means for some years one is above or below the other. However, if a second reconstructions is systematically colder (or warmer) that the MBH reconstructions, even if it is always within the uncertainty range, they are not compatible. In other words, the probability that the error made by MBH is always positive (or negative) is very small.

Marco said...

@Eduardo:
Fully random is perhaps too much to ask, but there definately are reconstructions that are above and/or below for parts or the whole curve:
http://tamino.files.wordpress.com/2008/09/fig3.jpg
But maybe I'm just looking at the reconstructions different than you do.