Andrew Montford‘s The Hockey Stick Illusion is one of the best science books in years. It
exposes in delicious detail, datum by datum, how a great scientific mistake of immense political weight was perpetrated, defended and camouflaged by a scientific establishment that should now be red with shame. It is a book about principal components, data mining and confidence intervals—subjects that have never before been made thrilling. It is the biography of a graph.
I can remember when I first paid attention to the ―hockey stick‖ graph at a conference in
Cambridge. The temperature line trundled along with little change for centuries, then shot through the roof in the 20th century, like the blade of an ice-hockey stick. I had become somewhat of a sceptic about the science of climate change, but here was emphatic proof that the world was much warmer today; and warming much faster than at any time in a thousand years. I resolved to shed my doubts. I assumed that since it had been published in Nature—the Canterbury Cathedral of scientific literature—it was true.
I was not the only one who was impressed. The graph appeared six times in the
Intergovernmental Panel on Climate Change (IPCC)‘s third report in 2001. It was on display as a
backdrop at the press conference to launch that report. James Lovelock pinned it to his wall. Al Gore used it in his film (though describing it as something else and with the Y axis upside
down). Its author shot to scientific stardom. ―It is hard to overestimate how influential this study has been,‖ said the BBC. The hockey stick is to global warming what St Paul was to Christianity.
Of course, there is other evidence for global warming, but none of it proves that the recent warming is unprecedented. Indeed, quite the reverse: surface temperatures, sea levels, tree lines, glacier retreats, summer sea ice extent in the Arctic, early spring flowers, bird migration, droughts, floods, storms—they all show change that is no different in speed or magnitude from other periods, like 1910-1940, at least as far as can be measured. There may be something unprecedented going on in temperature, but the only piece of empirical evidence that actually says so—yes, the only one—is the hockey stick.
And the hockey stick is wrong. The emails that were leaked from the University of East Anglia late last year are not proof of this; they are merely the icing on the lake, proof that some of the
scientists closest to the hockey stick knew all along that it was problematic. Andrew Montford‘s
book, despite its subtitle, is not about the emails, which are tagged on as a last chapter. It is instead built around the long, lonely struggle of one man— Stephen McIntyre—to understand how the hockey stick was made, with what data and what programs.
A retired mining entrepreneur with a mathematical bent, McIntyre asked the senior author of the hockey stick graph, Michael Mann, for the data and the programs in 2003, so he could check it himself. This was five years after the graph had been published, but Mann had never been asked for them before. McIntyre quickly found errors: mislocated series, infilled gaps, truncated records, old data extrapolated forwards where new was available, and so on.
Not all the data showed a 20th century uptick either. In fact just 20 series out of 159 did, and these were nearly all based on tree rings. In some cases, the same tree ring sets had been used in different series. In the end the entire graph got its shape from a few bristlecone and foxtail pines in the western United States; a messy tree-ring data set from the Gaspé Peninsula in Canada; another Canadian set that had been truncated 17 years too early called, splendidly, Twisted Tree Heartrot Hill; and a superseded series from Siberian larch trees. There were problems with all these series: for example, the bristlecone pines were probably growing faster in the 20th century because of more carbon dioxide in the air, or recovery after ―strip bark‖ damage, not because of temperature change.
This was bad enough; worse was to come. Mann soon stopped cooperating, yet, after a long
struggle, McIntyre found out enough about Mann‘s programs to work out what he had done. The result was shocking. He had standardised the data by ―short-centering‖ them—essentially
subtracting them from a 20th century average rather than an average of the whole period. This
meant that the principal component analysis ―mined‖ the data for anything with a 20th century
uptick, and gave it vastly more weight than data indicating, say, a medieval warm spell.
Well, it happens. People make mistakes in science. Corrections get made. That‘s how it works, is
it not? Few papers get such scrutiny as this had. But that is an even more worrying thought: how
much dodgy science is being published without the benefit of an audit by Mcintyre‘s ilk? As a
long-time champion of science, I find the reaction of the scientific establishment more shocking than anything. The reaction was not even a shrug: it was shut-eyed denial. If this had been a drug trial done by a pharmaceutical company, the scientific journals, the learned academies and the press would have soon have rushed to discredit it—and rightly so. Instead, they did not want to know. Nature magazine, which had published the original study,
went out of its way to close its ears to McIntyre‘s criticisms, even though they were upheld by
the reviewers it appointed. So did the National Academy of Sciences in the US, even when two reports commissioned by Congress upheld McIntyre. So, of course, did the IPCC, which tied itself in knots changing its deadlines so it could include flawed references to refutations of McIntyre while ignoring complaints that it had misquoted him.
The IPCC has taken refuge in saying that other recent studies confirm the hockey stick but, if you take those studies apart, the same old bad data sets keep popping out: bristlecone pines and all. A new Siberian data series from a place called Yamal showed a lovely hockey stick but, after ten years of asking, McIntyre finally got hold of the data last autumn and found that it relied heavily on just one of just twelve trees, when far larger samples from the same area were available showing no uptick. Another series from Finnish lake sediments also showed a gorgeous hockey stick, but only if used upside down. McIntyre just keeps on exposing scandal after scandal in the way these data were analysed and presented.
Montford‘s book is written with grace and flair. Like all the best science writers, he knows that the secret is not to leave out the details (because this just results in platitudes and leaps of faith), but rather to make the details delicious, even to the most unmathematical reader. I never thought I would find myself unable to put a book down because—sad, but true—I wanted to know what happened next in an r-squared calculation. This book deserves to win prizes.
Oh, and by the way, I have a financial interest in coal mining, though not as big as Al Gore has in carbon trading. Maybe you think it makes me biased. Read the book and judge for yourself.
The Hockey Stick Illusion is published by Stacey International, 482 pages, £10.99 Read what James Lovelock, Bjorn Lomborg, Ed Miliband and many other experts have to say
about climate change in Prospect’s Copenhagen special
Section: Web exclusive Subject: Environment Tags: Al Gore, best science, bird migration, canterbury cathedral, confidence intervals, droughts, evidence for global warming, gaspé peninsula, hockey stick, intergovernmental panel on climate change, James Lovelock, montford, principal components, sceptic, science books, sea ice extent, sea levels, spring flowers, summer sea ice, surface temperatures
Add Comment
Name*: EMail*: Website:
Comment*:
Type the two words:Type what you
hear:Incorrect. Try again.
Submit Comment
Comments (48):
1.
2.
Phillip Bratby says:
March 11, 2010 at 8:23 am
I entirely agree with Matt. I have now read The Hockey Stick Illusion twice; it is compulsive reading.
John Hewitt says:
March 11, 2010 at 10:54 am
I too have read this book twice. I was a believer in GW, simply because I had not done any research myself. This book finally revealed to me the shocking deception that is AGW and the lengths that its supporters are prepared to go to, to perpetuate the myth. It should be essential reading for all politicians especially Ed Milliband.
3.
Roddy Campbell says:
March 11, 2010 at 12:09 pm
Matthew, excellent review, excellent book. Did you see my CRU piece in March Prospect?
Your best line (since it agrees with my piece!) is ―There may be something
unprecedented going on in temperature, but the only piece of empirical evidence that actually says so—yes, the only one—is the hockey stick.‖
I don‘t know why people have such trouble with the ‗unprecedented‘ word being fairly
critical.
4.
TG O'Donnell says:
March 11, 2010 at 6:59 pm
Good heavens, what‘s happening down at Prospect Villas?
First some gentle questioning from Roddy Campbell, now several more direct and challenging assertions from Matt Riddley.
Do these straws in the wind herald a sea change in Prospect‘s hitherto smooth cycle of
appropriation, digestion, assimilation and ultimate regurgitation of AGW orthodoxy?
David Goodhart‘s editorial comment ―A proper scepticism towards the data is not only legitimate, but necessary, before we change the way we live‖ gives some grounds for
hope. Now perhaps Prospect will encourage debate on a topic of profound importance, where specialists can argue interpretations of the science and where the massive economic and human consequences of political action (or inaction) can be calmly evaluated. (Were this to happen, I could happily renew my subscription (cancelled in November in the wake of the 22 page non-feature on the Copenhagen Conference) in the anticipation of being informed, challenged and perhaps ultimately persuaded on this most urgent of issues).
Finally, a ringing endorsement for Matt‘s views on the Hockey Stick Illusion. This is a
splendid and (surprisingly) a most exciting book. It will make a profoundly depressing read for any scientist as also for those who harbour the view that the scientific establishment can be trusted to make the right decisions when their research funding, their pride or their peer status is at risk. TG O‘Donnell
5.
Comment via Facebook says:
March 11, 2010 at 7:49 pm
Sam Carson:
―And yet consensus isn‘t uprooted.
The article calls the hockey stick a central plank to communicating the science – then
says that has been under scrutiny by someone who isn‘t a climate scientist.
Therefore, one man has an opinion and that is counter to the vast majority who specialise
in that field. That‘s great. Good for him. … See More Climate scientists themselves aren‘t abandoning the thesis in great droves, the ―gotcha‖
point is little more than the perfectly plausible idea that the mechanics of the global
climate can‘t be easily explained. It‘s complicated and we don‘t know enough to make
any great claims. Certainly not enough to make that greatest claim of all, that we are at all safe from the effects of climate change. There is still reason to suspect that digging up sequestered carbon from under the earth and injecting it into the atmosphere at the rate of tonnes per
day could be of consequence.‖
6. 7.
ben says:
March 12, 2010 at 12:34 am
Superb piece, Matt. Just superb. McIntyre is a hero for his work.
Phil Howerton says:
March 12, 2010 at 3:22 am
Sam Carson: You need to realize that the Hockey Stick was created by mathermatical formulas and statistics. Steve McIntyre, a brilliant mathematician and statistician, is more than qualified to critique the works of Mann and his cohorts, none of whom are qualified statisticians.
8.
Girma says:
March 12, 2010 at 7:52 am
The hockey stick is not made of the same material. The handle is made of proxy data and
the blade of thermometer data. The main problem is that the recent proxy data don‘t show
the current thermometer warmings. If this the case, how can we thrust the proxy data along the handle of the hockey stick? Here is what Professor Michael E. Mann said about releasing their research data: WE DO SO AT OUR OWN PERIL! This is not from second hand source; it is from his own email:
http://www.tuxwerx.com/Climategate/mail/1076359809.txt
9.
early morning bird says:
March 12, 2010 at 7:59 am
Right, so have decided to go for the ―raw data‖ on this one. Pity my daughter who will
now almost certainly spend the weekend watching Snowwhite and the seven dwarfs over
and over again. Pre-ordered ―The Lomborg Deception‖ by Howard Friel in a vague
attempt to hear submissions from both sides. According to this book Lomborg has also
been fiddling the numbers! Oh how annoying. Why can‘t people use the data/information to get to the answer rather than start out with the answer and then ―chop toes until the shoe fits‖?
10.
RobertD says:
March 12, 2010 at 11:16 am
early morning bird, you‘ve ordered the wrong book then. Lomborg deliberately makes no
attempt to question the assertions of the IPCC – he assumes (and believes) that the change model they publish is absolutely correct, but then uses that as input to economic analysis. He argues that the *policy* conclusions of the IPCC science are wrong. By contrast, McIntyre began the process of demonstrating how fundamentally flawed the *scientific* process of the IPCC was.
The way Lomborg has been treated by publications like Scientific American, and by NGOs like the WWF is a scandal in itself that ought to bring great discredit on those institutions.
11.
Douglas Cohen says:
March 12, 2010 at 12:32 pm
You might find the article \The Big Science Poker Game\ useful as a way of seeing how the moral failings of Climate scientists are shared by other researchers.
http://climaterealists.com/index.php?id=5221
The news article
http://www.sciencedaily.com/releases/2009/06/090621195620.htm
shows that this sort of self-serving scientific sloppiness occurs many places outside of climate science — such as paleontology. After reading this description, consider how it could be possible that such an obvious check on the original paleontological finding could have been overlooked the way it was.at ar
12.
13.
Comment via Facebook says:
March 12, 2010 at 1:02 pm
Matt Rooney:
―Jesus. Prospect IS turning into the Spectator.‖
Comment via Facebook says:
March 12, 2010 at 1:03 pm
Martin O‘Neill ―Very depressing that Prospect is running drivel like this. Time to let my subscription lapse.‖
14.
Comment via Facebook says:
March 12, 2010 at 1:03 pm
Simon McGrath
―fascinating that the comments above don‘t actually address what the review is saying. Indeed they illustrate the point the the hockey stick is regarded as a given and can‘t be questioned.‖
15. 16.
Comment via Facebook says:
March 12, 2010 at 1:04 pm
Douglas Clark:
―I suppose it is what you expect from the ex-chairman of Northern Rock.‖
Comment via Facebook says:
March 12, 2010 at 1:05 pm
Simon McGrath:
―any thoughts on what the review said or is his business career enough to mean you can ignore it?‖
17.
Comment via Facebook says:
March 12, 2010 at 1:05 pm
Risto Pyykkö
―It is disappointing to see Prospect starting to peddle illiterate nonsense like Ridley‘s conspiracy theories.‖
18.
Steve Milesworthy says:
March 12, 2010 at 4:40 pm
This is not a review. This is an uncritical repetition of the book‘s claims. Surely a proper
review would look at the context of the book a little more. The context being an attempt to wholly undermine an area of science by picking at a few flaws.
It doesn‘t seem likely that the curve is an accurate representation of 1000 years, and
certainly Mann comes across as a little too belligerant. But in the 12 years since the curve was published there is no seriously researched evidence that the current period was globally cooler than the Medieval warm period.
19.
Dave says:
March 12, 2010 at 8:35 pm
Steve Milesworthy:
―in the 12 years since the curve was published there is no seriously researched evidence that the current period was globally cooler than the Medieval warm period.‖
Er, there is. What Steve McIntyre has done is to take the data that has been used to claim the world is warmer, and show that in fact it shows no such thing – in other words, that it
is ‗globally cooler than the Medieval warm period‘. If that data is sufficient to prove the
argument one way, it ought to be sufficient to prove it the other way.
You can argue that it‘s not ‘serious research‘, but, since he‘s a proper scientist, McIntyre
has published his full workings. You can check them yourself if your maths is good enough, or you can be satisfied that despite all his critics, no-one has been able to find
anything wrong with his maths. You can be sure we‘d have heard about it if they had. Instead, there‘s endless ad hominem attacks, and arguments that he has misunderstood
the science. Recaptcha: Political Petard – how appropriate.
20.
21. 22.
DANIEL_SEXTON says:
March 12, 2010 at 10:19 pm
Hope this helps:
http://www.realclimate.org/index.php/archives/2004/12/myths-vs-fact-regarding-the- hockey-stick/
DANIEL_SEXTON says:
March 12, 2010 at 10:22 pm
And this:
http://www.realclimate.org/index.php/archives/2005/02/dummies-guide-to-the-latest- hockey-stick-controversy/
Francis Voller says:
March 13, 2010 at 12:58 am
For all those that keep hamering home the notion that a majority of scientists are not leaving the AGW hypothesis consider this: The vast majority of these scientist study the effects of AGW, based on the considered
―rock solid‖ data sets that are availabe from places like CRU or papers like Manns‘s
