Saturday
Mar162013
by 
Bishop Hill
Bishop Hill OMG
Mar 16, 2013 Climate: MWP Climate: McIntyre McIntyre's latest post on the Marcott hockey stick is simply astonishing.
Mar 16, 2013 Climate: MWP Climate: McIntyre 
Buy
Books
Click images for more details
Currently discussing
-
May 7 - Etavian
on
Launched my very own online store. -
May 7 - fabby
on
The Benefits of Using Tutorbin for Custom Writing Services
Links
A few sites I've stumbled across recently....
-
Science blogs
-
Political blogs
-
Other blogs
Subscribe
Copyright © 2006-2013, Bishop Hill All rights reserved.
Reader Comments (157)
Climate Science ... doing its best to try and make phrenology look a little more reasonable.
In the same spirit (thanks Bruce).....
Climate Science ... accomplishing what alchemy never could.
or.....
Climate Science ... astrology for the nerds.
Climate Science ... psychoanalysis for frustrated physicists.
Climate Science ... when homeopathy is not the cure.
Climate Science ...
[I suppose we can go on like this for awhile]
Can't wait for Rob Wilson's twopenny worth
I hope that the editors of "Science" are having a nice weekend - would love to be a fly on the wall :)
It is interesting to speculate just how much more analysis will be required before the editor of Science does something, such as withdraw the paper, issue a statement etc. On the other hand, since this is climate science, we will probably be told that the errors, discrepancies and data-juggling actually make no difference to the published result.
Having not paid too much attention to this Marcott issue I started to look around at the headlines associated with this paper, thinking it may be a marginal thing, but see now that quite a lot "useful" PR was generated:
True face of climate's hockey stick graph revealed
We're Screwed: 11,000 Years' Worth of Climate Data Prove It
Global warming is epic, long-term study says
Global temperatures are close to 11,000-year peak
Scientists Find an Abrupt Warm Jog After a Very Long Cooling
In this latter NYT link, you can see a video of the co-author Shakun taking to Revkin saying:
However now looking at McIntyre's graph showing the disappearance of the up-tick when using a correct conventional dating system, it starts to look like this BOOM! is a petard to be hoisted upon ;)
Re: OMG
Surely the paper will be withdrawn? From the Climastrology perspective, don't they get it yet? The thirteenth sign of the Zodiac - The Hockey Schtick - is the universal sign of the cursed - northern hemisphere (Mann et al. 1998), southern hemisphere (Gergis et al 2012) or global (Marcott et al. 2013).
The Abstract states:
"Surface temperature reconstructions......suggest that recent warming is unprecedented in that time."
and:
"Current global temperatures of the past decade.........are warmer than during ~75% of the Holocene temperature history."
and:
"Intergovernmental Panel on Climate Change model projections for 2100 exceed the full distribution of Holocene temperature under all plausible greenhouse gas emission scenarios."
The Editor's summary states:
"The climate has been warming since the industrial revolution, but how warm is climate now compared with the rest of the Holocene? Marcott et al. (p. 1198) constructed a record of global mean surface temperature.....using a variety of land- and marine-based proxy data from all around the world."
and:
"Temperatures have risen steadily.......with a global temperature higher than those during 90% of the entire Holocene."
Are the Editor and Author's talking about the same paper?
I'm going out on a layman limb here but this smells like McIntyre et al have indeed spotted another Gergis. This time without offering the option for the authors to retract claiming to have spotted the error themselves;)
Thinking of the thesis of the book Outliers, I feel sure that Steve McIntyre is way past his 10,000 hours. He is now a certified genius at spotting this stuff. McIntyre has put in the time to get:
knowledge of the data,
knowledge of the correct techniques,
knowledge of the pathological tactics and thinking that drive this kind of work.
And yet his name is spat out (or ignored) like the curse of Voldemort by the "science" and media supporters.
Sweet! Keep it up you charlatans! ;)
"Climate "science", the grift that keeps on giving..." *
* Hockey sticks are crooked, after all.
Based on further comments over at CA it looks like my layman's take on what they did ('perturbation') was in essence to take the plots of their proxies, trace over the line with a fat Sharpie (more or less obliterating any confidence-interval boundaries in the process), then take the resulting scrawls and shuffle them individually back and forth on the timeline a bit until the product came out suitably Mannish.
Whether this was done for what they considered to be valid statistical reasons or whether it was done just to keep reviewers at arm's length ('that's too damn much work to replicate, we'll just pass it') it astonishes me that no one appears to have asked them to defend their methods.
a tangent, but this could point to some important further 'auditing' I suspect:
Look at how Marcott's co-author Jeremy Shakun describes their approach to studying proxies for their Shakun et al. (2012) in "Nature" last year. I think they will rue the day that Shakun described their approach to developing a complex multi-proxy study as "let's just slap them together" because it is such "simple science" (this says a lot about the research standards that Michael Mann and Phil Jones have promoted for multi-proxy studies, perhaps??) ---
Marcott co-author Shakun on their approach to the Shakun et al. (2012) study
[emphasis added]
Maybe I've misinterpreted McIntyre's assessment but would I be wrong in believing that his latest findings provide clear evidence of scientific fraud?
Dave Salt, I'm not sure Steve's posts imply more than careless or even reckless behavior perhaps, but not necessarily any intentional malfeasance. I know that Steve Mc is highly resistant to speculations about the F-word and prefers to detail what can be known from his careful analysis of data.
Skiphil, I know Steve runs a 'tight ship' over at CA and very much appreciate his neutral approach to assessing the data, so I don't believe he would ever make such a direct accusation.
However, I do wonder how much more 'evidence' it would take to demonstrate that this is more than just "careless or even reckless behaviour"?
The half-life of papers purporting to show recent warmth based on proxies is declining. In the good old days you could siphon off a few grants, go to a conference in Bali, and collect a Nobel, before anyone noticed that you had the numbers backwards...
I picture a whack a mole machine at a fair with hockey sticks instead of moles and McIntyre driving up with a steam roller.
@ David Salt
"However, I do wonder how much more 'evidence' it would take to demonstrate that this is more than just "careless or even reckless behaviour"?"
As they say in the TV whodunnits: "A confession now would be nice"
Never going to happen
And the Marcott paper will be quoted in the AR5 as there is no time now left for any to-and-fro rebuttal stuff in Science (even if the self-serving editor permitted that anyway)
And the Marcott paper will be quoted in the AR5 as there is no time now left for any to-and-fro rebuttal stuff in Science (even if the self-serving editor permitted that anyway)
Mar 16, 2013 at 9:55 PM ianl8888
It might be in time for inclusion but they won't use it if it's been comprehensively trashed on the net. The rebuttals would be out in time for AR5 to be published.
re: IPCC and Marcott et al (2013)
Here is a fascinating blog commentary on the Marcott study, because it appears to be by the same Prof. RA Brown who recently commented critically at CA (on a different matter), and who is at Dept. of Atmospheric Sciences, the Univ. of Washington. [h/t John M at CA]
I find it suggestive that he twice refers to the Marcott study as having an IPCC imprimatur (see where he says "IPCC study" and "IPCC report"). As a climate science insider, Prof. Brown may have an interesting perspective on why he portrays the Marcott study as already an "IPCC study" when AR5 has not been finalized and the Marcott paper has only just been published. Did he have specific information that that Marcott was already an "IPCC study" (whatever that means)??
RA Brown on Marcott et al. (2013) and the triumph of the IPCC hockey stick
[emphasis added]
As a fellow Canadian, just let me tell the world that it is real cool to bask in the reflected glory and watch expert rectal surgery performed by Steve McIntyre!
Let's hope that the authors, reviewers, Science, Mann, Revkin et al are also enjoying the experience.
The next few days will be fun!
I would like to see someone like Lawson or Monckton send this as an open letter to Sir Paul Nurse and ask him to comment...
Skiphil -
It is indeed odd that Marcott et al. is referred to as an IPCC study in that post. As pointed out above, the paper narrowly met the criteria for inclusion in AR5 WG1, both for submission and publication date. However, I've taken a peek at the leaked Second-Order Drafts, and Marcott is not mentioned therein. I suppose there is no rule to prevent inclusion of a new citation after the last review; there seem to be few real rules.
Well, I guess they weren't telling a lie when they said "not robust".
Like a stone through a wet paper bag.
Indeed it is astonishing. But most people won't understand what is going on because of his ironic style. The two most recent CA posts explain that (a) (Marcottian upticks post) the spurious upticks were simply generated by different series running out at different times; imagine taking the average of two flat data sets, one that goes
-1 -1 -1 -1
and another that runs out sooner
3 3 3
The average of the two series is
1 1 1 -1
generating a completely spurious downward spike.
(b) (dating service post) Obviously if the dates of the two series were adjusted slightly so that the higher one lasted longer, the spike would go up instead of down.
When the dust has settled there will be a need for somebody to explain all this simply and calmly and clearly, if we can think of anybody whois good at this sort of thing.
TLITB, I think this is much worse than Gergis. I have to agree with Dave. They cannot pretend they did not know exactly what they were doing - particularly in view of the Marcott thesis showing no spikes.
As promised I submitted a brief comment on the paper to Science, among other things calling for the withdrawal of the paper. I did this earlier this morning, before the two most recent CA posts.
Its got to rise to the level of fraud or something like that to get the attention of the lib MSM. The thing is that the story of the incontrovertible evidence of unprecedented modern global warming ran night and day it seemed for weeks. That's all the people are going to know. Insane.
I've watched you getting into a lather for the last week over Marcott et al and fail to see what the fuss is about.
I first saw graphs showing this Holocene temperature curve in the early 1970s. At the time I was a student counting pollen grains in peat samples for the paleobotanists.
Marcott et al have improved the resolution since my day, but there's nothing really new in it.
EM...so the anomalous differences between the proxy data in the thesis and in the Scence paper hold no surprises for you? Just how large would the differences be before you could get interested? Or are you still searching for temperature data where it is not being measured in deep oceans?
"I've watched you getting into a lather for the last week over Marcott et al and fail to see what the fuss is about."
EM disqualifies himself as a poster worth reading.
Yes - I suppose that's why the authors were so self-effacing & laid back about it....
(HT tlitb)
I'm not really interested in microanalysing the data handling of the paper. If his data analysis technique has improved since he wrote his thesis, that's what experience is for.
I'm more interested in the temperature pattern of the Holocene. The graph shows a pattern to be expected of an interstadial at this stage, except for the upward twitch at the present end.
Do you regard the shape of the temperature graph as wrong?
If so, what shape should it be?
On what evidence?
EM global temperature could be replotted very differently by using select temperature stations. You could probably even create a cooling world. Even the CET would give you 'normal' temperatures right now.
Merging proxies is bad enough but adjusting the dating the original scientists assigned them would require a strong justification. Personally I'd rather see proxies grouped in geographical locations eg Europe rather than an attempt to create a global temperature.
If you cared about the integrity of climate science you wouldn't dismiss the spurious appearance of a hockey stick. Unless it's unshakable the uptick is pure propaganda.
There's a whole lot of going on in Shakun (and pal).
"Do you regard the shape of the temperature graph as wrong?"
As noted elsewhere - it is the same shape as the hungry caterpillar. Nothing wrong with that.
TinyCO2
I have not yet had access to the full text of Marcott et al, so I cannot comment on the minutae. From the effort the sceptic propoganda sites are putting into trying to discredit the paper, its conclusions must have an important bearing on your case, though I fail to see the significance of a pretty standard Holocene temperature curve.
From the reports I have read, the conclusion of Marcott et al matches other sources and the work I was associated with in my youth.
I still dont understand the fuss. Had you all expected a flat temperature record or something weirder?
Skiphil Mar 16, 2013 at 10:12 PM
The only Rab I know is this one:-
http://www.youtube.com/watch?v=8k7VoFiagfs
Seems to be a kindred spirit of the one you refer to?
C U Skip!
TinyCO2
Ther is plenty of more detailed work on more local climates using individual proxies available in the published literature. This is one example.
http://www.sciencedirect.com/science/article/pii/S003358949791961X
If you want details of individual proxies used in the Marcott et al, there should be a considerable reference list at the end of it.
I'm more interested in the temperature pattern of the Holocene. The graph shows a pattern to be expected of an interstadial at this stage, except for the upward twitch at the present end.
Which shows you aren't interested in paying even the slightest bit of attention to the criticisms. You like Hockey Sticks, regardless of how much sense they make. Rather than defend them, which is impossible, you merely pretend that the matter isn't even interesting. It's what little kids do when they can't defend an argument.
The uptick is not at the present. Amazingly, it's pre-1950, and ridiculously large. Are you trying to say that you believe global temperature rose by over a degree between 1900 and 1950? You'll be pretty much alone in that view, since not even the biggest warmists hold it to be true.
Entropic man
"If you want details of individual proxies used in the Marcott et al, there should be a considerable reference list at the end of it."
No need to look too far, a very kind gentleman is rearrange them into what will probably end up being known as a well known phrase or saying:-
So, are we sitting comfortably? Then we shall begin:-
"Marcott Mystery #1"
http://climateaudit.org/2013/03/13/marcott-mystery-1/
"How Marcottian Upticks Arise"
http://climateaudit.org/2013/03/15/how-marcottian-upticks-arise/
The Marcott-Shakun Dating Service
http://climateaudit.org/2013/03/16/the-marcott-shakun-dating-service/
I doubt that the above will be exhaustive, so prudence, experience, scientific logic plus a little commonsense suggests that "radio silence" might well be the order of the day?
The uptick is not at the present. Amazingly, it's pre-1950, and ridiculously large. Are you trying to say that you believe global temperature rose by over a degree between 1900 and 1950? You'll be pretty much alone in that view, since not even the biggest warmists hold it to be true.
Mar 17, 2013 at 12:32 AM | Mooloo
The only graph I've examined so far has a horizontal scale of 4000 years to the inch. Could you link to your high resolution information, to allow me to examine the data.
Green Sand
Your links are all to opinion pieces on propoganda sites. I was referring to the original published papers from which Marcott et al sourced their proxy data
EM - If you'd been following the detail, you would know that the series ends at 1950 - it is not contaminated by any modern datapoints at all. You could be right if the data included 2000, but it does not.
If you refuse to look at sites where the proxies are collated, and the interesting average of the endpoints, you'll need to do your own legwork.
Entropic man
My links are all to somebody carrying out an investigation into the actual "details of individual proxies used in the Marcott et al" exactly what you suggest should be done!
Are you saying that only certain "approved" people should have or have such ability?
Is it your view that authority is all?
If so say so, let's know just how uncertain this "Entropic" is?
Dave Salt, even when scientists get angry enough to be scathing about this kind of thing, it seems to me that they seldom use the word "fraud" for it. "Scientific fraud" has a narrower connotation than merely being misleading, or even being clearly dishonest. E.g., I have never seen a scientist choose "scientific fraud" as a description of dishonestly paraphrasing the second law of thermodynamics, and I rather expect I never will. (And it's not because they don't get angry about it!) You might also find it surprisingly difficult to find scientists choosing "scientific fraud" as a term to condemn some kinds of misleading analyses in bad pharmaceutical studies. And for that matter, note that "fraud" isn't usually used by the people who use "denier". It's not just because of a desire to minimize how wrong the behavior is supposed to be. It seems to me that it's because scientific fraud has come to have a narrower connotation of messing with the original data or experiment at a fairly basic level. Someone who constructs a fake perpetual motion machine would be called a fraud. Someone messing with the truth at a higher level of abstraction might (if an apolitical critic chooses not to mince words) be called "openly dishonest" or "charlatan" or other things, but seldom "fraud".
(See the classic nuking of SJ Gould by Tooby and Cosmides, http://cogweb.ucla.edu/Debate/CEP_Gould.html in which we find characterizations like "transcend being completely wrong by aspiring to achieve the reverse of the truth (e.g., locating the orbit of the moon within the Earth's core)" but the word "fraud" does not appear. Or see http://lesswrong.com/lw/kv/beware_of_stephen_j_gould/ where "scoundrel" appears, but not "fraud". You could perhaps get those same authors to choose the term "scientific fraud" today, but I bet they'd limit it to the narrow fact (see http://blogs.nature.com/news/2011/06/did_stephen_jay_gould_fudge_hi.html from after Gould's death) that Gould published systematically biased measurements in order to support his case against one of the racists he attacked in _The Mismeasure of Man_.)
Trolls come in many shapes and sizes but one thing that they all have in common is enjoying working up a good lather.
Mr. Entropic,
You should start the next stage in your pedagogical mission by educating distinguished climate scientist Michael Mann on why he was wrong to trumpet the Marcott et al. (2013) as an "important paper."
Since he was so confused as to emphasize the importance of this paper he is one who stands greatly in need of your help:
Michael Mann trumpeted Marcott et al. (2013) as an "important paper"
[emphasis added]
It's time to stop being surprised or outraged at Team members and wannabes manipulating the data to serve their agenda.
They are activists; it's what they do.
AR5 needed a boost quickly; Marcott supplies it willingly.
What I fear is expressed in my response to Mindert Eiting at ClimateAudit:
"Thanks for bringing up data – as in “facts.” Work with proxies is so far from fact that the crucial issues in the Marcott controversy do not touch upon fact at all. Even the critics, first rate critics such as McIntyre and Telford, agree that the dates can be changed and legitimately. Clearly, then, the entire discussion is over what is “proper” in the relevant statistical methodology. Whenever proxies for temperature are the topic, scientists believe that they are quite justified in failing to tie their inferences to any factual ground at all. In my humble opinion, the lack of empirical science in the study of proxies is exactly why Warmists love them."
So the argument will come down to the width of the uncertainty bars and Warmists will have free reign to publish the "NSF Hockey Stick" without fear of embarrassment. Use of proxies for temperature is exactly as scientific as Freudian psychology. In the case of Mann's Hockey Stick, this was demonstrable from Briffa's worries about changes in the proxies, changes that were hidden.
The sort of thing any "interested" commentator can find with two minutes of google time:
//
4. Tabs containing raw data for each proxy record. Age models, proxy values, and calibrated temperatures are shown on the left side of each tab, and radiocarbon data are given on the right side.
//
http://www.sciencemag.org/content/339/6124/1198/suppl/DC1
Entropic Man,
You don't even have read anything Steve McIntyre writes. He posts the graph from the Science paper and then the same graph from Marcott's thesis.
Here:
http://climateaudit.org/2013/03/14/no-uptick-in-marcott-thesis/
McIntyre then shows the key parts blown up for those with impaired vision. The added blade of the hockey stick is immediately obvious. Just look at the pictures.
The authors have a lot of explaining to do.