Quote of the week – CRU is now a candidate for publication in the Journal of Irreproducible Results

From Andrew Orlowski, the Register UK:

In Parliament’s enquiries into the Climategate Affair, Graham Stringer MP was surprised to learn that the CRU team couldn’t produce the same result twice.

“When I asked Oxburgh if [Keith] Briffa [CRU academic] could reproduce his own results, he said in lots of cases he couldn’t,” Stringer told us. “That just isn’t science. It’s literature. If somebody can’t reproduce their own results, and nobody else can, then what is that work doing in the scientific journals?”

Full story (well worth the read, please support the online advertiser driven Register by visiting) here (h/t to WUWT reader “Rational Debate”).

In case you’ve never seen it, the Journal of Irreproducible Results actually exists, and can be found here. They have an interesting book,  Extreme Weather as it is linked to “global warming”.

Advertisements

54 thoughts on “Quote of the week – CRU is now a candidate for publication in the Journal of Irreproducible Results

  1. I’m pleased to be from Manchester when i see Graham Stringer MP fighting the lonely fight. We have a proud history of independent thought.

  2. I… I just don’t understand.

    I can see, if you do an experiment and take measurements to get a result, and then try again and can’t reproduce it, how that’s not reproducable.

    But how does that work in dendropaleoclimatology or whatever it’s called now? You have a bunch of measurements. You put them through some bunch of statistics. You get an answer. How can that be unreproducable? How can you get different results? Or does it mean that if they go out and sample more trees they get different answers?

  3. I thought you had to be as stubborn as a mule to publish there.

    I’m groaning on the inside.

  4. [snip – this is nothing more than trolling for donations, which is the first thing that pops up “in your face” on your website, further, your email address (ending with @twitter.example.com) is fake, so, I’m putting you on the automatic moderation list for double blog policy violation, further posts will not appear without the express approval of a moderator – Anthony]

    .

  5. “If somebody can’t reproduce their own results, and nobody else can, then what is that work doing in the scientific journals?”

    Whoa!

    That’s a brief, but adequate, description of much of the CAGW by CO2 concept, is it not?

  6. “Or does it mean that if they go out and sample more trees they get different answers?”

    Probably, that’s what he’s referring to. That would be the reason why they have to cherry pick the right data to get the ‘right’ results.

  7. Ah yes,

    Tea leaves and tree rings – and shoes and ships and sealing wax – and other fanciful things…

    Mike

  8. Andrew Orlowski has been writing interesting bits on climate for El Reg for a while (he writes some other very good stuff on IT as well) and it is interesting to see how the comments have changed over the past couple of years. From unmitigated bile (well, it is a site that encourages flames) to a lot more support for the skeptical viewpoint.

    As always, commentors on web-sites are a self-selecting bunch so this is hardly a representative survey, but the change is note-worthy.

  9. Results in this field are strongly influenced by the choice of datasets and choice of processing of those datasets.

    If you aren’t meticulous in documenting (and archiving copies of) your choice of data, the code you use to manipulate that, the tunable parameters to those routines that you might supply at runtime, and any manual adjustments you decide are justified (even if that justification is thin), you’ll run into trouble with reproduction.

  10. I think that’s the idea. Run your model until you get the results you like then publish the results. Hide the data, call anyone who can’t reproduce your results an evil, shill for the nasty oil companies. Then cash your grant checks from WWF, Green Peace, NOAA, BP, etc.

  11. And herein lies the true undoing of the scientific “consensus.” What, precisely, does the consensus believe – the data (not the models) are open to revision to align with expectations? Such an acceptance further classifies that “literature” as fiction.

    Data can neither be created nor destroyed*

    *Note: Except at the CRU where full exoneration and affirmation of the resulting science are obligatory to maintain the sanctity of consensus.

  12. To the person who said (in the leaked CRU emails):

    “OH [frig] THIS. It’s Sunday evening, I’ve worked all weekend, and just when I thought it was done I’m hitting yet another problem that’s based on the hopeless state of our databases. There is no uniform data integrity, it’s just a catalogue of issues that continues to grow as they’re found.”

    Someone should buy that person a beer as that person has personal integrity – about the only person who has show any from that group. That and the person who leaked the emails ;)

  13. “If somebody can’t reproduce their own results, and nobody else can, then what is that work doing in the scientific journals?”

    The simple explanation is outright it’s fraud as evidences in the Climategate data dump!

  14. CRU seem to be good at losing data, as Dr Jones admitted. I expect that, one winter day, somebody threw some logs on the fire……oops, we wanted to keep those tree-rings – now we won’t be able to reproduce the results. Damn!

  15. I have come across a very important technology called heterogeneous computing on that website, theregister, that many WUWT readers may be interested in. A great video explaining and demonstrating the technology can be watched here:

    http://channel9.msdn.com/posts/AFDS-Keynote-Herb-Sutter-Heterogeneous-Computing-and-C-AMP

    Briefly it provides some convincing arguments that we are now entering the second phase of the third revolution in computing, which is a super computer on every desk, and even every pocket.

    I will be heavily invested (intellectually) in this next year when the enabling software, called C++ AMP, comes out next year, and already have most of the skills required to get started.
    We might even be able to soon run our own climate model software on our home super computers, correctly adjusted to the proper values of CO2 climate sensitivity of course.

  16. But ladies and gentlemen… it’s post-normal science, so reproducibility doesn’t matter! What matters is that the political end justifies the “climate science” means–just make up any old thing and if critics don’t like it, call ’em “deniers”, or “flakes”, or whatever else comes to their pea-brained minds.

    About time this is getting a lot of attention. It has taken a long hard struggle by many struggling longer and harder than I have (kudos to those many salient folks), but it appears the battle toward re-establishing normalcy is beginning to bear fruit.

  17. Dave et al says:
    “If somebody can’t reproduce their own results, and nobody else can, then what is that work doing in the scientific journals?”

    Before jumping to conclusions about the “work”, don’t forget there is an equally valid conclusion: that the real culprit is that the journals are not the “scientific journals” they proclaim to be.

    “Why?” I ask myself do these “Scientific” journals need to scrape the bottom of the barrel and print this kind of rubbish? What is it that has happened to “science” whereby there isn’t enough decent science left to kick this rubbish out?

  18. Darn, they used up the “One Tree To Rule Them All”, and so can’t ever reproduce the results without that tree. Nobody expects an exact replication of things via proxy measurements, but an approximate independent replication is the minimum necessary for the data/theory to withstand the rigors of the scientific method. Maybe I will design a “One Tree To Rule Them All” coffee mug and put it up on my Zazzle store with all profits going to Anthony. It is “Oli Si Sai Veli Saes Eir” in Elvish.

  19. Dave says:
    July 7, 2011 at 7:39 am

    “If somebody can’t reproduce their own results, and nobody else can, then what is that work doing in the scientific journals?”

    The simple explanation is outright it’s fraud as evidences in the Climategate data dump!

    __________________________________________________________________

    or uninitialized variables in a program (and several other newbie programming errors) which depending upon programing languages used etc… can lead to no two runs producing the same results.

  20. It’s what ‘science’ is now. There must be many real scientists turning in their grave. Shame. isn’t it. After all, science is (was) the last refuge. Over here in the UK (and I suspect in many countries) our anti-drug policies are based on ‘feeling’ rather than science. When the chief scientist to advise our government on the mis-use of drugs pointed out that cannabis is less harmful than alcohol and tobacco, he was promptly sacked!

  21. When will the warmists admit, now that NAS has inadvertently confirmed coal does not contribute to AGW, they were wrong? Is Obama still intent on bankrupting the coal mining industry.

  22. Clearly, not only is the climate chaotic but so is the “science” at CRU. I want my tax payers money back from them. Breaking news in the UK is that Rupert Murdoch has set the standard for dealing with disfunctional organisations, he’s just closed down the News of the World (at last) following the exposure of their culture of phone hacking private citizens. I hope the University of East Anglia can learn something from that example concerning the future of CRU.

  23. I remember an article in the Journal of Irreproducible Results from the 70s that postulated North America would slide into the ocean because of the weight of all the National Geographic magazines in people’s basements. I tend to give that more credence than CAGW.

  24. Steve from Rockwood:
    That was “Harry”. Lots of other choice quotes from him, too, including pasting data sets from A into gaps at B, making values up, hard-coding adjustments to values, and much more.

  25. Back in the early ’70s, I was involved in the development of an economic (computer) model that used random numbers to generate frequency distributions in order to simulate what today is called ‘chaos’ – i.e. behaviour associated with complex non-deterministic systems. From reading posts on this website, I understand that today’s climate models do much the same thing. Now, back then, to generate a random mumber, you fed a seed to some function and got a set of random numbers. If the model was re-run with the same seed, the output was identical. However, today there may be a way to generate random numbers such that a given random number set cannot be re-generated at will. If this is the case, then it could well be the case that a given model run (and hence a group of model runs) cannot be reproduced.

  26. American Patriot says:
    July 7, 2011 at 9:44 am

    When will the warmists admit, now that NAS has inadvertently confirmed coal does not contribute to AGW, they were wrong? Is Obama still intent on bankrupting the coal mining industry.
    Of course! But now the focus will change from environmental concerns to “Safety Concerns” for the miners………..

  27. Richard Day says:
    July 7, 2011 at 10:18 am

    I remember an article in the Journal of Irreproducible Results from the 70s that postulated North America would slide into the ocean because of the weight of all the National Geographic magazines in people’s basements. I tend to give that more credence than CAGW.
    I rather liked the “Periodic table of Elephants” myself, although the “Cosmological Quasiconstant for the Standard Model” is one of my favorites.

  28. Gary Mount says:
    July 7, 2011 at 8:27 am

    “…a super computer on every desk, and even every pocket.
    ___________________
    Already have a 3.2 TFLOPS rig on my desk work bench, but hope to step up to a really fast machine next winter.
    Only run it during the winter- helps keep the house warm and will spin the electric meter at a rapid rate.
    Anyone can duplicate my machine, right now.

  29. son of mulder says:
    July 7, 2011 at 10:11 am
    (sorry Anthony – off topic a bit)

    BUT Murdoch has admitted that NOTW (News of the World) is/was a corrupt organization.
    He seems to think that the management of that corrupt organization is not in itself corrupt.

    BUT

    it is a modern day (political class) paradigm that MANAGEMENT sets the tone; ethos and morality of a company.

    How come if the NOTW is corrupt the then editor and current head honcho in News International is not tainted ?

    It’s sort of like finding that an academic organization can not replicate results; and blaming one individual instead of understanding that the whole organization must also be corrupt.

  30. I wonder how much lipstick they can possibly have left. ‘Cuz that thing sittin’ there still looks like a pig.

  31. H/T to Lewis Carroll – The Walrus and the Carpenter

    “The time has come,” the Walrus said,
    “To talk of many things:
    Of shoes–and ships–and sealing-wax–
    Of cabbages–and kings–
    And why the sea is boiling hot–
    And whether pigs have wings.”

  32. Brian_H:
    Some days I wonder if Harry leaked the emails. I am alone in that thought?

  33. Hmm. I thought that JIR had changed their name to “The Annals of Improbable Research”. The staff claim they are proud to be “AIRheads”. Perhaps that same description might be applied to CRU denizens?

  34. H/T to Mosher and Fuller whose book ‘Climategate: the CRUtape letters’ was posted to each member of the Science & Technology Committee of the House of Commons. Only Graham Stringer read it…

  35. Since this dendro stuff is all about collecting data and then running it through some statistical manipulations, I interpreted Briffa as meaning not that he couldn’t run the data through the stats and get the same results, but rather that the data no longer existed.

    That’s a serious problem, but it’s a different problem than the way some people are interpreting it here.

  36. @ The Ghost Of Big Jim Cooley

    “Over here in the UK (and I suspect in many countries) our anti-drug policies are based on ‘feeling’ rather than science. When the chief scientist to advise our government on the mis-use of drugs pointed out that cannabis is less harmful than alcohol and tobacco, he was promptly sacked!”

    And it was a very good thing indeed that he was sacked. It is not the job of scientists to decide whether or not mind altering drugs should be legal. We have enough problems with misuse of alcohol and with tobacco, which does not really have any legitimate use (deliberately inhaling smoke is not natural). Legalising cannabis would encourage its use and only a moron would want to add to existing problems.

    Roy

  37. Tom said: “You have a bunch of measurements. You put them through some bunch of statistics. You get an answer. How can that be unreproducible?”

    I think mrsean2k hit it squarely and it’s consistent with HARRY_READ_ME.TXT. The data is so badly documented that every use causes new interpretations of exactly which data is where or means what. Nor do they document these new interpretations. They can’t do it again the same way because they don’t know what they did the first time.

    Unbelievable? Yes, of course it is. But that appears to be the way CRU has operated.

  38. Kids do visit your website Anthony…should provide a PG-17 warning next time around as the Chemical link is semi-pornographic.

    =8-)

  39. One of the best Journal of Irreproducible Results articles … “The lab coat as a status symbol”

  40. I have wondered from the beginning of this brouhaha how it came to pass that the CRU was the only repository in the world for collecting worldwide surface temperature data. Fundamental computer security guidelines require off-site backup and generation logging. For data that’s this important and this difficult to collect, the original plan should have included two completely separate repositories all along. With completely separate management.

    How much money has been spent on this travesty? Let us hope that when the data is finally pried loose from their soiled hands it is found to be so useless that the whole lot of them will be out of work.

    And what does this say about the BEST Project? How can they reproduce long-lost metadata, not to mention the lost “unadjusted” data itself?

  41. “When I asked Oxburgh if [Keith] Briffa [CRU academic] could reproduce his own results, he said in lots of cases he couldn’t,” …

    Stringer [then said}: “That just isn’t science. It’s literature. …”

    Priceless … (as also noted by JohnWho above).

    .

  42. There are a few sympathizers (amazingly- but perhaps they are not scientists) with cru and their irreproducible works. I believe the problem with much of the stuff,not just out of cru is that they start with a sketch of the graph they require and then produce it with bits of trends, twisted and inverted proxies, clipping off unwanted divergences and selecting choice tortured segments of the instrumental record. This is macrame not science and is hard to reproduce.

  43. no doubt a lot of people would have liked the MSM to interview the CRU computer programmer, Ian “Harry” Harris. funny how his full name never even came up, hey? nor did it at the many so-called inquiries.
    wonder if any of these folks interrogated him?

    Jan 2010: Bishop Hill: ++++Statement from Norfolk Police++++
    Norfolk Constabulary continues its investigations into criminal offences in relation to a data breach at the University of East Anglia. During the enquiry officers have been working in liaison with the Office of the Information Commissioner and with officers from the National Domestic Extremism Team
    http://bishophill.squarespace.com/blog/2010/1/8/statement-from-norfolk-police.html

  44. mrdarcy_pemberley says:
    July 7, 2011 at 7:37 am
    Data can neither be created nor destroyed*

    You just described Global Warming gimmicry: Thermodatamimicry.
    The travesty is that the original purpose was to hide the perceived lull in rising temperatures, which ended up lasting longer than previously imagined. The lid came off the pressurized vessel of secrecy.
    Someone just recently said that the best way to maintain a secret is to not have one.

  45. Re post: Roy says: July 7, 2011 at 3:06 pm

    @ The Ghost Of Big Jim Cooley

    “Over here in the UK (and I suspect in many countries) our anti-drug policies are based on ‘feeling’ rather than science. When the chief scientist to advise our government on the mis-use of drugs pointed out that cannabis is less harmful than alcohol and tobacco, he was promptly sacked!”

    And it was a very good thing indeed that he was sacked. It is not the job of scientists to decide whether or not mind altering drugs should be legal. We have enough problems with misuse of alcohol and with tobacco, which does not really have any legitimate use (deliberately inhaling smoke is not natural). Legalising cannabis would encourage its use and only a moron would want to add to existing problems.

    Roy

    Might I gently point out that surfing the internet and typing comments to a blog isn’t natural either. Wearing clothing isn’t natural. Neither is living in houses. Norr using air conditioning and heating. Or using lights at night or anywhere for that matter. Or wearing shoes. Or driving cars, shopping, wearing glasses to correct vision, or prosthetics to help with physical problems, or using crutches or wheelchairs etc., or having surgeries of any sort. Or refrigeration for our foods, let alone eating regularly without having to personally hunt and/or gather ones own food out of the wild. Or a million other things that are common every day aspects of our lives.

    As to marijuana – I don’t even use the stuff and my entire career involves security sensitive areas so I’ve always been subjected to pre-hire drug screens and random drug screens etc., but I’m sick of our government sticking it’s nose into people’s personal lives, wrecking havoc, and costing the entire nation a fortune in the process – a fortune that could be put to far better more productive uses. Prohibition in the USA of alcohol certainly didn’t work, and frankly it doesn’t seem that prohibition against marijuana has been any more successful – but it certainly does create a criminal class, clog up our court systems and jails, and so on.

    Besides, if it isn’t any more harmful than alcohol or cigarettes, then it really isn’t reasonable for the government to legislate against it and suck all the related money out of our taxpayers and legitimate business enterprises. Obviously they’ve got to have and enforce laws against public intoxication regardless of whether that’s from alcohol or marijuana, and against things like driving while intoxicated – beyond that, they ought to get out of the business of wasting money and unnecessarily creating a very large criminal class, members of which may very well have their lives ruined if they are simply caught smoking a joint even in their own home.

  46. I agree, and I am someone who has never smoked anything remotely dodgy! Neither have I used anything worse than aspirin. But we should look at things rationally and logically, not in an airy-fairy way. I’d legalise it for a trial period, say five years, THEN use rationale to decide policy. Roy doesn’t see that.

  47. I have the tiniest shred of sympathy for East Anglia and its Climate Research Unit – or at least for the likeable Harry of the Read_Me file – because he reveals the sort of mess I get into myself when I have not left enough records to remind myself where I got the data from or what, precisely, I did during the processing last time.

    I am put in mind of a research department of a University which bumbled along OK when it was just one PhD thesis at a time, but which gets rapidly out of its depth when told to guard a depository of international data. It has no infrastructure to do this – no equivalent of a finance audit trail or computing production department.

    Which does not, of course, justify any cover-up or skulduggery, but makes me think that there, but for the Grace of God, go I.

  48. If these guys cannot even predict the past, such as getting repeatable proxy results from a single tree, how the hell do they expect to predict the future? Do they read chicken entrails, tea leaves, or do they just throw dice?

  49. Roy says:
    “Over here in the UK (and I suspect in many countries) our anti-drug policies are based on ‘feeling’ rather than science. When the chief scientist to advise our government on the mis-use of drugs pointed out that cannabis is less harmful than alcohol and tobacco, he was promptly sacked!”

    And it was a very good thing indeed that he was sacked. It is not the job of scientists to decide whether or not mind altering drugs should be legal. We have enough problems with misuse of alcohol and with tobacco, which does not really have any legitimate use (deliberately inhaling smoke is not natural). Legalising cannabis would encourage its use and only a moron would want to add to existing problems.

    I’ll leave your arguments about legalization aside and just ask you for clarification on one point:

    You really think it’s good that a scientist should be fired for pointing out a fact?

Comments are closed.