Facts to fit the theory

On 12 Oct 2009, climatologist and “hockey-team” member Kevin Trenberth wrote:

The fact is that we can’t account for the lack of warming at the moment and it is a travesty that we can’t. The CERES data published in the August BAMS 09 supplement on 2008 shows there should be even more warming: but the data are surely wrong.

Eyebrows have quite rightly been raised over this quote. It is indeed a travesty that AGW theory cannot account for the lack of warming, and bears out what I and other AGW critics have been saying for years about the fallaciousness and lack of predictive power of AGW models.

But the second sentence is actually far more damning. “The data is surely wrong.” This is how and where most scientific fraud begins.

Scientific fraudsters are not, in general, people pushing theories they know to be false. Outright charlatanism is not actually common, because it’s relatively easy to detect. Humans are evolved for a social competitive environernt
and are rather good at spotting lies, except when they’re fooling themselves because they want to believe.

In general, scientific fraudsters are people who are overinvested in a theory that they believe. Because they know it must be true, they interpret predictive failures as “The data is surely wrong”. It is only a short step from “The data is surely wrong” to fixing the pesky data until it looks right — see my previous post for an immediate example.

It’s only slightly longer step after that to destroying the inconvenient data that fails to fit your theory — something one of the hockey-teamers actually called for and there is strong reason to suspect they actually did.

Sometimes, actually, the data is wrong. Occasionally, experimental error will appear to falsify a theory that is actually correct. But research groups are entitled to the benefit of that doubt only when they meet the most rigorous standards of full disclosure about the “wrong” data. Not when their reaction is to conceal and destroy it.

108 thoughts on “Facts to fit the theory

  1. Dear Eric,

    As a joyful user of FreeBSD (& a joyful non-user of emacs) thank you for your service to freedom & software. Happy Thanksgiving!

    To JonB: I would council caution as well, but damn! this vindication feels good. Sixteen years (well, for me) of trying to nail jello to the wall and we finally get a break. The website to which you link tries to play it off as “cherry picking”. If only there were _some_ _other_ _damning_ _information_ in those files to substantiate the perfidy contained in “the data are surely wrong.” Oh, well. Maybe next time. ;)

  2. Wonderful. They’re trying to use an instrument that’s not meant to do the job they’re asking it to do. No wonder there’s wailing and gnashing of teeth. They should have been using ERBE, which was specifically desgned to return the kind of data for this application. I spent 40 years in the business. Never fails to amaze me when “scientists” make elementary mistakes.

  3. Eric: Are you interested in the truth, or are you just throwing mud around to see if any of it might stick? In an earlier thread you asserted that climate scientists ought to make their models open source, but you’ve ignored the comments that point out that much of it is already open source. In this thread you’ve picked a quote out of its context, and ignored the fact that the context explains the concern is about uncertainty over heat transports that cause inter-annual variability, which has nothing to do with long term warming.

    Does the quest for truth and understanding interest you at all? Can you even admit it when you’re wrong?

  4. Steve, I looked over your your blog. It seems you have quite a bit invested in this being a non-story. Of course you would say that. This is just like some of your friends who assume anything critical of AGW is funded by the oil companies. If you (and other AGW researchers) are funded by true believers who stand to make trillions, doesn’t that disqualify you from commenting as much as the skeptics funded by oil companies?

  5. The thing is, the CRU scientists might actually be making things up out of whole cloth, which would certainly fit with their declared policy agendas (in and of itself dangerous for scientists to push) as well as with the internal discussions they were having. But for other places only loosely affiliated with these guys, there is still an incentive to massage the data to fit what the CRU people found: the CRU people and Mann (and Mann and CRU were certainly coordinating) were declared the gold standard for climate science through the actions of the IPCC.

    Richard Feynman tells the story of how the mass of the electron was determined. Someone very widely respected (I forget who) measured the mass, and it was considerably lower than the value now generally used. Over the next fifty or so years, as more and more people measured the electron’s mass, the measurement crept slowly upwards. The people measuring it were getting the right values, but they couldn’t believe that the original guy was that far off, or that they were worthy to challenge his results, so they massaged their data – finding one after another justification for why it should have been lower, and then adjusting accordingly.

    Seems like that might have been going on in NZ at least, and possibly elsewhere.

  6. Busterdog: Huh? I’m funded by people who stand to make trillions? I get my salary from a public university and a small research grant from the federal government. I’m not an AGW researcher. I’m a computer scientist, who decided a few years ago to study the software development methods of climate scientists. As I’ve got to know many climate scientists, and understand their work, I’ve become increasingly alarmed that their work is being attacked by a bunch of people who are completely ignorant of (and not even interested in) the basic science. If that makes me “quite a bit invested” then, yeah, I care about science and truth. If you can point to any mistakes in my work, I’d love to hear them. Or are you also just throwing shit around?

  7. Also, if you follow the subsequent thread in the stolen emails, you’ll find a colleague points out that if you correct for ENSO, there is significant warming.

    This whole episode is a quite extraordinary attack on the personality of a few scientists among thousands working in this field, as if discrediting a few individuals by innuendo and quote mining invalidates a scientific theory. People are unguarded when talking to colleagues and say things that a hostile audience may misinterpret. Big deal. Publish all your own emails if you assert this doesn’t happen and let us judge for ourselves.

    If you look at the totality of attacks on climate scientists including the Bush administration trying to silence James Hanson and editing media releases to reverse the sense of them, hauling scientists before Senate enquiries, and attempting to label an entire field as suspect based on nitpicking the work of individuals, you have to go back the the Inquisition to find a comparable attack on science. I’m surprised people who consider themselves libertarian side with this massive attack on freedom of speech.

    Show me the alternative theory that fits the data. Then we’re talking. You can be pretty damn sure that theory doesn’t exist, otherwise why personal attack politics? Einstein overthrew Newton’s theory because he had a better one, not because he told everyone in the pub that Newton was a jerk.

  8. Selecting which data to use happens in science all of the time and it is a legitimate practice.

    A classic example is the Piltdown Man hoax. At first it was pretty convincing, but as scientists dug up more “data” and the picture of early man got clearer, Piltdown Man started to make less and less sense when compared to the full spectrum of data. It was an outlier and scientists had every right to ignore that data until new data came to light that made sense of it. Eventually Piltdown Man was able to be fully dismissed when the picture got clear enough that scientists were able to expose it for the hoax that it was.

    Scientists are rightly able to dismiss data because their professional intuition (talent + training + experience + team) says “something is not right here so let us keep working with what makes sense and maybe that outlier data will make sense later”.

    That is one facet of how science works…

    Side Note: Creationists *REALLY* need to stop taking credit for exposing the hoax. Scientists exposed the hoax fair and square because the data simply did not fit with the rest of the mountains of evidence for evolution.

    ..Ch:W..

  9. >you’ve ignored the comments that point out that much of it is already open source.

    That’s right, because the data and models relevant to this case are not open source.

    Fine, you established that some climatologists are not villains. It’s good to know that. Applause to you. It’s not a point on which I required persuasion, however. I didn’t even think most climatologists are villains, let alone all of you.

    Now let’s get back to the actual issue at hand — which is that, on the evidence, the hockey team are villains who have led the field into a gigantic error cascade. And we need to make open source a requirement — not a pat-yourself-on-the-back option, a requirement — so this doesn’t happen again.

  10. A few questions:

    1) Over the years, to what degree has the data from the CRU formed the basis for the IPCC reports and the current “consensus” about AGW? Has their data been 10% of the evidence, or 90%, or what? If it’s a small portion of the total evidence, the AGWers may have a case that the leaked documents are being blown out of proportion.

    2) If the climate models the CRU was using have been predicting dangerous levels of warming over the coming decades, but can’t account for the last ten years of stable/declining temperatures, to what degree can we count on their predictions at all?

    3) Why shouldn’t there be some some of independent, scientific audit of all the CRUs data, leaked or not?

  11. 1) It’s a damn large piece. The first revision of the IPCC report relied heavily on the Mann Hockey stick, both as evidence and as marketing tool.

    2) Not a whole hell of a lot, in my opinion. Kevin Ternberth apparently agrees — “it’s a travesty that we can’t”

    3) Absolutely there ought to be an independent audit of the data, the code, and everything else. With all of it available for public inspection. The hockey team and the rest of the AGW boosters should be insisting on this as a way to save their reputations and salvage AGW theory. That is, assuming they have nothing to hide. And why should they, if they’re honest and innocent researchers being unfairly maligned?

  12. @PapayaSF

    1) Paleoclimate was one chapter out of 11 in the Working Group 1 section of the IPCC AR4.

    2) We can count on their predictions over multiple decades, where their models do have predictive power. Intra-decadal predicitions not so much.

    3) Cost. The expertise required for such an audit would be better spent elsewhere.

  13. Pete, how do we know that their models have predictive power over decades? They haven’t been around for decades to check. The only timescale on which they can effectively be tested with data that hasn’t already been used to calibrate the models is a span of 10-20 years, and on that scale, their predictive power is crap. Why should we trust them in the range where we can’t check it when we know it fails in the range where we can?

  14. Eric, I’m not sure you’re being fair to Kevin. “The data are surely wrong” could very well be better said as “The data don’t agree”. And the reason that the data are wrong is because while they predict greater warming, the fact is that there hasn’t been any warming.

  15. >1) Paleoclimate was one chapter out of 11 in the Working Group 1 section of the IPCC AR4.

    You conveniently ignore the huge role of the hockey stick in AR1 and its promotion. The only reason it was de-emphasized in AR4 was because McIntyre and McKittrick had shot it full of holes.

    >2) We can count on their predictions over multiple decades, where their models do have predictive power. Intra-decadal predicitions not so much.

    Neither their long-term predictions nor their short-term ones are worth a fart in a windstorm. There’s been no time for a long-term check, and the models can’t even retrodict the historical data without strenuous data-munging to “contain” (as Mann euphemistically put it) the Medieval Warm Period and the Roman Climactic Optimum.

    Hell, I have a better short-term predictive record than the hockey team. When GAT measures nosedived at the beginning of 2008, I can produce email to prove that I offered a bet to some AGW true believers I know. The bet was that if the solar minimum continued over the next year, GAT would move in parallel with the change in insolation rather than the rising CO2 level. Those friends were too chicken to take me up on it, which turned out to be wise of them.

    Meanwhile, Kevin Trenberth has to admit that they can’t account for the lack of warming and it’s a travesty. Idiots! I can account for the lack of warming with the simple hypothesis that variations in solar forcing swamp greenhouse forcing, and I used that hypothesis to make a testable and correct prediction. Now, it’s possible that my hypothesis might be falsified over longer timescales — but whether that’s true or not, this was better performance than the hockey team or the IPCC models have ever managed to turn in.

    >3) Cost. The expertise required for such an audit would be better spent elsewhere.

    When trillions of dollars of deadweight loss from carbon taxes and regulation to a world economy already in recession are at stake? Hogwash.

  16. And never was a blogpost headline truer to its subject matter. Except here there are no facts at all, other than out-of-context emails and (snigger) typos in code comments.

    I wonder whether you self-professed anti-AGW types are at least a little embarrassed to be following the leader like this? Seems to me that denialism is a silent requirement for membership of this little libertarian club …

  17. >>3) Cost. The expertise required for such an audit would be better spent elsewhere.

    >When trillions of dollars of deadweight loss from carbon taxes and regulation to a world economy already in recession are at stake? Hogwash.

    I suggest you read (or re-read) “The Mythical Man Month”. Some problems (and science is one of them) can’t be solved just by throwing money at them, because the bottleneck is elsewhere. A politically motivated “audit” would be a huge distraction to scientists, and all we’d get would be a huge deadweight loss in scientific output. There is sufficient interest in paleoclimate that the usual scientific process will achieve the ostensible goal of such an audit (e.g. recent work by von Storch, Zorita, Christiansen, Lee and others. Google (scholar) for “pseudoproxy”), without achieving the denialist aim of preventing climate scientists from doing their work.

    >I can account for the lack of warming with the simple hypothesis that variations in solar forcing swamp greenhouse forcing.

    No climate scientist would disagree with you that in the short term solar variation (and plenty of other things, such as ENSO style configuration changes) has a greater influence that changes in CO2. But solar variation stays within certain bounds, while atmospheric CO2 can keep increasing.

    Consider a function f(x) = sin(x) + x/100. Which term “swamps” which?

    >You conveniently ignore the huge role of the hockey stick in AR1 and its promotion.

    Presumably you mean AR3, where paleoclimate was part of one of fourteen chapters? Or are you confusing the assessment reports with An Inconvenient Truth?

  18. @Alsadius >Pete, how do we know that their models have predictive power over decades? They haven’t been around for decades to check.

    Hansen made his predictions in 1988.

  19. Eric – I believe you to be a very intelligent person, and I respect your work. I get my students to read your papers. You have great insights into how open source communities work and why they are important. But on the topic of climate change, you don’t appear to be willing to engage your critical thinking skills. So, I understand you have a passionately held political viewpoint. I suspect it differs from mine, but that’s quite irrelevant. The problem is that your political viewpoint appears to have led you to reject a vast body of basic atmospheric physics because you don’t like the implications. That’s just nuts. Your comments in these blog posts all show a remarkable confirmation bias – you want the scientists involved to be wrong, so you’re only finding bits of evidence that confirm this, stretching your interpretation of the wording to fit your conspiracy theories, and ignoring all the relevant context. Nowhere have you actually looked at the actual science. You talk about what the IPCC reports are based on, but it’s absolutely clear from the way you talk that you’ve never even looked at the IPCC reports at all (they are freely available online: http://www.ipcc.ch/). You can understand most of the WG1 report if you have a good grasp of high school physics. Go look at the science, rather than speculating about what a few phrases in private emails may or may not mean.

    ESR: “Fine, you established that some climatologists are not villains. It’s good to know that. Applause to you. It’s not a point on which I required persuasion, however. I didn’t even think most climatologists are villains, let alone all of you.”

    You missed the point. A number of climate models and nearly all the datasets are open source. But it doesn’t help, because you need a PhD in climatology to understand how to use them and how to make productive changes to the code. Requiring them to be open source is fine, but it would change nothing, until people are actually willing to get themselves educated about the basics of atmospheric physics, and try understanding the science, rather than trying to discredit scientists and score political points. Open source communities only work when a number of people agree to work together constructively towards a goal. They don’t work when some people (in this case, *you*) are not interested in constructive engagement.

    You also missed the comment above where I clearly stated I’m a computer scientist, not a climatologist. I’ve been looking at the code and the development processes that climatologists use to build it for several years, and it’s quite clear that much of the code is poorly structured, and very hard to modify. My initial hypothesis was that a bunch of good programmers could come in and sort it out for them. Unfortunately, my initial hypothesis was completely wrong – every case where experienced programmers have attempted to do this has been a disaster – it results in code that is useless as a scientific instrument. I’m happy to explain at length why this is, if you’re interested. Unfortunately, from your tone, you don’t appear to be.

    ESR: “Now let’s get back to the actual issue at hand — which is that, on the evidence, the hockey team are villains who have led the field into a gigantic error cascade.”

    Sorry, I must have completely missed the point where you had evidence of an “error cascade”. Unless you actually go and look at the IPCC reports, you have no idea what role the original “hockey stick” graph played in developing the science (hint: none at all – it was developed as a way of *visualizing* some of the key ideas for a lay audience). You need to calm down, stop exaggerating, and start thinking scientifically about evidence and what it shows. Your passion and intellect would be much better applied to furthering our understanding of the physics and better ways to use software in that endeavour. Whatever your personal perspective is, this topic is far too important to be treated as an excuse to make wild accusations in place of careful critical thinking.

    ESR: “”And we need to make open source a requirement — not a pat-yourself-on-the-back option, a requirement — so this doesn’t happen again.””

    And once again, that unthinking belief that an open source requirement fixes everything. I completely agree it should be a requirement. But if it made any difference at all, then we would be able to see it by now in the models that are already open source. Come on, use your insights into what makes open source communities tick to figure out how to help climate scientists do good work. It’s much better than jeering from the sidelines.

  20. So, Eric, have you read Trenberth’s paper?

    I see a lot of comments here from people who aren’t up on the subject, including Eric. I wouldn’t presume to judge without a fair bit of knowledge on climate science.

    BTW, you’re welcome to download the source code and all necessary input data for one of the IPCC AR4 models, CCSM3.0, at

    http://www.ccsm.ucar.edu/models/ccsm3.0/#src

    Instructions are located within that page.

  21. The whole situation brings to mind Thompson’s “Reflections on Trusting Trust”. All climate research uses HadCRUt (and GISStemp, where Gavin Schmidt seems up to his elbows in this). Jones and Schmidt certainly have the opportunity to manufacture consensus by fiddling the data that everyone uses as a basis of their research.

    Did they do this? Who knows?

    Will they release their data and methods (collection, adjustment)? Not so much. Reflections on Trusting Trust, indeed.

  22. @Philip Machanich:

    Show me the alternative theory that fits the data.

    What if you can’t trust the data? What do you do?

    How do you show that you can trust the data?

  23. @Pete:

    Hansen made his predictions in 1988.

    Boy, howdy, I’ll say:

    While doing research 12 or 13 years ago, I met Jim Hansen, the scientist who in 1988 predicted the greenhouse effect before Congress. I went over to the window with him and looked out on Broadway in New York City and said, “If what you’re saying about the greenhouse effect is true, is anything going to look different down there in 20 years?” He looked for a while and was quiet and didn’t say anything for a couple seconds. Then he said, “Well, there will be more traffic.” I, of course, didn’t think he heard the question right. Then he explained, “The West Side Highway [which runs along the Hudson River] will be under water. And there will be tape across the windows across the street because of high winds. And the same birds won’t be there. The trees in the median strip will change.” Then he said, “There will be more police cars.” Why? “Well, you know what happens to crime when the heat goes up.”

  24. >How do you show that you can trust the data?

    That question, at least, has well defined answer.

    The data has to be open source, and it has to have a provenance attached. It has to be open source so that, for example: if NASA has datasets purporting to describe early 20th-century temperatures in Colorado, they can be checked against Colorado state records.

  25. Steve,
    I didn’t say you were funded by them, just that from your website you are very invested in AGW, emotionally if nothing else. I was discussing incentives. AGW researchers are funded by governments. Govt’s are always looking for more reasons to expand their power. Cap and trade, Copenhagen and other solutions, whatever the science behind them, all require a massive expansion of govt power. I am suspicious of AGW up front because the researchers always seem to come up with what their funding source wants them to. These emails show people under an incentive to produce what they are paid to. Critics of the anti-AGW think nothing pointing out the incentives of their critics, but get upset if anyone implies incentives in their own funding sources.
    The facts on the ground right now is that we are getting cooler for the last 10 years and the AGW crowd has no explanation for it. The fact is that is has been much colder before and it has been much hotter. The fact is that carbon dioxide is plant food, and not a dangerous chemical. For life to have lived here as long as it has, there have to be some powerful feedback mechanisms in place to keep it livable. Through volcanoes, asteroid strikes and other natural catastrophes temperature manages to right itself.
    Foul language doesn’t advance your argument at all.

  26. “If you look at the totality of attacks on climate scientists including the Bush administration trying to silence James Hanson…”

    The moment I hear someone claim Hanson was “silenced” I know they’re pushing an agenda. Someone who has been silenced doesn’t give dozens of interviews, as Hanson did.

  27. “Show me the alternative theory that fits the data. Then we’re talking. You can be pretty damn sure that theory doesn’t exist, otherwise why personal attack politics? Einstein overthrew Newton’s theory because he had a better one, not because he told everyone in the pub that Newton was a jerk.”

    And yet the “hockey team” wallows in ad hominem attacks on the “denialists”.

  28. Finding the program that fakes the hockey stick is like finding the file-marks on the orangutang teeth in Piltdown Man’s skull. Clear evidence of intentional fraud.

    This will be in the history books.

    ESR says: Stop right there. I do not believe this is the source of the Mann hockey stick. The timing is wrong, the provenance is wring, and anyway McIntyre showed years ago that you get a hockey stick if you feed Mann’s code data that is entirely random.

  29. Numerous scientific experiments have shown that the beating of tom-toms causes the sun to reappear after an eclipse. ;-) The climate “scientists” were using the same sort of post hoc reasoning to try to prove their greenhouse gas hypothesis. Their “models” are not models in any sense of the word that a physical scientist would understand, i.e., a simulation of physical processes. They are statistically munged extrapolations. And from the code and emails, not even honest ones. The fact that the underlying data are also questionable is only icing on the cake.

    The computational requirements to actually simulate with any reasonable grid size and time step energy flow between space, the air, land, and water are *huge*. Ten years ago certainly challenging even for the fastest, largest supercomputers. In retrospect, it should have been obvious that climate “models” weren’t doing anything physically meaningful. Unfortunately, I never really thought about it until now, and I don’t think any of the “critics” did either.

  30. Steve Easterbrook Says:
    > A number of climate models and nearly all the datasets
    > are open source. But it doesn’t help, because you need
    > a PhD in climatology to understand how to use them and
    > how to make productive changes to the code.

    See Steve, here is the problem. You say that you need a PhD in climatology to understand the data. However, the very essence of understanding is the ability to predict. The simple fact is that climatologists have demonstrated again and again that they just can’t predict with any degree of accuracy. Which is to say, PhD’s notwithstanding, evidently they actually don’t understand what is going on after all. This seems to me to be mainly true because only a tiny fraction of the factors affecting the climate are actually measured, and so, frankly, prediction is fraught with peril and signal swamping error. We do not keep good global records of butterfly wing flaps yet.

    “The data seems wrong”. Perhaps this might be out of context, I don’t know, I haven’t studied the emails. Perhaps it was careless banter between two experts. Again, don’t know, haven’t not read the context. However, any serious scientist would get hives before typing or thinking such a thing. Frankly, “the data can’t be wrong, but it doesn’t map to my model” is the source of more life changing discoveries in science than any other.

    This I know for sure: if “big oil” sponsored studies that produced results saying burned hydrocarbons had not effect on the climate, and if they did so without producing the raw data for their predictions, or the methodology for their predictions, they would be laughed off the stage. If some Russian hacker stole their emails and details and published it showing the types of scientific dishonesty evident in the CRU emails, then the hackers would be praised as heroes, and the putative scientists would never work in science again. However, when “big government” sponsors such a study, with exactly the same conflicts of interest, the same standards are simply not applied.

    Why exactly have we barely heard a word about this explosive topic on the mainstream media? I don’t know why, but I think it is deeply troubling that we have not.

  31. I am happy to announce that I can lay this entire controversy to rest by open-sourcing my heretofore top-secret climate model. I developed this at my evil lair on the well-known Commodore VIC-20 supercomputer. Please feel free to download the source and run it yourself to verify my results:

    10 INPUT “What is the name of your planet”, WHATEVER$
    20 INPUT “How much money do you have”, RANSOM
    30 INPUT “What is the name of a UN climate bureaucrat”, MARXIST$
    40 FOR T = 1 to 100000 : NEXT T
    50 PRINT “The only way to save planet “; WHATEVER$
    60 PRINT “is to send “; RANSOM; ” dollars to “; MARXIST$
    70 PRINT “and do whatever he says.”
    80 END : REM of Western Civ, Scientific Method, etc.

    I hope this helps clear things up.

  32. You are doing a great job, Eric. The panicked responses from all of the self-interested alarmist political factions shows that you are touching a nerve. Keep it up!

    Some of these emails are indeed being taken out of context. But if so, Mann, Jones et al have themselves to blame for evading FOIA requests for so many years, and threatening journal editors not to stray from the talking points, and doing everything they could to control the narrative of climate alarmism. When the things they tried so hard to keep private suddenly become public, of course people will assume the worst. In this case, such assumptions may fail to reach the depths of motivation involved.

    Trenberth was talking about a particular type of warming that wasn’t manifesting itself, but the quote clearly and accurately transmits his apparent willingness to skew the data to fit the desired model results.

  33. ESR: “The data has to be open source, and it has to have a provenance attached. It has to be open source so that, for example: if NASA has datasets purporting to describe early 20th-century temperatures in Colorado, they can be checked against Colorado state records.”

    Okay, one more try to get through, and then I’ll go and find something more productive to do. ALL of NASA’s datasets are open. So are their models. I’ve given you the links. Yet I don’t see you going off and doing the checks you’re suggesting. Instead, you’re just banging on about some small portion of climate data that isn’t open. Which means what you’re offering is nothing but a strawman argument – if you really believed in what you are saying you’d already be working on the code and data that *is* open source.

  34. Einstein did not overthrow Newton. What Newton developed is still true. Relativity extends the description, it does not replace it.

  35. Steve Easterbrook:

    I would welcome your input (and advice) on my blog – Data Against Demogogues That may be a more appropriate venue (Eric has proof-review privileges there, FYI)

    In particular, I’m trying to assemble a ‘written for laymen’ post on what proxies are being used for climate and isolating the sources of noise within the data. EG “Dendrochronology in carefully selected stands of trees can be a good proxy; it runs into noise issues based on free nitrogen in the soil, moisture, and CO2 levels, which are corrected for by methods A, B and C.”

    I’m a good technical writer; I’ve got a friend who used to code climate models. I could use your insight as well. Please feel free to come on over.

    Please note that while I’m a skeptic, I mean that in the true definition of the term: Someone who wishes to know more and understand the issues. I acknowledge that agenda driven reporting exists on both sides of this debate. I want to CUT THROUGH that and present the facts in as clear a way as I can. I will also write things in defense of the AGW crowd – for example, I feel that the political use of the Soon and Baliunas paper and its timing and interesting publication picture justified the censure of Climate Review by Phil Jones and Michael Mann.

    This entire issue (from either side) is too important to be left as the province of sound bite reporting.

  36. Maybe I’m way too much of a history geek, but this matter open sourcing the science, and the hysterical reaction to the suggestion, reminds me of the way that “saint” thomas more had people burned at the stake for passing around copies of a bible translated to English. When the bible was only available in Latin, you had to have a priest tell you what it said. When you could read it yourself, you’d learn that Jesus wasn’t always for the rich guy.

  37. @Steve Easterbrook:

    You mean links like like this one, based on the highly-criticized methods of one James Hansen? The same James Hansen who is complicit in the e-mail messages contained in the 61 MB CRUGate files?
    The same analysis that has been criticized by minds no lesser than that of physicist Freeman Dyson?

    Surely you can come up with something better than that.

  38. >ALL of NASA’s datasets are open.

    Ken and I will post on this shortly. There’s been a development…

  39. All of NASA’s importable data-sets (post massage) have been available for a decade. I do not know – but would not be surprised – if a number of their models were publicly available and open sourced. On the other hand, Steve Easterbrook points out quite well that they probably don’t conform to good software engineering standards, so I’m not sure that even source code access would yield quick analysis.

    However, James Hansen has said, in an blog post on Newsweek:

    “Question: Do the e-mails indicate any unethical efforts to hide data that do not support the idea of anthropogenic global warming, or to keep contrary ideas out of the scientific literature and the Intergovernmental Panel of Climate Change reports?

    Hansen: They indicate poor judgment in specific cases. First, the data behind any analysis should be made publicly available. Second, rather than trying so hard to prohibit publication of shoddy science, which is impossible, it is better that reviews, such as by IPCC and the National Academy of Sciences, summarize the full range of opinions and explain clearly the basis of the scientific assessment. The contrarians or deniers do not have a scientific leg to stand on. Their aim is to win a public-relations battle, or at least get a draw, which may be enough to stymie the actions that are needed to stabilize climate.”

    Emphasis added by me. Peer review and science are an incremental process, and I laud Hansen’s position on this – including trying to be even-handed with his colleagues. If he denounces them utterly, he loses his job. If he stands behind them unconditionally, he takes the easy way out. If he says “let the data speak for itself”, he’s taking the ethical road.

    We’re on the first steps towards greater openness on this. I’m hoping that with more eyes looking over the data, this will become interoperable openness with metadata attached.

  40. I went and read some of the material that Steve Easterbrook pointed to. My first impression was shock at the speed with which the climate science community is willing to enter into ad hominem attacks against critics. For example, in public statements about the Soon and Baliunas paper, the pair were identified as AGW “deniers” before any technical content was even discussed! The climate science community clearly believes of its detractors that their belief state has significant impact on their conclusions; why, then, do they profess immunity for themselves?

    As a physicist, I have twice in my career been involved with results that were pushed by firm believers and have since been debunked thoroughly. In both cases, the believers really did believe that their effect was real, while making subtle statistical errors that seemed reasonable on the surface but on deeper inspection gave incorrect results. And the “quality” of those results was stunning! In one case, it resulted in a 10-sigma effect that looked incredible. It turned out that the entire effect was caused by the fact that the experimenters could choose when to start and stop each data run. They would look at the data coming in, and, if it didn’t look “right,” they would abort the run prematurely and tweak a few parameters, then start again. That’s all they did wrong, yet they got a statistically huge result.

    So when a group of scientists tells me to take it on their word that [i]all[/i] the data agree on some hard-to-measure effect, my scientific antennae prick up. When they quickly and consistently resort to ad hominem attacks against their critics, my skeptical senses start to really quiver. When I then go back to review some of the relevant literature (in this case, Mann’s original dendrochronological paper) and see basic statistical errors that went several years before being corrected (after being pointed out by a so-called denier), my opinion of the climate science community is reduced. And when there is evidence that a group tried to subvert the peer-review process, I go into full unbelief mode.

    If this were regular science, I would be content to let things sort themselves out over the long run. But this isn’t regular science. We are being asked to completely transform the world’s economies, at an enormous cost in both money and lives. Skepticism is not only rational at this point, but it is essential.

    The facts must be faced by the climate science community: scientific trust has been broken, and the burden of proof is on them to prove that the anthropogenic effect is real. Sorry, folks, but that is how science works. I know many climate scientists, including those in the emails, are very anxious that the entire structure of the global economy be transformed right away, but, unfortunately, science does not always proceed at the pace politicians would like.

  41. Ken,
    The metadata problem is huge. There are two major projects (one in Europe and one in the US) attempting to develop appropriate meta-data standards for climate models and data. It’s slow going. Take a look:
    http://www.easterbrook.ca/steve/?p=56

    Personally, I don’t think these efforts will be very fruitful, because they ignore the sociological and epistemological problem surrounding the definition of scientific terms. At the EGU meeting earlier this year, I saw a talk on an attempt to develop a single geological survey dataset for Europe, from dozens of individual national surveys. They got bogged down trying to find a common definition for the term “bedrock” (ironic, huh?). More here: http://www.easterbrook.ca/steve/?p=333

    Once again, openness is irrelevant on it’s own. What matters is a knowledgeable community who can do useful stuff with it. Never underestimate the level of expertise needed to work with raw climate and meteorological datasets.

  42. What I see here in the comments from the AGW true believers are attempts to discredit what skeptics are pointing out about specific files in the large number that have been made public.

    But when examined all together, the totality of evidence is very damning. The “hockey team” conspired to control the peer review process on climate science so that only papers which agreed with AGW could get published. Then they used that as a bludgeon against critics. “Pay no attention to the critics! Their work hasn’t been published in these journals, therefore it is crap.” Easy to do when all the peers doing the reviewing are communicating and coordinating in advance so they know what to accept and what to reject. That is right there in plain language in those e-mails.

    Real peer review is supposed to be done anonymously, by several others conversant in the field of the paper presented. If there is anything they don’t understand, the reviewers should consult with other people who are knowledgeable about those parts.

    The presenter of the paper must provide ALL of the data used and the methodology used to obtain it. That would include both source code and compiled programs.

    On top of that are the statements about blocking Freedom Of Information Act (both the US and UK laws) requests for data, and warnings to not ever mention to anyone that the UK even has a FOI law.

    CRU did release some data, but only as a bunch of numbers without any information stating exactly where and when it was collected. It could be graphed and plotted but without a time/location reference – completely useless, as CRU intended.

    It’s bloody obvious to anyone who has ever been taught the scientific method that none of the “hockey team’s” work went through such a process. It’s quite easy to discover what the Scientific Method is for testing a hypothesis – Google it.

    Fudging data with artificial adjustments is fudging data with artificial adjustments, even if “only for a piece of cover art”. If your theory needs a faked-up graph for its cover art…

    The HARRY_READ_ME.txt is a real hoot of a read, even for people who don’t know the first thing about programming. Whomever “Harry” was, he clearly was not having a fun time trying to make version 3 of the program he was working on fit the outcome CRU wanted it to have. He certainly wouldn’t have written 15,000 lines of notes for a “cover art” program.

    What we’ve seen from CRU and the IPCC is “outcome based science”. First decide on what you want the outcome to be, then do whatever is needed to make that outcome “true”. (Look up outcome based education, wherein to reach a desired number of students graduating, the standards are reduced until that number do graduate, no matter how poor their actual education.)

    One really large travesty of this mess is that many honest scientists got dragged along because CRU, IPCC and company were pitched as the “gold standard”, a ready source of solid, “peer reviewed” science and data. “They’ve already run the data through the mill, so why do I need to repeat all that hard work?” (Because you’re a *scientist* ya lazy bum. Repeating experiments many times is one of the core parts of the Scientific Method.)

    P.S. I’ll leave it to others to look up, 4:44 AM here and I’m for bed. There’s an e-mail in that archive where one of the “hockey team” wishes that the warming was real, no matter what the consequences, just so he could be proven right. Another e-mail suggests polluting the troposphere with sulfur dioxide (SO2) as a method to combat the warming. Yet another asks what causes some 11 year spikes in the data and wonders if it might be something to do with the sun.

  43. I have to chime in with Easterbrook here on the issue of experience and credentials. Some fields are like that. I once got interested in computational fluid dynamics by way of a fascination with space vehicle propulsion. How hard could it be? I thought. After a while, it dawned on me: it’s really damn hard. I looked around at what it would take to get job experience. Guess what? Hardly anybody hacks CFD code professionally without a PhD, more usually with a dissertation that at least fractionally advanced the state of the art. (I found an exception once, a guy with an MS, but all he was doing was mesh generation.) You could open-source absolutely everything in CFD, and hackers who thought they were making the code “more elegant” and “better structured” might only slow it down on vector machines by a factor of 5, while “refactoring” computations in a way that would make any numerical analyst weep bitter tears about error propagation.

    Yes, climate science is very hard. A fact which is … too bad, for climate science, because it makes it easier to reinforce the perception of climate science as an insular scientific priesthood. And if it turns out their problems with near-term prediction lead to major credibility issues for them, and policy is formed around that damaged credibility, even though in the longer term their models actually turn out to be reasonably accurate, it could mean losing a decade or more — possibly a critical decade. Here’s something an amazing number of people don’t understand: ending human GHG emissions tomorrow won’t end warming caused by all previous unabsorbed emissions, not for a very long time. This has always been the flaw in the argument that we need to be absolutely sure before doing anything that could cost much. This is why I was in favor of exploring policy options even if the IPCC had said there was only a 25% chance that AGW was real — once you step on the brakes, you’ll still skid for decades before you come to a halt. And you might come to a halt in the middle of onrushing traffic.

    In fact, worst-case economic costs of AGW countermeasure policies seem quite moderate, but worst-case costs of neglect could be staggering. Especially if you imagine nuclear-armed tin-horn dictators and corrupt elites in sham democracies, running various countries comprising 3-4 billion impoverished, benighted subjects drenched in anti-Western propaganda, always hearing “THEY did this to the climate, because THEY never gave a rat’s ass about YOU.” Political environments not so different from, say, Pakistan today. Seriousness about possible AGW is, in fact, a genuine political legitimacy issue for liberal democracy over the long run. And you don’t have to be some unquestioning True Believer Warmist to think so.

    Not that I’m a huge fan of liberal democracy. Like Churchill said: it’s the worst, except for all the other schemes tried. I just believe now it’s probably the best we can do, and it’s one of the world’s sad inequities that too many people still live under other systems that actually do leave them poor and stupid. If AGW might be a significant threat to the best we can do, I’ll support efforts to model AGW, and policies (tentative, at least) to prevent AGW. Yes, even if climate modelers can’t get their near-term predictions to satisfy every crank, curmudgeon, contrarian, arm-chair statistician, amateur climatologist, or demagogue. You can’t satisfy those people anyway.

  44. “specially if you imagine nuclear-armed tin-horn dictators and corrupt elites in sham democracies, running various countries comprising 3-4 billion impoverished, benighted subjects drenched in anti-Western propaganda, always hearing “THEY did this to the climate, because THEY never gave a rat’s ass about YOU.”

    This statement tells me that your political naivete is astounding.

    When have “tin-horn” dictators lacked for a reason to demagogue? You think if the West commits economic suicide it won’t propagate to the impoverished countries with consequent demagoguery?

    Puh-lease. Get a clue.

  45. Turner/Easterbrook

    Linux isnt as complex as those climate models eh ?
    Heh bet 2 bucks to a doughnut the code is as crappy as in the email/code dump.
    But dont give us no “the rubes are too stupid to figure it out” garbage.
    Open source climatology will work just as well as it did for linux,
    And the pool of expertise out here is just as disporportionally superior to that closed fudging sloppy clique that deleted the raw temp records to prevent Steve and Anthony from exposing the apparent fact that the so called warming is all sourced from adjustments to the raw data, that never show their warming trend when you look at long term rural stations.

    Its that raw data that Phil and crew talked about deleting rather than Steve get it,
    The same data that “surprise!”, is now said to be “Lost”.

    Its time to get some friggin rope …

  46. OK, the all-important next step in the unraveling of this fraud is to identify all the policy proposals that rely on the CRU’s output, either directly or indirectly. Those proposals need to be stopped in their tracks immediately.

    As many people have said already, if this were occurring in the business or financial world, a fraud of this magnitude would result in prison terms for at least the major actors.

  47. The “lost”(deleted!) dataset was the raw numbers for HadCRUT.
    The focus of the FOI request,

  48. It it true that critiquing a finite element analysis, non-trivial heat transfer, or anything involving fluid dynamics is quite complex. That tends to be why people doing such things actually hire competent computer scientists and engineers to write the code, and involve extensive validation processes before being used in production.

    On the other hand: We hoi polloi can recognize a freaking overflow error in code that really should be using an established numerical library. And some of us do actually write the numerical libraries anyway.

    This is the very essence of the Cathedral versus the Bazaar. There is just no reason to be rolling a custom least-squares algorithm in the first damn place. Even dabblers can have crucial facts that the non-computer scientist writing the code might not know. “Why didn’t you use XYZ here?” The bulk of the code involved in the discussion of interest isn’t overwhelmingly complex in scope: Least squares, interpolate missing points, spacial gridding of available data.

    And yet they manage to have an “incomprehensible” overflow into negative numbers. Wait, a silent overflow. Delightful.

  49. There is only one truth in the universe….If Al Gore is proposing it, it is wrong.

  50. Michael Turner Says:
    > Some fields are like that. I once got interested in
    > computational fluid dynamics

    See here is the thing about computational fluid dynamics: it is not soft science. You can do your math, then you can test it out. I guarantee you that people with expertise in the field can, for example, accurately calculate the lift on a wing, or the turbulence at the trailing edge. They can predict it, test it and prove their calculations right.

    People in climate science can’t do that. They have a blackboard full of complex equations that only their little group can understand, however, they can’t use it to make any accurate predictions. Why? Most likely because their models are profoundly inadequate not factoring in many variables that have a very direct impact, but that they don’t or can’t measure.

    So the case Easterbrook is putting forward is that climate science is too hard to understand for we rubes, and he is right, what he fails to say though is that climate science is apparently too hard to understand for all the climate science PhD’s too.

    Not entirely on topic, there is a far more important question in play. Simply this, even if the catastrophic models are true, it really doesn’t matter. This is a problem that can’t be solved by politicians. If there was one world dictator with all encompassing power, he might be able to solve it, but there isn’t, thank god. Good gravy, the world agrees that Iran getting a nuclear bomb would be a terrible thing, but we can’t even stop that! How are we going to convince the world to damage their economies severely in order to combat a phantom? All politics can do here is damage the economies of major powers (read cap and trade) so that politicians feel self righteous and get re-elected. However, these changes merely move greenhouse gas emissions elsewhere (“the cap and trade bill” should be called “shift the last remaining vestiges of American manufacturing to China bill”.)

    What we can do though is to deal with any consequences as they come up. Build bigger sea walls. Recapture land from the sea, build better flood control systems and so forth. We can do these things on an as needed basis, rather than invest in a putative threat who’s predicted outcome has error bars that would make a high school science student blush.

  51. One would believe the data from the last decade to be the MOST accurate. Are we to believe just the opposite is true? How sad a state of affairs this is turning out to be.

  52. what he fails to say though is that climate science is apparently too hard to understand for all the climate science PhD’s too.

    So what? That’s almost the definition of science. Investigations into the unknown. No body understands quantum mechanics, or how to unify it with classical mechanics. That only indicates that more time and energy should be spent on scientific investigation.

  53. drewster Says:
    > So what? That’s almost the definition of science. Investigations into the unknown.

    I am with you 100%. I am very much in favor of continued research into climatology. However, I think you missed the point I was making, namely, that the level of understanding is so dramatically low that their predictions will be (an have been) dramatically unreliable. Basing major public policy on science that is dramatically unreliable is nuts. Using words like “denial” and “heretic” in reference to critics of the mainline position, when the mainline position is so unreliable is gross scientific dishonesty. Good gravy, those are words used in religion not science.

  54. “That only indicates that more time and energy should be spent on scientific investigation.”

    Yes, the keyword there being “scientific”. Basing political decisions with profound economic impact on insufficiently “investigated” science seems to be rather unwise.

  55. Jessica: didn’t anyone tell you that climate models are based on computational fluid dynamics? What do you think is at the heart of a GCM?

    Feel free to call yourselves “rubes” if you like – I certainly wouldn’t. If you look back to my earlier comments, I explicitly stated I think you’re all very intelligent people. But you’re confusing intelligence with domain knowledge.

    Let’s take Eric’s comparison with Linux. I explored this comparison in detail when I studied the Hadley centre’s model last year. Their model is currently about 1 million lines of Fortran, and the size has grown linearly over about a 15 year period. The Linux kernel is currently, about 9 million lines of code, and has also grown linearly over a similar period. So, let’s say for sake of argument the Linux kernel is about an order of magnitude bigger (gotta be careful with the LOC measures, but what the heck). The number of developers is similarly related. The last few releases of the Linux kernel each had about 1,000 developers contributing (with a cumulative total around 4,000). The last few releases of the Hadley model each had about 100 developers contributing, with a cumulative total around 400. Both projects work on a 2-3 month release cycle. (BTW Hadley use SVN and Trac, and have continuous integration with an automated overnight test harness).

    Now, why is Linux so successful? Yes, yes, it’s all because of the power of open source communities. Plus one more ingredient. All those developers are domain experts. They know operating system kernels inside out. The level of domain knowledge in the Linux development community is phenomenal, but we often don’t realise it, because knowledge of the principles of operating systems is built into our blood as a community.

    The Hadley model is so successful because it’s built by a team of people with a similar level of domain knowledge. Virtually every single one of them has a PhD in numerical analysis, atmospheric physics, meteorology, or a closely related discipline. The numerical core contains some of the most efficient computational fluid dynamics routines ever written.

    But maybe you’re still skeptical about the level of domain knowledge involved. So let me bring in a third comparison, also a system I’ve studied extensively: The flight control software for the space shuttle. When I worked for NASA (in the mid-90’s), it was about 500,000 lines of code (smaller than the Hadley model, but not dramatically so). It is also built by domain experts – people who know the design of the shuttle inside out.

    Now, NASA could open up the shuttle flight control software as open source, but it would do them no good, because there is no community of experts out there who could help them with it. It would take years of training to bring people up to speed with the necessary knowledge of guidance, navigation and control etc before they would be useful. It’s hard to imagine how such an expert community (outside of NASA’s rocket scientists) would come about.

    So, all three systems are successful software because they are built by people who have had years of experience in the relevant domain. They are built by domain experts, not by amateurs. The Linux kernel has a huge advantage here because there is a much larger pool of talent with the requisite expertise in operating systems than there is in numerical weather simulation or shuttle flight control.

    Now, Eric. Are you still confident that it can’t be any harder than Linux?

    PS. In case it’s not clear by now, I think your obsession with a few data handling routines in the email files from CRU is completely misplaced. These routines have about as much to do with climate modeling as Gcalctool has to do with the Linux kernel. You did read chapter 8 of WG1 of the IPCC AR4, right? So you do know what software the IPCC analysis is based on? If you really want to debug the software on which the IPCC reports are based, you need to look at these GCMs. And you’ll need a PhD in numerical methods if you want to understand how to improve the code.

  56. Curious. Why would anything believe that the code had anything to do with cover art?

    Cover art can be *drawn*. It doesn’t need tens of thousands of lines of code interpreting, modifying, adjusting, changing axis deviations, nor even a single “Artificial Adjustment”.

    That has to be one of the silliest claims imaginable. The code is cover art? Good Grief.

  57. What I wonder is why CRU even needed to write their own software simply to create graphs of historic temperature data? There are plenty of commercial and free data graphing programs. Even Microsoft Excel could do the job.

    Take the *original*, unmolested temperature record data and graph it. How difficult is that, if all you want is a direct chart?

  58. Galane:

    I do medium to advanced Excel atrocities. Excel is the swiss army knife of software tools. It has serious limits when graphing large arrays.

    As to ‘Take the *original* unmolested temperature record data and graph it…’ let’s look at the current New Zealand kerfluffle:

    From 1900 to 1927, the Wellington NZ weather monitoring station was at 3 meter of altitude and about 500 meters from the coast. The land was sold and the station moved.

    In 1928, the place where it was moved to was about 125 meters in altitude and farther inland.

    In 1960, a second monitoring station was put in at the Wellington airport, at about 5 meters of altitude, and farther inland from the original one.

    All three of these data sets are combined to determine the ‘average temperature’ for this Wellington station.

    When gridding data out on these data sets, you’re dealing with unbelievably noisy data collection and discontiguous time series, with lots of things that muck up the signal. Taking the *original* unmolested temperature record data and graphing it won’t tell you much. There are known techniques for how to solve this problem; most have been in use in science (from particle physics to astronomy, not just climate science) for decades; those techniques get refinements.

  59. Steve Easterbrook Says:
    The Linux kernel has a huge advantage here because there is a much larger pool of talent with the requisite expertise in operating systems than there is in numerical weather simulation or shuttle flight control.

    Linux has a larger pool of talent now, when it started however it certainly didn’t. When linux was first anounced the OS domain experts told him that his design was a bad idea (if you want an example of this do a search for The Tanenbaum-Torvalds Debate). The easier it is to grab and run the model (and I see no reason why, medium-long term, it couldn’t be as simple as the standard configure/make style build process that unix has long rested on) the lower the barrier to entry and the easier it is to grow a pool of talent.

  60. # Steve Easterbrook Says:
    > Jessica: didn’t anyone tell you that climate models are
    > based on computational fluid dynamics? What do you
    > think is at the heart of a GCM?

    Come on Steve, lets get serious here. I have stated this several times already. To say that you can model the airflow over an aircraft wing and make accurate predictions is one thing, to use the same tools to model the entire climate of the planet is an entirely different thing. Your attempt at a cheesy gotcha is hardly worthy of you.

    The bottom line once again is this: does your model actually work in a testable, falsifiable way? Linux works every day for people, if you are reading this message, Linux is working for you. The space shuttle launches and gets to orbit most of the time, your telescope can verify this. How about this million line climate model? Does it deliver results that are proven accurate as time plays out? Is it falsifiable? Exactly how many times does it have to be wrong before it is thrown on the trash pile and a new, better model used in its place?

    Again my question is simply this: do your domain experts have enough domain expertise to write this software? Or, perhaps more importantly, is the data of sufficient quality and completeness to write the software; does it include all the variables that have significant effect, or just the ones that are easy to measure?

    And the next question is even more important: even if the model is accurate, why does that translate into stupidity like Kyoto that won’t solve the putative problem, but will provide a naked power grab by political forces around the world?

  61. Jessica – you are conflating the solutions of the problems with the data gathering and modeling problem.

    You are also conflating the actions of the most publicity grabbing 10 or so climatologists with a field of about 2000 of them.

    Most of these models are being continually refined. The data sets that we have are, barring a time machine, the data sets we have to work with. The discussions on how to correlate those data sets for historical series (at least with thermometer temperatures) took place in the 1950s with incremental adjustments.

    For example, if you have a 60 year span of temperature records and tree ring growth rates where X temperature corresponds to Y mm of tree ring growth, corroborated with carbon sampling, and a correlation factor of 0.75, that’s a pretty solid mapping of that temperature proxy series for the period of overlap.

    Is using temperature proxies always a perfect match? No. For dendrochronology, there are factors like nitrogen in the soil, total CO2 and rainfall patterns (droughts cause thinner tree rings, which are hard to distinguish from colder temperatures, for example). Not all dendro chronology cores are suitable; some are in isolated microclimates.

    Climatology is a science in its infancy. Barring a time machine, we have no way of getting better data than what we have. The modeling techniques with CFD have gotten enormously more robust, to the point where one of the ‘exit paths’ from climate science for the people who do computer programming is getting a job in the aerospace industry doing wing load modeling.

    Here’s a quick test for you:

    What predictions does the MBH ’98 or ’99 papers make about the temperatures in 2010?

  62. A further expansion on the “why the naked power grab” – because politicians, particularly of the stripe that wants to ‘save the world’ are always looking for a crisis to drum up support for a movement. It’s called opportunism. The reason it keeps happening is because the kind of person drawn to that sort of work tends to NOT be a libertarian in outlook.

    There are other factions in climate science. Roger Pielske Sr, and Roger Pielske Jr both agree that the warming trend we experienced in the last quarter of the 20th century is caused by human activity, and that activity is much more broadly based than CO2 emissions.

    They contend that the data is dirty enough that the severity of what’s happening is hard to gauge. They also feel that our best use of efforts is going to be on adaptation mechanisms, rather than curbing emissions, because the incremental costs of adaptation are lower, even if the total cost may be higher…and they don’t doom 40% of the world’s population to abject poverty.

    In the worlds of Pielske, Jr. “Multi-variate problems rarely prove amenable to single variable solutions, no matter how stridently said solutions are called for.”

    The Pielskis are both considered anethema to the AGW-alarmists. They both detest being lumped into the ‘AGW-denier’ camp, because they don’t deny human factors influence climate, and greatly. (For example, the glaciers of Kilimanjaro are vanishing due to the overharvesting of timber in the local microclimate.) Neither has come out and said that we picked the wrong normalization period (approximately 1960 to 1978) for these models, but you can see that it’s a possibility they’re willing to contend with.

    EG, why does the activity of 10 climate scientist publicity hounds – who are now in the process of debunking themselves – make you want to throw out the work of thousands of climate scientists, and apparently want to throw out the entirety of the historical climate record?

  63. Ken Burnside Says:
    > Jessica – you are conflating the solutions of
    > the problems with the data gathering and modeling problem.

    No I’m not. I’m simply saying that people who claim authority to predict the climate over the next 50 years, and expect massive public policy changes based on their predictions had better have a pretty good track record. They do not. If, as you say, climatology is a science in its infancy, then I suggest they take the time to mature before we take their opinions too seriously There is nothing wrong with being a two year old, however, there is plenty wrong with letting a two year old drive your car. If the data is what it is (as you also argue), I accept that as true, however, just because you have data doesn’t mean your data is useful, or can form the basis for serious prediction.

    So, I recognize that it is difficult. I recognize the the guys doing is are smart. I recognize that many of them are giving their best good faith efforts. But none of that matters. All that matters is this: how reliable are their predictions? If they cannot make a very strong case that their predictions are good (that is to say, look we predicted this and it happened), then we should wait until they can before taking any action on it, certainly so if the actions are as radical as they propose.

  64. And if their predictions are good in a 25 year time scale, but time frames shorter than 15 years are too chaotic?

    As mentioned, there are climatologists who feel that we should be working on strategies for adaptation and remediation rather than socialistic attempts to remake the world economy. Rather than whining about the socialists, why not work to support the people who’re trying to get an alternative message out?

    PS. As I said to Al, if you have a time machine sufficient to let us get better climate data to run the models from, I have a number of lucrative proposals for you. :)

  65. JessicaBoxer, what would you consider to be an accurate prediction from a climate model?

  66. # Gary Strand Says:
    > JessicaBoxer, what would you consider to be an accurate prediction from a climate model?

    What a curious question Gary. The people who advocate things like the Kyoto treaty and no doubt some extra nonsense in Denmark are taking steps that will deeply undermine the world economy and your question is not “why”, but “why not”?

    To put it another way it is the responsibility of the Kyoto crowd to convince me (and my peers), it is not our job to convince you. If you want to convince me that the climate model is reliable enough to extrapolate fifty years, and that the appropriate response is deep cuts in the world’s wealth, then I need you to give me a hundred, no make that a thousand irrefutable examples of how these models are substantially reliable at predicting the future.

    Extraordinary claims require extraordinary evidence.

  67. Ah, Jessica. So it shall always be a moving target for you.

    Thank you. That saves the trouble of trying to convince you.

    Please note that the climate models, as they currently exist, try to explain an existing data record that is incomplete and noisy. There’s a reason why I asked you what predictions MBH ’98 and MBH ’99 made about the climate in 2100.

    The answer is – they make none. The hockey stick graph, flawed as it may be, is not a projection into the future. It’s one (of many) explanations of what has come out of climate science. There are several other interpretations of the data. All of the ones that use statistically rigorous methods of bridging noisy data sets show a warming trend in the last quarter of the 20th century, and it was the steepest such trend in the data set.

    It makes for good, scary press, because the natural thing to do with a graph going to the right edge of the margin is to read into it the fears of the unknown. A lot of political chicanery has gone into this stupid graphology trick, and you’re right to be upset about stupid tricks used to alter the political climate.

    Nearly all of the “Climate to 2100″ predictions are politically driven, not science driven. The scientists know the models aren’t complete, which is why they’re talking about climate in the past and growing understanding of it. Unfortunately, the politicians can use the scariest of the results to mobilize people to vote.

    Which heads back to the question:

    What would convince you, personally, that a climate model was valid?

  68. Let’s leave aside all the stuff about Kyoto and Copenhagen and the like. They’re not relevant to my question.

  69. >Extraordinary claims require extraordinary evidence.

    But you’re the one making the extraordinary claim: despite the preponderance of current evidence, you claim that our carbon emissions are not dangerous.

    The adjective “extraordinary” applies not to the claim itself, but to the relationship between the claim and the present context of what is “ordinary”.

    e.g. when Einstein made extraordinary claims about General Relativity, extraordinary evidence (gravitational lensing, orbit of mercury) was required. But nowadays claims against GR would be considered extraordinary, and would require extraordinary evidence.

    We can defer a few percentage points of GDP growth in order to insure ourselves against the potentially unlimited downsides of climate change. Your side of the argument is going to need some damned extraordinary evidence if you don’t want that insurance.

  70. We can defer a few percentage points of GDP growth in order to insure ourselves against the potentially unlimited downsides of climate change. Your side of the argument is going to need some damned extraordinary evidence if you don’t want that insurance.

    Yes, we (first-world nations) probably can. We have at least some of the infrastructure already in place to get much of our power from hydroelectric dams, wind farms, nuclear plants, etc. However, there are other countries that don’t have that infrastructure–places that depend absolutely for power, heat, and transportation on fossil fuels. If we ban their use for power, or severely curtail it, then what do these countries do? I would say that the ‘extraordinary claim’ here is that we absolutely must hugely overhaul our economy within the next several years in order to avoid massive natural disasters; such a claim is completely unprecedented, yet you assume it and insist that saying that the status quo is acceptable is the ‘extraordinary claim’.

  71. >Yes, we (first-world nations) probably can.

    It’s the first world countries that are dragging their heels.

    >I would say that the ‘extraordinary claim’ here is that we absolutely must hugely overhaul our economy within the next several years in order to avoid massive natural disasters;

    We’re not talking about a massive overhaul. Just a price on carbon.

    >saying that the status quo is acceptable is the ‘extraordinary claim’.

    Saying that the status quo is acceptable is an extraordinary claim because it goes against the evidence we have. It requires extraordinary evidence, because the evidence for a safe status quo must be more convincing than the evidence we already have.

  72. >We’re not talking about a massive overhaul. Just a price on carbon.

    I fear the power conferred on politicians by a CO2 tax far, far more than I fear elevated CO2 levels.

  73. >I fear the power conferred on politicians by a CO2 tax far, far more than I fear elevated CO2 levels.

    They already tax all economic activity. Allocating carbon permits is hardly a massive extension of their powers.

  74. Jessica: Your breathtaking ignorance of climate science is quite remarkable, given the certainty of your opinion of it. Now, before you get all insulted, I don’t mean “ignorance” here in a derogatory way; I mean it as utter lack of domain knowledge.

    Yes, climate models are falsifiable. Climate scientists are perfectly aware of which climatological phenomena the models reproduce well, and which they reproduce poorly. I’ve studied in depth the software validation processes of all three types of software I described above, and it’s my professional opinion that climate models are better tested and validated than any of the others (yes, better even than the space shuttle flight software, which is reputedly the most expensive software per line of code ever written). I should point out that this was a complete surprise to me – I expected to find that climate scientists were doing a lousy job with their software V&V. But I don’t imagine for one minute you’re likely to believe me about this.

    If you actually read any of the science, you’d know how the models are tested. Please, please, please go and read chapter 8 of IPCC AR4 WG1. Then do me a favour, and read my paper on the development processes (http://dx.doi.org/10.1109/MCSE.2009.193). Once you’ve actually gone and found out what climate models are, and how they are tested, I’d be quite happy to answer any questions you have about the details.

    Alternatively, you could put your fingers in your ears and shout “la la la, I can’t hear you”. Your choice.

  75. ESR “I fear the power conferred on politicians by a CO2 tax far, far more than I fear elevated CO2 levels.”

    Eric, how about you humour me, and perform the following thought experiment. Imagine for a moment that we do have incredibly strong scientific evidence that continued emissions of CO2 on our current trajectory would make the majority of the planet uninhabitable by the end of the century. If a carbon tax (or any other form of governmental regulation) is completely unacceptable to you, what would we do about it? (To do the thought experiment properly, you have to pretend that the risk is *very* real – no cheating and saying “but it isn’t”!!!)

  76. >Eric, how about you humour me, and perform the following thought experiment.

    I’m not going to humor you. The scenario you describe is impossible, and attempting to rhetorically back me into a corner by proposing that I was including it in “elevated CO2 levels” is either dishonest of you or outright crazy. I might just as well ask you how you would respond if the moon turned into gorgonzola and pieces of it fell on your head.

  77. >Eric, how about you humour me, and perform the following thought experiment.

    I’ve already had an answer from esr on this, it’s towards the end of this thread.

  78. Pete: do you mean this one:
    [ESR: "The standard libertarian answer starts with treating pollution like a tortious assault. CO2 is particularly susceptible to this kind of market internalization because, while it’s not toxic at the concentrations industrial processes emit, it’s associated with other combustion byproducts that definitely are.

    To get a Coase effect going, you need a second step: a market in causes of action. So, say I’m Greenpeace and I want to sue a polluter to clean up their act. What I do is, I announce what I’m interested in doing, and ask individuals to sell or assign me their right to recover pollution-related damages. I collect a million or so of those, then then go to the court system. From there it’s all process issues about how you enforce the law; the answers vary depending on whether you’re a minarchist or an anarchist."]

    So the only solution is to let the pollution happen and then sue for damages? That’s not a valid response to my thought experiment, because my thought experiment was about levels of pollution that make much of the earth uninhabitable. You can’t let it happen and then sue, that’s just stupid. So it comes down to this then does it? Libertarians deny the science of global warming because their political theory cannot handle risks of this magnitude?

  79. >Libertarians deny the science of global warming because their political theory cannot handle risks of this magnitude?

    No political theory can handle risks of the impossible magnitude you describe. The ones that claim to are fantasizing.

    I was allergic to junk science long before I was a libertarian.

  80. Steve Easterbrook Says:
    > Eric, how about you humour me, and perform the following
    > thought experiment. Imagine for a moment that we do have
    > incredibly strong scientific evidence that continued emissions
    > of CO2 on our current trajectory would make the majority
    > of the planet uninhabitable by the end of the century.

    Imagine if you will that crazy loons around the world were on the point of obtaining world destroying weapons to accompany their Messianic eschatology. Wouldn’t that insane, world destroying scenario pull all the peoples of the world together to overcome such a threat? Surely the political will can be found to solve that crisis?

    Imagine the time was 1975. NASA had just deployed a new telescope, and had discovered a 50Km wide asteroid speeding toward the earth with 75% certainty of collision in the next year. Would we have the technological know how to solve the problem?

    If we did not have the technological know how, would you propose a global law that required all the peoples of the world to have a bow and arrow to fire at the asteroid when it entered earth’s atmosphere?

    Imagine someone built a model of the stock market. This model was the most beautiful piece of software that was ever written. Its millions of lines of unit tests ran and reran analysis of all historical stock data in history, and accurately predicted the historical prices. However, when run in extrapolation mode, guessing future prices, it was always off by a profit killing margin (mainly because it only modeled the stock prices themselves — which are easy to measure — not the extrinsic effects on stock prices which are harder to measure or quantify.) Would anyone consider this a good piece of software? Even if Don Knuth published it in a book and paid you $16,384 dollars to find a bug?

  81. ESR: “…impossible…”

    Impossible you say? For it to be impossible then something must be wrong with either Fourier’s experiments on the role of greenhouse gases, or the laws of thermodynamics. For taken together, they surely demonstrate it’s *possible*. Perhaps you meant “unlikely”?

    But I think I will leave you at this point. I think I may have been mistaken in believing you to be intelligent.

    ESR says: Yes, I say “impossible”. There’s no way humans could add that much CO2 to the atmosphere if they were *trying* to do so.

  82. > I think I may have been mistaken in believing you to be intelligent.

    When argument fails, get snotty!

    Really Steve, you can do better than that.

  83. Steve, a couple of points, you seem to argue that there’s no use in open sourcing climate modeling software because most people wouldn’t be able to understand them. However, that’s besides the point, the real issue is that there exist other sufficiently knowledgeable people who would like to run the models, examine the code and results, and come to their own conclusions. The current CRU flap is precisely because they weren’t willing to have their results second-guessed in such a way, which is one of the reasons why it’s such a big deal. As for your comparisons of the code complexity of linux vs Hadley vs shuttle flight control software, I believe the point Jessica and esr are trying to make is that the complexity of the actual climate is many orders of magnitude more complex than what linux or shuttle software try to accomplish, to the point where it’s likely impossible to model with any precision. As linked by Morgan earlier in this thread, many eminents, such as Freeman Dyson, have criticized climate models for this same reason, that they’re nowhere near being complex or robust enough to be worthwhile.

    As for the doomsday scenarios mentioned by you and others here, where are those coming from? Even if I were to accept AGW, which I don’t, my understanding is that the worst-case predictions are of heating of 1-6 degrees C. That’s nowhere near the extinction event that some are now making it out to be. Finally, you keep asking people to read all the reports and data before discussing. However, that’s fairly disingenuous if you believe that one must have a very specialized PhD to really engage with all of it. The fact is most people aren’t going to go to all that trouble, we don’t have the time or interest. We will sniff at what’s going on, looking for various telltale scents of a scam that we’ve learned to recognize over years, and when it smells bad, many of us are fine with writing AGW off as junk and moving on. Those who are predisposed to believe AGW will do the opposite and ignore the smell. If you want to get anywhere, you need to convince all the people who are capable of engaging deeply with the theory and modeling that your predictions of dire consequences are true. However, the AGW camp has failed at this so far, as there is a significant minority of skeptical scientists that disagrees. Maybe you will be able to convince all of them through your evidence someday, but I doubt it. Also, there will always be a group of people that doesn’t think climate modeling can ever work, likely including Jessica, and who will never come around.

  84. A very interesting discussion. I’ll add a few thoughts.

    >Gary Strand asks: “what would you consider to be an accurate prediction from a climate model?”:

    How about: one that can do better than a naive model; i.e., for each year’s prediction of the global average temperature, use last year’s actual temperature. My understanding is, the naive model has tremendously less error than any other proposed CO2-based model.

    >pete says: “We’re not talking about a massive overhaul. Just a price on carbon.”

    My understanding is AGW proponents are talking about taxing carbon to the effect of reducing world GDP by 1% (and in some cases, 2%). They use these figures – 1 to 2% – as a way of showing how slight the effect is. In reality this is a tremendously expensive endeavor. If world GDP grows by 5% a year for 50 years, the world’s economy will be ~11.5 times larger than today’s. At 4% a year, it will be ~7 times larger.

    For 100 years out it’s even worse: ~131 times larger vs. ~50 times larger. I think the extra money would be better spent on mitigation and adaptation (if needed) than any expensive plan today that is based on an unproved/uncertain connection between CO2 levels and global warming.

    >Steve Easterbrook asks: “Eric, how about you humour me, and perform the following thought experiment. Imagine for a moment that we do have incredibly strong scientific evidence that continued emissions of CO2 on our current trajectory would make the majority of the planet uninhabitable by the end of the century. If a carbon tax (or any other form of governmental regulation) is completely unacceptable to you, what would we do about it?”

    I’ll take a stab at it. The idea behind AGW is that CO2 increases will increase global temperatures. Since CO2 is not actually a pollutant, it is the increased GT that would “make the … planet uninhabitable.” The connection for CO2 driving temperature is tenuous and in dispute; however, the negative consequences of higher temperatures is … well, less in dispute. Therefore, we should spend tens of millions of dollars on strategies that could reduce GT, rather than trillions of dollars on strategies that reduce CO2.

    I think the hard part will be a global consensus on what GT is optimal. Probably, it is ~1-2% C warmer than today, but that – somewhat – depends on where you live.

    Cheers. -Ted

  85. “There’s no way humans could add that much CO2 to the atmosphere if they were *trying* to do so.”

    Upon what grounds do you base this claim?

  86. >Upon what grounds do you base this claim?

    Do the math. Currently atmospheric CO2 is at 300-400ppm. Toxic reactions start at about 50,000 ppm (5% of atmospheric volume) and become a prompt danger at about 250,000 (25% of volume). That’s more than two orders of magnitude difference. This means that to get to Steve Easterbrook’s doomsday scenario humans would have to dump about five hundred times as much CO2 as they do now. Where’s it all going to come from? Even burning down all the world’s forests wouldn’t do that much.

  87. Why do you assume that “uninhabitable” means “everything is dead”? The Atacama Desert, big chunks of the Australian outback, the Sahara, and so on, are pretty much uninhabitable. Doesn’t require CO2 concentrations to be toxic, either.

    ESR says: I was following out Steve Easterbrook’s premise exactly as he expressed it. It’s a nutty premise, and changing the subject now won’t make it any less so.

  88. >ESR says: “I was following out Steve Easterbrook’s premise exactly as he expressed it. It’s a nutty premise, and changing the subject now won’t make it any less so.”

    Oh, I read that differently. I thought Steve was talking about the CO2 making things uninhabitable due to increased heat, flooding, etc. At least, that’s the idea I was responding to. -tc

  89. I agree with Ted – let’s assume a worst-case scenario. Coastal cities have to be evacuated, current agricultural regions lose a lot of productivity due to precip and temp changes, and more-northern areas lose their permafrost and become basically big areas of muddy goo. I don’t know if the loss of those areas’ habitability constitutes a “majority”, but I think we can all agree that such a situation would be bad, especially considering that there’s ~6.7 billion of us.

  90. Gary Strand Says:
    > The Atacama Desert, big chunks of the Australian outback, the
    > Sahara, and so on, are pretty much uninhabitable. Doesn’t require
    > CO2 concentrations to be toxic, either.

    This is a point worth talking from. It is true that these deserts are largely uninhabitable by humans, but it doesn’t have to be that way. For example, Las Vegas and Los Angeles were both built on uninhabitable deserts too. Much of Holland is built on reclaimed land from the sea. Washington DC was built on a malarial swamp (cough, cough, I mean “wetland”.) Heck Disneyworld was built on a place only alligators and snakes love. Engineering can be applied to make these things work, and for vastly less dollars, and vastly less risk than hamstringing the manufacturing industries of the west. Most importantly, we don’t need to do any such work until the problems actually arise, which is to say, we don’t have to speculate on dicey science.

  91. “It is true that these deserts are largely uninhabitable by humans, but it doesn’t have to be that way. For example, Las Vegas and Los Angeles were both built on uninhabitable deserts too.”

    Both Vegas and LA depend on Colorado river water, and LA depends on mountain snowpack. If the Colorado’s flow drops and more mountain snow falls as rain, the potential water crisis is significant. Likewise, draining underground aquifers faster than they replenish is unsustainable.

    “Much of Holland is built on reclaimed land from the sea.”

    Holland can afford it, and the area reclaimed is small. The same cannot be said for, say, Bangladesh.

  92. Gary Strand said: “Both Vegas and LA depend on Colorado river water, and LA depends on mountain snowpack. If the Colorado’s flow drops and more mountain snow falls as rain, the potential water crisis is significant. ”

    This is true; it would be a ‘water crisis’ (if all these horrible AGW effects came to pass). But I zeroed in on this point – water availability – because it is a particular pet-peeve of mine from the media. The issue isn’t that these places would run out of water … rather, they would run short of *cheap* water. In other words, they would have to build huge coastal desalination plants, and reconfigure transport, and probably change a lot of practices (such as having lawns and swimming pools).

    We can do all these things now. In 50 or 100 years – if we need to do them – then we can, in a much cheaper and more efficient way. *If* we allow our economies to continue to grow, and we put money in research and development, etc. Mitigation and adaptation is pennies on the dollar compared to the dubious strategy of spending trillions to reduce CO2 emissions by some fraction.

    What do you think Bangladesh would prefer? That we spend $140 billion dollars a year (1% of our GDP) to reduce CO2 by a bit; or, say, spend $1 billion or so to give them food, shelter, infrastructure, micro-loans, dikes, etc.?

  93. @Ted >My understanding is AGW proponents are talking about taxing carbon to the effect of reducing world GDP by 1% (and in some cases, 2%). They use these figures – 1 to 2% – as a way of showing how slight the effect is.

    That 1–2% is 1–2% of 2050 GDP; i.e. in 2050 the world economy would be 1–2% smaller than it would have been without global warming.

    >Since CO2 is not actually a pollutant.

    Don’t forget ocean acidification.

    >Oh, I read that differently. I thought Steve was talking about the CO2 making things uninhabitable due to increased heat, flooding, etc.

    You read it correctly. esr’s just being disingenuous.

    ESR says: pete is pretending he can read my mind.

  94. “The issue isn’t that these places would run out of water … rather, they would run short of *cheap* water. In other words, they would have to build huge coastal desalination plants, and reconfigure transport, and probably change a lot of practices (such as having lawns and swimming pools).”

    Desalination isn’t cheap, nor energy-cheap. Is that really preferable?

    “*If* we allow our economies to continue to grow, and we put money in research and development, etc.”

    That assumes that for each dollar that the economy grows richer, it costs less than a dollar to mitigate climate change. If the balance is the other way, no matter how much the economy grows, climate change will bite us. Hard.

    “…say, spend $1 billion or so to give them food, shelter, infrastructure, micro-loans, dikes, etc.?”

    Do you think the dike system to protect Bangladesh would amount to $1 billion, at most? That’s amazingly low.

    The irony, of course, is that the Bangladeshis did nothing to cause it, but they have to pay anyway.

  95. >ESR says: pete is pretending he can read my mind.

    ESR, in contrast, is actually reading Steve’s mind.

  96. Gary Strand Says:
    > Desalination isn’t cheap, nor energy-cheap. Is that really preferable?

    Thanks Gary. You have correctly moved the discussion from religion to economics. It is not “The end of the world is nigh!!!” it is instead, here are several alternatives, which is most cost effective. That is exactly the point I was making, and it is a much more reasonable discussion. We can properly discount the crazy AGW scenarios for risk (and include the discounted benefits of putative AGW too.)

    Enough of the political eschatology.

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>