The “pause” that probably isn’t

Matt Ridley has a particularly silly article in the Wall Street Journal called Whatever happened to Global Warming. In his article he says

Well, the pause has now lasted for 16, 19 or 26 years—depending on whether you choose the surface temperature record or one of two satellite records of the lower atmosphere. That’s according to a new statistical calculation by Ross McKitrick, a professor of economics at the University of Guelph in Canada.

It has been roughly two decades since there was a trend in temperature significantly different from zero. The burst of warming that preceded the millennium lasted about 20 years and was preceded by 30 years of slight cooling after 1940.

This is based on a new paper by Ross McKitrick in which he determines what length of period prior to 2009 is required so that the trend plus 2σ uncertainty interval in the temperature record does not intercept zero. Richard Telford has already done a wonderful take-down of this work by showing that 15-20 year periods are quite possible even if the underlying long-term trend is constant and rising. What Ross McKitrick and Matt Ridley are almost certainly doing is making a Type II error. He’s accepting the null hypothesis (we’re not warming) when we almost certainly are. The comments by Dikran and Chris Colose on Richard’s post are also worth reading.

Matt Ridley goes on to say

This has taken me by surprise. I was among those who thought the pause was a blip. As a “lukewarmer,” I’ve long thought that man-made carbon-dioxide emissions will raise global temperatures, but that this effect will not be amplified much by feedbacks from extra water vapor and clouds, so the world will probably be only a bit more than one degree Celsius warmer in 2100 than today. By contrast, the assumption built into the average climate model is that water-vapor feedback will treble the effect of carbon dioxide.

Well, here Matt has completely ignored the ocean heat content data and the continually reducing ice mass, both of which indicate that we’re accruing energy at a rate that is consistent with what we’d expect. The surface warming is only associated with a few percent of planetary energy imbalance, and so it’s not at all surprising that it shows significant variability.

And then, quite remarkably, concludes with,

But now I worry that I am exaggerating, rather than underplaying, the likely warming.

Don’t worry Matt, I think many people still regard you as an out and out denier. Of course, I would never call you that, but your concern that you’ve been a little alarmist is entirely misplaced.

When I see this kind of thing it makes me realise that there’s no real point in discussing this with such people. We’re talking a completely different language. If you want to consider anthropogenic global warming, you really should consider all the evidence. You can’t consider a subset of the evidence and then draw conclusions about whether we’re warming or not. Additionally, these discussion invariably end up being ones in which you’re challenged to show where there’s an error in their calculation. There’s not necessarily an error, but context is crucial. It’s a little like (and I exaggerate) someone doing a simple calculation (2 + 2 = 4), claiming that they’ve shown that Einstein is wrong, and then insisting that you can’t prove them wrong until you find the error in their calculation. I’m sure Willard would have some term that described such forms of argument.

I’ll finish this post by mentioning a relevant article by Richard Betts on Climate Revolution called Pooh sticks, pauses, and predictability. It’s a good post, and it reminded me that – at university – we had a student magazine that once included the classic drawing of Pooh and Piglet playing Pooh sticks, but in which they were facing the other way. My mother claimed that it had ruined Winnie-the-Pooh for her. Richard finishes his post with

Claiming that long-term warming won’t happen because the ‘pause’ was not specifically predicted is like saying that you can’t be sure the river will flow downhill because you can’t predict the winner at pooh sticks.

Advertisements
This entry was posted in Uncategorized and tagged , , , , , , , , . Bookmark the permalink.

76 Responses to The “pause” that probably isn’t

  1. “And some people say climate scientists need to work more closely with statisticians.”

    I do, but then I am (a sort of) statistician! ;o)

    I like working with climate scientists, they have interesting data to play with and interesting physics to learn about.

  2. I like working with climate scientists, they have interesting data to play with and interesting physics to learn about.

    That’s how it’s meant to be. The insistance that one group should work with another only really works if they actually want to work together.

  3. I was sure I’d written a comment on this post, but it appears to have completely disappeared, or I’ve commented on the wrong post and can’t find it. I’m actually on the way back from a meeting at which I gave a talk that was mainly aimed at showing that what another group has been doing is wrong (well, heavily over-interpreted) and also to partly show that what some in my research community thinks is plausible, isn’t. An advantage of all this climate blogging (the only one, maybe 🙂 ) is that it’s got me thinking a bit more fundamentally about physics again. So, in my talk I focused on why what was being suggested by the other group was physically implausible and – from some of the comments afterwards – I think it was successful. However, one of those whose research I was “criticising” was in the audience and their response was to comment that even if I was right, what it really meant was that we had to do more to understand this particular area. Rather than pointing out that this was mainly because they’d gone and muddied the water by publishing work that was heavily over-interpreted, I decided to simply agree. It made me think that we were simply engaged in a round of ClimateballTM.

  4. Don’t worry Matt, I think many people still regard you as an out and out denier. Of course, I would never call you that, but your concern that you’ve been a little alarmist is entirely misplaced.

    I’m sure Willard would have some term that described such forms of argument.

    I guess it are these kind of sentences, scientists love to write, that made Anthony Watts once complain about the formulations on my blog. 🙂

    I love them.

  5. Victor,
    Indeed, they’re great. I have no idea why people have problems with them. There’s nothing specifically wrong with them, and they seem to be pedantic about everything else 🙂

  6. Willard,
    I knew you wouldn’t let me down 🙂

  7. So Matt goes on a cherry picking expedition through the satellite data and then pretends not to be a climate change denier.

    *Yawn*

  8. Robert,
    In fact, in this case it appears that he’s worried that he’s been too much of an alarmist because we aren’t even warming as fast as he thought we were. I’m not quite sure how to describe that …. incredibly confused is all I can think of at the moment (well, I can think of others, but I’ll refrain from actually saying them).

  9. Tom Curtis says:

    1) In Feb of this year, Matt Ridley wrote “When the sceptic David Whitehouse first pointed out the current 15 to 17-year standstill in global warming (after only 18 to 20 years of warming); now the science establishment admits the “pause” but claims to have some post-hoc explanations.” If he thought the explanations were ad hoc, he did not think they were valid. Therefore his claims that he previously thought the “pause” to be just a blip are, therefore, disingenuous.

    2) In response to Telford, McKittrick emphasizes that the tests conditions that “…that the trend CI includes zero across the entire time subsample and applied in both the NH and SH”. In fact, significant subsamples are statistically significant for all time series checked by McKittrick (and the two major ones he left out of the test), as shown by comparison of the longest period satisfying the test applied by McKittrick as determined on the SkS trend calculator starting from 2013 and working backwards, compared to the longest period with significant warming starting from the start year from the above test and working forward:

    RSS: 1990-2010: significant; -2013: not significant (24 years)
    UAH: 1993-2010: significant; -2013: significant (21 years)
    HadCRUT4: 1995-2011: significant; -2013 not significant (19 years)
    NCDC: 1995-2007 significant; 1995-2013 not significant (19 years)
    GISS: 1996-2010 significant; -2013 not significant (18 years)

    This asymmetry in the test, where a statistically significant forward interval is considered irrelevant, would be (I think) impossible to justify theoretically, and renders the V-F (2005) test irrelevant.

    3) If we applied the V-F test to determine how long the reduced warming in the 50s and 60s lasted, it would likewise be biased to find no-trend in periods where the same terminal year but a later initial year showed statistically significant trends. The net effect is that the test is heavily (outrageously) biased to show no trend over trend.

    4) In any event, more recent literature has shown the V-F(2005) test to perform poorly in testing for lack of a trend compared to a simple t-test. Therefore, all that McKittrick has shown is that using an inferior test with a heavy bias towards finding no trends results in finding long periods with no trend. Telford then shows his finding to not be statistically significant, a conclusion McKittrick simply brushes aside.

    5) Finally, it is well known that tropospheric temperature measures (RSS, UAH) have a greater response to ENSO fluctuations than to surface temperature measures. The finding that they show a greater period of “no trend” in this flawed test is an indirect (albeit weak) confirmation that the apparent lack of trend is a product of ENSO variations, and hence not a feature of the long term trend.

  10. Marco says:

    I am sure the Auditor will demand McKitrick provide him with all his data and intermediate calculations, and write something disparaging about those economists and their erroneous understanding of statistics. Right?

  11. Lars Karlsson says:

    “What Ross McKitrick and Matt Ridley are almost certainly doing is making a Type I error. He’s accepting the null hypothesis (we’re not warming) when we almost certainly are. ”

    So we can add Ridley to the list of people who don’t understand basic concepts of statistics.

  12. Ridley has given the world yet another set of denier “own goals” – at least two of them in his one article.

    Own Goal (1): He gleefully promotes the recent Chen and Tung paper in order to try to support his claim that global warming stopped. But this paper and its authors outside this paper say the opposite. Ridley wrote,

    “Putting the icing on the cake of good news, Xianyao Chen and Ka-Kit Tung think the Atlantic Ocean may continue to prevent any warming for the next two decades. So in their quest to explain the pause, scientists have made the future sound even less alarming than before.”

    Icing on the cake? Good news? In this article (a good one that all should read)

    “New Study Provides More Evidence That Global Warming ‘Pause’ Is A Myth”
    http://thinkprogress.org/climate/2014/08/25/3475168/global-warming-atlantic/

    we see this quotation from the Chen and Tung paper itself:

    “When the internal variability that is responsible for the current hiatus switches sign, as it inevitably will, another episode of accelerated global warming should ensue.”

    As reported in articles like this one

    UW Study: Despite ‘Hiatus’ In Rising Temps, Oceans Show Globe Still Warming
    http://www.kplu.org/post/uw-study-despite-hiatus-rising-temps-oceans-show-globe-still-warming

    Tung himself says that global warming has not stopped. This just cited article quotes Tung as saying,

    “This is not denying that there’s global warming. The global warming is there, but sometimes it’s reinforced by natural cycles, and sometimes it’s partly canceled by natural cycles.”

    In other words, a roughly staircase-shaped graph fluctuates around the long-term upward trend that is just as steep as before.

    Own Goal (2): Some deniers deny the deep ocean data that show not only no slowdown in overall global warming but actually an accelerating increase. This next article addresses this data:

    http://www.skepticalscience.com/The-Oceans-Warmed-up-Sharply-in-2013-We-are-Going-to-Need-a-Bigger-Graph.html

    Quote:

    “Long-term the oceans have been gaining heat at a rate equivalent to about 2 Hiroshima bombs per second, although this has increased over the last 16 or so years to around 4 per second. In 2013 ocean warming rapidly escalated, rising to a rate in excess of 12 Hiroshima bombs per second – over three times the recent trend. This doesn’t necessarily mean we are entering a period of greatly accelerated ocean warming, as there is substantial year-to-year variation in heat uptake by the oceans. It does, however, once again dispel the persistent myth of a pause in global warming, because the Earth has actually continued to warm faster in the last 16 years than it did in the preceding 16 years.”

    Yet Ridley cites a study using this data that some deniers deny.

    As I understand it (perhaps someone can verify this), this ocean data that Ridley now accepts says that if all the heat that went into the oceans during so-called no global warming during the last roughly 15 years went instead into heating the atmosphere, we would have seen during this time a global warming increase at the surface of roughly 40 degrees Fahrenheit. He needs to explain how this is no global warming.

    Side note: I think that in order to win the public’s mind against denier propaganda that tries to make this ocean heat increase sound like nothing, it would be a good idea to give strong information that the public can relate to such as this would-have-been surface warming of roughly 40 degrees Fahrenheit in just the last 15 or so years.

  13. Marco says:

    K&A, probably Richard Tol will now show up and say that it wasn’t that Chen and Tung paper that Ridley meant. Not even if Ridley hotlinks to that paper – no, he was referring to another (left unidentified) paper.

  14. Michael 2 says:

    [Mod: I don’t want to have a discussion about this word, thanks. It has been done numerous times already to no avail.]

    “it would be a good idea to give strong information that the public can relate to such as this would-have-been surface warming of roughly 40 degrees Fahrenheit in just the last 15 or so years.”

    Yes, please do that. But you can do much better. Since you are proposing to move heat from where it went to where it didn’t; infuse a cubic centimeter of matter with all that heat and describe the result. Others here could probably do the math mentally. I’m sure it would make an impressive temperature, much more impressive than quoting a number of Hiroshima bombs which I doubt tells very many people very much. The word “millions” needs to be in there somewhere — millions of people dead, millions of degrees. It’s a nice word.

    To think the IPCC was forecasting only 9 degrees. Wimps!

  15. Michael 2 says:

    [Mod: Off-topic]

  16. jsam says:

    Is M2 seems to be playing a different game ,with fewer pieces. Listen to the rules again, https://www.youtube.com/watch?v=VrwUAFV706k&feature=youtu.be&t=1m14s

  17. BBD says:

    M2

    BBD’s conversion, quoting BBD in part “Or at least, some of us do,” reveals (in my opinion) his actual social motivation. He wishes to belong to a tribe, and speak for it (“us”) as an equal member. This is the “tell” that reveals fake skeptics. In my opinion, a “true” skeptic isn’t a member of any tribe because he is skeptical of all of them.

    For future reference, I prefer to be quoted in context and with a link to the source:

    I discovered that I was being lied to. This simply by comparing the “sceptic” narrative with the standard version. Unlike my fellow “sceptics” I was still just barely sceptical enough (though sunk in denial) to check both versions. Once I realised what was going on, that was the end of BBD the lukewarmer (NB: I was never so far gone as to deny the basic physics, only to pretend that S [the climate sensitivity] was very low). All horribly embarrassing now, of course, but you live and learn. Or at least, some of us do. …

    Always check. Fail to do this in business and you will end up bankrupt and in the courts. I failed to check, at least initially, and made a colossal prat out of myself. Oh, and never underestimate the power of denial (aka ‘wishful thinking’). It’s brought down better people than me. …

    There wasn’t a single, defining eureka moment, just a growing sense of unease because nothing seemed to add up. … Once I eventually started to compare WUWT [Watts Up With That] with RC [RealClimate] and SkS [Skeptical Science], that was it, really.

    It’s much easier to understand what I was – and was not – actually saying.

  18. Rob Nicholls says:

    I’m always fascinated by McKittrick’s papers. Needless to say the stats always go above my head and I’ve not generally heard of the statistical methods that he uses. Perhaps nitpicking, but I was going to ask whether it’s surprising that McKittrick managed to get this paper published when it talks about “an apparent levelling off of globally-averaged temperatures over the past 15 years or so” (with the word “surface” often omitted) but then I read the link to Richard Telford’s blog post and the links within that about the quality of the journal that published McKittrick’s paper, so perhaps it’s a silly question.

    Then the title of another Richard Telford post caught my eye: “Statistics is not a substitute for physics.” Then the comments caught my eye and I thought they were really interesting and I almost fell off my chair at Ross McKittrick’s comment: “There are any number of null hypotheses one can test, but I think the one of most interest is trend=0. At least that’s the one I was interested in.”

  19. anoilman says:

    Rob Nichols: There’s nothing new about all that… after all Richard Tol substituted real scientific papers for statistical papers. He steadfastly refuses to actually locate his mythical papers despite being offered cold hard cash many times. In fact. No one in the denial community can seem to actually find those papers. A discrepancy they are collectively and intentionally blind to.

    Yup… This is cutting edge stuff if you believe in fairies, magic and Father Christmas.

  20. Tom Curtis says:

    Must M2 infest every post on this blog with comments? I know there are often very intelligent replies to his comments, but the aggregate effect still must lead to a loss of IQ points in any reader.

    In this particular instance his comments are not even on topic, and the second does not respond to anything in the prior comments. That should be a strong clue that his sole purpose on this blog is to disrupt intelligent discussion with his own brand of nonsense. It is trolling in the purest and simplest form. Having discovered a site expressing opinions he disagrees with, he wishes to kill it with endless blather in the comments.

    I do not know why he is still tolerated here.

  21. Rachel M says:

    Thanks, Tom. I’ve already deleted three of M2’s comments today 🙂

    It’s quite nice to have some different voices though otherwise people complain that this is an echo chamber. People are free to ignore his comments if they want to.

  22. BBD says:

    Rachel and Tom C

    I agree with the different voices bit. That’s essential. But the really blatant misrepresentation that M2 ventured above is irritating. So I’m with you both, which might be a minor contribution to keeping it civil.

  23. Rachel M says:

    … a minor contribution to keeping it civil.

    Thanks, BBD. That’s very diplomatic of you. Maybe you should be the moderator 🙂

  24. John Mashey says:

    My management training (both explicit and implicit) was from pretty savvy people.
    1) One theme was to draw out opposing or minority poi8nts of view, get them articulated as strong as possible, see if viewpoints could be combined, synthesized or generally gotten to contribute to progress. Decide if more data was needed, but if necessary, get decisions made and move on. This might be modeled after the classic 12 Angry Men. I’ve often run program committees this way and some of my bosses have been superb at such things.

    2) On the other hand, I was taught to recognize the sort of person who tries to make every discussion revolve around them (whether their expertise is central or not), whose arguments tend to divert from the task and at, or fuzzify them or generally turn brains into mush, as per Billy Madison. For these:
    a) First, if they worked for you, try to change their behavior, and if unable, get them to move somewhere else, or if all else fails, fire them, following the rules, of course.
    b) If they don’t work for you, but are in same organization, maybe they can be offered a chance to join some prestigious strategy task force that will need to study something for a year and then write a report. Another possibility is that they be given enough rope to overreach in empire-building, and be asked to resign.

    Silicon Valley companies tend to be a little more hair-trigger on 2) than Bell Labs, but I had occasion to implement 2) in both places. There are plausible analogies for blog moderation.

  25. Steve Bloom says:

    “People are free to ignore his comments if they want to.”

    Or go elsewhere, or participate less here. There’s a cost, Rachel. That was Tom’s underlying point (he should correct me if not), and your response to it wasn’t very responsive.

    One thing to bear in mind is that when casual readers hit a tl;dr compendium of blather like M2’s most recent comment they may just stop reading the thread.

  26. Steve Bloom says:

    Footnote that last with this.

  27. Rachel M says:

    Ok, I accept the criticism. The last comment was off-topic and so I’ve moderated it accordingly. I am the one who approved it so it’s my fault.

    I’m not really supposed to be here. I was planning to take this month away for personal reasons but could not manage to stay a few days away without reading comments. But perhaps I will try to slink away again for a time.

  28. Sorry, I’ve not really been paying attention for the last day or so. Got back really late on Friday night (partly because I bumped into some friends outside a pub while heading home from the train station and decided that since it was already late, being a little later wasn’t going to make any difference 🙂 ) and then was out gain yesterday. Will probably be back up to speed in the next day or so. Rachel, thanks for holding the fort.

  29. Steve Bloom says:

    Thanks, Rachel.

    KeefeAndAmanda, 40F is obviously not right. What’s the correct figure and the source for it?

  30. Steve,
    If you assume the land/atmosphere has a mass of around 1019kg and a specific heat capacity of 1000 J kg-1 K-1, then that suggests you need 1022J to increase the temperature by 1 degree. A ballpark figure (I think) for the OHC is around 1023J per decade and so the last 15 years would have increased the land/atmosphere temperature by around 15 degrees Celsius if it had all gone into heating the land/atmosphere (which it can’t of course). These are all just ballpark figures, so it might be 10 rather than 15, but it would certainly be 10s of degrees fahrenheit.

  31. Steve Bloom says:

    Right, I missed the “all the heat” bit.

  32. ATTP, thanks for the corrective calculations.

    It’s increasingly clear to me that an effective and important way to counter the denier community’s incessant chanting of “No more global warming! No more global warming!” is simply to incessantly counter with the fact that this chant is a flat out lie or at least the moral equivalent of one and to prove that it is such by showing how much energy has actually accumulated during this time of supposed no warming with numbers that the public can relate to, numbers like how much in terms of the Fahrenheit measure the atmosphere would have warmed had all the energy accumulated during this time gone into heating the atmosphere. This is not to have the public think that such a thing is physically possible, but simply to have them see just how big this lie really is. I mean, the amount of energy to cause close to a 30 degree Fahrenheit increase in the average surface temperature in just 15 years reflects a lot of global warming. No more global warming? Please!

    (Side note: Some claims of fact can be proved in science at least to the degrees of proof used in courts of law such as proof beyond a reasonable doubt, or else we could not speak of scientific facts, which are claims of fact that have been proved to such degrees by science. Isn’t for instance the claim that the Earth revolves around the sun a scientific fact proved beyond all reasonable doubt?)

  33. > when casual readers hit a tl;dr compendium of blather like M2’s most recent comment

    SteveB’s now an authority on casual reading.

    The same argument can be used against any comment one does not like. For instance, how many casual readers do I need to invoke against playing the ref? What about piling on? And food fights:just imagine casual readers cringing over food fights.

    Let’s do some blind tests with casual readers. Let’s make them read M2’s comments and SteveB’s. Let’s see where this leads us.

    The only thing we need to solve is how to make them read blinfolded.

  34. Tom Curtis says:

    Rachel, my comment was not intended as a criticism of your moderation, which I consider excellent, but rather of the rule under which that moderation operates. Specifically, you are a part-time, volunteer moderator so we do not expect that you will be able to moderate every comment live and immediately – or even within 24 hours of posting. Therefore the rules of the site should allow you to place explicit restrictions on repeat offenders such as a requirement that they restrict their comments to a particular thread, or a declared no tolerance to off topic posting from them, or a declared requirement that they do not raise a topic exhaustively discussed on one thread on other threads; with the implicit assumption that if they do not self moderate as directed they will lose all posting rights.

    In M2’s case, he has raised his idiosyncratic claim that “true skeptics™” are rugged individualists while those who in fact value science suffer from a herd mentality (rendered unfalsifiable, I note, by a “no true Scotsman fallacy”) on several threads and discussed it exhaustively. Therefore it would be entirely appropriate to direct him to not raise it again on any other thread on this site without written moderator approval on pain of banishment.

    There is no risk of becoming an “echo chamber” under that policy. There is no risk, IMO, in any event in that the regular commenters here are critical thinkers who do not accept things on other peoples say so; but even if that was a risk, having threads which do not end up as yet another exhaustive examination of M2’s self serving hypothesis hardly makes the site an echo chamber. Further, if M2 feels a desperate need for the L^Lth iteration of his theme (where L is some very large number), he is always able to do so on a thread in which he has already raised it and is permitted to continue doing so. Or on his own blog, even (and we can see how many readers such a blog would attract).

    The “written exception” possibility would be to cover cases like the discussion of the definition of a “pause” or “hiatus” on this thread where it is obviously on topic but may have been done to death on other threads (and hence placed under restriction). In such cases it should always be possible to include a single sentence referral to that prior discussion along with a link without being considered to have breached moderator guidelines. (That is, we don’t want to restrict the expression of relevant ideas, only their discussion ad nauseum.)

    (For the record, Steve Bloom did capture my concerns.)

  35. Joshua says:

    FWIW –

    I don’t always read M2’s comments. When I do read his comments, sometimes I skip over parts of them.

    The fact that he posts here makes it more likely that I’ll read and participate in these threads – because there have been occasions where I’ve read discussions between him and others that I found informative and/or interesting. As just one example, sometimes I’ve seen him present an oft’ seen “skeptical” argument and watched it debunked in ways that make sense to me.

    One of the reasons why I like coming here is that I can read discussions between people to help me understand a variety of views.

    (Did that sound like something the Dos Equis’ most interesting man in the world might say?)

  36. John Mashey says:

    For the nth time:
    1) Good moderation takes time.
    2) Blogging software, as a whole, is weak in support for efficient moderation,
    3) For example, a common choice for a moderator is:mostly binary:
    Accept (possibly with edits, as often seen at Stoat or RealClimate)
    Reject
    4) Some have (more or less manually), added “move to Borehole, Stoat Burrow, Rabett Hole or equivalents, leaving behind just a note to that effect.
    5) Better would be:
    Accept(with/without edits)
    Reject
    HIDE, with reason code … which would show the author, date, and reason code, and either move the comment elsewhere, with link, or perhaps cleaner, hide the comment in place unless someone clicks, as is sometimes done by some commenting software (like in Amazon) where a comment gets too many downvotes.
    6) Even better would be coherent interaction with browsers, for example, a flag overriding HIDE cases and just seeing everything. Of course, some of us wistfully recall the old USENET days when KILLFILEs really worked well.

  37. Joshua says:

    Of course, I should acknowledge that simply because a certain practice makes it more likely that I’ll read and participate here is not a reason, necessarily, why that practice should be implemented. 🙂

  38. AnOilMan says:

    M2 turned me into a Newt. (It got better.)

    Steve captured my concerns….as did John Mashey’s Billy Madison reference.

    It’s not that I mind M2’s garbage so much as the sheer quantity. I don’t mind a dissenting opinion, and I have learned from reading those. (But not from M2.) I’d rather see less garbage, including my own snarky posts.

    Willard, I don’t think M2 is here to participate. I feel that he’s here to gum up the works. If the quality of his responses are anything to go by, then perhaps he needs a new playground.

    I feel that if M2 to or some other of his ilk would like to contribute, it would be nice if they actually looked up or referenced some on topic material. One particularly nasty individual shared a link questioning carbon trading, and I learned from it. But as we all know it’s pretty rare to get citations out of them, and even rarer that they understand their own citations.

  39. Tom Curtis says:

    Willard and Joshua, I cannot speak for casual readers but now if I see an M2 post after the first one or two on a thread, I don’t bother reading. I don’t bother reading the responses to M2 either. If most of the comments are by or in response to M2, I simply abandon reading the thread. I figure I may miss some informative commentary that way, but chances are I have come across the same ideas better expressed (on both sides of the argument) elsewhere.

    This is a marked contrast to how I treat comments by Richard Tol (for example), who when he is not being snarky typically presents his case well. I make it a point to read all his comments. It is not a agreement/disagreement thing but the ratio of dross to substance that interests me.

  40. Joshua says:

    Tom –

    Sometimes I have precisely the same response to M2 as you described (although I have a different response to reading Tol’s argument than you – as I almost always find his logic to be much worse than M2’s. Maybe it’s just that I haven’t seen his arguments where he was being non-snarky).

    I don’t doubt that you are basing your response, at least to a large degree, on what you feel is the content (dross/substance ratio). I have read you engage in good faith to substantive arguments from those who disagree with you – so I’m not saying that I think that your response is only because you differ in perspective from him.

    But there is, nonetheless, a subjective element to how characterizations such as “troll” are applied. The questions related to “substance” and “dross” are not easily delineated and inextricably linked to point of view/political outlook/personal identifications. So I think when you’re treading into that ambiguous territory, caution is warranted. If other people think that M2s comments are so substance free, if they think that there’s nothing he says that is worthy to respond to, then they will just not read his comments or maybe read some of them and not respond.

    My perspective is colored by the fact that so often I have the same logic you’re using applied against me – at places like Climate Etc. but here also (from Steve, for example). Now maybe my comments are substance free – but then the answer is why do people respond? Holding me accountable for the decisions they make seems to lack of accountability, IMO.

  41. Tom Curtis says:

    Joshua:

    ” …why do people respond?”

    I have very often responded to absolutely rubbish arguments/theories on Skeptical Science not because I think they have substance but because I do not presume that all readers will have either the knowledge to see that – or if they do, have the logical acumen to lay out why the argument is rubbish. This has gone to the extent of refuting an “argument” that global warming is caused by radar (I kid you not).

    I am prepared to do that on SkS because its role on the web is as a teaching site, and is used as such in several college courses. Therefore I assume rightly or wrongly that it will be read by interested but relatively uninformed readers.

    I see this site as quite different. I think it is read by interested and relatively well informed people who can detect dross for themselves. Therefore I ignore what I consider to be rubbish. I assume, however, that others are less likely to do so either out of habit picked up on other sites, or because they are particularly argumentative. Or who knows, they may even find something interesting among the dross I do not read anything into their responding because it could be for a variety of reasons.

    Allowing, however, it is out of interest. The fact remains that M2’s comments (regardless of other merits) are repetitive, extensive and repeated on thread after thread. No matter how interesting you find his theory that Libertarians alone are virtuous, rugged, individualists while all acceptors of AGW are people seeking a group identity because they are not up to Libertarian virtues, just how many threads does he need to argue that point? And if you are interested, how are you harmed by confining that discussion to just one or two threads?

  42. Tom,

    I only skip M2’s comments that respond one-liner for one-liner. This is a common practice. It always degenerates. I asked him to stop this when I moderated. He persists.

    That’s a general comment. It applies to every Climateball players. For more example like this, see Lucia’s, who is now writing a novel. (Good for you, genial one!)

    It leads to an exponential discussion.

    ***

    When M2 concentrates on one topic, I think he fares quite well. I disagree with many thongs he say, but he frames them that provides a challenge. They’re oftentimes verbose and too general. I skip these parts. (Just like casual readers do – there are studies on that.) I’m sure many do that with my comments too.

    I think the frustration comes from the difficulty of replying to his narrative. He’s a worthy opponent. He’s not here to bring peace. Nobody is.

    The problem with John’s suggestion is that it’s not a working environment, and nobody but AT rules. Also, there has not been tangible efforts to come up with the best arguments contrarians. SkS falls short on this. The Contrarian Matrix tries to fill that gap.

  43. Steve Bloom says:

    Tol’s snarky comments generally have the advantage of being very short.

    Joshua, I did at one point flame you for lengthy and (IMO) uninformed bloviation, but IIRC I didn’t suggest you be placed on moderation for it, nor did you continually re-post the same material across multiple threads. Those are major differences.

    I love you too, Willard.

  44. Steve Bloom says:

    I think the frustration comes from the difficulty of replying to his narrative. He’s a worthy opponent.

    Now this makes me curious. In what way?

  45. And of course SteveB omits the sentence that precede the two he quotes.

    In a nutshell, M2 speaks from his heart, offers a personal twist to the usual memes, concedes points from time to time without bending over, dodges when needed, sticks to his game plan, pushes all the right buttons, knows when to start a food fight, expresses incredulity like few do, accepts moderation gracefully, etc.

    What I write under my name is my honor. I feel M2 abides by the same principle. I trust M2 not to do anything whatsoever to get his way.

    I wondered for a while what would be the equivalent of a Moshpit for the establishment. I wonder no more. I’d field a whole team of M2s before picking any SteveBs.

  46. anoilman says:

    Willard, There comes a point in your life when you have to grow up and stop going to Moshpits.

    I’m not sure anyone wants this to be a moshpit. Anders? Is this what you want? I’m pretty sure no one wants to hang out at one. (I personally came here to avoid that.)

    Furthermore… there are others here with dissenting opinions that don’t fill the threads with ignorance and hear say. Or at least not nearly as much as what drivels from M2’s keyboard.

    If I could get a compromise, I’d prefer to see fewer replies from M2. I’d prefer to see something intelligently researched however biased it may be. I feel that might actually be enlightening to see for a change.

    As it is now, we could easily replace M2 with a very small Python Script.

  47. Vinny Burgoo says:

    Willard: ‘I wondered for a while what would be the equivalent of a Moshpit for the establishment.’

    I’ve always found it hard to remember which was which. Sorry, Steve B. It’s the shared terseness and aggression. Mosher tends not to do complete sentences, but that’s no help when you are trying to remember which Steve is in which tribe.

  48. I don’t think this site suffers badly from excessive spamming or from too many stupid comments. Those may take over during periods short of more substantive discussion, but disappear in the noise when more valuable discussion gets in speed.

  49. Michael 2 says:

    AnOilMan says “I personally came here to avoid that”

    Yes. Often enough I see useful and interesting math formulas and sometimes even understand it. The examples that I do not understand I usually save for future study. There’ll be a lot of time for that once my last teenager has left the nest.

    “I’d prefer to see something intelligently researched however biased it may be.”

    Yep, same with me. I don’t expect to see it in the *comments*; rather, the comments are where I gain a sense of what is considered interesting. I try to understand the thinking/feeling of those who fear global warming. I have a sense that I’m missing something, or might be.

    “As it is now, we could easily replace M2 with a very small Python Script.”

    There’s that “we” again. Do you not have sufficient skill to do this by yourself? Creating a functional Turing machine would be quite an achievement especially if it can replicate me. But then, perhaps I *am* a Python script. Be impressed!

    While I am not a logician, I have a sense for logical flaws (or simply untruths) in an argument such as the one you have just made. It helps me solve computer problems but in Borgish society it does not serve me that well.

    Each blog owner decides whether the blog is going to be filled with adoring sycophants agreeing with every word, merely trying to find more clever ways of agreeing (DailyKOS being an example), or a battleground of ideas where only the fittest survive.

    So far it appears ATTP is a (civilized) battleground of ideas where only the fittest will survive. Ideas must be challenged. Claims, such as yours replicating a human being with a Python script, ought to be challenged — if you can do that it would be an astonishing feat worthy of great praise, but if you cannot, then how scientific and/or truthful shall I consider your words?

  50. Michael 2 says:

    For ATTP: I am a libertarian (small L). That means I respect your “property”, this blog. If the owner of it wishes me to leave then that is what will happen, without rancor or complaining. It is a privilege to be here.

  51. M2,
    I appreciate your last comment and is one reason why I find myself somewhat uncertain about this situation. I have no reason to think that you aren’t well-meaning and you’ve never complained about the moderation. You also seem willing to take criticism even if it isn’t particularly complementary. On the other hand, I do understand why some people get a little frustrated. As far as I’m concerned, you’re welcome to continue commenting as long as you don’t object to whatever moderation may be applied. Technically, that applies to everyone.

    To everyone else, I thought it best to post M2’s two comments since he has been discussed in quite some detail. However I don’t see any value in really continuing this, so maybe we can draw this to a close. I realise there are some differing views, but all I can say is that I’ll do my best to moderate as appropriately as possible.

  52. Rob Nicholls says:

    An Oil man: “There’s nothing new about all that.” You may well be right, and I guess I’m the only person who is surprised but I had been making the assumption that Ross McKittrick knows a lot about statistics even if his views on climate science seem to me to be contradicated by the evidence (But what would I know?) To get a null hypothesis the wrong way round in a hypothesis test seems such an elementary error, it really surprised me, and it was one of my favourite moments from nearly 5 years of watching climateball…I guess I need to get out more.

    (Please correct me if I’m wrong about it being an elementary error). It’s quite possibly a lack of familiarity with Ross McKittrick’s output on my part that causes me to be surprised although I’ve read a few of his papers, mainly because I’m fascinated to see complex statistical techniques being used to support the suggestion that IPCC WG1 has got things very wrong. I find Scafetta’s work enjoyable for the same reason. (I’ve been struggling to understand PCAs for a few years now, so I’ve not really been able to form an opinion of McKittrick’s statistical expertise based on his writing about PCAs in the MBH 98 Hockey Stick, although the fact that many more recent papers have found similar hockey sticks makes it all academic in my opinion).

  53. Rob,

    To get a null hypothesis the wrong way round in a hypothesis test seems such an elementary error, it really surprised me, and it was one of my favourite moments from nearly 5 years of watching climateball…I guess I need to get out more.

    I think it’s somewhat subtler than that. Let’s consider some time interval in which we want to ascertain if it’s been warming or not. So, we determine a trend and a 95% confidence interval and we discover that the trend plus 95% confidence interval includes no trend (i.e., no warming). Does that mean there’s been no warming? No. One thing it might mean is that we can’t rule out that there’s been no warming, but that doesn’t mean that there’s been no warming. What is our best estimate for the warming? Well, it would be the mean trend. Also, the 95% confidence interval allows us to actually quantify the likelihood of the trend being positive or negative. If the trend is 0.1 degrees per decade +- 0.2 degrees per decade (2 sigma) then that means there is a 83% chance that we’ve been warming (i.e., 1 sigma is 2/3 and 2 sigma is 95%). So, there is a chance that we’ve been cooling, but it’s much more likely that we’ve been warming. So, doing a null hypothesis test as Ross McKitrick has done doesn’t mean that the data is trendless (as he seems to be suggesting), it simply means that we cannot – with confidence – reject that it’s been trendless.

    There’s also another way to look at this. Imagine Ross McKitrick redid his analysis but using a null hypothesis that it’s been warming at 0.2 degrees per decade. For the surface temperature datasets, a period where warming of 0.2 degrees per decade is consistent with the data extends back to the early 1970s, and for the satellite record there is no period where we could reject that null. That doesn’t mean that it has been warming at that rate, simply that we can’t – in a statistical sense – reject that it has.

    There’s also an extra issue – I think. The variability in the data is real, rather than being measurement error, or some kind of random noise. It’s because the surface temperature really does vary. Therefore, one could argue that the most likely trend is the mean trend, because the uncertainty interval reflects real variability, rather than an uncertainty in the actual trend.

  54. M2 writes: “While I am not a logician, I have a sense for logical flaws (or simply untruths) in an argument ….” and also writes, ” I am a libertarian (small L)…”

    Yet, when told that the US gov’t can’t go bankrupt because it controls its own currency and can just print more money he immediately played the Weimar card – which has been shown to be completely irrelevant by the past 7 years when US fiscal and monetary policies *should have* -by libertarian thought – resulted in hyperinflation, a debased currency, and skyrocketing interest rates.

    So M2’s assertion that he can spot logical flaws seems flawed. Indeed we see that he cannot adjust his viewpoint even when reality steps up and slaps him in the face. It’s little different from the various pseuodskeptic theories that do not take OHC into account or paleo-climate. They’re one-trick ponies that can’t handle multiple lines of evidence. Considering that the basics of an economy at the ZLB have been understood since Keynes more than 80 years ago we can probably look forward to M2 spoutingg the same climate and economic nonsense for decades to come – regardless what facts show him to be wrong.

  55. Michael 2 says:

    ATTP says “it simply means that we cannot – with confidence – reject that it’s been trendless.”

    Agreed. I do not know how long one must measure in order to decide there’s a trend rather than some really long period phenomenon mimicking a trend (or masking it). Reading these comments suggests 30 years is a minimum time of observation to decide that the trend in climate change has itself changed (at least a full cycle of AMO and PDO). The Nyquist theorem seems relevant but that’s for periodic functions which don’t exactly seem to pertain. Still, I accept it as a guide to estimating claims of trend change.

    Having said all that, I do find it useful as a type of “speed brake” on poorly considered government programs and get-rich schemes. I am also cautious about the quality of the data even when I do not (and cannot) dispute the mathematics involved in processing the data.

  56. Michael 2 says:

    AnOilMan wrote “it would be nice if they actually looked up or referenced some on topic material”

    ATTP is a physicist. Almost everyone here knows more than I on mathematics and physics. I cannot imagine contributing something on those topics that you do not already know vastly more than I. I have saved many ATTP post and comments, as well as that of many other readers here.

    It is a FLOOD of information and when things settle down a bit (real soon now!) I’ll take some time to delve into it, which is also the case with Willis Eschenbach. Incredibly, I actually knew something on the topic he was studying since computers are my specialty. Out of that exchange I finally learned about wavelets and their utility in studying climate change; or more precisely, their utility in revealing why Fourier Transform isn’t very useful.

    These comments sometimes provide information directly but more often simply provoke an interest whereupon I go study something. That in turn reveals a lack of knowledge which must be “back-filled” before I can proceed. Meanwhile y’all are discussing libertarians in strange ways which strikes me as I need to at least suggest a correction on some of it, maybe taking a more nuanced approach that recognizes many more kinds of people than just “us” versus “them”.

  57. Michael 2 says:

    Joshua: “seen skeptical argument and watched it debunked in ways that make sense to me.”

    Yes! Debunk with intelligence and usually a bit more fact and citation, more weight than just “I’m right and you’re wrong.”

  58. Michael 2 says:

    ATTP says “If you assume the land/atmosphere has a mass of around 1019kg and a specific heat capacity of 1000 J kg-1 K-1, then that suggests you need 1022J to increase the temperature by 1 degree.”

    Thank you, sir. THIS is why I come here. Science and math in bite-sized pieces each individually verifiable without much difficulty. The aggregate then is also verified.

    I love math but my computer work almost never needs it. It’s been more than ten years since my last calculus class (which I took three times in a row to “get it”) so I appreciate having something that stretches my mind but doesn’t completely wash over me like a tsunami. I really am a person with one foot in “John Q Public” and the other in math and science, not expert at either.

  59. Steve Bloom says:

    In a nutshell, M2 speaks from his heart, offers a personal twist to the usual memes, concedes points from time to time without bending over, dodges when needed, sticks to his game plan, pushes all the right buttons, knows when to start a food fight, expresses incredulity like few do, accepts moderation gracefully, etc.

    What I write under my name is my honor. I feel M2 abides by the same principle. I trust M2 not to do anything whatsoever to get his way.

    All style, no content. Yep.

  60. Michael 2 says:

    Rachel M “I’ve already deleted three of M2’s comments today :)”

    No problem. Occasionally I wish I hadn’t sent a message but no recall (or edit) exists. It can be very difficult to resist a taunt especially when its a straw-man, it isn’t even my claim that is being ridiculed.

    The person for whom I intend the response will get an auto-emailed copy so my purposes, answering a taunt or providing information, is achieved either way and you’ll have a cleaner blog.

    I will take into serious consideration all of the critique and try to return to topical responses only.

  61. Rachel M says:

    No problem. Occasionally I wish I hadn’t sent a message …

    Oh thanks, M2. I could kiss you. But only because I’ve had one beer and I’m now pissed. I’ll probably regret saying this too 🙂

  62. Michael 2 says:

    Tom Curtis “In this particular instance his comments are not even on topic, and the second does not respond to anything in the prior comments.”

    Sorry. I have been reading the emailed notifications most-recent-first and that means I haven’t even SEEN your prior comments or questions that very likely “moot” the comment I just made.

    I’ve just now gotten to your message. This is a bad way to do it (reverse order).

    I wish I hadn’t sent that long post to Steve Bloom. It was arrogant of me.

    I wish for a recall/edit facility. The next best thing is a wise moderator. Thanks for deleting some and I regret the burden I add to you.

  63. Michael2 wrote “Agreed. I do not know how long one must measure in order to decide there’s a trend rather than some really long period phenomenon mimicking a trend (or masking it).”

    The way to do this is to perform an analysis of the statistical power of the test (statistical power is the probability of rejecting the null hypothesis when it is false, which is where the alternate hypothesis should enter the test). ISTR a paper by Santer et. al. performed such an analysis and found that for OLS trends you need *at least* 17 years, and this will be much longer if you take autocorrelation into account, which broadens the error bars considerably. The power of the test also assumes that you are looking at a randomly selected period from the time series, the power will be much lower if you cherry pick to find the longest trend, or if you only choose to perform tests *after* observing an apparent “hiatus” (as the design of the test is influenced by what you have seen in the period under test).

    There are two ways that you can use statistcal hypothesis tests to enforce a degree of self skepticism, either you can take the null hypothesis to be the thing you are arguing against, or you can argue in favour of the null hypothesis and perform a test of statistical power (as in that case it is the statistical power that provides the hurdle, not the test itself). In this case, the onus was on MacKitrick to perform the test of statistical power, which he didn’t do.

  64. A better way of understanding whether it is a real trend or just an artefact of the noise is to use physics rather than statistics. It is not the case that we know nothing about the physics of climate change, or of sources of internal variability (the noise), but that is what a simple hypothesis test for the tren assumes!

  65. dikran,

    use physics rather than statistics.

    Indeed. Statistics is clearly very useful, but applying statistical tests without considering our understanding of the physical processes, is typically rather meaningless.

    I notice that Ross has not returned to Richard’s blog. Would be interesting to see how he addresses the various points that have been made.

  66. Rob Nicholls says:

    ATTP, thanks for responding in detail to my comment; very helpful and much appreciated.

  67. M2 writes: “The Nyquist theorem seems relevant but that’s for periodic functions which don’t exactly seem to pertain…..

    Sampling rate is not the issue here. Besides, the Nyquist-Shannon sampling theorem was developed for digital signal processing and most climate data is analyzed using time series analysis.

    For some reason there’s a large contingent of pseuodskeptics with Electrical Engineering or electronics backgrounds. As a result they often make the mistake believing that a signal cannot be reconstructed unless the sample rate satisfies the Nyquist-Shannon sampling theorem. They know – just absolutely *know* – that this is true. And as Mark Twain famously said, “It ain’t what you don’t know that gets you into trouble. It’s what you know for sure that just ain’t so.”

    Tamino has dealt with the fallacy of the Nyquist frequency as a universal sampling limit in several posts:
    Sampling Rate
    Sampling Rate, Part 2
    Through A Picket Fence

    To be clear, there is nothing wrong with the Nyquist-Shannon sampling theorem. What happens is that many with an electrical engineering or computer science background who think they understand it actually don’t – at least not fully – and as a result usually misapply it to climate data.

  68. Rob Nicholls says:

    Dikran Marsupial – “There are two ways that you can use statistcal hypothesis tests to enforce a degree of self skepticism, either you can take the null hypothesis to be the thing you are arguing against, or you can argue in favour of the null hypothesis and perform a test of statistical power (as in that case it is the statistical power that provides the hurdle, not the test itself). In this case, the onus was on MacKitrick to perform the test of statistical power, which he didn’t do.”

    Wouldn’t that be a fairly basic flaw in a paper? (Although to be honest I’ll have to defer to others on the lack of a test of statistical power; although I can’t find an obvious statistical power test in McKitrick’s paper, I couldn’t follow a lot of the maths, as usual.)

  69. Rob Nicholls, yes, the reviewers ought to have spotted that it is not meaningful to describe such periods as “trendless” on the basis of a failure to reject the null hypothesis, it is a common error, but one that should have been picked up in the review of a paper in a statisics journal (IMHO).

  70. “For some reason there’s a large contingent of pseuodskeptics with Electrical Engineering or electronics backgrounds. As a result they often make the mistake believing that a signal cannot be reconstructed unless the sample rate satisfies the Nyquist-Shannon sampling theorem. “

    Thank you Kevin, You are one of the people that gave me some good guidance on this issue in the past.

    Consider a phenomena such as ENSO, where we have data for every month over a number of years. In this case, we will see the seasonal effects come through in the data, which will be manifested as a peak or valley during a favored time of the year. However, if we have data that is only taken once per year, or that is averaged over one year, that yearly periodicity becomes very difficult to reveal.

    I think that is the essence of the Nyquist theorem, in that you need at least two data points every year to pick up unambiguously the seasonal period. I think this is Tamino’s picket fence effect, where if you blink at a periodic rate when traveling along a fence, one only sees the white and not the spaces between the pickets.

    What is both intriguing and frustrating to those of us that have been indoctrinated to the Nyquist rule is when we find an analysis tool that instructs us that there are sinusoidal components at frequencies much higher than the sampling rate that the data is delivered with. Apparently this is also observed during variable star analysis.

    It is definitely a topic worth pursuing because one still has to be able to explain what is happening. For example, I am looking at yearly ENSO proxy data that goes back 100’s of years and the machine learning tool I am using (Eureqa) is picking up frequencies that are closer to that of lunar tides, which is less than a month period.
    http://contextearth.com/2014/08/26/soim-fit-to-unified-enso-proxy/#comment-67537

    I have no control over what the Eureqa tool is doing yet understand that an aliased period will give the same result, and so have to scratch my head as to how the discrimination actually occurs.

    BTW, I agree the # of EE’s is problematic and they seem very belligerent and not open to new ideas. I is one and work to correct this 🙂

  71. John Mashey says:

    EEs and computer scientists: can anyone point to credible, peer-reviewed papers by good social scientists (say like Ed Maibach) who’ve done well-constructed surveys with proper demographics, to calibrate the % of dismissives according to different disciplines? And if so, can they point at data that shows EEs rank especially high on this, compared to the general population or other engineers or other technical disciplines? How do they compare with economists? With petroleum engineers? (“Many” is not very interesting, since there are large numbers of EEs and CMPSCs around, and unsurprisingly, they tend to be computer-literate, so may well appear more often online.)

    For some history, see discussion at RC.

    *I* don’t have any real population data, although of course living in Silicon Valley for 30 years, I know and have worked with many EEs/CMPSCs .. .but then, maybe SV is different, Long ago, I met someone with a BS EE who certainly is no dismissive, fellow named Michael Bloomberg. Well, OK, he has been doing other things than EE for a while.

    Anecdotes are not very useful, because this is as silly as “we had a cold winter in our state, so AGW is over.”

  72. Michael 2 says:

    Kevin — I read the links you provided, Tamino’s explanation of the Nyquist Theorem, aliasing and the picket fence. The first page is stuff well known to me but I’ll admit to being surprised and intrigued at the second page, in effect tricking the Fourier transform because of inadequate sampling.

    “But you can’t reconstruct the signal exactly — not even assuming it’s band-limited — unless you have an infinite amount of data.”

    Agreed. The start and stop of sampling introduces high-frequency artifacts and produces spectrum spread.

    DFT and FFT are (imo) mainly useful in climate science to show what it is NOT (ie, completely and predictably periodic).

  73. Pingback: Fraudulent? | …and Then There's Physics

  74. Pingback: Matt Ridley responds to Tim Palmer | …and Then There's Physics

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s