Mathturbation

There seems to be a rash of trying to explain global warming by theories that either ignore, or flatly contradict, the science called “physics.”


Often they involve some mysterious “cycle” (usually a 60-odd year cycle) which is claimed to be “obvious” but doesn’t stand up to analytical scrutiny. This is suggested as a cause of the global warming that’s been observed over the last century or so — or at least, the cause of so much of it that it enables one to minimize the warming due to man-made greenhouse gases.

They may even attempt to suggest a physical relationship to some other so-called “cycle,” often the AMO or the PDO, which also fails the rigorous test for genuine periodicity. These models are built on the correlation-and-to-hell-with-causation principle. You might as well ascribe global warming to Leprechauns. The “physics” behind such a relationship is tenuous at best. Let’s face it people, temperature increase is heat, heat is energy, but there’s paltry reasoning about where the energy is coming from, and why it has decided to surface now.

Then there are the “natural variation” theories — those which imply that the globe’s warming is just “business as usual.” You have to ignore a helluva lot to believe this. Like the rapid, dramatic, and unprecedented decline of artic sea ice, the evidence from paleoclimate studies, the dwindling of most of the world’s glaciers, the breakup of ice shelves that existed for thousand of years, rise of sea level coincident with the industrial revolution, change in the timing of cherry blossoms in Japan, etc. etc. etc. etc. etc. You have to wear those blinders very tight, not to see that modern climate change is far from “business as usual.” And perhaps the most idiotic of the “natural variation” variants is the idea that global temperature is just a random walk. Such ideas are often mathematically subtle and fascinating, but seem to fail rigorous statistical testing, while suggesting them as the real explanation of global warming is an insult to physics.

In fact it’s astounding how much known physics has to be ignored. Carbon dioxide is a greenhouse gas, that’s no longer a matter of opinion. Sulphate aerosols cool the planet, we’ve seen it happen (from both factory emissions and volcanic eruptions), and we can model it with impressive accuracy. Solar changes have had a modest impact, and no recent trend, because solar changes have been modest, with no recent trend. And yes, Virginia, feedbacks really exist, like water-vapor feedback which is every bit as undeniable as CO2, while a notable increase in water vapor content has been observed.

I suggest there are two motives for the proliferation of such “theories.” One is the obvious: those who have an ideological objection to holding business responsible for its actions, will bend over backwards to convince themselves global warming is not real, or it’s not caused by man, or it’s not dangerous. The other is the natural tendency for smart people to play with numbers and ideas, find interesting relationships, then persuade themselves they’ve happened on a key insight. This often requires ignoring the existing literature of climate science — otherwise you’d know how much is already known, and how little (or none) your new “pet” theory adds to understanding.

Some (not tamino to be sure) would say this is how “crackpot” ideas are born. I have, however, decided to give a name to such theories: mathturbation.

About these ads

86 responses to “Mathturbation

  1. Charming title. As though we didn’t have enough to deal with already reading through the tripe they produce you’ve now left us with an extra residual queasiness.

    Vaguely connected, do you know of a good single source summing up the state of knowledge about the attribution of the global temperature record during the first 70 years of the 20th century? I’ve seen lots of posts and articles that touch on different parts of it, but can’t find anywhere that it’s all summed up and the relative contributions are assessed in context. If not, would you consider doing this?

    [Response: Well, that's the kind of thing that the IPCC reports are about. I don't know of a short article or blog post which condenses it. Anybody?]

  2. Very apropos, since “turbation” would derive “from the Latin turbatio, turbulence.” For some reason, it’s a more common term in German, referring to “mixing processes of soil horizons or sediment.” “Turbid,” of course, is a good English word, which when used of liquids means “clouded as with sediment; “a cloudy liquid”; “muddy coffee”; “murky waters”.”

    So “mathturbation” would denote using math to muddy the waters.

  3. The motive? Delay.

    I also suspect that the very wealthiest of the public deniers are not deniers in private. Just like Enron believed that they were the smartest people in the room and would thrive in the chaos, the wealthiest believe that they’re the smartest people in the room and will similarly thrive. Particularly if they’ve already gamed the political system.

    • Not just the wealthiest. This seems to be a libertarian trait, particular among those libertarians who are techie types who grew up on a certain style of, say, science fiction ala Heinlein, in which the protagonist always comes out on top of the heap in an anarchist/libertarian wet dream world.

      I loved such stories as a kid. As I got older and read history (in the American, informal sense, not the British university sense) I understood that in an disordered world, smart people tend to be the first to be lined up against the wall…

  4. Yes, temperature increase is heat, heat is energy. And almost all of that heat is in the oceans, not the atmosphere. Need more focus on oceans.

    • And the temperature of the oceans is trending upward. The question is how deep the heat penetrates on a timescale of 30-100 years. I believe this is more likely to increase sensitivity estimates than decrease them.

  5. I detect a theme in your post titles.

  6. Tamino – I speak as an admirer of this blog. You seem to have a chip on your shoulder about natural variability. While some may be attempting to discount AGW due to periodicity and natural variability (I have met them, be they leprechauns or not), that doesn’t – by itself – discount the observation that surface temperatures (regional or global) (driven by variable energy inputs and feedbacks) seem to demonstrate periodicity over selected time scales.

    Now you have stated several times that these apparent cycles don’t hold up under rigorous analysis. I’m interested. But as they say in writing workshops and a current commercial: Show, Don’t Tell So if you find Tisdale, or Vauhgn, or even my own preliminary pokings naive and misdirected, I’d be happy to read your criticisms. But this post isn’t a critique – it is a (fairly mild and for blog standards polite) rant.

    Of course, ultimately, the energy which drives all this derives from the sun (with Earth based energy input probably negligible in the last several million years). Solar variability seems to be fairly small compared to climate variability. The question of how these small changes in input result in large changes in climate is interesting.

    Now as I (a naive amateur student) see it, there are three, not necessarily exclusive, paradigms to examine trends: First order + noise. Natural cycles/variability + Anthropogenic forcing/feedbacks + Volcanic forcing. Stochastic shifts. Should we just accept “CO2 forcing + Noise” as the sum total of what we can learn from Holocene climate proxies and instrumental records? I doubt it.

    Meanwhile … (these are both paywalled so I haven’t seen them yet) …
    Possible solar forcing of interannual and decadal stratospheric planetary wave variability in the northern hemisphere: An observational study

    On the observed relationship between the Pacific Decadal Oscillation and the Atlantic Multi-decadal Oscillation

    [Response: I've blogged about cycles before, here and especially here. It seems to me that those who claim the existence of cycles never produce any evidence beyond "seems" or "looks," but even when it's examined analytically it's all too easy to fall into traps -- demonstrating genuine periodicity is a hell of a lot harder than most people imagine. I say that not just from a firm theoretical foundation, but from decades of experience thinking I'd found periodic behavior in astrophysical systems, only to be proved wrong by subsequent observations.

    Your speculation that "small changes in input result in large changes in climate" assumes that the primary driver of modern climate changes is solar variability. I wonder whether you're making the mistake highlighted in this post, of ignoring the physics. Do you think greenhouse gases have no impact? Do you believe sulphate aerosols don't affect climate? Have you modeled the impact of these factors? And when they're accounted for, how much "natural variability" is left to explain by solar variations and/or "cycles."

    The fact is that climate models which do not include mystery cycles or exaggerated solar influence, do a superb job of explaining temperature changes. How can those other factors be so influential, when they have no physics I can see and to include them you'll have to explain why the known physics which succeeds in modeling climate, should be changed or abandoned? It ain't broke -- why are so many people trying to fix it? I have no chip on my shoulder regarding cycles or natural variability. I do, however, object to claims about them (and their dominant but of course hitherto-ignored influence) which don't come with evidence, and/or which fly in the face of the laws of physics.

    And I object to the all-too-common practice of elaborate curve-fitting with which you can model anything that exists in any way you want. It's so common for those who do to be convinced they've made a great discovery, but just like all those "cycles" I thought I'd found, they don't stand the test of time.

    If you want to speculate, good -- I encourage that. If you think an idea has real merit and importance, great -- publish it. Publication isn't just the way science is communicated, it's also a crucible forcing people to justify their ideas and subject them to outside scrutiny.

    Finally, I have to wonder about the merit of referencing two papers which you haven't yet seen.]

    • Hi Ron,

      I’ve enjoyed your work on your blog and even discovered a few references there in your posts that I hadn’t previous found, so thanks for your posting on your interesting efforts.

      Have you had a chance to read Burroughs, ‘Weather Cycles: Real or Imaginary?’ (Cambridge, 2nd edition)? There is a review of it here:

      http://iceman2.umd.edu/www/preprints/burroughs03_rev05.pdf

      The reviewer summarizes one of Burroughs’ arguments thusly: ‘Only a few cyclicities survive Burroughs’ analysis: the annual cycle, the quasibiennial oscillation, the El Nino-Southern Oscillation (ENSO), and perhaps a bidecadal cycle.’

      None of this is to say we can’t learn more about natural variability from proxies than we can glean from the short instrumental record (we can and we do) — what is interesting is that we don’t have to look too far back with proxy data to observe that many of these (AMO, PDO) don’t appear to be true oscillations — or even necessarily have preferred timescales of variability.

    • Ron Broberg: Preprint version of Wu & Liu’s “On the observed relationship between the Pacific Decadal Oscillation and the Atlantic Multi-decadal Oscillation”:
      http://ccr.aos.wisc.edu/publications/pdfs/CCR_998.pdf

  7. Ron,

    First off, I think natural variability is very important. The second paper you cited (with Shu Wu and Zhengyu Liu as the first two authors, whom I worked with over the summer in Wisconsin’s Center of Climatic Research) have a lot of good work on this. Wu is interested very much in the problem of decadal-scale prediction, to which understanding both the internal variability and the radiative forcing are important. It’s a shame REAL work on the AMO and PDO, etc are ignored in favor of WUWT-type drivel. I can assure you that those authors do not discount the importance of CO2 in driving modern global warming. During the summer, I also showed the very intuitive result that if you pick some point on the globe and draw a box around it, the relative magnitude of inter-annual variability vs. decadal variability vs. the “trend” changes as that box becomes bigger, with variability usually winning out initially and forced trends dominating once the box is quite big.

    The thing with CO2 is that it is growing and the forcing is persistent, so eventually the signal emerges from the background noise of natural variability. There’s not a single simulation in models that spontaneosuly generates global temperature changes as large as a doubling of CO2 under holocene-like conditions, so it’s very hard-pressed to just give “natural variability” the leeway to do whatever you want it to. If you’re also going to argue that it’s all natural and no anthropogenic, you need to convince people why CO2′s well-understood radiative effects are not working. That’s another matter all together.

    In any event, this is why we have attribution experts who look at the spatio-temporal signatures associated with climate change and with various possible causes under differing magnitudes.

  8. Thank you for your considerate reply.

    … but even when it’s examined analytically it’s all too easy to fall into traps

    A danger indeed. But some people have to go through the school of hard knocks.

    … assumes that the primary driver of modern climate changes is solar variability.

    My apologies, but I was thinking over Holocene, Quaternary, and Pliocene time scales (although the later is long enough to include geographic and geologic drivers).

    [Response: Thanks for the clarification. I agree that over the holocene, solar variation is probably a primary driver (although volcanism is also important). I would caution against reading too much solar variation into the Plio-Pleistocene climate variations unless you're referring to insolation variations due to orbital/axial (Milankovitch) changes rather than genuine solar flucutations. I have a "pet theory" that the sensitivity of Plio-Pleistocene climate to Milankovitch influences is actually an argument for stability of solar output itself, but this is just idle speculation on my part.]

    The fact is that climate models which do not include mystery cycles or exaggerated solar influence, do a superb job of explaining temperature changes.

    I don’t have enough experience or education with the models to evaluate that claim extended to Holocene time scales. But I have seen papers that claim that models with sufficiently detailed ocean slabs can produce patterns similar to AMO, PDO, and even ENSO.

    [Response: I suspect that patterns like AMO and PDO (and even ENSO) are not only genuine, they exhibit time scales similar to those usually given as "periods" of oscillation. I strongly doubt the periodic nature of those oscillations. And what I've seen of the primary literature (admittedly, not enough) usually implies "time scales" rather than "periods."

    I don't know to what extent GCMs have been applied to the holocene rather than just the last century or so. I would guess (just a guess) not much because the computational workload may be a limiting factor. I am aware that at least a few non-GCM models have success post-dicting millenial-scale climate variations (I think Crowley did one such study).]

    It’s so common for those who do to be convinced they’ve made a great discovery,

    I’m not here to overturn science; my journey of discovery is personal.

    [Response: I didn't have you in mind with that remark.]

    Finally, I have to wonder about the merit of referencing two papers which you haven’t yet seen.

    Two purposes. One point rhetorical: To show that these discussions are not confined to skeptical bloggers. The other in hopes that you would be able to access them and maybe even blog about them.

    A final meta-point. Watts and D'Aleo raised charges of fraud against NOAA in relation to the surface temperature records. New to the blogging, Hausfather, Stokes and I came out to criticize their claims analytically. As did you and Barnes/Jones (ccc). On the 'skeptic' side, so did Mosher, Id, and RomanM. Stubbornly neutral was Chad. And others. While problems exist (eg station location data), neither the data (GHCN-M) nor the methods (GISTEMP, CRUTEM) are fraudulent. We managed to have Watts rewrite his 'white paper.' I consider the last year to be my freshman year. I'm pretty sure that 'natural variability' is on the curriculum for the sophomore year (re Spencer, Curry).

    [Response 2: Just to be clear: I think that your approach (try a lot of ideas, see what happens, but don't read to much into the results without solid foundation and much review) is the *right* one. The *target* of this post is those who read too much into their results while ignoring what's already been learned -- especially those who follow indulgent curve-fitting with extrapolation that almost inevitably concludes we're headed for several decades of imminent global cooling.]

    • Ron,
      I would say that you are doing some interesting work, but I think it is important to understand that in a system with many complex interactions, you can get some very odd and interesting time dependence.

      It is important to understand not just the sorts of patterns you can get, but also the sorts of forcings and influences that can drive them.

      One of the simplest is a linear trend (a stable trend with zero slope would be a subset of this). For all practical purposes, you simply will not get a linear trend unless you have a linearly increasing forcing. We can also ask whether a trend is linear or quadratic or exponential. It will take a whole lot more than 30 years (probably more like 60) to discern this.

      The next simplest sort of change is one that is truly periodic. To get such a trend, you absolutely need a periodic forcing–e.g. the diurnal or annual variation in insolation.

      In addition to periodic variations, there are many other types of variations that might appear cyclic. In most cases these are systems in which you have a sort of bistability. The system is stable in a two given configurations, but not when perturbed from those configurations. The cycles evident in such systems tend to occur because the probability distributions of perturbation size tend to be distributed such that on average it takes a certain amount of time to get a large enough perturbation to get the system to flip. The best example of such a system I know of is the heliodynamo–which is behind the 22-year solar cycle (that is, 2 11-year cycles). I suspect that ENSO, AMO, PDO and most other climatic oscillations fall into this class. These sorts of systems can take many cycles before you really understand their dynamics–and really, unless you have some idea of what drives them, you might never really understand thier true time dependence or even whether they exist.

      This is why I tend to be a serious skeptic of the sorts of “Fun with Fourier” analyses we tend to see applied by amateurs. Linear trends are pretty easy–given enough data. Periodic trends can be trusted after enough complete cycles to see if they really are periodic, and once the source of the periodicity is identified. Again, relatively easy if you have enough data.

      The Cyclic and Quasi-cyclic systems are REALLY tough. So, it’s not really a travesty (despite what Trenbreth says) that we don’t understand everything about natural variability. It’s to be expected, because natural variability is really tough to understand. However, natural variability tends not to kill you. What this implies to me is that most significant variability occurs on relatively short time scales (~decadal) and that longer “cycles” tend not to drastically affect climate.

      [Response: There are some systems that are strongly periodic even without periodic forcing, because the system has periodic modes which are easily excited and can tend to be self-sustaining. Pulsating variable stars come to mind. They do, however, require energy flow to maintain the pulsations.]

      • Tamino,
        Yes, that is true. I understand that these essentially work by absorbing the Fourier component at their resonant frequency out of the forcing, which is nonzero, even for white noise (though that might be a gross over simplification). However, I don’t know of any such influences in the climate–and in order to avoid conjuring Scafetta and West, I didn’t want to mention them.

    • Gavin's Pussycat

      > We managed to have Watts rewrite his ‘white paper.’

      I’ve heard rumors to that effect. How did it happen? Was the original version retracted in any high-profile (or any-profile) way (and especially, were those slandered by the original, in any way apologized to)? How can your (“We”) claim to have been instrumental in this, be substantiated? Were you acknowledged for it?

      So many questions…

  9. Some (not mashey to be sure) might think you’ve been picking up Rabett-isms :-)

    I object to the leprechaun theory, which is clearly insufficient.
    That is:
    a) One needs both leprechauns, which provide a non-physical connection that changes the Earth’s energy content, somehow.
    b) But one also needs “gremlins” that nullify greenhouse gas effects, quantum mechanics, laws of thermodynamics, etc.

    But the mathturbation term is great. In a recent look at the newly-famed “tallbloke” (Roger Tattersall) website, I found a cornucopia of cycles and anti-physics, of which a small sample is:

    http://tallbloke.wordpress.com/2011/02/21/tallbloke-and-tim-channon-a-cycles-analysis-approach-to-predicting-solar-activity/ http://tallbloke.wordpress.com/2011/02/20/miles-mathis-democracy-in-science/
    http://tallbloke.wordpress.com/2011/02/17/roger-andrews-the-solar-sst-relationship-part-ii/
    http://tallbloke.wordpress.com/2011/02/13/roger-andrews-the-solar-sst-relationship/
    http://tallbloke.wordpress.com/2011/02/12/sunspot-asymmetry-barycentre-to-solar-centre-link/
    http://tallbloke.wordpress.com/2011/02/02/the-sun-climate-link-brief-synopsis/
    http://tallbloke.wordpress.com/2011/01/18/millennial-cycle-in-solar-system-dynamics/
    http://tallbloke.wordpress.com/2011/01/14/gerry-pease-barycentric-orbital-periods/
    http://tallbloke.wordpress.com/2011/01/13/miles-mathis-un-unified-field-new-book/
    http://tallbloke.wordpress.com/2011/01/10/giorgieva-et-al-planetary-tidal-effects-on-solar-activity/
    http://tallbloke.wordpress.com/2011/01/09/wolff-and-patrone-a-new-way-that-planets-can-affect-the-sun/
    http://tallbloke.wordpress.com/2011/01/06/a-test-for-planetary-solar-connection-jupiter-perihelion/
    http://tallbloke.wordpress.com/2011/01/02/galactic-motion-and-solar-barycentric-motion-and-lod-a-puzzle/
    http://tallbloke.wordpress.com/2010/12/26/vukcevic-sets-a-christmas-puzzle/
    http://tallbloke.wordpress.com/2010/11/30/miso-rise-in-co2-due-to-plants-not-coal/
    http://tallbloke.wordpress.com/2010/11/25/more-support-for-svensmark-effect/
    http://tallbloke.wordpress.com/2010/11/25/more-support-for-svensmark-effect/
    http://tallbloke.wordpress.com/2010/10/04/length-of-day-solar-activity-link-confirmed/
    http://tallbloke.wordpress.com/2010/09/22/john-dodds-planetary-gravitational-explanation-for-cyclic-warming-on-earth/
    http://tallbloke.wordpress.com/2010/09/15/headbutting-the-rebutters-gerlich-and-tscheuschner/
    http://tallbloke.wordpress.com/2010/08/13/roy-martin-new-planetary-solar-cyclicity-hypothesis/
    http://tallbloke.wordpress.com/2010/08/11/roy-martin-venus-earth-jupiter-solar-cycle-analysis/
    http://tallbloke.wordpress.com/2010/08/08/interesting-correlation-sunspots-vs-specific-humidity/
    http://tallbloke.wordpress.com/2010/08/05/thunderbolts-opens-the-door-to-the-electric-universe/
    http://tallbloke.wordpress.com/2010/08/05/volcanos-dont-cause-global-cooling/
    http://tallbloke.wordpress.com/2010/07/30/barycentric-motion-spacequakes-and-questions/
    http://tallbloke.wordpress.com/2010/07/28/gray-stevens-planetary-effects-on-solar-activity/
    http://tallbloke.wordpress.com/2010/07/27/nasif-nahle-nails-the-radiative-physics-of-co2/
    http://tallbloke.wordpress.com/2010/07/21/nailing-the-solar-activity-global-temperature-divergence-lie/
    http://tallbloke.wordpress.com/2010/07/15/solar-system-simplified/
    http://tallbloke.wordpress.com/2010/07/15/understanding-the-solar-system-as-a-true-system/
    http://tallbloke.wordpress.com/2010/06/21/willie-soon-brings-sunshine-to-the-debate-on-solar-climate-link/
    http://tallbloke.wordpress.com/2010/06/14/venus-earth-jupiter-alignment-and-the-solar-cycle/
    http://tallbloke.wordpress.com/2010/06/04/new-paper-from-nicola-scafetta/

    • John:

      As the author of one of the Tallbloke threads on your list, specifically
      http://tallbloke.wordpress.com/2011/02/13/roger-andrews-the-solar-sst-relationship/, might I respectfully suggest you read it? You will find that my admittedly crude conclusions support not only a solar influence on climate but also an anthropogenic impact that’s much in line with the IPCC’s estimate.

      And if after having read it you still have a problem with what I said, please tell me about it. I’m not an anti-physicist or a leprechaun. I’m just another guy trying to unravel the mysteries of climate change, and I’m always open to new ideas.

      [Response: OK I looked at your post. This here post is about you.

      Anyone wishing to argue about it (including you) can do so on tallbloke's blog. Not here.]

      • 1- I find it interesting that prominent skeptics are ready to accept that sea-surface temperatures have been (and still are) under-estimated by .3C after 1945. Especially when I consider the amount of vitriol poured on scientists when automatic procedures adjust land temperatures by much smaller values.

        2- I think you may have missed some passages from Thompson et al., which suggest that the problem is confined to the mid-20th century, and that data “after the mid-1960s are not expected to
        require further corrections” for this particular bias (but may be affected by a smaller, different bias). Maybe I just misunderstood something though.

  10. The only “obvious” fact is that if the current warming were due to some ocean cycle, then the oceans would have to be cooling as the air warms. Ocean cycles don’t create heat, they just move it around. But since we know the oceans are warming too, it can’t be an ocean cycle that’s causing the current warmth.

    First law of thermodynamics. Indeed, the first law of all of science.

    • kap55,

      To be fair, I think the more intelligible claims of the possible climatic importance of such cycles suggest that they somehow alter cloudiness and hence the earth’s albedo, thus affecting the radiative balance. Haven’t seen much evidence presented for this but I don’t think the concept can be dismissed out-of-hand via the First Law.

  11. Perhaps unsurprisingly and the opposite of the above, it is becoming the latest meme that in comment-thread useless argument, citation of papers gets the “nice try with paywalled links, those articles could say anything” when it wouldn’t even matter if the article were freely available for a large proportion of those making pronouncements.

  12. But don’t worry: they’ll then shout that scientists are confusing correlation with causation when it comes to greenhouse gasses.

  13. I lurk here more than I post because any stats I did was eons ago and much out of date.

    I just wanted to say that I’ve always thought of ‘natural variability’ in regard to climate, as a catch phrase meaning ‘variability for which there is currently insufficient information to attribute specific causes’.

    Does it have a more specific meaning (eg relating to chaos)?

    • David B. Benson

      Sou | February 27, 2011 at 2:48 am — The phrase ‘natural varialibity’ is a bit ambigous. Some use it to refer to variation unforced by just anthropogenic causes; other include variations unforced by solar variations and at least volcanic aerosols [although both these are certainly natural]. There are such variations. The most striking, to me, is the ~3.75 year period Rossby/Kelvin wave in the North Pacific. It exists becuase the North Pacific is resonant at that period and somehow wind+tidal is enough to keep it energized. Related to that is the quasi-periodic oscillation (QPO) called ENSO which has no predicable period other than in a statistical sense. Indeed, the repetition rate of the El Nino component of ENSO varies over long periods of time:
      Variability of El Niño/Southern Oscillation activity at millennial timescales during the Holocene epoch
      http://www.nature.com/nature/journal/v420/n6912/full/nature01194.html
      That paper commits, IMO, an error of ascribing a ~2000 year period to the proxy data. Looking at it carefully only the last half of the spectral analysis shows a 2000 year period band; the first half shows a band around a ~1800 year period. Also, it is difficult to even imagine what could be resonate at such a low frequency. So I’ll just say, for now, that ENSO in all of its manifestations is natural variability at all time scales.

      Similarly, there is reason to think that other measures of variability such as the PDO and (properly defined) the AMO are forms of natural variability in the climate. Both seem to be QPOs with long enough quasi-periods that there is no obvious resonances and certainly no extrenal forcings to explain this variability, i.e., oscillation. So such, are, at this state of knowledge, simply chalked up to ‘internal variability’.

      To match these, the ocean components of some AOGCM climate models have added codes which minic these sorts of behavior. [I'm uncertain the extent to which this is necessary for the much more thoroughly studied ENSO].

      I’m not sure this answers your question, but internal variability has nothing to do with deterministic chaos as far as we now know.

  14. “Natural variability” means changes that we can do nothing about, and since they are “variable” are likely to just change back the other way, no worries nothing to see etc. The first line for the don’t change a thing crowd.

  15. Thanks, David and PP. I suppose I can accept the definition of ‘natural variability’ as that not having a human cause. Although I find it difficult to come to terms with, largely because I (personally) view the climatic response to, say, rising atmospheric CO2 as a natural response rather than an artificial response. That is, the response (rising temperature, ocean acidification etc) is in keeping with the physics and chemistry. I also view (personally) that humans are natural not artificial. I accept it’s semantics – and, if you like, an artificial distinction to aid in describing different aspects of climate so it’s perfectly acceptable in that regard :) (Not that what I think has the slightest bearing on the matter – lol.)

    To clarify then, variations in solar intensity, volcanic eruptions etc would be ‘natural variability’. ENSO events are natural variability but probably heightened by anthropogenic forcings. Changes in ocean and atmospheric circulation could be either natural variability or anthropogenic or a mixture of the two. Would that be fair comment?

  16. I’ve just re-read David’s post and realised I’ve oversimplified things in my response. I’m starting to get the gist, though, and from the sound of things, the terms natural variability / internal variability etc can have different meanings attributed to them.

    • David B. Benson

      Sou | February 27, 2011 at 6:27 am — The distinction is not between ‘natural’ versus “super-natural’ but between “natural’ and ‘artificial”. The latter is the wrok of man and the former then just means ‘non-artificial’. Ideally natural variability then ought to be fourther subdivided into forced via external causes and internal variability. So solar and volcanic variations are external, natural forcings leading to variability while ENSO is (largely) due to internal variability (in winds, SSTs and variations of sea level from the geodestic).

      It is surmised that future (artifically induced) warming may led to a state called ‘permanent El Nino” or “Le Padre’. So yes, ENSO may be influenced by artificial forcings.

      [As a warning, not all writers are so careful in these distinctions and one must read, and then re-read, with care. But this has alerted me to the desirability of writing more carefully in the future.]

      • Appreciate your further comments, David, which seem to remove the need for any notion of ‘natural variability’ altogether.

        To me as a layperson, I’d find it easier if the term ‘artificial’ was replaced by ‘anthropogenic’. Variability can be broken into variability within the weather-related factors influencing weather/climate such as pressure, humidity, SST pattern etc (‘internal’ variability) as you’ve said, and forces on the weather/climate that are external to weather and climate (volcanoes, sun, anthropogenic forces such as deforestation, carbon emissions, and secondary factors such as forest fires etc).

        External forces can obviously affect the patterns of internal weather/climate variability. In other words, external forces and internal variability are not mutually exclusive and ‘natural variability’ has less meaning, because it can be replaced by the notion of ‘internal variability’.

        At least thinking through this should help me interpret writings that use these various rather fuzzy (to me) terms :)

      • David B. Benson

        Sou | February 28, 2011 at 9:42 am — Natural variability is not the same as internal variability as it includes the response to natural external forcings.

        But yes, with some care in writing the word ‘natural’ can be avoided altogether. Unfortunately not everyone does so.

  17. Tamino,
    although I think it is clear that the post 1975 climate is dominated by anthropogenic forcings, there is a strong argument to be made that some of the warming prior to that has been the result of centennial to millennial scale climate change. The Bond cycles come to mind but also Viau et al (2006) which identifies a millennial scale cycle with a periodicy of around 1000 years.

  18. I’ll just note that sun is 98 % of sun is H or He w/ a rotation period of 25.05 days and the volcanic activity hasn’t been proven to have any cycles. Man, they can’t even predict singular eruptions or solar flares!

  19. The Vaughan Pratt analysis is very amusing. Pratt must be a smart man in his field of study, but he is getting awfully close to Professor Irwin Corey territory with his applied math. I spend a lot of time trying to extract some practical applications from abstract and formal mathematics, but Pratt’s efforts on doing this himself makes me think twice about applying that
    strategy.

    Pratt’s 62.5 year “filter” is puzzling. Why he thinks that a 62.5 year filter suppressing a few large fluctuations is so surprising confounds me. This is numerology at best.

    I now understand the title of your post and the detailed explanation. You are a decent man not to name names, but I thought it important to tell the rest of the story.

  20. tamino (and any others) ,
    i was reading a thread over at climate etc on chaos:
    http://judithcurry.com/2011/02/10/spatio-temporal-chaos/
    and was wondering how you would correctly contextualize chaos (temporal and spatiotemporal) in the climate system.
    That is can we model the entire climate system chaotically, some parts of it (ENSO,PDO,AMO etc), artifacts of poor analysis or insufficient data and understanding to tell?

    some comments by Fred Moolten and Arthur Smith made the most sense to me but my knowledge is insufficient to even speculate on the accuracy of such claims.
    Hans von storch seems to dismiss the idea as having no merit here(comment 5): http://klimazwiebel.blogspot.com/2011/02/climate-models-reading-material-for.html but i see no argument.
    also i watch a presentation from Jeffrey Weiss here:
    http://online.itp.ucsb.edu/online/climate08/weiss/ where he seem to say stochastic models of climate do a good job of replicating ENSO, Storm tracks, etc and that non-equilibrium thermodynamics may have something to say about climate variability.

  21. Of course proponents of these ‘theories’ are the first to leave out known natural variations if doing so makes their arguments seem more persuasive; such as leaving out ENSO which, when taken into account, makes the ‘no warming since 1998′ fiction vanish all by itself. Which Tamino (thank you) showed clearly enough in his Sharper Focus post.

  22. Great idea, an excellent description of those who suffer a severe form of denial cognititive dissonance and aversion to facing the reality of a warming world.

    May I ask a dumb and stupid question, has anyone thought of defining this particular term for the online “Urban Dictionary”, as found here http://www.urbandictionary.com/ ?

  23. Marion Delgado

    If Market God had wished persons (corporate and inferior) to use “physics,” He would have given us all pocket protectors at birth when Mammon, Ricardo and Mises created the heavens and the earth.

  24. tamino: have you by any chance looked into Muller&Rohde’s work under “Current Research”: that finds a strong 62Myear cycle in biodiversity?
    You may also want to look under the “Nemesis” topic for more cycles.
    Some cycles may be real, some may not.

    Muller is recently in news due to BEST effort.

    • David B. Benson

      John Mashey | February 28, 2011 at 1:28 am — Did you really mean a 62 million year ‘cycle’ in biodiversity? This is too absurd to bother following up [especially, for me, consiering the source].

      [Response: There was a theory once, that mass extinctions happened about every 60 million years or so because of an undiscovered object (perhaps a brown dwarf) in a huge orbit around the sun. Once every orbit it would approach close enough to the "Oort cloud" of comets to perturb some of their orbits and send a large number into close-approach orbits, leading to impact events and extinctions. Now, however, I don't think most mass extinctions are attributed to impact events (just the K-T extinction).]

      • Yes, 62M cycle, attributed to the “Nemesis” that I mentioned.
        Muller’s webpage says:
        “No, I have not abandoned the Nemesis idea.”

        This is the 3-page letter in Nature which claims the 62My cycle is very strong, lists 7 possibilities for how it might arise. The math question is: is the 62My cycle real?

      • Gavin's Pussycat

        > The math question is: is the 62My cycle real?

        Indeed.

        Richard Cowen strongly thinks it is not (as do I):

        http://mygeologypage.ucdavis.edu/cowen/HistoryofLife/muller2005.html

        I have been looking for an authoritative (peer reviewed) debunking of this Muller & Rohde 2005 paper, with no luck. And it was in Nature! Doesn’t anybody care? This is ‘mathturbation’ at its best, and with the usual over-confidence of physicists… doesn’t bode well for BEST.

    • Gavin's Pussycat

      BTW Robert A. Rohde is the guy behind ‘Global Warming Art’… seems to be a straight shooter. Hope he doesn’t get burned

  25. http://onlinelibrary.wiley.com/doi/10.1111/j.1745-3933.2010.00913.x/full
    Nemesis reconsidered
    Adrian L. Melott, Richard K. Bambach
    17 AUG 2010
    DOI: 10.1111/j.1745-3933.2010.00913.x

  26. David B. Benson

    I see, an atronomical cycle is hypothesized

    Unfortunately, not a scientific hypothesis as there is no way to demonstrate it.
    (i) The solar system is chaotic so comets might drop out of the Oort cloud at any time. [Indeed Mercury may well move to an orbit outside of the orbit of Mars; one long run of the approximate dynamics suggest this possiblity as well as another in which Mercury wanders out of the solar system.]
    (ii) There is evidence to suggest that the K-T event (a) didn’t finish off all the dinosours [the remainder hanging on for some hundreds of millennia] and (b) the number of dinosaurs had been declining for about 10 million years and (c) maybe the number of dinsour specieas in a similar decline. To the extent that additional evidence for (b) or (c) is forthcoming, I’ll posit virus evolution, parasite evolution and the evolving mammals eating dinosaur eggs.
    (iii) Other mass extinctions appears to be related to the formation of traps, as in Siberian traps; massive basalt flows. These need not be continental; there is reason to expect that oceanic lava flows might be even more detrimental to the climate and hence life. Few such oceanic formations have been well dated; there are certainly plenty in the vicinity of the Indonesian Islands.

    Too many hypotheses; not enough data. So far impossible to exclude the ones I’ve mentioned but sure, thow a posited Nemises into the ring as well.

  27. Pete Dunkelberg

    How does this paper fit into the mathurbation picture?

    [Response: I note they claim to establish 1500-year periodicity in 2000-year long data sets.]

  28. Horatio Algeranon

    You can lead a mathturbator to Mother Nature, but you can’t make him lay a hand on her.

  29. Then there are the “natural variation” theories — those which imply that the globe’s warming is just “business as usual.” You have to ignore a helluva lot to believe this.

    Thanks for pointing this one out. It involves a confusion of logical type of the terms ‘natural’ and ‘random.’ For example, species extinction is ‘natural’ in the sense that it can be caused non-anthropomorphically. But extinctions must have causes; just as speciation must have causes. Yes, mutations are random, but whether the mutations persist in the genome is highly non-random. The Pleistocene is ‘natural.’ The Permian and Cretaceous extinctions were ‘natural.’ While an asteroid hitting the Earth, is in some sense ‘random,’ its trajectory can be plotted using Newtonian mechanics, and resulting species extinctions would certainly not be called ‘random’ — they were caused by the freaking asteroid !!!

  30. To me, invocation of the term ‘natural variation’ often falls into a category of special pleading akin to asserting that the speed of light has changed over time as opposed to remaining a constant. At minimum, proponents of a ‘natural variation’ should propose a specific method of causation and a way in which this mechanism can be tested and falsified. ‘Natural’ is not a method of causation.

  31. Horatio Algeranon

    Just as Horatio suspected, Venus has a hand in this( although it’s a long-distance relationship)

    “How the exact timing between Venus and Earth correlates with the 100,000-year weather cycle is as yet unclear [but when did that ever prevent wild speculation?] One thing is clear. There is an unspecified lag time involved between the two cycles. [one can not rule out that the lag may be 49,000 years]“

    Even the number “62″ (or more precisely “62.5″) makes a prominent appearance (see “Molar Extractions”)

  32. The other is the natural tendency for smart people to play with numbers and ideas, find interesting relationships, then persuade themselves they’ve happened on a key insight.

    A certain climate scientist from UAH with the initials RS springs to mind.

    • One needn’t be smart to have a fascination in playing with numbers. A certain person with the initials JD springs to mind.

  33. Then there are the “natural variation” theories — those which imply that the globe’s warming is just “business as usual.”

    If you talk to bedrock geologists studying microfolds of microfolds in Ordovician metapelite strata in Waterville, Maine who are looking for any remnant graptolite fossils, mass invertebrate extinctions are “business as usual” for Earth on a 450 million year time scale.

    But this is a long leap from saying that extremely short term perturbations, like the Wisconsinan Ice Age, is just “business as usual” on the same scale as the Atlantic Ocean opening and closing and leaving a chunk of West Africa stuck onto Harvard University in Cambridge. At a large enough scale, all Earth change is just ‘business as usual,’ including the formation of life itself.

  34. I was trained as a mathematician, and I think ‘mathturbation’ is a very good term for modeling exercises with too little regard for the actual concstraints that apply. That they are often carried out with an agenda, makes everything worse. I don’t know if I would call all of it anti-physics, rather it looks like a-physics to me, it is akin to returning to older kinematical descriptions when dynamics is available. They tend to use physics fragments whenever it seems handy. And, having an elephant in the room – the sun – everything may of course be explained invoking sufficiently advanced elephant theology.

  35. I think it is unfortunate that ‘natural cycles’ seem only to be defined as ‘quasi sine waves with a defined period’. I would suspect that most natural cycles would NOT be of this simplistic a nature (pun intended).

  36. Tamino, thanks for this post. I have now more clear picture about some frauds against climatology. Yes, my math is bad, I am lousy at statistics. But, I your post reveals one very very important issue which I’ve missed for a while being mesmerised by statistical tricks: There must be sound phisycal process behind warming and if one want to prove that current warming is due to some “natural” cycles, one should prove us that internal variability of climate system is actually capable of doing such changes with some decent probability of being correct (and as I read here, also being a falsible theory).
    However, I’ll point out some other issue: For us, non-scientific people, it is hard to distinguish between “at first glance correct” arguments and actually correct arguments – speaking for myself, I do not have enough knowledge and skills to make a fine rebuttal or confirmation of some claims in climate science (and I am nevertheless not most uneducated person in this world). However, I can understand some errors in argument you point out. For this, I am most grateful.

  37. Patrice: I would suggest that unlike, for example stock market prices, physics has conservation laws, to which tamino alluded. Conservation if Energy is The Law. Then the total variable heat content if the Earth ( most in oceans) goes up if the outgoing energy us less than the incoming energy. Thus is what the gremlins have to nullify if they want to let the leprechauns do it some other way. Heat may slosh around, but it isn’t magically created.

  38. From the 2010 paper I linked above:
    http://tamino.wordpress.com/2011/02/26/mathturbation/#comment-48943

    “… We examine the evidence for the previously proposed periodicity, using two modern, greatly improved paleontological datasets of fossil biodiversity. We find that there is a narrow peak at 27 My in the cross-spectrum of extinction intensity time series between these independent datasets. This periodicity extends over a time period nearly twice that for which it was originally noted. An excess of extinction events are associated with this periodicity at 99% confidence. In this sense we confirm the originally noted feature in the time series for extinction. However, we find that it displays extremely regular timing for about 0.5 Gy. The regularity of the timing compared with earlier calculations of orbital perturbation would seem to exclude the Nemesis hypothesis as a causal factor….”

  39. PS, Cowen’s page also linked above is cautionary: “Now let’s move to what Rohde and Muller actually did. First, they threw away more than half of Sepkoski’s data: those genera that happen to have had a short history and/or a small number of fossils. These discarded data are real data, of course, and throwing half the data away may very well bias the analysis, for example against times when marine life was evolving quickly….”

    When the BEST group publishes their sources, screening algorithms and methods it will be interesting to see what they keep and what they discard and why.

    • Gavin's Pussycat

      > Cowen’s page also linked above is cautionary

      Yep. And what’s more, of the three ‘big’ extinction events that produce much of the periodicity, one (the K-T) was very likely due to asteroid impact, while the other (P-T) has been commonly attributed to massive volcanic activity leading to greenhouse warming. Two very different physical meachanisms producing different peaks on a single sine curve — really?

      It’s a pity we only have Cowen’s write-up. This should have been taken down authoritatively.

  40. Gavin's Pussycat

    Replying to myself, there is this:

    http://arxiv.org/abs/math-ph/0608014

    • Horatio Algeranon

      After visitingthis site by the author of that arxiv paper (M. Omerbashich) a second (or third) opinion might be good. One should careful, at the very least.

      Some (including Horatio) might think some of the ideas found there fall pretty squarely in the “cracked pot”.

  41. The problem with trying to find genuine periodicity in those “mysterious cycles” is twofold.
    Firstly they seem to be looked upon by the pro-AGW commentaters as if they are expected to explain where the heat energy is coming from, rather then as a means of identifying where it is being transported to.
    By that I mean from the oceans and land surfaces where solar energy is initially absorbed, through the atmosphere from where it can be eventually lost back to space or deposited elsewhere.

    Secondly, and most importantly they are analysed as if they act in isolation to all other forces.
    The assumption seems to be that any oscillation must be witnessed as being regular otherwise there is no cycle. The expectation seems to be that for such a cycle to exist it must manifest itself similar to the regularity of the day/night cycle, or the annual seasonal cycle, in order for it to be explainable by simple physics.

    Unfortunately, not everything is so simple, how the various forces manifest themselves in combination becomes very complex, more so than apparently what one analyst alone can get his mind around, even more so than what the entire scientific population combined can comprehend, at least at this stage.

    Taking the viewpoint from Australia which is a relatively simple situation, it is subject to forces generated in different oceans which have there own identified ocean atmosphere systems and ocean currents, and it is only now that some small understanding is being reached about how the two seemingly major influences interact with each other and so influence the Australian climate.

    Each ocean is both independent in that the systems oscillate separately, but as with all such systems, at times offset, and other times reinforce what is happening in the adjacent regions.
    The nett observed result may look chaotic, but the apparent chaos has more to do with the limitations of understanding amongst those looking for perfect order than any real chaotic behavior of the forces involved.

    All could explained by known physics if not for the aforementioned limitations, the finite capacity of the human mind.
    The power of the human mind is similar to the power of an electrical system, for a given capacity, if the pressure increases, the volume will drop.
    In the human mind, as the depth of understanding increases, the breadth decreases. It is very true that ultimately, some people do exhibit that they have reached the point of knowing everything about very little.

    Given the extreme complexity of the earths climate, it may be that the greatest understanding lies with those whose broader perspective leaves them knowing a little bit about everything.
    At least they may be better able to understand how the many systems interact rather than trying to find perfection in genuine periodicity in one of those “mysterious cycles”.

    • Nice strawman, johnd. Was it too much bother to read what Tamino has actually written on the subject and reply to that?

      • Ray Ladbury

        The really sad thing about Johnd’s post is that he thinks somehow that he’ll persuace a website chock full of science with his handwaving. Do the math, John.

      • At that, I don’t know quite what John is trying to say.

      • Ray Ladbury

        I think John is of the school that thinks if they can state their position vaguely enough, no one can definitively disprove them–the “not even wrong” school.

      • Rattus Norvegicus

        Ah, the Judith Curry school of thought!

    • Firstly they seem to be looked upon by the pro-AGW commentaters as if they are expected to explain where the heat energy is coming from…

      No, John, that would be the exact opposite of reality. It’s the cranks who are trying to invoke oscillations as an alternative to CO2 forcing.

    • Given the extreme complexity of the earths climate, it may be that the greatest understanding lies with those whose broader perspective leaves them knowing a little bit about everything.”

      Wow, a denier meme two-for–”Climate’s too complex!” “The warmists are ignoring other factors!”–all in one sentence.

      Now, that’s economical.

  42. I am delighted to report killfile is working here again (with Firefox 4 beta 12).

  43. Horatio Algeranon
  44. Rattus Norvegicus

    Related to epicycles and defiantly an example of mathturbation is this paper from Douglass, et. al.

  45. There a new “human finger print” on climate change. It involves some fancy modeling and is unlikely to convince deniers, but it is still rather interesting.

    http://www.sciencedaily.com/releases/2011/03/110311140706.htm

    NASA Study Goes to Earth’s Core for Climate Insights

    ScienceDaily (Mar. 11, 2011) — The latest evidence of the dominant role humans play in changing Earth’s climate comes not from observations of Earth’s ocean, atmosphere or land surface, but from deep within its molten core.