RealClimate logo

Note 3/23/2021: we had a few hiccups with comments after moving the site to https/SSL. Hopefully they're fixed now. Please let us know if there are remaining issues.

Unforced Variations: March 2012

Filed under: — group @ 1 March 2012

This month’s open thread – for appetizers we have: William Nordhaus’s extremely impressive debunking in the NY Review of Books of the WSJ 16 letter and public polling on the issue of climate change. Over to you…

617 Responses to “Unforced Variations: March 2012”

  1. 451
    dbostrom says:

    That has been a problem lately; too few people are actually using data in their anaylses.

    [edit – We have no wish to encourage these people to start spamming the comment threads here. Let it just be recognised that there are fringes beyond which we have no desire to go]

  2. 452
    Dave says:

    I posted this in another thread and was advised to post here.. I hope someone can answer.

    I’m a professor of physics at a leading European University. I mention this not to establish any type of authority. Rather it should be taken to mean that I can grasp a fairly technical explanation if it is appropriately phrased for a scientist working in a different field.

    I consider myself to be a sceptic of all research (including my own). However, there are certain areas of research in which working theories of nature have been established and can be relied upon to an extremely high degree to predict future experimental results. Examples include general and special relativity, and quantum mechanics. These theories gained acceptance because they passed “classic” scientific tests. They made predictions which would have allowed them to be falsified and discarded. And they passed these tests. We still test these theories today because we know they are not the “final word” and we hope that through any discrepancies with data we can learn more.

    When it comes to climate science its difficult for me find the equivalent of, eg the general relativity tests of mercury’s perihelion or binary pulsar motion. Have there been any “classic” tests of climate models that would falsify the hypothesis of large warming over the long term ? I ask this question since I’m more than fed up with hearing about a consensus. I don’t doubt that one exists but its not an argument that I would use to argue for the correctness of general relativity or quantum mechanics. Instead I would describe the numerous experimental tests which could have falsified these theories. Is there a series of experimental tests which could have falsified the climate models which presently imply drastic long term warming i.e. X degrees over Y years or something similar whereby X and Y would have been defined at the time of the test and would correspond to substantial long-term warming ? By “falsified” I mean that a paradigm shift would be needed rather than further model optimisation.


  3. 453
    MartinJB says:

    Holy cr@p! TVMoB is a birther!

    “I have watched Sheriff Arpaio’s press conference in AZ and have examined some of the evidence directly. It is clear – as Alex Jones rightly said on the day when Obama first put up his faked “long-form birth certificate” on the White House website – that a fraud has been committed, and that, absent a valid official record of Obama’s birth or a very good explanation of the anomalies in the published version, he is not qualified to stand for re-election as President.[…] This is beginning to look like a widespread, high-level fraud.” – from Susan Anderson’s link in 450

  4. 454
    dhogaza says:


    Holy cr@p! TVMoB is a birther!

    That’s rather funny seeing as how TVMoB has falsely claimed to be a member of the House of Lords for so many years …

  5. 455
    David B. Benson says:

    Dave @452 — I’m not quite sure what you are looking for. But to begin, note the physical chemistry of tri-atomic atmopspheric molecules; pass optical bands and absorb/emit IR. Therefore the (poorly named) greenhouse gas effect is real. As for how much effect, will correlating CO2 concentration with a global temperature product do? If so, visit Barton Paul Leveson’s web site where the exercise is completed.

  6. 456
    Craig Nazor says:

    Dan H @443,

    You’ve got to be kidding.

    Assume: Suppose to be the case, without proof.

    I am not assuming – there is overwhelming evidence (from copious data, much of which can be found on or linked to from this web site) that global temperatures are rising at a rate that may soon seriously disrupt human civilization, and that the best explanation for the cause of that projection (based on even more data) is human-driven, rising atmospheric CO2 levels. You just deny this, and then complain: “too few people are actually using data in their analyses,” without providing any support whatsoever for that statement.

    So who is assuming?

  7. 457
    Dan H. says:

    The short answer currently is “no.” That is largely due to the nature of the test that is required; the earth’s response to increasing levels of greenhouse gases. No experiment can mimic that outcome, so validation of the theories will require time. In all likelihood, optimization of parameters will occur, rather than falsification. Although, we can never be truly sure until the test is actually run.

  8. 458
    Dave says:

    Thanks but that’s not quite what I’m looking for. The underlying principles are well grounded. My questions concern specific falsification tests of the quantitative long term predictions of global temperatures i.e. is it possible to do now with a range of observables ?

    [Response: Well of course it’s possible if you wait long enough – and there is plenty of confirmation (stratospheric cooling predicted in the 1960s, not observable until the 1980s, confirmed; ocean heat content increases predicted in the 1980s, not really observable until the 2000s, confirmed; etc.). However, the central question in climate is deciding in what is likely to happen before it actually does. Thus one has to have a different strategy – predicting things that may have already happened but that you either didn’t know about, or for which the data have not yet been analaysed. There are also plenty of examples where models have correctly suggested that different data sets were inconsistent (satellite vs. surface in the 1990s, tropical ice age ocean temperatures vs. land temperatures in the 1980s etc.) which were resolved in favor of the models. One just needs to be more creative in looking for interesting tests. – gavin]

  9. 459
    Charlie H says:

    The “Updating the CRU and HADCRUT…” article reminded me of a question that might be slightly off-topic there, so I ask it here.

    1998 seems to have been remarkably warm, seemingly much warmer than any preceding year and only {tied or barely exceeded – take your pick} since. Is there a general agreement on what factors caused 1998 to be so warm? What were the factors?

  10. 460

    #459–“Is there a general agreement on what factors caused 1998 to be so warm?”

    It’s generally ascribed to an unusually strong El Nino.

  11. 461
    Hank Roberts says:

    > The short answer …

    “For every complex problem there is an answer that is clear, simple, and wrong.” — H. L. Mencken

    Search using Dave’s phrase
    shows these words often used at wattsup and the like, but the search will also find some interesting work, e.g.

  12. 462
    Ray Ladbury says:

    As you have moved to the open thread, perhaps you didn’t see my query as to your research specialty. It would help us to understand how much your background might allow you to comprehend of climate science.

    As an analog to your question, one could ask what tests might cause us to radically reassess the atomic theory of matter or nucleosynthesis as the model for energy generation in stars? Once a field is fairly mature, evolution of the model is much more likely than is a complete revolution.

  13. 463
    Charlie H says:

    #460, Kevin McKinney: Thanks.

  14. 464
    Dan H. says:

    Here is a plot of the GISS data. This is the data that I used for my analyses. If you are using other data, or arrive at a different conclusion, please explain.

  15. 465
    Jarmo says:

    I have a question regarding the recent claims of future sea level rise.

    Many recent studies (e.g. Hansen & Sato) have claimed that future rise in global average temperature (GAT) will create a much greater effect on sea level than IPCC AR4 predicts. They use paleo proxies to make comparisons with the present. For example, Hansen & Sato argued that since GAT during the Eemian (last interglacial before the present) was only slightly higher (less than 1 degree C) and sea levels 4-6 meters higher, a 2 degree rise in GAT in the near future will result flooding very quickly.

    The Eemian warming was due to higher solar insolation than today ( 13%) on the high northern latitudes during NH summer (SH winter). A recent study on the GIS melt during the Eemian argues that temperature rise alone produced 55% of the melt and the rest was caused by higher solar insolation and feedbacks. See link below:

    However, increase of solar insolation in the high latitudes did not warm the tropics and so GAT rose relatively little. Most studies argue that there was tropical cooling compared to the present.

    Today the situation is different since we have less solar insolation up north in the summer but more on low latitudes. Furthermore, CO2 is supposed to warm the planet more uniformly so will get temperature increases also on lower latitudes and the equator. GAT will probably rise much more than during the Eemian.

    Despite these differences in forcings and their latitudinal distribution, studies continue to argue that the past ratio of arctic temperature rise and GAT, as well ice sheet melt rates, will also be valid in the future. See the link below:

    What’s your take on this?

    Btw, this is my second try to post this. First time failed for some reason

  16. 466
    Hank Roberts says:

    > evolution of the model

    For example — when I took physics four decades ago, the environment around an atom wasn’t mentioned as affecting nuclear decay (heck, the physical/ microwave environment around molecules wasn’t discussed as affecting chemical reactions)

    Nowadays? Hm.

    Perturbation of nuclear decay rates during the solar flare of 2006

    So would past variations in our local star — or a long-ago local supernova — produced a pulse of warming? Detectable eventually at the surface of the planet? And how big, compared to what we know CO2 produces?

  17. 467
    Hank Roberts says:

    Reminder — don’t revise everything or proclaim proof “it’s the Sun” quite yet; work continues:

  18. 468
    Jim Larsen says:

    432 SA said, “The scalability of solar energy technology is indeed one of its best “qualities” — along with the ubiquity and inexhaustibility of the energy source.

    I think it’s entirely possible that within a much shorter time frame than most people imagine possible, the overwhelming majority of the electricity consumed in the USA will be generated at the point of use, and large, centralized power plants will be a much less important part of the mix than they are today”

    I’d add solar’s lack of externalities and its matching of peak power demand to its best “qualities”. Wind, hydro, biomass, and solar thermal all work best when centralized, so I think centralized power plants will be the dominant sources for a long time. Now, if PV drops to a penny a kwh…

  19. 469
    Jim Larsen says:

    434 Walter said, “Dan H. will have a hard time finding a comparable 30-year period to the most recent one.”

    The “alternative” view is that it’s a 60 or longer year cycle with huge underlying longer cycles, so we’re at the peak. As predicted, temperatures crested over the last decade, and are poised to fall as the cycle inevitably follows its course. Some think these cycles overlay a small (<1C/doubling) increase from man. Some think there's a natural warming in the background from an ongoing recovery from the little ice age. All rely on 1998 being the peak.

    Should, by some miracle, temperatures start to rise again, their theories all become instant rubbish. Seriously. This is the end of logical debate amongst non-scientists. Either it's a cycle and we're cresting, or it's a linear trend and we're about to warm significantly. Zealots do what zealots do, but assuming that the weather over the next five years is median, the backlash should make things perfect for beginning a massive buildout of renewables. By then, hopefully, batteries will be solved and wind and solar will be truly competitive. We'll certainly know a lot more about smart grids. Maybe technology and public opinion will come together over the next five years to enable the beginning of a well-planned rebuild of our energy infrastructure.

  20. 470
    Ray Ladbury says:

    Jim Larsen,
    Ah, but when temperatures continue to rise, all the denialists will do is hallucinate another “cycle” with a slightly longer period. After all, they have no evidence for their 60-year period cycle. Why should any be required for an 80-year cycle? See that’s the wonder of Fourier series: You can match any data series except near the endpoints. Just keep moving the endpoints and the fun never ends.

  21. 471
    Hank Roberts says:

    > Dan H. … if you used …

    If you read:
    “… trend-lines … can be fairly dangerous…. Depending on your preconceptions, by picking your start and end times carefully, you can now ‘prove’….
    … If you look at the trend data, you can see the current trends in °C, between 0.14-0.16°C/decade, or, if it continues at the same rate, between 1.4 and 1.6°C per century.”

    And most recently added to the Notes there:
    “The trends from the data dump are as follows:”

    Source—-Trend °C/century
    GISTEMP DTS—-2.06

  22. 472
    Jim Larsen says:

    437 Ray said, “Is this series periodic? NO. The y values are the digits of the base of Napierian logarithms, while the x values are thier ordinal position. Therefore it cannot be periodic.”

    That there happens to be a precise “solution” to a set of hypothetical temperatures is irrelevant. A periodic plus other natural variation can certainly achieve the series you proposed. The first number is time in six month intervals, and the second is temperature in unspecified units.

    Maybe it’s me (I’m often dense), but even though your posts are usually clear and right on, this one seems clear and wrong…. explain?

  23. 473
    Pete Dunkelberg says:

    Monckton is coming to town.

    From comments here: (Rabbet)

    Monckton will be giving a presentation at the Joan B. Kroc Institute for Peace and Justice Theatre, 5998 Acala Park, University of San Diego, Saturday Mar 24 at 7PM.

    Followup re: Monckton in San Diego.

    On-line reservations here:

    Total auditorium capacity is about 300 — currently about 113 seats left.

  24. 474

    #471–Amusingly enough, Dan’s Woodfortrees graph confirms what I thought he had done–simply fitted a linear trend to the entire timespan.

    Of course, the fact that this is possible to do says nothing whatever about whether that trend is the best fit to the data (or even a moderately good one.) Thus–as many here know, possibly including Dan?–you can’t use that linear trend by itself to conclude anything in particular about the internal temporal structure of the time series.

    As, for example, “the trend has not changed…”

  25. 475
    Jim Larsen says:

    470 Ray said, “See that’s the wonder of Fourier series: You can match any data series except near the endpoints. Just keep moving the endpoints and the fun never ends.”

    I’m not an expert, but I assume it works by huge and/or ever-increasing numbers of waves of very large timeframes. Given that not just one endpoint, but the vast majority of the data (paleoclimate) is constricted (it wasn’t boiling anytime in the last billion years), and the number of waves at huge magnitudes is strictly constrained by believability, I don’t think math is going to save deniability.

    Sure, there will be people who will believe it’s all natural until they die, but if 90% of the population is convinced, policy will change drastically.

  26. 476
    Dan H. says:

    Yes, those are the land-only trends. I was referring to global trends.

  27. 477
    Hank Roberts says:

    Hm, the “trend data” and “data dump” links only work from inside the woodfortrees site — to get those click the links above them.

    The “last:360/trend” refutes Dan H.’s attempt to be the explainer.

  28. 478
    Hank Roberts says:

    Jarmo says: 21 Mar 2012 at 12:59 PM

    “… studies continue to argue that the past ratio of arctic temperature rise and GAT, as well ice sheet melt rates, will also be valid in the future. See the link below:

    Well, is Jarmo right? Click the link read it — it’s

    “suggesting that Arctic warming will continue to greatly exceed the global average over the coming century, with concomitant reductions in terrestrial ice masses and, consequently, an increasing rate of sea level rise.”

    Jarmo, where did you get that idea and fake cite for it?
    Searching for the key words:

    The top two hits are septic sites. It may be that you were misled.

    Be skeptical, do actually read what they claim as a source.
    Don’t be fooled again.

  29. 479
    Craig Nazor says:

    Dan H:

    In addition to the comments about a “linear trend” that I and others have mentioned, I noticed that you used LOTI data. NASA has this to say about LOTI data:

    “Note: LOTI provides a more realistic representation of the global mean trends than dTs below; it slightly underestimates warming or cooling trends, since the much larger heat capacity of water compared to air causes a slower and diminished reaction to changes.”

    So I suppose that is why you used it – because, in addition to the falllacy of a linear trend being meaningful, the LOTI data will underestimate that trend… ?

    No bias there!

  30. 480
    Ray Ladbury says:

    Jim Larsen,
    Sorry the post was obtuse. The y values of the series are the first 10 digits of e=2.7182818284…, the base of Napierian logarithms. This number is transcendental, like pi. It has no repeating patterns and so cannot be periodic. This is an example I use to illustrate that unless you have several periods of data or a known periodic forcing of the system, it is simply silly to attribute periodicity to the system.

  31. 481
    Pat Cassen says:

    Jarmo (#465)- Interesting question. The conclusion that the Greenland ice sheet melting was significantly enhanced by the increased N. Hemispheric insolation during the Eemian affects projections of future (near term) sea level rise insofar as Greenland melt contributed to the Eemian sea level rise.

    Estimates of the Greenland contribution generally require a significant fraction of the total from Antarctica: The Greenland contribution is variously estimated:
    2.7 – 4.5 m out of 5 – 6 m
    4 – 5.5 m out of ? (“more than 5 m”)

    McKay et al. conclude that “…4.1 to 5.8 m of sea level rise during the Last Interglacial period was derived from the Antarctic Ice Sheet.”

    Regarding the ratio of arctic to global average temperature, your last link presents evidence that it was not affected by the distribution of insolation during LIG. That paper gives a good discussion, thanks.

    Hank (#477) – Miller et al. really don’t say much about sea level rise, beyond the bit you quoted.

  32. 482
    Dan H. says:

    You are correct, there is no bias in my analysis.

  33. 483
    Ray Ladbury says:

    Dan H.: “You are correct, there is no bias in my analysis.”

    And no sense either.

  34. 484
    Brian Dodge says:

    For DanH and others interested in warming trends and linear fits, y’all might want to consider this, or this, (Figure 5.9). There’s also this,
    Partitioning Recent Greenland Mass Loss, van den Broeke et.a,lScience, 13 November 2009: Vol. 326 no. 5955 pp. 984-986; DOI: 10.1126/science.1178176
    “A quadratic decrease (r2 = 0.97) explains the 2000–2008 cumulative mass anomaly better than a linear fit (r2 = 0.90). Equation S1 implies that when SMB-D is negative but constant in time, ice sheet mass will decrease linearly in time. If, however, SMB-D decreases linearly in time, as has been approximately the case since 2000 (fig. S3), ice sheet mass is indeed expected to decrease quadratically in time.”

  35. 485
    John E Pearson says:

    462 Ray said about reassessing old physics:

    “what tests might cause us to radically reassess the atomic theory of matter or nucleosynthesis as the model for energy generation in stars?”

    SPot on. What tests would lead us to overthrow O18/O16 as a determinant of temperature. What test would lead us to throw away our understanding of cosmic ray physics on which rides C14 statistics. What tests would lead us to conclude that the Navier-Stokes equations were fundamentally flawed? What tests would lead us to conclude that the consensus view of the absorption spectra of CO2 and H2O is fundamentally flawed? Etc etc etc. Yawn.

  36. 486

    #482–Yes. And an exponential fit is best for Arctic sea ice volume, as well, per Piomass.

  37. 487
    Hank Roberts says:

    Pat, this is the comparison I made:

    Jarmo claimed the paper supports this: “the past … ice sheet melt rates, will also be valid in the future.”

    The paper says this “… concomitant reductions in terrestrial ice masses and, consequently, an increasing rate of sea level rise.”

    Jarmo: past … rates, will also be valid in the future
    Paper: an increasing rate

    Jarmo’s stated a claim that isn’t in the paper he cites for support.

    I’m curious where he got the claim and cite.

  38. 488
    Dan H. says:

    As Jim pointed out, unless you have enough data to prove otherwise, the premise cannot be refuted. Obviously, two periods are not enough to prove periodicity. However, should this oscillation continue for another cycle, one would have to seriously consider its applicability, especially in light of the steadily increasing CO2 emissions. The data is more involved than just the first ten digits of e; it comprises 1350 months of temperature data.

  39. 489

    Dan H., there is no “analysis” in what you wrote.

    If you ask yourself:

    1) Are there time series which cannot be fit with a linear trend?
    2) Are all time series unvarying over their length?

    …then you might be approaching the fringes of analysis.

  40. 490
    wili says:

    I understand that it takes some time to crunch the numbers to determine whether a particular major weather event (or long-lasting and widespread series of events, in this case) can be attributed with high probability to GW. Who does these determinations? How long before someone can give a definitive statement on this spring’s extreme weather across much of the US and Canada?

  41. 491
    dhogaza says:


    Dan H.: “You are correct, there is no bias in my analysis.”

    And no sense either.

    Nor honesty … call it what it is. He knows what he’s doing.

  42. 492
    David B. Benson says:

    John E Pearson @484 — Well, for the d18O proxy the tests have been performed and the proxy is deficient in that it also depends upon sea level. See, for example, chapter 1 in Ray Pierrehumbert’s “Principles of Planetary Climate” for a summary.

  43. 493
    Hank Roberts says:

    > Dan H.
    > land-only

    nope, not true for most of the data sets.

    You use the guy’s site the way he warns readers to be wary of — to distort what’s there, to pick points to claim it shows what you wish were true.

    Woodfortrees shows you different takes on global data.

    Pick another set:

  44. 494
    Hank Roberts says:

    PS for anyone coming new to the site — start with the explanation.
    It helps to get it first hand from the site, not second or third hand:

  45. 495
    Craig Nazor says:


    Since you seem unable to appreciate sarcasm: you have supplied no evidence that a linear trend is a best fit. This web site is FULL of evidence that historic temperature change in relationship to forcings is not linear.

    Those who produce the LOTI data have clearly stated that it underestimates warming or cooling trends.

    These are both evidence that your statement is biased at best, likely irrelevant, and that your continued denials and evasive responses suggest dishonesty. One has to wonder why.

  46. 496
    Jarmo says:

    Re 487,

    When I try to quote papers my posts end up spam. Too many spam words, the computer says.

    Anyway, the papers I mentioned included recent Hansen and Sato paper on paleoclimate and future warming where they argued that since the Eemian was less than one degree warmer (GAT) than today and had 4-6 meters higher sea levels, the set 2 degree limitation for future warming was too little. 2 degrees would mean almost Pliocene warmth and 20 meter sea level rise.

    The Miller paper actually has a graph (Figure 4) where past interglacials, their warming, and Arctic warming were plotted on a chart. 1 degree GAT warming will result 3-4 times more Arctic warming was the message.

    However, the third paper modelled GIS melt during the Eemian and concluded that 45% of the melt was due solar forcing, 55% due temperature rise. They claimed that GIS melt in the future would be less than thought because of smaller solar forcing today.

    In summary, the Eemian solar insolation increase in high northern latitudes produced pronounced Arctic warming and some GIS melt. However, the tropics cooled and hence GAT increased only little.

    Today the solar insolation is less up north and more elsewhere. CO2 will warm the planet more uniformly and also warm the tropics.

    On the basis of this, unless CO2 produces dramatic Arctic warming and tropical cooling (or no change)in the future, the future GAT will rise more than during the Eemian but will not have proportionally similar Arctic temperature or ice melt increases as during the Eemian. I

  47. 497
    Dan H. says:

    I you even read my posts, you will know that I have said that a linear trend is not the best fit to the data. However, the data does appear to oscillate about the linear increase. But don’t take my word for it, read was others have to say:–Xiuan-MeteorAtmosPhys-2007-d1227bc1-3183-456f-a935-69c263af1904.pdf

  48. 498
    Jim Larsen says:

    480 Ray said, “unless you have several periods of data or a known periodic forcing of the system, it is simply silly to attribute periodicity to the system.”

    But it’s also unreasonable to rule it out. Since the things which show that this isn’t the crest of a cycle are complicated and perhaps need caveat, we’re performing an experiment. It’s taken 14 years so far, but when warming (instead of cooling) cranks up again, the Monoskeptics will shed membership and prestige. Given that by then we’ll have three(?) more years of science, cohort replacement, and ice melt, odds are that enough of the public will “go AGW” to initiate a big push to reduce carbon emissions. This mind-shift will be enhanced as renewables drop towards or below parity – it’s nice to save money by saving the world.

    So I’m cheering for some nice warm weather over the next few years. We’ll get the message across!

    “AGW works mostly in winter. This is what winters are going to be like!”

  49. 499
    Hank Roberts says:

    > the data does appear to oscillate about the linear increase.

    Ah, “does appear” — such careful language.

    Anyone can claim something appears to be there.

    Dan H. presumes a linear increase, and presumes an oscillation.

    Neither presumption holds up — those aren’t in the data.

    Tamino, last November, quoted Dan H.’s misstatements about this, and did the math.

    Readers should examine those ideas closely.

  50. 500