RealClimate logo

Fun with correlations!

Filed under: — gavin @ 9 May 2007 - (Türkçe)

We are forever being bombarded with apparently incredible correlations of various solar indices and climate. A number of them came up in the excoriable TGGWS mockumentary last month where they were mysteriously ‘improved’ in a number of underhand ways. But even without those improvements (which variously involved changing the axes, drawing in non-existent data, taking out data that would contradict the point etc.), the as-published correlations were superficially quite impressive. Why then are we not impressed?

To give you an idea, I’m going to go through the motions of constructing a new theory of political change using techniques that have been pioneered by a small subset of solar-climate researchers (references will of course be given). And to make it even more relevant, I’m going to take as my starting point research that Richard Lindzen has highlighted on his office door for many years:

That’s right. Forget the economy or the war(s), the fortunes of the Republican party in the US Senate are instead tied closely to the sunspot cycle.

“Oh yes”, the sceptics might say “but that’s just a couple of cycles and doesn’t use up-to-date numbers. What happens after 1986?”

Well, that is a little problematic, however, the good early correlation is obviously still important (r=0.52! 1960-1986) and so we should be able to refer to it over and over again without noting that it breaks down subsequently (cf. Svensmark, 2007 referring to Marsh and Svensmark (2000)). But more importantly, it just demonstrates that the theory needs a little adjustment.

Let’s look at the second half of the record. Well, there’s another strong correlation for that period as well (r=-0.63, 1988-2006). Only this time the correlation is inverted, but that shouldn’t be surprising to anyone – solar-senator effects are complicated!

If we now put it all together, we can see that there is a reasonable match over the whole period…. well, except that break in the period 1984 and 1988 and, unfortunately, last year’s elections didn’t fit the pattern either. But 1984-1988 was Ronald Reagan’s second term and clearly no theory of Republican senators can ignore that. We therefore propose that the ‘Ronald Reagan second term phase shift’ combined with the change of sign of the Hale solar magnetic cycle in 1986, obviously changed the dynamics. This kind of phase shift is frequently seen in solar studies (cf. Landscheidt and many others), where it is rarely taken as a sign that two time series with decadal spectral power are in fact completely independent. Finally, it is permissible to leave off the more recent data points (cf. TGGWS) for “graphical convenience”. So after just a little work, we have managed to rescue the original theory to match a much longer amount of data:

Some readers may scoff and suggest that in the absence of any mechanism, these powerful correlations are numerological artifacts arrived at using post hoc fallacious reasoning that have no predictive capability. That might appear to be a valid argument. However the ultimate test will of course be experimental. On the basis of these intriguing results, we propose exposing Republican senators to varying levels of cosmic rays in a basement and monitoring their electability. Any refusal by the funding agencies or ethical review panels to support this would simply be confirmation that the political science establishment are scared of what this research would imply for their so-called “consensus”.

Convincing, eh?

The data for sunspots and senators can, I’m sure, be manipulated even more effectively than I’ve done here. I’ve made no use of various lags or filters (which can be altered as you go along cf. Friis-Christensen and Lassen (1991)), or of partial detrending (cf. Marsh and Svensmark (2003)), or of splicing of unconnected data sets (cf. Svensmark and Friis-Christensen 1997, Nir Shaviv). More ideas could be taken from “New evidence for the Theory of the Stork” (Höfer et al, 2004)”. A special RealClimate commendation for anyone who can do better!

356 Responses to “Fun with correlations!”

  1. 51
    Matti Virtanen says:

    Great post! And what a debate – I especially love the pirate vs. temperature correlation, a graph of which I have on my office wall. Maybe you should change the name of the blog to Real Humor.

    [Response: Maybe, but I’m not going to give up the day job… – gavin]

  2. 52
    Jim Manzi says:

    Re: 38, 16, 45

    Re: 38, Bruce:

    Thanks for the excellent post. I agree that we are discussing a spectrum of theory validation ranging from replicable tests of a roughly isolatable phenomenon like dropping a ball in a near-vacuum to test F = MA at one end and something like econometric models predicting global GDP growth at the other.

    I find your astronomy analogy interesting and useful (but I used to do astrophysics, so I’m probably biased!). The big distinction, I think, between astronomy and climate modeling is that there are many, many “natural experiments” in the observable universe. Therefore, while we can’t conduct controlled trials (as you say, the scientific gold standard for asserting causality), there is an opportunity to make numerous observations of different events and compare them to post hoc controls.

    I agree that “doesn’t meet the highest asserted level of possible scientific proof” does not equal “valueless”, and that it would be foolish to ignore the information provided by climate science just because it is imperfect, though I do think that trying to understand this uncertainty, and ideally bound it, is an incredibly important activity to support policy development.

    Re: 16, Margo / Gavin:

    I think it’s only fair to include Hansen’s own review of the predictive performance of these models in the NAS paper that he authored in 2006:

    Hansen’s take, expressed in this article, was that “a 17-year period is too brief for precise assessment of model predictions”.

    Gavin, we’ve had this discussion before, and I know that you (reasonably) draw a distinction between an assessment and a “precise assessment”. I won’t (re-)bore everyone with a repeat of the model evaluation discussion. Margo, you strike me as a very sophisticated analyst of model performance, and I’m sure you’re competent to read the base papers and come to your own conclusions.

    Re, 45, Don:

    That strikes me a great, commonsense point. An issue, though, arises if the series under consideration is autocorrelated. Such a series will usually have sub-series trends (e.g., temperatures increasing) that will run for some amount of time and then reverse to create a sub-series running in the opposite direction, rather than simply being random walks. This kind of a series – think of certain parts of the stock market – is notorious for fooling people into thinking they have built predictive models that are “right” for a while until they are suddenly, unexplainably wrong.


  3. 53
    Bryan Oakley says:

    Gavin great comment.
    Similar to the global warming/pirate graph circulating as well.

    enjoyed listening to you on “TWIS” as well.

    Keep up the good work!

  4. 54
    Chris Shaw says:

    Hansen et al 1998 predictions seem to be off now by a factor of 2.

    His scenario A prediction was for a 1.0C increase (from 1958 to 2007) and his scenario B prediction was for a 0.8B increase.

    Actual emissions have increased pretty close to the scenario A level although the trace gas assumption for scenario B is closer to what happened with that group.

    Temperatures have only increased by 0.4C (lower than scenario C in which GHG concentrations stabilize by 2000.)

    So it is good that a prediction was published but the models should adjusted now that we know they are off by a factor of 2. I’m assuming that has happened.

    [Response: This is incorrect. The projections were from 1984 onwards and scenario B, which was within 10% of the actual forcings over that period were 0.23 +/- 0.06 deg C/decade compared to observations of 0.23+/-0.04 or 0.20 +/- 0.03 deg C/decade (different datasets). Error bars are just for the linear fit -i.e. the weather noise. Under no circumstance can you describe that as a factor of 2 error. This is however something that is not well appreciated and so I will do a post specifically on this at some point soon (including all the numbers so you can test it for yourself). -gavin]

  5. 55
    Timothy Chase says:

    Re #11

    Observational vs. Historical Sciences

    You might hear practitioners of “historical sciences” refer to their science as “historical,” or practitioners of “observational sciences” refer to their science as “observational,” where one is able to “control the variables” in the observational sciences. At this point, one should point out that even the term “observational science” is itself fairly misleading, at least to a novice. “Observation” is normally thought of as being passive, but in this context it is being used to refer to active methods rather than passive methods of study, suggesting what is in fact the opposite of what it is intended to imply.

    The distinction has some utility in terms of the division of cognitive labor between “historical” and “observational” sciences, although it should properly be understood as not as a difference in kind but degree, with different sciences or scientific theories lying along the same continuem. However, creationists (for example) almost inevitably try to treate the distinction as a dichotomy, then claim that any science which is observational is real science consisting of knowledge, whereas any historical science is simply a matter of belief no different from matters of faith.


    What must be remembered is simply that was we are dealing with is a continuum where the the most essential distinction which is being made is between active and passive methods of discovery, or more accurately, the degree to which one is actually able to manipulate the objects objects of study. Beyond that, some sciences are more historical than others, but that too is a matter of degree. Additionally, the subject matter of “observational” and “historical” sciences will overlap, with principles from “observational sciences” being appealed to by the “historical sciences,” or the so-called “observational science” of sub-atomic particle physics speaking of deep-time phase transitions which broke the symmetry between the four fundamental forces, further blurring the distinction. Then there is the even more basic point that when refering to earlier experiments in which one “controlled the variables,” in logic it is still possible that there are some variables that one didn’t account for and thereby hold constant.


    For an example of where it gets properly applied, you might look at:

    Edwin C. Allison Center for Historical Science

    … a center which is devoted to the “historical sciences” of “paleontology, paleoclimatology, geochemistry, sedimentology and organismal biology.”

    Outside of the lingo of specialized disciplines, there are usually only two sets of occasions in which I hear of the distinction itself: one is when the distinction is being applied only in order to demonstrate that it is not a dichotomy or even particularly helpful, and the other is when creationists use it to try to argue that evolutionary biology isn’t hard science, or that for example, the soft historical science of evolution might argue for life having a natural origin, but hard observational science demonstrates otherwise.

    It has a basis in the philosophy of science, but this is of little more than historical value, much like reading Descartes in a standard philosophy course. For one thing, there is no hardfast dichotomy between observational and historical science: all science is a continuum. Difference branches of science use both “observational” (by which the speaker actually means “controlled experimental”) procedures and historical methods: for example, while evolutionary biology might be thought of as essentially historical, we can observe how viruses and bacteria mutate into new species and even affect their environment so as to bring about this change, and may occasionally observe speciation at the multicellular level, particularly in plants, such as the creation of species through an active process of hybridization or polyploidy.

    Likewise, astronomy would seem to be “observational” insofar as we “observe” stars and galaxies, but it generally cannot be performed in a lab, we are usually unable to set up experiments where we control the variables and thus manipulate the object of study, and what we are actually observing is what took place thirteen billion years ago or eight minutes ago in the case of the sun. So in this sense, it would it seem to be historical. But is it? At one point the moon was something we could simply observe from a distance — but now it is a place we can visit, we send probes places like Titan or Europa. Moreover, when one says that “observational sciences” make predictions in the sense that they are with regard to future events, these are oftentimes passive in the sense that one does not control the variables as one might in a lab. Furthermore, even historical sciences using “historical” methods make predictions of a sort: postdictions where they predict things which will be found.


    Empirical science itself is a unity because reality is a unity. There exists degrees of justification, but greater and lesser degrees of justification exist throughout all of empirical science, particularly if there is any area of active study, and no one branch is truly privileged over another. There necessarily exists a cognitive division of labor due to the limitations of individual human awareness, but the criteria employed for establishing the permeable boundaries between disciplines are ultimately pragmatic in nature.

  6. 56
    James says:

    Re #42: [I can do you a theory which ties mineral oil production to isotope signals — I don’t think the conventional CO2 theory does this very well as isotope signals appear in 1850 and AFAICT the real effect of greenhouse warming doesn’t show up until later.]

    Before the early 20th century, virtually all the isotope signal would have been coming from coal, not oil. The amount of greenhouse warming depends on the total amount of CO2 in the atmosphere, not on the amount being emitted at any given time. So you emit a little bit each year (and keep increasing the size of that bit), and over 150 years, it adds up to enough to produce measurable effects.

  7. 57
    Donald E. Flood says:

    #52: Jim,

    “Time will tell,” I suppose. IMHO, this is a case where Pascal’s Wager would be appropriate! Agreed?


  8. 58
    Margret says:

    Does not matter which party is in power, it is the people who do not get elected every four years that run the Governemnt like the tri laderal commisioners and the Generals in the Armerd Forces and ofcourse the Pentagon

  9. 59
    Hank Roberts says:

    While y’all are having fun with the weird correlations, I’d also ask that you go down the hall (or over the web) to the other NASA-related groups that are doing solar cycle work and see if you can get a thread going on watching what effects can be observed on and near Earth — because prediction is interesting, and because predictions of a small effect followed by observations that can test whether the effect is indeed present and of the expected order of magnitude are fascinating.

    Who else could we trust, eh?

    If you look at these two NASA links, the latter for the next cycle, the former for the one after that — they’re expected to be quite different. If you all can get some info from the people watching whatever measures might correlate in interesting ways and make it an ongoing thread (or point us to one elsewhere) it’d be good to watch the science as it’s actually being done. (Updated 2007/05/03)

  10. 60
    Trent says:


    Re #16

    You offer Hansen, et al. 1988 as evidence of the predictive ability of climate models. In that paper, the predicted temperature rise from 1987 to 2007 is about 0.55 ºC. This is for scenario A (continued increase in GHG emissions), which is what the world has followed since then. In fact, the actual temperature rise has been about 0.1 ºC.

    I do not see how this paper supports your point.

    [Response: See inline response above. Scenario B is the relevant one since that was both the ‘most plausible’ (and described as so at the time), and had net forcings that are with 10% of the actual forcings. The projection was from 1984 and temperature trends are around 0.2 deg/decade in both obs and modelling. -gavin]

  11. 61
    Steve Horstmeyer says:

    RE #55

    “The amount of greenhouse warming depends on the total amount of CO2 in the atmosphere, not on the amount being emitted at any given time.”

    Careful about the semantics of this argument. If all emitted carbon dioxide (and other greenhouse gasses) could easily be absorbed and then sequestered by Earth’s oceans and other systems would there be any measurable global warming due to GHGs?

    Is it not true that the component of global warming due to GHGs occurs because the rate of emission of GHGs exceeds the rate at which the GHGs can be sequestered?

    The global average temperature at a point in time is a function of the concentration of GHGs while GH warming (the change of the global average temperature) is a function of the accumulation of GHGs.

    Any one have any thoughts on this?

  12. 62
    Hank Roberts says:

    Steve, I think you’re confusing the greenhouse effect — which keeps the planet warmer than it would be without the greenhouse gases —- with “global warming” (increase in the planet’s temperature until a new equilibrium is reached, that follows the change in the level of greenhouse gases above the baseline).

  13. 63
    tamino says:

    Re: #61 (Steve Horstmeyer)

    The global average temperature at a point in time is a function of the concentration of GHGs while GH warming (the change of the global average temperature) is a function of the accumulation of GHGs.

    This would be true if the planet responded to GHG forcing instantaneously. But because of the thermal inertia of the oceans, it takes decades for the full response to a given GHG level to be reached. According to Hansen’s research, we have another 0.6 deg.C warming already in the pipeline, even if GHG gas levels instantly stabilize today.

  14. 64
    Ulric von Bek says:

    ‘Excoriable’? Surely you mean ‘execrable’.

  15. 65
    Timothy Chase says:


    The rate of greenhouse warming at any given moment will be a function of the amount of greenhouse gas that is in the atmosphere and the current temperature (neglecting the spatial distributions, convection, etc.), not the rate at which greenhouse gases are emitted by sources or sequestered by sinks. However, if you wish to calculate these interdependent variables over time, ideally you would be employing partial differential equations, tracking the system to its new equilibrium state – but this is highly unrealistic. Absent the differential equations or exact solutions to them, you would employ numerical models (distributions and all), calculating each state in a sequence of states from the previous state – which is I believe how it is actually done.

  16. 66
    fieldnorth says:

    The article Marsh and Svensmark 2000 requires a subscription, so I’m not sure what the object of your attack is.

    In the PDF Fig 3. there’s a correlation of low cloud cover with two sunspot cycles. Is that what you were referring to? If so what’s the problem with it?

    [Response: The fit to the second cycle is solely because of an added trend to the cloud cover data which has no justification in anything. See this comment – gavin]

  17. 67
    Steve Horstmeyer says:

    RE #62

    I think we are all saying the same thing. That is we all understand the concept but how we say it either clarifies or confuses.

    My point is that for any given time period the concentration of GHGs determines how much infrared radiation is retained by the atmosphere. During the next time period if the concentration of GHGs increases, all other factors being equal, the temperature can be expected to be higher. Both time periods have a greenhouse effect but the second time period has an enhanced greenhouse effect, because of the accumulation of GHGs.

    So the change of greenhouse effect is “global warming” in this example. If we detected no increase of average temperature there would still be a greenhouse effect but no global warming.

    So global warming ocurs over time because during that time GHGs accumulate.

    RE #63
    I have to think about your comment. Nothing happens instantly and because the response time of the climate system has a measurable lag doesn’t seem to change my thinking.

    I made the point because I was listening to a “progressive” radio talk show (The Ed Schultz Show) as I was reading the comments here and a caller made several points including citing Milankovitch (“…that name I cannot pronounce…”, he said),Hertzberg (the explosives expert) and the lag of the build up of cabron dioxide after warming begins. He had many snippets of information but was confused by it all and certainly did not have a coherent picture of the various processes. From his tone though I thought he was earger to understand. I therefore posted my comments to make the point that clear language is essential if the public is to be educated on global warming.


  18. 68
    Lawrence Brown says:

    Very Interesting! As a registered democrat(in the U.S.) I would like to see a minimum of sunspot activity in the coming decades for the sake of our planet. This analysis corroborates the old saw ” there are liars, damned liars and statistics”. It seems as though there are lots of the two former categories on the skeptic side of the aisle.

  19. 69
    DocMartyn says:

    “But because of the thermal inertia of the oceans, it takes decades for the full response to a given GHG level to be reached.”

    Well there seems to be an easy test, take a pair thermally isolated tank of salt water, some 30 meters deep and irradiate the surafce, open to the atmosphere with additional IR source masked by a CO2 shroud. Measure the temperature of each tank at half meter depths and monitor any temperature difference. You might also wish to monitor the depth of the tank.
    This model is a very good test for the impact of elevated levels of CO2 on the ocean temperature.
    If the surface of one of the tanks is blasted with an additional 8w/m2 of IR radation, what will be the steady state temperature difference in the two tanks after a year?
    My estimate is as close to zero as the noise allows, although the debth will have changed.
    What is your prediction?

  20. 70
    John Wegner says:

    In reponse to #50 and #60 Gavin, you note that from Hansen et al 1998 – “net forcings are within 10% of the actual forcings”

    What exactly does that mean?

    I note that the emissions have grown somewhere between Hansen’s Scenario A and Scenario B.

    Scenario A’s forecast had temperatures increasing 0.85C from 1984 to 2006 and Scenario B had temperatures increasing 0.6C from 1984 to 2006.

    By my reckoning, the Hadley Centre’s global temperature estimate is an increase of 0.25C from 1984 to 2006.

    So the forcing estimate was within 10% but the temperature prediction was off by more than 100%??

    [Response: This isn’t going to be clear until I show the figures and analysis, so give me a few days to put it together properly. The numbers are as I stated above. – gavin]

  21. 71
    James Killus says:

    There is another point about statistics that seldom gets mentioned and is widely misunderstood, and that is that attribution of variance is not the same thing as an attribution of a trend.

    If one does an analysis of continuously measured temperatures, for example, the strongest terms would be the daily and seasonal cycles. Whether it is day or night, summer or winter, are the strongest determinants of temperature. If one did such an analysis even over a hundred years of steadily rising temperatures, that would still be the case. So then there would be some that dismiss the temperature rise as being irrelevant to the temperature outside; after all, time of day and season of year are the main contributors to variance. So why even consider that there is a warming trend?

    Actually, I’ve seen this sort of reasoning used pretty regularly on hurricane data. Cycles dominate the variance statistics, so trends must not matter.

  22. 72
    Bruce Tabor says:

    Re 52 Jim,

    Thanks for the kind comments and the reference to the Hansen paper, which is fascinating, particularly with it’s warning of increased likelihood of El Ninos, a major concern in our part of the world (Australia).

    I agee with your comments. I have a chemical engineering – biomedical engineering and biostatistics background, the last of which often has me asking about the quality of evidence.

    To take a medical example, we don’t have experimental evidence (the gold standard) that smoking causes lung cancer in humans (imagine trying to enrol people in an randomised trial). Yet the observational evidence (from “natural” experiments) is so overwhelming that only a tobacco industry executive could fail to see it.

    The science of climate change is perhaps one rung lower on the quality of evidence scale, but the moral issues are overwhelming. In essence we have an N=1 observational study. But unlike the smoker, the outcome of this natural experiment has significant likelihood of harming those with little or no control over the “manipulated variable”.



  23. 73
    Jimmy says:

    Just saying,
    The correlation of sunspots to global temp actually fits much better than this, especially over time.

    [Response: For very similar reasons…. -gavin]

  24. 74
    Bruce Tabor says:

    Re #69 Doc Martyn,
    “Well there seems to be an easy test, take a pair thermally isolated tank of salt water, some 30 meters deep and irradiate the surafce, open to the atmosphere with additional IR source masked by a CO2 shroud. Measure the temperature of each tank at half meter depths and monitor any temperature difference. You might also wish to monitor the depth of the tank.
    This model is a very good test for the impact of elevated levels of CO2 on the ocean temperature.
    If the surface of one of the tanks is blasted with an additional 8w/m2 of IR radation, what will be the steady state temperature difference in the two tanks after a year?
    My estimate is as close to zero as the noise allows, although the debth will have changed.
    What is your prediction? ”

    I’ll do a crude calculation to give you my prediction:
    Assume the area of the 30m tank is 1 m2. First the energy added:
    8w/m2 in one year is 8 J/sec *3600sec/hr+24hr/day*365 days/year = 252 MJ
    Now the water’s response. The sepcific heat of water is about 4.2 J/g.C, which is 4.2 MJ/tonne.C. 30 metres of water with cross section 1 m2 is 30 tonnes.
    Assuming the water is well mixed and the heat distributes evenly it will take 30*4.2=126 MJ to raise the water by 1 degC
    So with 252 MJ we have a temperature rise of 252/126=2.0 deg C.
    Sounds like a good high school experiment. You could measure it with a lab thermometer. Certainly well above the noise assuming you can properly isolate the test and control tanks.


  25. 75
    Eli Rabett says:

    In 70, John Wegner states:

    “Scenario A’s forecast had temperatures increasing 0.85C from 1984 to 2006 and Scenario B had temperatures increasing 0.6C from 1984 to 2006.”

    To understand the kind of sophistry John is practicing you have to look at the graph of the predictions vs. the obervational record.

    1984 in the model was a very low point, due to variability in the model and a bit from El Chichon. By measuring from a low point, John exaggerates the predicted increase in temperature.

    “I note that the emissions have grown somewhere between Hansen’s Scenario A and Scenario B.”

    I challenge John to document that. He could start from the figures and the discussion here and here and here which discusses methane in detail Then he might want to go on with with a discussion of how A and B (and C until 2000 when it goes flat) nail the CO2 forcing. He might also want to discuss how the difference between forcing in A and B (and C until 2000) is principally in the trace gases with A having more forcing than B after 1990. In FACT B somewhat overestimates ACTUAL emissions/forcing in the principal trace gases, CH4 and CFCs.

  26. 76

    It feels ominous, somehow, that the number of Republican senators actually has an impact on the sun…

    When it comes to GW, however, you may only look at the number of cell phones in the world and the change in global temperatures over the last 20 years to realize what’s going on. Cell phones emit microwaves. Microwaves heat water. You connect the dots.

  27. 77

    [[The reduction of trend is because the Southern Hemisphere is warming less rapidly than the Southern Hemisphere]]

    <snark>Parallel worlds must be involved.</snark>

  28. 78
    Marian says:

    Gavin, are you 100% certain that CO2-emissions from humans are the reason for Global Warming?

    I have trouble understanding your views since you report and write about all things so completely unbiased and devoid of a particular agenda.
    I need to see the black & white of this, since all political decisions regarding GW are very particular and decisive, and affects so many people, and more.. Right?

  29. 79
    Dan Johnson says:

    By the way, a conspiracy to keep the “Milankovitch effects” out of North American textbooks may have been unearthed, as noted in a presentation before the US House Subcommittee on Climate Change on March 20, 2007. Tim Ball, a retired geographer, has discovered the importance of this factor and noted that it is not included in most textbooks. This may explain why you have probably never heard of the Milankovitch effect, unless you have taken an introductory course, or read a book, that included discussion of ice ages, climatology, paleoclimate, geography, geology, or related issues, or unless you are aware of how to use libraries or the internet.

    Ball also noted that the cosmic effect is largely due to “the gravitational pull of the planet Jupiter”, which apparently is one of the major forces that pulls earth’s orbit into “an extreme elipse”, although luckily it is currently almost circular, whew.

    Tim Ball’s testimony -US House Subcommittee on Climate Change,

    From the talk (which begins with ice ages, etc.):


    “… The explanation for that melting is primarily given by these factors, which is called the Milankovitch effect, and interestingly enough, this is not included in most of our textbooks across North America today, I’ve checked them out. What it shows in the lower right, is the orbit of the earth around the sun, as an almost circular but slightly eliptical orbit. That’s the situation right now. But the orbit is changing every single year, pulled by the gravitational pull of the planet Jupiter, and what you see on the lower left is the orbit of the earth as it was 22,000 years ago, an extreme ellipse. So the orbit is changing every single year. And in the center of the diagram you see that the tilt is shown at 23 and a half degrees. It isn’t; it’s just close enough for government work, but it also constantly changes from 21.4 to 24.8, …” etc., etc.

    end quote

    Powerpoint Slides:

    Could this choice of the Universities and publishers to keep mention of Milanokovitch cycles from us, combined with sloppy government measurement of the earth’s tilt, be the real reason that climate models appear so compelling?

    Elsewhere, Ball explains the rest of the effect with this observation:

    “TB: Basically, when sunspots are active, the earth is warmer and, when they are less active, it is colder. “

    It is too late to tell them about the secret Milankovitch effect, but the clear correlation of climate warming with sunspots should not be kept from students. So, I’ve posted the NOAA data in a spreadsheet for them:

    (PS. Ball doesn’t mention whether the effect of Jupiter altered by the impact of Comet Shoemaker-Levy 9, but if I ever again have the unpleasant experience of being the speaker following him at a public meeting, I will put this bug in his ear at coffee time.)

  30. 80
    PHE says:

    Gavin: I agree with you 100%! Spurious correlations are the enemy of good science. Let’s hear more.

  31. 81
    B Buckner says:

    Re: Eli #75
    Eli – that is good stuff you referenced in post 75. I am a bit new to the site and have not seen such data and analysis before. Interesting to see how this plays out, and I await Gavin’s post on the Hansen predictions. One question, any reason the observed temps stop at 1998? Temps have dropped a bit since and seem to diverge from the Hansen predictions.

  32. 82
    Nick says:

    Entertaining article, Gavin. I actually like use of humour. Cheers!

  33. 83
    Lawrence Brown says:

    Here’s what the British Meteorological Office( has to say about solar activity and climate change:

    “There are many factors which may contribute to climate change. For example, over the last million years most of the long-term changes in climate were probably due to small but well understood changes in the Earth’s orbit around the Sun. Over much of the last 1,000 years most of the variability can probably be explained by cooling due to major volcanic eruptions and changes in solar heating.

    However, the situation in the 20th century is more complicated. There is some evidence that increases in solar heating may have led to some warming early in the 20th century, but direct satellite measurements show no appreciable change in solar heating over the last three decades. Three major volcanic eruptions in 1963, 1982 and 1991 have led to short periods of cooling. Throughout the century CO2 increased steadily and has been shown to be responsible for most of the warming in the second half of the century.

    The final piece of the jigsaw is that as well as producing CO2, burning fossil fuels also produces small particles called aerosols which cool the climate by reflecting sunlight back into space. These have increased steadily in concentration over the 20th century, which has probably offset some of the warming we have seen. Only when all of these factors are included do we get a satisfactory explanation of the magnitude and patterns of climate change over the last century.

    The bottom line is that changes in solar activity do affect global temperatures. However, what research also shows is that increased greenhouse gas concentrations have a much greater effect than changes in the Sun’s energy over the last 50 years.”

  34. 84
    CC says:

    Tim Ball’s argument is simply amazing. (see post 79). Has he read ANY intro textbook? All the standard intro texts discuss Milankovitch orbital cycles in their discussion of ice ages. Given the long time scales of the rather modest orbital forcing, I’m not sure how they could explain substantial warming on decadal time scales. Then again, he “nicely” conflates Milankovitch forcing with sunspots.

  35. 85
    James says:

    Re #78: [Gavin, are you 100% certain that CO2-emissions from humans are the reason for Global Warming?]

    I’m not Gavin, but a simple answer from a non-expert is that, like a lot of people, you’re getting things backwards. CO2 has an insulating effect – we know this because people have measured it. Humans have been putting more CO2 in the atmosphere – we know this too, because we’ve measured it. Putting those two facts together, we deduce that the Earth should get warmer as a result of the CO2. We look at temperature records, and lo and behold, we see that the Earth is in fact getting warmer.

    Where exactly is the problem in this? It seems really simple to me, about at the level of figuring out that you’re feeling too warm because you’re wearing an extra sweater :-)

  36. 86
    stuart says:


    Aha, but you are just being fooled by a coincidence. If fact while in a lab CO2 will absorb energy and act as an insulator, when it is in the atmosphere pixies cast magical spells on the CO2 so that the radiation passes straight through it. Of course we would have realised this long ago if it wasn’t for the fact that, completely coincidentally, as human CO2 output has risen unicorn wave emission from the sun has increased at the same time. Now unicorn waves have the strange property that they curve away from detectors (they are shy), so the satellites and other systems that monitor the sun continuously don’t see them, and this is actually what has caused the warming.

    Of course on its own this would still be detectable, as the upper atmosphere would be warming up as well as the lower parts, showing us that the GHG model of climate was flawed. However by complete fluke there has been a massive increase in cold dragon activity in the upper atmosphere in recent decades. These dragons of course keep themselves completely invisible to humans, so we can’t tell that the are massively cooling the upper atmosphere, more than offsetting the warming of the unicorn waves, and hence making us unaware of the actions of the pixies at messing up our understanding of how physics works in the real world.

  37. 87
    Jonathan Gilligan says:

    Nice work. See also, “Political Innumeracy: Encounters with Coincidence, Improbability, and Chance,” by Carol Mock and Herbert Weisberg, Am. J. Pol. Sci. 36(4) 1023-46 (1992), which discusses the effect of astrological birth sign on political party affiliation. Perhaps there’s a Ph.D. dissertation in unscrambling the connections between astrological sign, control of senate, and climate. Then we could shortcut the expensive and laborious scientific research that goes into understanding climate and just consult a handy horoscope instead. As all the Republican hopefuls for 2008 are scrambling to appropriate Reagan’s mantle, this would be a timely research topic.

  38. 88
    Chuck Booth says:

    Re # 78

    100% certainty? Black and white?

    I’m afraid very little new scientific knowledge comes with absolute certainty – rather it comes with degrees of confidence, colored in shades of gray.

  39. 89

    #79 Dan, Mr Ball has not been observing the sun for its spots lately, especially during the warmest winter in history just past, someone must also tell him that there wasn’t many spots. You may do him a favor, and perhaps he will revisit his “basic” sun spot correlation theory….

  40. 90
    Philippe Chantreau says:

    RE #81. Temps have not dropped since 1998. 1998 was exceptionally warm, the following few years were not quite as warm but still very consistent with the general trend if you exclude 1998. That same trend can still be seen up to 2005, which was almost as warm as the exceptional 1998, or equally warm depending what data set you look at. What is important to consider is that 2005 was warmer than 1997 or 1999. From what I read, 2006 is consistent with the trend as well, although I don’t know if all the data have been compiled (haven’t seen it yet). RC has analyzed this and Gristmill also has a good discussion on it.

  41. 91
    Eli Rabett says:

    In answer to M. B. Buckner #81. No the updated GISS global temperature series, shows that although 1998 was exceptionally hot due to a strong El Nino, the smoothed five year average continues to rise. The image next to the temperature series shows that the global rise masks much stronger local ones, esp at far northern latitudes. The images come from Global temperature change, James Hansen, Makiko Sato, Reto Ruedy*, Ken Lo*, David W. Lea, and Martin Medina-Elizade Proceedings of the National Academy of Sciences September 25, 2006, 10.1073/pnas.0606291103

  42. 92

    This was just Slashdotted:,1518,481684,00.html
    Don’t we still expect superstorms?

  43. 93
    cce says:

    I just checked the current and former Earth Science text books (by different publishers) used in the high school where I work, and both of them describe Milankovitch cycles as a cause of ice ages, although not mentioned by name.

    It’s just another example of Ball talking out of his posterior.

  44. 94
    Eli Rabett says:

    Thanks to B. Buckner, I went back and found the PNAS article which also had an updated comparison of observations, vs. Scen. A-C.

  45. 95
    Timothy Chase says:

    Marian (#78) wrote:

    Gavin, are you 100% certain that CO2-emissions from humans are the reason for Global Warming?

    Chuck Booth (#88) wrote:

    100% certainty? Black and white?

    I’m afraid very little new scientific knowledge comes with absolute certainty – rather it comes with degrees of confidence, colored in shades of gray.

    Quite right – and I have little doubt that Gavin would claim otherwise.

    However, when you have a large number of independent lines of investigation justifying a given conclusion, the justification that the conclusion receives is far greater than that which it receives from any one given line of investigation alone.

    Empirical science never reaches absolute certainty, but it can approximate it. Afterall, I suspect that you are reasonably confident that the ocean won’t suddenly reverse current trends and freeze within the next five minutes. Likewise, we are reasonably confident that neither carbon dioxide nor water vapor will suddenly become transparent to infrared radiation, or for that matter, that Planck’s law describing the spectrum of radiation emitted by a warm body will suddenly be suspended.

  46. 96
    Hank Roberts says:

    > Josef H. Reichholf

    Hmmm. His expertise is in mites that live in the feathers of birds, according to Google Scholar.

    He also has opinions. Has anyone read his book?

    Josef H. Reichholf
    Evolution. What is true? Facts and answers
    ISBN 978-3-451-05779-3
    The biblical story of the creation – big bang or intelligent design? Which laws does evolution follow? Which power is active in the cosmos and in nature? Is everything just chance or is there a plan behind it all? ….

  47. 97
    Steve Bloom says:

    Re #86: Very nice, Stuart. IMHO it’s perfect for use in public debates.

  48. 98
    Arvella Oliver says:

    My son reviewed Milankovitch cycles in his 8th grade AP Science course last month. He was so fascinated he talked about them for a week. My daughter got a simpler version in 5th grade. Ball clearly has no idea what’s going on with textbooks.

    Gavin, thank you for the excellent post. And I’d like to thank Ike Solem, Hank Roberts, Tamino, Eli, and Barton Paul Levenson. I always learn from your posts.

  49. 99
    Timothy Chase says:

    Rionn Fears Malechem (#92) wrote:

    Don’t we still expect superstorms?

    The expert they are undoubtedly getting this from is Hans von Storch who they interviewed only a couple of months ago:

    SPIEGEL: And what about the monster storms that will supposedly be rushing in our direction in a greenhouse climate?

    Storch: A false alarm, so far, even though it’s become warmer by almost one degree since the beginning of industrialization. According to the computer models, we do expect high winds in northern Germany to increase by one percent per decade. But this is such a weak phenomenon that we won’t even notice it at first.

    March 16, 2007,1518,472200,00.html

    It is only within the past few years that climate models have achieved the resolution needed to “see” the formation of hurricanes. The models currently predict a moderate increase in the strength of the the more severe hurricanes. In contrast, what we have seen is a fairly strong response to increased oceanic surface temperatures. Part of what appears to be happening is that the water which is now warmer extends further down than it did previously, further fueling the hurricanes – and as of yet we aren’t taking this into account in our models.

    In any case, Hans von Storch views climate change as largely inevitable, with whatever Germany does as having little effect, particularly given that the major players (the United States, China and India) are doing very little to stop it. He views mitigation as Germany’s best strategy. (See the interview cited above.) Given this, perhaps he has chosen to downplay the effects of climate change to some degree. Of course, what he says in Germany is heard in the United States…

    In any case, if you are particularly interested in this topic, you might want to check out the following posts:

    Hurricane Spin
    Michael Mann and Gavin Schmidt
    24 Apr 2007

    21 Cat 4-5 Storms for 2006?
    Posted on: January 4, 2007 6:55 AM, by Chris C. Mooney

    Unlike the previous few yeares, last year was slow for the Atlantic due to dust blowing off of north Africa reducing the temperature in that region, but it was busier than ever in the Pacific.

    You might also try the search box at the top: hurricanes are a recurring subject here.

  50. 100
    Richard Ordway says:

    Think people, seriously, …what would happen if Gavin…or anyone else submitted this “article” to the peer-reviewed Journal SCIENCE…Now how about if submitted to a “journal” that was funded by a special interest group? Could it be used politically?