RealClimate logo


Technical Note: Sorry for the recent unanticipated down-time, we had to perform some necessary updates. Please let us know if you have any problems.

A warming pause?

Filed under: — stefan @ 6 October 2009 - (Español)

The blogosphere (and not only that) has been full of the “global warming is taking a break” meme lately. Although we have discussed this topic repeatedly, it is perhaps worthwhile reiterating two key points about the alleged pause here.

(1) This discussion focuses on just a short time period – starting 1998 or later – covering at most 11 years. Even under conditions of anthropogenic global warming (which would contribute a temperature rise of about 0.2 ºC over this period) a flat period or even cooling trend over such a short time span is nothing special and has happened repeatedly before (see 1987-1996). That simply is due to the fact that short-term natural variability has a similar magnitude (i.e. ~0.2 ºC) and can thus compensate for the anthropogenic effects. Of course, the warming trend keeps going up whilst natural variability just oscillates irregularly up and down, so over longer periods the warming trend wins and natural variability cancels out.

(2) It is highly questionable whether this “pause” is even real. It does show up to some extent (no cooling, but reduced 10-year warming trend) in the Hadley Center data, but it does not show in the GISS data, see Figure 1. There, the past ten 10-year trends (i.e. 1990-1999, 1991-2000 and so on) have all been between 0.17 and 0.34 ºC per decade, close to or above the expected anthropogenic trend, with the most recent one (1999-2008) equal to 0.19 ºC per decade – just as predicted by IPCC as response to anthropogenic forcing.

GISS temperature trends

Figure 1. Global temperature according to NASA GISS data since 1980. The red line shows annual data, the larger red square a preliminary value for 2009, based on January-August. The green line shows the 25-year linear trend (0.19 ºC per decade). The blue lines show the two most recent ten-year trends (0.18 ºC per decade for 1998-2007, 0.19 ºC per decade for 1999-2008) and illustrate that these recent decadal trends are entirely consistent with the long-term trend and IPCC predictions. Even the highly “cherry-picked” 11-year period starting with the warm 1998 and ending with the cold 2008 still shows a warming trend of 0.11 ºC per decade (which may surprise some lay people who tend to connect the end points, rather than include all ten data points into a proper trend calculation).


Why do these two surface temperature data sets differ over recent years? We analysed this a while ago here, and the reason is the “hole in the Arctic” in the Hadley data, just where recent warming has been greatest.

Mean temperature difference between the periods  2004-2008 and 1999-2003
Figure 2. The animated graph shows the temperature difference between the two 5-year periods 1999-2003 and 2004-2008. The largest warming has occurred over the Arctic in the past decade and is missing in the Hadley data.

If we want to relate global temperature to global forcings like greenhouse gases, we’d better not have a “hole” in our data set. That’s because global temperature follows a simple planetary heat budget, determined by the balance of what comes in and what goes out. But if data coverage is not really global, the heat budget is not closed. One would have to account for the heat flow across the boundary of the “hole”, i.e. in and out of the Arctic, and the whole thing becomes ill-determined (because we don’t know how much that is). Hence the GISS data are clearly more useful in this respect, and the supposed pause in warming turns out to be just an artifact of the “Arctic hole” in the Hadley data – we don’t even need to refer to natural variability to explain it.

Imagine you want to check whether the balance in your accounts is consistent with your income and spendings – and you find your bank accounts contain less money than you expected, so there is a puzzling shortfall. But then you realise you forgot one of your bank accounts when doing the sums – and voila, that is where the missing money is, so there is no shortfall after all. That missing bank account in the Hadley data is the Arctic – and we’ve shown that this is where the “missing warming” actually is, which is why there is no shortfall in the GISS data, and it is pointless to look for explanations for a warming pause.

It is noteworthy in this context that despite the record low in the brightness of the sun over the past three years (it’s been at its faintest since beginning of satellite measurements in the 1970s), a number of warming records have been broken during this time. March 2008 saw the warmest global land temperature of any March ever recorded in the past 130 years. June and August 2009 saw the warmest land and ocean temperatures in the Southern Hemisphere ever recorded for those months. The global ocean surface temperatures in 2009 broke all previous records for three consecutive months: June, July and August. The years 2007, 2008 and 2009 had the lowest summer Arctic sea ice cover ever recorded, and in 2008 for the first time in living memory the Northwest Passage and the Northeast Passage were simultaneously ice-free. This feat was repeated in 2009. Every single year of this century (2001-2008) has been warmer than all years of the 20th Century except 1998 (which sticks out well above the trend line due to a strong El Niño event).

The bottom line is: the observed warming over the last decade is 100% consistent with the expected anthropogenic warming trend of 0.2 ºC per decade, superimposed with short-term natural variability. It is no different in this respect from the two decades before. And with an El Niño developing in the Pacific right now, we wouldn’t be surprised if more temperature records were to be broken over the coming year or so.

Update: We were told there is a new paper by Simmons et al. in press with JGR that supports our analysis about the Hadley vs GISS trends (sorry, access to subscribers only).

Update: AP has just published an interesting story titled Statisticians reject global cooling, for which they “gave temperature data to four independent statisticians and asked them to look for trends, without telling them what the numbers represented”.


456 Responses to “A warming pause?”

  1. 401
    Hank Roberts says:

    > When I copy a post from here and paste it into MS Word
    Use “Paste Special, Unformatted” (Safe Text) when pasting into Word.

  2. 402
    Ray Ladbury says:

    Tim Bone, Your post is sufficiently vague that it is effectively meaningless. What data? What sensors? What research? Vague accusations are not helpful and merely establish you as an ignoramus who is too unfamiliar with the research to be specific.

  3. 403
    Geoff Wexler says:

    Mark

    Misunderstanding?

    I don’t agree with parts of your comment. Definitions cannot be ‘wrong’. Considering them again:

    (a) Populist. Just one parameter= slope. No use for establishing that a stable change is occurring and yes to quote you, very poor for making extrapolations.

    (b). This includes the existence of an additional parameter (statistical sig. or error bars). Depending on the values of this parameter this addition can lead to a trend with useful properties.

    I did not attempt to justify why (b) can provide evidence of sustained change. The comment was never intended to reproduce the excellent discussion by Grumbine or the various versions by Tamino. If you haven’t yet done so, I suggest you go there before replying to me again, even if you think that you know enough about the topic. Unfortunately Tamino has hurt his hand and we have to wait for part 2 of his essay on the Kullback-Leibler divergence ; perhaps that may provide a better handle for separating out noise from a linear change. Why call it noise? Because it makes no progress and does not involve sustained change.

  4. 404
    Hank Roberts says:

    Of course definitions can be wrong, if you want to be understood.

    There are definitions of ‘trend’ that don’t mean a statistically detectable likelihood of a change. But they’re not used for good reason here.

    ——
    `I don’t know what you mean by “glory”,’ Alice said.

    Humpty Dumpty smiled contemptuously. `Of course you don’t — till I tell you. I meant “there’s a nice knock-down argument for you!”‘

    `But “glory” doesn’t mean “a nice knock-down argument”,’ Alice objected.

    `When I use a word,’ Humpty Dumpty said, in rather a scornful tone, `it means just what I choose it to mean — neither more nor less.’

    `The question is,’ said Alice, `whether you can make words mean so many different things.’

    `The question is,’ said Humpty Dumpty, `which is to be master — that’s all.’
    ——
    http://www.sabian.org/Alice/lgchap06.htm

  5. 405
    Geoff Wexler says:

    Re: #404 ; Hank.

    “There are definitions of ‘trend’ that don’t mean a statistically detectable likelihood of a change. But they’re not used for good reason here.” [My italics]

    First the italicised sentence is not strictly true ; see e.g. #121 which provoked my earlier criticism and also my final example below. Secondly I may have not made myself clear. I was not welcoming the use of one word (trend) for both usages , just the opposite. Finally, in the words of an earlier article

    “Don’t be such a scientist” .

    The problem is that the populist usage, which I called type (a) is used by many of the lay public (and most deniers) all of the time as well as some experts some of the time (often with caviats).

    Consider Figure 1 at the top. The term ’10 year trend’ happens to be correct because it is equal to the 25 year trend which can be justified statistically. On the other hand Stefan also points to the slope of a least squares fit to an 11 year set. Even if this fit is not stat. sig. his point would be worth making, but my point is that it would have been better if the the stat. sig. of the result 0.11 degs.C /decade had been quoted.

    Otherwise we cannot complain when deniers draw their own least squares fits and fail to tell us that they are stat. insignificant which they do .

  6. 406
    Hank Roberts says:

    > I was not welcoming the use of one word (trend) for
    > both usages, just the opposite.

    Thanks, that makes more sense, I hadn’t understood that.

    > 121

    There Lawrence quotes the MetOffice’s “Quick Guide” — I found the word “trend” there only once, though my search may be missing something

    http://www.google.com/search?q=site%3Ahttp%3A%2F%2Fwww.metoffice.gov.uk%2Fclimatechange%2Fguide%2Fquick%2F+trend

    Seems to me they have to use the word most people use to explain what’s happening in the world — then try to explain what the scientists are doing.

    > don’t be such a scientist

    Sorry; I haven’t gotten nearly good enough to quit trying yet.

  7. 407
    Deep Climate says:

    A new contrarian voice?

    Andrew Revkin interviews Canadian futurist Vaclav Smil, who claims that a looming (unspecified) pandemic is a much greater threat than global warming.

    Oh, and did you know that there has been “no global warming in the past ten years”, and James Hansen should have been able to predict that global warming would “basically stop for 10 years”?

    See:
    http://deepclimate.org/2009/10/19/vaclav-smil-no-global-warming-in-past-ten-year/

    Revkin’s highlihjts are at DotEarth (but don’t include the “no global warming” comments):
    http://dotearth.blogs.nytimes.com/2009/10/19/smil-on-hummers-hondas-meat-heat

  8. 408
    Mark says:

    Geoff: “Even if this fit is not stat. sig. his point would be worth making, but my point is that it would have been better if the the stat. sig. of the result 0.11 degs.C /decade had been quoted.”

    Uh, people don’t WANT the error bars.

    Denialists NEVER USE error bars. Though they do admit that the IPCC doesn’t tell everyeone about their errors, then they crow about how the IPCC has admitted to an error (without ever saying that they’ve just proven their earlier assessment wrong).

    The trend line DOES pass through 0.11C. Please prove that wrong.

    It’s impossible because the definition of the trend line is mathematical and the maths is based on all the data available used correctly.

    Missing out the error bars has been fixed before.

    Do you know what the denialists did?

    They took the lower bound and said “SEE! The IPCC themselves say there’s nothing to worry about” because the lower bound was less than zero at the time.

    When you’re dealing with simpletons, you have to make things simple.

    So please prove the 0.11 trend in the short period data is wrong.

    You can’t.

    You may be able to prove it’s not the actual trend, but you have to use more data than the short trend the denialists want to use and incorrectly (because the trend you get is positive) state it’s cooling.

    And if you DO use more data, you’d find the cooling is not on the cards at all.

  9. 409
    Mark says:

    “(a) Populist. Just one parameter= slope.”

    Nope. Wrong.

    Just one parameter made incorrectly is not a trend.

    You were talking about TRENDS not slope.

    If you want to talk about slope, you need two numbers: gradient and intercept.

    “(b). This includes the existence of an additional parameter (statistical sig. or error bars).”

    Nope, still wrong. You have error bars around an incorrectly calculated number. It still isn’t a trend. If you want a slope, you need two numbers for the slope and you’d need another number for errors. Except, since you have the WRONG trend, the error in the future number from the calculated slope that isn’t the trend will always be higher than the standard error you calculated on the graph and therefore your error bars are also wrong.

    The trend is “calculate the best idea of where the numbers in the future are going to go”. You can check your figures by having someone give you a smaller selection of numbers and then asking you to guess the value of the missing numbers.

    What are the missing numbers of this series?

    1, 2, ?, 8, 16, 32, 64, 128, ?, 512

    ?

    Now if you did YOUR “correct” calculation, we’d say the slope was 51.2.

    IS that the trend?

    Would that make the two missing numbers?

    Would the next number be between your calculated error bars or outside the three sigma limit? Heck, it’s probably outside the five sigma limit. If not, the number after that would be.

  10. 410
    Geoff Wexler says:

    Re: #407 and Vaclav Smil

    Your links are most informative, Re: Vaclav Smil.
    I don’t think you need a question mark at the top of #407.
    He has been cherry picking his sources of reading matter. It is just possible that his forecasts on non climatological matters might be less lazy, but I prefer to read experts who demonstrate that they can be trusted on important matters. So I shall not be buying his book. Unfortunately there might well be a disastrous pandemic but we don’t need to use him as a source of information on it.

  11. 411
    Mark says:

    Geoff, is the slope 51.2? Would any standard deviation hold for the next five numbers? If it cannot, how can it be the standard deviation?

  12. 412
    Mark says:

    PS there’s a difference between error bars and standard deviation.

    It’s one reason why climatology is 30 years and that a decade or more of flat lining is no proof of an end to global warming.

    Each year, the temperature varies from the year before by ~1C.

    THAT is an error bar. I.e. you could see 1C difference between any two measurements.

    After 30 years the average difference between the average from that sample and the true sample (given there is no trend) the standard deviation will be 1/sqrt(30) C. This is 0.2C and warming over this period is approximately 2-3 times that, so above the noise level.

    That 30 year figure of trend with standard deviation is not the trend with error bars. It is saying “the trend we get from the data we have is THIS, but the real trend could be anywhere from HERE to THERE”.

  13. 413
    Geoff Wexler says:

    Mark.

    #409 “When you’re dealing with simpletons, you have to make things simple.”

    First not all contrarians are simpletons, secondly not all insignificant trends are short term, thirdly climatological articles are not just addressed to contrarians, fourthly you should follow your own advice about simplicity.

    If you use a repertoir which includes an “actual trend”, a “short trend” (line 4 up) , an “incorrectly used trend” and a mathematically defined trend based on correctly used data, you will then need a few extra lines on each occasion to explain and justify what you talk are talking about.

    To really make things simple restrict the discussion to 25 years (say) or more. But that would be insufficient to rule out insignificant long term estimates based on other kinds of cherry picking. It would also ban Figure 1 above with its reference to 10 year trends. I originally suggested using two different words. If you don’t want to mention the stat.sig.or the error bars the very least would be a reference to a “statistically insignificant 11 year estimate of the trend”. I don’t like your ‘mathematically defined estimate’ , its almost meaningless and would open the way to the abuses which I have seen by various professors.

    #409. Yes I forgot to mention the intercept.

  14. 414
    Mark says:

    “#409 “When you’re dealing with simpletons, you have to make things simple.”

    First not all contrarians are simpletons,”

    Never said so. But are NO contrarians simple?

    No.

    So the simple explanation has to be made.

    The less simple explanations can work if you have a minimum pass level for intelligence.

    Blogs don’t do that too well.

    Proper journals do.

    Ta.

  15. 415
    Hank Roberts says:

    > “statistically insignificant ….”

    —–
    “Statistically Insignificant”? Watch out!

    Experts can detect non-experts just by the way they use technical terms. Something is said that no one trained in the field would actually say. It gets one’s teeth grating to hear it…
    —–
    http://financialrounds.blogspot.com/2005/06/statistically-insignificant.html

  16. 416
    Mark says:

    “I don’t like your ‘mathematically defined estimate’ , its almost meaningless and would open the way to the abuses which I have seen by various professors.”

    You don’t have to like it.

    It IS.

    But saying you don’t like the mathematical definition of trend doesn’t mean you can take two points and draw a straight line between them and call it a *trend*.

    Because a trend is defined mathematically.

    Now, is the slope of that graph y=51.2x?

    Or would both of your definitions of trend (which is a slope, not a trend) be wrong?

  17. 417
    Geoff Wexler says:

    Mark

    I do not have time to check those parts of your comments which have nothing whatsoever to do with my suggestions concerning the spec.or name of a trend. There is a danger that the number of misunderstandings on this sub-thread could diverge. You could always try displaying your non-linear convex sequence at another thread e.g that following Tamino.

  18. 418
    Geoff Wexler says:

    #415

    I plead guilty. Please don’t report me to the chief of police for statistics , especially as I referred to him earlier.

  19. 419
    Mark says:

    It does have a lot to do with how both your (a) and (b) options were wrong.

    There already IS a name for trend and a meaning for trend and as Hank points out, the meaning for trend in science and the meaning for trend in everyday use are the same.

    Neither requires error bars or standard deviation.

    And denialists aren’t creating a trend when they say “it’s been cooling for the last 10, sorry, 11 years!” they’ve created a slope between TWO years and called it a trend.

    Neither of your options, not even the option b you were pushing is a trend and would still be abused by denialists drawing a slope between two points and calling it “cooling trend”.

    It just isn’t a trend.

    And neither were your two options.

    Hence the name for those two different slope definitions are “wrong” and “also wrong”. Replacing the wrong definition of denialists by another also wrong definition of trend is not going to solve anything. We use the right one.

  20. 420
    CTG says:

    The point of all this is not whether you can find any trends within the overall data set that in themselves are statistically significant.

    You can take any 11 points from the overall data series, do a linear regression on them, and come up with a fit that you could call statistically significant. But is it?

    You have to bear in mind that those 11 points are not independent – they are part of a time series. The time series as a whole expresses two modes of variance. If you take a series of 11 points, then all you are looking at is the short-term mode of variance. You cannot say anything about the long-term variance of the whole data series, because an 11-year trend is not measuring the long-term variance.

    The shortest trend for which most or all of the variance comes from the long-term variance is 20 years. Bob Grumbine will soon be posting my Java applet that shows the effect of trend length. You can also see it at Hot Topic here.

  21. 421
    Geoff Wexler says:

    “Robust trend” and trend; That would be a possible compromise. Stefan used those terms in an earlier article.
    ——————–
    Re: #420. Thanks CTG I’ll look out for your work when I get the time.
    ————–
    Re: #419. The signal strength is falling to zero. You are not discussing my comments but a fantasy. Perhaps if we were in the same room we could be more efficient.

    I previously suggested that you go to Tamino or Grumbine. But I see that you have been to Open Mind already and dhogaza warned what can happen when the communication channel gets blocked:

    http://tamino.wordpress.com/2009/08/13/do-you-believe-ian-plimer/#comment-34285

    On that occasion one of the best articles about Plimer was followed by a thread which got diverted into 40 or more comments centred on a single trivial misunderstanding. When you avoid pitfalls like that you have some useful points to make.
    —————-

  22. 422
    John says:

    I love the “trend lines” when you have an R^2 of about .5!!!!! It means absolutely nothing – you draw a line through random numbers and then extrapolate out 50 years.

  23. 423
    Ray Ladbury says:

    John,
    You obviously don’t have much data analysis experience. An R^2 of .5 simply means you have a lot going on BESIDES the increasing trend. This is where physics comes in handy. You really ought to try it sometime–you know, actual science.

  24. 424
    Mark says:

    What’s trend, then? What’s the difference between robust trend? And how are these differences going to change anything?

    The trend is still the trend and it isn’t the “draw a line between two points 11 years apart and call it a trend”. Doing the calculation on all 11 yearly averages doesn’t produce a “robust trend”. It produces ***a trend***.

    A robust trend would be the trend you’d get from an 11 year trend (done properly) that you could expect to be replicated in the next 11 years.

    But that isn’t a trend and the draw-a-line-between-two-points is still not a trend and isn’t made a trend by calling the real trend “robust”.

  25. 425
    Mark says:

    “The shortest trend for which most or all of the variance comes from the long-term variance is 20 years.”

    CTG, and this falls out because of the known cyclical elements controlling weather but not climate.

    PDO, ANSO, Sunspots, etc.

    Now you can draw *some* inference of the long term result if you remove the effects you know change the weather but not the climate. It’s been done the other way around once before by denialists (that has been a long thread on a previous RC thread): they’ve removed the effect of CO2 and found that the majority of what remains is caused by the PDO.

    Denialists then drew the conclusion that the PDO was the main climate driver.

    Go figure.

  26. 426
    Hank Roberts says:

    > robust trend

    You can’t just pick words you like and decide you want to use them in explaining things to the public without looking up how they are being used in current practice, or you will make people even more confused about meanings.

    ‘Robust’ has a meaning. You need to understand what people mean by it.

    http://scholar.google.com/scholar?q=statistic+robust+trend

  27. 427
    Richard says:

    “It’s nice to know that a lot of people have been pushing efforts to make planet earth a better place. With biofuels, and all other alternative resources for energy, we might be able to curb the growing problem that is global warming and at the same time preserve the ecosystems of the world. The Earth’s natural forces such as wind, hydro and solar power are limitless and effective ways to gain energy. So unless the winds stop blowing, the seas stop roaring or the sun dies out, there’ll always be a place to go for energy. Kudos mother earth!”

  28. 428
    Patrik says:

    Well, here is one lay person who claims that you may use GIS Temp and start the trend calculation from any year during 2001-2007 to now, and get a downward trend.
    It’s not 10 years, but it sure is 8 years telling the same chilling tale for AGW.

  29. 429
    Sekerob says:

    Patrik says: 22 October 2009 at 12:47 PM

    Think you should draw the trend line based on monthly data from 1.1.2008. Might put some shiffers up your spine.

  30. 430

    Patrik:

    you may use GIS Temp and start the trend calculation from any year during 2001-2007 to now, and get a downward trend.

    It’s not a “trend” unless it’s statistically significant and you use an appropriate time scale. The appropriate time scale for climate is 30 years. Eight years tells you exactly nothing.

  31. 431
    Dan says:

    re: 428
    All of which is quite irrelevant to the long-term (i.e. well-defined as 30+ years) warming trend re: AGW. “…8 years telling the same chilling tale for AGW” is essentially a non sequitor.

    As has been said before, the idea that someone who is not an expert in climate science somehow knows something that literally thousands of peer-reviewed climate scientists and every major professional climate society/organization across the world do not is nothing more than scientific arrogance.

  32. 432
    Hank Roberts says:

    Patrik, this one lay person can’t be not more than 9 years old.

    Here’s the applet
    — move the slider and tell us how old you have to be:

    http://hot-topic.co.nz/keep-out-of-the-kitchen/

    Choose between data series with the button at the top. Blue lines show negative trends, red ones positive trends.

  33. 433
    dhogaza says:

    Well, here is one lay person who claims that you may use GIS Temp and start the trend calculation from any year during 2001-2007 to now, and get a downward trend.

    Here’s one lay person that notices that you’ve left out 2008. Why would that be? I couldn’t possibly guess.

    It’s not 10 years, but it sure is 8 years telling the same chilling tale for AGW.

    Yes, it tells us that CLIMATE SCIENTISTS ARE RIGHT then they tell us that the warming signal overlayed on natural variation in temps can lead to flat or even declining stretches of several years.

    This is why, decades ago, meteorologists settled on 30 years as being a reasonable length of time for discerning real climate trends.

    Not 1, 2, 3, 4, 5,6, 7, 8, 9 or 10 years. 30 years.

    Isn’t it nice to see another basic tenet of climate science shown to be true?

    Wouldn’t it be nice if you actually understood it? You can, it’s not that hard … click on the “start here” link at the top of the page and have at it. Learn some science, then come back and join in the fun.

  34. 434

    #431 Dan

    Actually, it is not arrogant to speak of something that is well known in ones field. You may be trying to point out that it is not considerate of the lack of knowledge that others have.

    That is an odd argument. One should not feel offended by not knowing, but rather happy to learn. Also, you should consider the fact that when you visit another country, the customs and the culture may be different. You don’t just barge in and tell them that they are all wrong for living the way they live. You observe and learn about the new environment and then it is easier to blend in. Asking questions is a good way to learn though. Making statements about how you know its cooling based on 8 years of data only shows that you are in the wrong country, or in the right country and speaking arrogantly about your own understanding of the culture.

  35. 435
    G. Karst says:

    Here is a nice video by Nasa illustrating how ice thickness varies with ice extent. Enjoy.

    http://www.nasa.gov/mov/391782main_sea_ice_concept.mov

  36. 436
    Mark says:

    “Here is a nice video by Nasa illustrating how ice thickness varies with ice extent.”

    Please.

    Show us where thickness is driven by extent.

    Where there was NO ice and now there is, well, obviously.

    But if that ice comes from a break off of another sheet of panice, it’s also thinned just as much (or more) somewhere else. Yet despite no new ice, the extent has increased.

    You have not yet shown that sea ice thickness depends on sea ice extent. And the movie makes no such connection either.

  37. 437
    G. Karst says:

    Mark:

    You are aware that “varies as” and “driven by” are two different statements. Maybe most of your comments are “driven” instead of “reasoned”.

  38. 438

    Mark: I’ve viewed the movie (and watched Tom Wagner’s associated interview) and it seems to be showing that the thickness of ice under the non-melted areas adjacent to areas that melt and re-freeze each year is thinned by the melting action.

  39. 439
    common_man says:

    Arctic ice extent and thickness is cyclical.

    n early May, a paper appeared in Nature that created quite stir in the media by showing how by including long term ocean cycles in models the recent global cooling or at least lack of warming may continue to 2020. The same week, a story by NASA’s Earth Observatory reported on the flip of the Pacific Decadal Oscillation to its cool mode. “This multi-year Pacific Decadal Oscillation ‘cool’ trend can intensify La Niña or diminish El Niño impacts around the Pacific basin,” said Bill Patzert, an oceanographer and climatologist at NASA’s Jet Propulsion Laboratory, Pasadena, Calif. “The persistence of this large-scale pattern tells us there is much more than an isolated La Niña occurring in the Pacific Ocean.”

  40. 440
    John Millett says:

    Evidence of global warming in the form “ n of the last m years have registered record temperatures” where n approaches m in magnitude is not very useful. It would be true for even the slightest warming trend and such a natural trend is to be expected at this stage in climate history. Measuring the magnitude of changes in temperature trends is more instructive. Trends in the temperature record between 1850 and 2007 changed direction five times. While the forces changing temperature trends from cooling to warming have weakened over time, those effecting the opposite change have strengthened. It takes more force to switch the climate system to a cooling trend than to a warming one. If emissions are the main climate force and if emissions have increased over time then they are a cooling force not a warming one.

  41. 441
    Ray Ladbury says:

    John Millett,
    Your post is just about the dumbest thing I’ve read on the internet this week. Congratulations. I would refute it, but looking over it, I see only a bunch of unsubstantiated assertions devoid of any evidence or logic, so you’ve already done my job for me. Thanks!

  42. 442
    Mark says:

    “Evidence of global warming in the form “ n of the last m years have registered record temperatures” where n approaches m in magnitude is not very useful.”

    Why not?

    When looking at whether there are unsafe practices at a construction site, they will use the number of deaths per worker-year compared with the industry average. A workplace with more than the last n deaths in the last m years is considered a priori unsafe and is investigated.

    “It would be true for even the slightest warming trend and such a natural trend is to be expected at this stage in climate history.”

    Only if your n and m is small and the n much smaller yet. But when the last 8 warmest years on record are in the last 10, then it’s pretty solid.

    But if you can prove this is not reliable, please do so.

    “Trends in the temperature record between 1850 and 2007 changed direction five times.”

    Far more than that. Roughly 314 times. Each march it started getting warmer. Each October it started getting cooler.

    “While the forces changing temperature trends from cooling to warming have weakened over time, those effecting the opposite change have strengthened.”

    The opposite change from what?

    “It takes more force to switch the climate system to a cooling trend than to a warming one.”

    Nope. It takes more energy. But that’s because it requires the addition of energy to raise the temperature of anything. Tautological BS.

    “If emissions are the main climate force and if emissions have increased over time then they are a cooling force not a warming one.”

    No they aren’t. And this statement does not follow on at all from your earlier sentences, so where you get this idea from is anyone’s guess.

    But if you can tell us how CO2 and methane emissions can become a cooling forcing, I’m all ears.

  43. 443
    Mark says:

    “Mark: I’ve viewed the movie (and watched Tom Wagner’s associated interview)”

    I don’t have sound here, so all I had was the images.

    Which didn’t show thickening of the ice when it expanded in extent.

    “You are aware that “varies as” and “driven by” are two different statements.”

    Yes I do.

    But what does that have to do with saying increased ice extent shows there’s more ice?

    If my comments are more driven than reasoned it’s because there’s no logic behind your arguments, therefore no logical basis to argue from in their relation.

    Ice extent measurements doesn’t measure ice volume.

  44. 444
    Hank Roberts says:

    hmmmm. There’s a John Millett of Neutral Bay who writes to The Australian a lot — same kind of thing, pure copypaste word salad stuff, e.g.
    http://www.theaustralian.news.com.au/story/0,25197,26226979-7583,00.html

  45. 445
    John Millett says:

    #441, #442, #444: Ray Ladbury, Mark, Hank Roberts: My method and numbers.

    The system switches from warming to cooling trends at maxima; and vice versa at minima. Trend change is the sum of the separate trends, measured as degrees Celsius per year.
    Temperature maxima: 1877, 1942, 1998; period: 65, 56 years
    Trend change: 0.0138, 0.0183, 0.0188 (assuming flat trend since 1998)
    Minima: 1909, 1974; period: 65 years
    Trend change: 0.0258, 0.0204
    Warming to cooling trend changes are rising over time; and vice versa. The two series are converging. They are likely to intersect within the next period, when cooling trend changes would exceed warming trend changes and temperature would peak. If rising emissions are contributing warming to this pattern, it must be minor. Alternatively, the contribution could be a cooling one? The data are HADcrut.

  46. 446
    John Millett says:

    #442 Mark

    “But if you can tell us how CO2 and methane emissions can become a cooling forcing, I’m all ears.”

    The natural greenhouse effect is defined as the difference in the whole-of-sphere average surface temperatures of an imaginary, atmosphere-free planet and the real one. Instead, consider day and night hemispheres separately. The imaginary day hemisphere, lacking a radiation reflecting and attenuating medium, would be hotter than the real one. That is, the day-time greenhouse effect is negative. The imaginary night hemisphere would begin and end with higher temperatures than the real one because the imaginary day hemisphere is hotter than the real one. The temperature of the imaginary night hemisphere, which lacks an external energy source, can’t fall below the end temperature. That is, the imaginary night hemisphere is hotter than the real one and the night-time greenhouse effect is also negative. What roles do CO2 and methane play in this cooling atmosphere? None, probably.

  47. 447

    Re 446–

    “Atmosphere-free planet”=NO greenhouse effect. (Not negative–the temperature patterns described may be correct but are irrelevant to “greenhouse.”)

    “No atmosphere”=”no cooling atmosphere”=no CO2 or methane.

    Analysis fail. (Or maybe it’s just so badly written as to be incomprehensible.)

    And note that we have a real instantiation of this “imaginary” planet–one which is well-studied, and of which climatologists are very well aware:

    http://www.asi.org/adb/m/03/05/average-temperatures.html

  48. 448

    On another topic, predictions are always dangerous, and perhaps nowhere more so than in the fickle domain of Arctic sea ice extent. And the larger significance of short-term events is always questionable.

    However, with those caveats, it’s interesting to note that there appears to be a real chance that November 2009 sea extent may feature some all-time lows, as the fall freeze-up continues to be quite slow. (Something I certainly didn’t see coming a month or two back!–FWIW.)

  49. 449
    Mark says:

    “That is, the imaginary night hemisphere is hotter than the real one and the night-time greenhouse effect is also negative.”

    This is incorrect. The real imaginary hemisphere is hotter than it would be if there were no CO2/methane. Therefore the greenhouse effect is positive.

    “What roles do CO2 and methane play in this cooling atmosphere? None, probably.”

    Then it probable doesn’t answer the query posed either:

    “But if you can tell us how CO2 and methane emissions can become a cooling forcing, I’m all ears.”

  50. 450
    Mark says:

    ”The system switches from warming to cooling trends at maxima; and vice versa at minima.”

    Tautology.

    A local minima is always followed by an increase. Hence the term “minima” if it dropped after a minima, it would not have been a minima, at most it would have been an inflexion point and that would be very hard to attain.

    What the clucking bell are you going on about? This is inane babbling, not even streamofconsciousness BS.

    Temperature maxima: 1877, 1942, 1998; period: 65, 56 years

    No, that’s not the period, that’s the interval.

    And you missed out the following maxima:

    August 1878, August 1879, August 1880, …

    Period: 1 year +/- 1 month.


Switch to our mobile site