Measuring the pause in global warming

I keep writing about the pause or ‘hiatus’ in global warming because it is at the heart of the  debate about global warming, as I mentioned in passing yesterday. And I continue today because in some sense this is a historic moment. The failure of the global temperature anomaly to increase began to worry the orthodox ten years ago, and the longer it went on the more worried some of its members became.

At length, after ten years of nothing had happened, a group of the good and great led by Benjamin Santer wrote a paper explaining that ten years was just too short a period in which to distinguish the signal of human-generated temperature changes from the noise in the system: ‘interannual noise’, they explained, had too much effect on ‘decadal trends’. And they finished their abstract with this strong summary, much quoted since: ‘Our results show that temperature records of at least 17 years in length are required for identifying human effects on global-mean tropospheric temperature.’ Now Santer is a much honoured climate scientist, with strong connections with the University of East Anglia’s Climate Research Student (he got his PhD there) and the IPCC. So here were the leaders of the orthodoxy talking.

Well, 17 years has passed without there being a resumption of warming of any consequence. There has been no response from Santer et al, though others have jumped to their aid, suggesting that even after 17 years you might not see much more than a smidgin of a signal, and that it might take more years — 20, 30, who can tell — for the signs to be there strongly. The trouble with that defence is that in the meantime there hasn’t been any warming, which makes it impossible to show that it is CO2 that does all the work, and extends indefinitely the time it might take to reach the (in my view quite unsupported) claim that we must prevent temperature rises reaching 2 degrees C above the present. And it suggests that ‘climate sensitivity’, the notion that clouds and water vapour will greatly amplify the warming due to increases in CO2 alone, cannot be very high, and might even be zero.

Someone has now done the work to show all of the standard datasets, and the change in 17 years, using the useful Woodfortrees instrument, which is available to anybody who has a bright idea and wants to try it out on temperature data. Here is the first graph.



The background line is the regular Remote Sensing Systems (RSS) plot of the global temperature anomaly, displaying the great el Nino peak of 1998 and its smaller cousin of 2010. Just eye-balling tells you that any net increase in temperature in this time has got to come from those two el Ninos and that, after 1998 at least, there can’t have been much increase, if any. The trend lines for the datasets are the straight lines, and that for RSS is flat, two others are almost flat, and the other three display small increases only. None of them shows statistically significant warming. That of UAH, the steepest, shows a trend that produces warming of less than 1 degree C by 2100, which is hardly a worry.

The next graph looks at how long a flat trend there is for each of the datasets.  They are all of course parallel — it is their length that matters. The longest is that of RSS, which you would expect from the previous graph, and only that of UAH (the other remote-sensing dataset) is less than ten years.  If present trends continue HadCRUt3, from the University of East Anglia itself, will reach the 17-year mark shortly.


The rising line shows the rise in the amount of carbon dioxide in the atmosphere. The jagged appearance of that line comes from the fact that atmospheric carbon dioxide is greatly affected by plants, which take up as much of their food as they can get in their growing seasons — and most of the plants are in the northern hemisphere.

Altogether, these two graphs are abundant visual evidence that there is something very wrong with the view that human activity in the burning of fossil fuels, making cement and clearing forests is the predominant cause of global warming. And, I’ll say it again, if it hasn’t been able to do so in the last 17 years, there is every reason to suspect that earlier periods of warming were also affected by whatever the elements of natural variability are. In short, the fundamental building block of the AGW scare is shown to be made of cardboard, if not tissue paper.

Join the discussion 19 Comments

  • Malcolm Miller says:

    Keep on giving the message – though who will be convinced by it is rather dubious.

  • Phillip Williams says:

    Yes – I find it astounding that even the group of business leaders on Q&A last night all stated they believed in climate change. Especially amazing was Elizabeth Proust, who said she didn’t know anyone who didn’t believe in it. What a narrow circle she moves in.

  • David says:

    Common Don

    Look at where you commence your date line, 1996!. If you go back to say the 1900 you will see another much lager “pause” from 1950 to 1980 where temperature actually declined!!!, before again increasing. Overall temperature has clearly risen steadily over the last 100 years.
    Your line of argument is embarrassing!
    I picked graph at this at random. There are many other available

  • Colin Davidson says:

    In response to David,
    It is perfectly correct that it matters where you start. If you start in 4000BC, there has been a fall in temperature. Same for 0AD or 1000AD. Different if you choose a cool year.
    But Don is correct – many Climate Scientists have claimed in the past that 5 years is too short a time, wait for 10 years. When 10 years are up, they have said 15 years. And so on…But for warming: why only 1 year will do!
    There are a number of points to make:
    1. What would be a falsification of the Hypothesis that CO2 has a large effect on Temperature?
    2. How is the identical slope (DegC/yr) of the 1890s, 1930s and 1980s periods to be explained, given that the rates of increase of CO2 were markedly different.
    3. Given the cooling periods in the 20th century, and the hiatus latterly, what is the assessed amplitude(s) and frequency(s) of the natural variations?
    These questions cannot now be ducked, or swept away with almost certainly false claims that the heat is hiding somewhere deep in the ocean where we haven’t been measuring it.
    And it certainly does not help the pro-hypothesis case that the loudest and most strident proponents are prominent activists – hardly an objective or careful source of wisdom.

    • David says:

      What would be a falsification of the Hypothesis that CO2
      has a large effect on Temperature?

      This is a great question.

      Prof Richard Muller was a sceptic who was commission by
      US Coal industry to test the AGW. His dataset consisted of 14 million
      observations drawn from 44,555 sites dating back to 1753. He concludes that the Earth has warmed and that humans are the cause

      But Muller has not proved anything. All he did was present evidence, consistent with AGW, with a data set consisting of 14 million observations and 250 years of data.

      Fundamentally Muller’s or any other climate model will look like this.

      Global Temperature = f (CO2, X, Y, Z)

      Where X, Y, Z are other explanatory covariates such as sun spot activity etc.

      So, to falsify the AGW hypothesis you would need to find a new plausible variable, which when variable was inserted into the equation above it would render the coefficient on CO2 statistically insignificant.

  • David says:


    The global temperature 10,000 years ago is not central to the AGW debate. The AGW hypothesis is that since the beginning of the industrial revolution (approx. 1750) CO2 emission have begun to increase global temperature. So while we might be interested in what the temperature in the distant past what is germane is (a) how temperature have changed during the last 500 years and (b) what is our capacity to cope with further changes.

    To answer these sorts of questions you would obviously not analyze time series data of only 20 years duration nor would you solely focus on time series data of 10,000 years duration.
    It is just not correct to imply that the AGW hypothesis is spuriously based on 1 year of data. NASA, BOM and Australia’s Chief Scientist not to mention the UN support AGW, and by any measure, they do collectively embody objective and a careful source of wisdom.
    Regards David

  • 3d1k says:

    This article from the WSJ may be of interest – the sun (and sunspots) not behaving as they should.

  • colin davidson says:

    In response to David.
    Sorry for this tardy reply.
    You are correct that there has been warming if you start in 1970 or 1900. Don is also correct in stating that there has been no warming since 1996. And I am correct that the world is cooler than 1000 years ago.
    So it all depends where you take the start point. There are two observations I would like to make:
    1. All the evidence we have on temperature is consistent with a truly chaotic system with at least two major “strange attractors”. Such systems are characterised by possessing multiple NEGATIVE (and no positive) feedbacks. If the climate system is chaotic then we cannot predict the future state. Full stop.
    2. The archaeological evidence is that there is a warm period every 1000 years. (Despite the denials by prominent warmistas that the MWP was confined to the North Atlantic, the archaeological evidence from NZ, Australia and Peru is clear cut. Also see for the science on this.) And recent temperatures with warming followed by cooling/hiatus also suggests a sinusoid. These observations are consistent with a system with at least 3 oscillations (caused by negative feedback with loop gain =1 at the relevant frequencies): a long term very cold to mildly warm oscillation (ice ages), a cold-mildly warm 1000 year oscillation (Roman WP, Medieval WP, now…), and a 50year cool-mildly warm oscillation.
    In neither case does a linear relationship such as the claimed effect of more CO2 feature. Natural systems (and indeed most systems designed by humans) are usually not linear. In general the negative feedbacks generate oscillations. Positive feedbacks in such systems (eg, when a hurricane is starting out) are short-lived, never long-term features.
    I have looked in vain for discussion of this by warmistas. Perhaps you can provide a link to where this sort of discussion is taking place.

  • […] to point to the correlation between rising CO2 levels and rising temperature. But after 1998, and the failure of temperature to keep rising, the orthodox had to find reasons for the failure of air temperature to rise. The current orthodox […]

  • Colin Davidson says:

    In response to David,
    Proof of the assertion that climate/weather is not chaotic requires much more than assertion. A proof would be that the drivers, system and start state are known to sufficient accuracy to permit accurate forecasts on any time period.
    None of those preconditions are present and neither are forecasts accurate. We must assume that the system is chaotic until we can make reliable accurate forecasts.
    The evidence you have cited for the temperature of the MWP is not the best available evidence. It comes from the Hockey Stick/UEA/Climategate group of scientists, a group which relies on tree ring methods which are shaky (at best!), and who ignore the great body of other, better evidence from archaeology and geology. It is for example undeniable that the Peruvian indians farmed higher terraces than is possible today, and that the Norse dug 6ft graves in what is now permafrost in Greenland,
    But enough anecdotal stuff! Here is the reference, which cites squillions of peer reviewed papers from people who study stalagmites, lake sediments, vegetation, snowlines, soil pollen, archaeology etc.
    The quantative result of analysing all those papers is that the MWP was higher in temperature than now by a mean of +0.75DegC, SD1DegC (eyeballed from chart).
    With regard to the start point, you are trying to establish the hypothesis that mankind’s increased CO2 emissions, insignificant as they are compared with the natural emissions, are the cause of increased temperature.
    To establish the hypothesis requires at least three proofs:
    1. That the rise in temperature in the period is not natural. That means that you have to show that.. a. the drivers which caused the Roman Warm Period, the Dark ages Cold period, the Middle Ages Warm Period and the Little Ice Age are not responsible for any of the warming..b. that the evident short term sinusoid in the temperature record of the last hundred-odd years has not also contributed to the warming..c. that the observed correlation of temperature with sunspots since 1600 is no longer happening.
    You cannot establish that the rise in temperature is not natural without examining the long-term temperature record in great detail, ie you need to look at far-back start points (My starting point).
    But also if the observed recent stasis in temperature (Don’s starting point) is due to natural causes, you equally need to quantify that and modify your hypothesis accordingly.
    2. That the observed rise in CO2 concentration above Hawaii is not due to natural causes. See (Mathematics alert!)
    and that the observed distribution of highest CO2 concentrations above the earth’s rain forests rather than industrial sites does not indicate the dominance of natural sources.
    3. That the observed rate of warming is identical to the calculated theoretical effect. My understanding is that the warmista argument relies on POSITIVE feedback, another hypothesis which is not established and is in raging dispute in the journals. Recent papers are all pointing to less sensitivity, ie lower than previously claimed temperature rise due to CO2.

  • […] will happen. Neither of these possibilities seems very plausible to me. As I have pointed out in recent essays, the most recent trend is towards no further warming, or even cooling, and that possibility might […]

Leave a Reply