Confidence tricks between IPCC AR4 and AR5

Patrick Carlberg writes:

I can’t assess the validity of the values from the climate models presented in AR5 (mostly due to lack of time). But I can tell you there are tricks used in the way they interpret the statistical inference.

One of the key points in the new IPCC report is that the CO2 forcing level of confidence was increased from “High” to ”Very High”. The way this was done was by increasing the uncertainty. The means are about the same in AR4 and AR5 but the uncertainty interval has increased from 0.34 to 0.7.

clip_image002

In the same graph (In AR5) an estimate of the total anthropogenic radiative forcing is included. Given the uncertainty in the calculations there is no statistical difference between 1980 and 2011.

clip_image004

My guess is that if the 1950 values were included at the same level of confidence then there wouldn’t be any difference between 1950 and 2011 either.

About these ads

43 thoughts on “Confidence tricks between IPCC AR4 and AR5

  1. see also SteveMc’s work at climate audit. As described on ‘notrickszone’ this report is an obscenity to science.

  2. From memory it always used to be 1850 not the interglacial minimum of 1750 that was used as a baseline.

  3. Is this no more than a way for alarmist ‘science’ to deal with the great heresy of natural climate cycles?

  4. From what I have noticed,it seems that when any statistic from AR4 now looks a bit dodgy, they have simply cherry picked a different date range to cover up the uncertainty, then expressed higher confidence in the new date range.

  5. This group of wastrels has taken something they cannot measure and elevated it to the significance of something they can. Uh, how did they manage to do this, and write 5 AR’s to boot?

  6. IPCC AR5: Now 95% certain global warming is mostly anthropomorphic*.

    *They saw something natural and non-human and thought it had human attributes, like seeing a smile on a snake, or altruism from politicians and other scavengers and parasites.

  7. For radiative forcing by CO2, the “Very High” refers to the level of scientific understanding, not the statistic. The reason for the wider CI’s is explained in footnote 14:

    “In the traditional RF concept employed in previous IPCC reports all surface and tropospheric conditions are kept fixed. In calculations of RF for well-mixed greenhouse gases and aerosols in this report, physical variables, except for the ocean and sea ice, are allowed to respond to perturbations with rapid adjustments. The resulting forcing is called Effective Radiative Forcing (ERF) in the underlying report. This change reflects the scientific progress from previous assessments and results in a better indication of the eventual temperature response for these drivers.”

    IOW, more degrees of freedom are allowed in calculating the response, so necessarily the variability is greater. But it covers more things.

  8. Since release of AR5 SPM, well known German climate alarmist propaganda TV stations (we have a lot!) almost every hour repeat mantra-like
    “Global warming has accelerated more than predicted!”

    No joke!

  9. It is mind boggling to understand why a group of so called scientists want to continue to cheat and lie in everything they do. What is their hidden agenda, and why would professional people want to sully their personal reputations, which they have, in order to promote a cause that cannot be supported with true science? Why do they wish to work together with a collective aim of perpetrating such disgusting behaviour on the worlds population and at such a devastating cost? When the truth is finally accepted, it is to be hoped that the ring leaders of this fraudulent behaviour are brought before their professional bodies to account for why they attempted to bring genuine climate science to such low repute by their totally unfounded assertions.

  10. Nick: Radiative forcing was originally defined as the change in flux across the tropopause (after the stratosphere equilibrated). Climate models suggest that every W/m2 of radiative forcing produces different amounts of warming (at equilibrium everywhere). However, the relative radiative forcing for CO2 was chosen as the standard forcing and is defined as unity. So it isn’t at all obvious why its uncertainty should rise.

    Patrick: There is no doubt that atmospheric CO2 has risen since 1950 and 1980. The forcing associated with CO2 increases in proportion with the logarithm of the CO2 concentration. There is some uncertainty associated with the proportionality constant, but such constants do not change with time. Therefore there is no doubt that the radiative forcing from CO2 has risen since 1950 and 1980-.

  11. Surely the most obvious confidence trick is the claim that a pause which is now 30% of the overall time-frame of interest is “not statistically significant”?

    That’s like saying the first foot of a yardstick should be ignored because it isn’t big enough to be significant.

  12. This might be a suitable analogy:- It is found as the result of a survey that 90% of a group of golfers can hit a particular green from 100m. A new survey is made later with the same group of golfers. Now 95% of them hit the green. How is this? Have their skills improved? Investigation reveals the answer. The green has been made bigger.

  13. I’m 100% certain that the loose change in my pocket is between zero and 1 million dollars. I’m 95% certain that it’s between one and 3 dollars. I’m 90% certain it’s between one and two dollars.

    I’m 100% underwhelmed by AR5 with 100% certainty. I’m sure 97% of you agree with me ;o)

  14. Frank says: October 1, 2013 at 1:52 am
    “However, the relative radiative forcing for CO2 was chosen as the standard forcing and is defined as unity. So it isn’t at all obvious why its uncertainty should rise.”

    These aren’t relative values. They have a box on the matter in Ch 8 (the first one). Of ERF, which they are using now:
    “Land-surface properties (temperature, snow and ice cover and vegetation) are allowed to adjust in this method. Hence ERF includes both the effects of the forcing agent itself and the rapid adjustments to that agent (as does RF, though stratospheric temperature is the only adjustment for the latter). In the case of aerosols, the rapid adjustments of clouds encompass effects that have been referred to as indirect or semidirect forcings (see Chapter 7, Figure 7.3 and Section 7.5), with some of these same cloud responses also taking place for other forcing agents (see Chapter 7, Section 7.2). Calculation of ERF requires longer simulations with more complex models than calculation of RF, but the inclusion of the additional rapid adjustments makes ERF a better indicator of the eventual global mean temperature response, especially for aerosols.”

    If they are including things like cloud response, I think the uncertainties just have to be higher.

  15. George Lawson says:
    October 1, 2013 at 1:24 am

    It is mind boggling to understand why a group of so called scientists want to continue to cheat and lie in everything they do. What is their hidden agenda, . . . .?

    EU bureaucrat Connie Heidegaard let the cat out of the bag on that recently when she said that it would be worth taking mitigation measures even if AGW weren’t real. I’ve suspected this all along. Lots of people get warm fuzzies when they think of “renewable” energy, aid to the 3rd world, less “pollution,” more “environmentalism,” more global cooperation, a bigger role in public affairs for “science” and academia, less of a “consumer culture,” and “sticking it” to big corporations, big oil, and “right-wingers.” What’s not to like? Starting from that POV, the “science” practically writes itself.

  16. @Nick Stokes – “IOW, more degrees of freedom are allowed in calculating the response, so necessarily the variability is greater. But it covers more things.”

    IOW they created the conclusion and then back filled the numbers to fit it.

  17. Frank — “The forcing associated with CO2 increases in proportion with the logarithm of the CO2 concentration.”

    Out of morbid curiosity. Assuming this is true, and that it is included in the models: Then how do the models turn a log curve into an exponential one for temperature prediction?

  18. Here is the original IPCC predictions (it is completely unethical for the IPCC to just restate their previous predictions lower – its like welching a golf bet, except this time the money on the line is into the billions).

    And then versus the spread of the models in AR4, with predictions starting about 8 and a half years ago.

  19. Thanks, Patrick. Very good, simple analysis. Congratulations!
    This is the trick that produces ‘confidence’ out the IPCC’s hat.
    Foootnote 16, “No best estimate for equilibrium climate sensitivity can now be given because of a lack of agreement on values across assessed lines of evidence and studies.” says it all.

  20. I could say that this is ‘on the topic of alarmist confidence tricks’, but to be honest I’m not particularly on-topic here.
    The appropriate article for me to attach this comment to is Willis Eisenbach’s ‘the real Canadian Hockey Stick’ http://wattsupwiththat.com/2013/07/13/the-real-canadian-hockeystick/ Sadly, that article no longer accepts comments.
    Anyway, the confidence trick is that the tax Willis wrote about was written up here, claiming it was an environmental and economic success story. http://www.sustainableprosperity.ca/dl1026&display
    ‘[S]cepticalscience.com’ as of 22 August is linking to that article without acknowledging any of the points Willis brought up.
    Did anyone get any of those points into comments on the article in later issues of Canadian Public Policy? If not, does anyone have access to those comments, and can they do so?
    Great set of articles, by the way, Willis.

  21. PETERMUE: do you have any sceptic blogs such as WUWT in Germany, or are they purely Anglo Saxon phenomena?

  22. George Lawson October 1, 2013 at 1:24 am says:

    “It is mind boggling to understand why a group of so called scientists want to continue to cheat and lie in everything they do. What is their hidden agenda, and why would professional people want to sully their personal reputations, which they have, in order to promote a cause that cannot be supported with true science?”

    I agree. Part of it is financial but this cannot be the full story. Reluctantly I have to conclude that it is stupidity. They really do not have the ability to think critically and understand how it all hangs together. As a result, thousands of technician-level papers get published that do not amount to a scientific advance. James Watson ran into the same problem when working on the double helix. This is how he explains it: “One could not be a successful scientist without realizing that, in contrast to the popular conception supported by newspapers and mothers of scientists, a goodly number of scientists are not only narrow-minded and dull, but also just stupid.”

  23. The real real confidence trick is the IPCC regarding these intervals as symmetrical. Thus, widening the interval to include the real world at the bottom end ends up making the high end “worse than we thought”, or at least “solidifies” support for wackily-high worst-case scenarios.

    That way, they’re able to take what should be a loss (models unrealistically high, worst-case scenarios ridiculously high) and turn it into a win (models within bounds, worst-case scenarios within bounds). Normally, having to widen your confidence interval is also something of a loss, since you have to admit that things were more uncertain than you originally believed, but they’ve managed to turn it into a double-win for themselves.

  24. JQuip wrote: “Out of morbid curiosity. Assuming this is true, and that it is included in the models: Then how do the models turn a log curve into an exponential one for temperature prediction?”

    With lots of positive feedback in the models resulting from small increases in temperatures. Which raises the question, if the total earth climate system is a positive feedback system exquisitely sensitive to small temperature increases, why didn’t all of us burn up in the Roman and Medieval Warm Periods. That we didn’t strongly suggests a system dominated by negative feedback.

  25. It appears for all the world that William Connolley of Wikipedia shame was the final editor/arbiter of this entire report. This has his MO all over it. The entire document deserves a place on the junk science shelf of irreproducible results.

  26. Nick Stokes,

    I read your words, but not sure they make sense. Their “scientific understanding” is higher, yet the know the number less well than they did in AR4? The only way that makes sense is if they are more certain the number lies in that range now that it has been expanded. But they are more confident of a less certain number (which is fine in one way but not so good if people don’t know the whole story and just see it as more certain). Sorry, no other way to spin it.

  27. Nick Stokes comments

    For radiative forcing by CO2, the “Very High” refers to the level of scientific understanding, not the statistic.

    And

    IOW, more degrees of freedom are allowed in calculating the response, so necessarily the variability is greater. But it covers more things.

    By implication, AR4 left out a number of things due a lack of understanding. A comparison of AR4 with AR5 shows this, now acknowledged, lack of understanding is far from trivial. The “degrees of freedom” in absolute terms have more than doubled for CO2 and more than quadrupled CH4. For halocarbons they increased more than five-fold. More remarkable is the forcing per unit of gas in the atmosphere. CO2 has reduced by 10%, CH4 almost doubled (due to including other effects) and Halocarbons halved.
    Most important of all, is that a few years ago we were being told that the lack of warming compared to the models was to the negative forcing from aerosols. The role of these is much reduced, and the “degrees of freedom” now include the possibility that aerosols have a net positive effect.
    The details are posted up here.
    With hindsight, will the IPCC now be admitting that they had too high a level of confidence with their figures in 2007? Given that overconfidence in the past, is it not prudent to wait for another few years to see if there is similar over-confidence this time around?

  28. manicbeancounter says: October 1, 2013 at 2:07 pm
    “Most important of all, is that a few years ago we were being told that the lack of warming compared to the models was to the negative forcing from aerosols. The role of these is much reduced, and the “degrees of freedom” now include the possibility that aerosols have a net positive effect.

    With hindsight, will the IPCC now be admitting that they had too high a level of confidence with their figures in 2007?”

    In the AR4, the number for direct aerosol effect was rated as med/low confidence, and indirect effect low confidence. The range did not include zero, but came close. The SPM rhen said that aerosols “remain the dominant uncertainty in radiative forcing”.

    Bill says: October 1, 2013 at 1:40 pm
    “which is fine in one way but not so good if people don’t know the whole story”

    They have given the whole story, mostly on the same page. What more can they do?

  29. Nick

    I did not say that it was a high level of confidence in AR4, but “too high a level of confidence”. For those who are no familiar with Mancunian, this means that the confidence level was too high. In AR4, there was high confidence in CH4 and halocarbons which have dramatically changed.
    As for aerosols, in 2007 I would alledge it was a fudge factor. I worked for over 20 years in accountancy, analyzing and compiling figures that do not fall out to a nice round figure like in a high school maths or physics problem. Climate is the same. Figures that come to nice round figures on real data are extremely rare. In auditing, a test of manipulated figures is patterns, or figures that tell a precise story. Working out the uncertainty bands on different forcings will lead to varied and apparently random figures. Still more, if anyone is dumb enough, as I was in beancounting mode, to add together the range of uncertainties for all the different forcings, there would be irregular figures. For AR4 summary table 2.4 the positive anthropogenic forcings uncertainty range add to 40% of the total, and the negative anthropogenic forcings uncertainty range add to 200% of the total. Change any number by +/-0.01 and the figures do not fall out. Further, the central net forcings of 1.60 Wm-2 was almost identical (given the huge uncertainties) to the central net forcing of 1.66 Wm-2. So the PR angle could be conveniently shortened from “anthropogenic greenhouse gas emissions” to “rising CO2 levels”.
    So not only is AR4 contradicted by AR5 (including for figures with “High Confidence” in AR4), it would appear the AR4 figures have been reverse engineered.
    Maybe the UNIPCC have got high quality science this time, that the lay public can have confidence in. But given the track record, it is advisable wait and see if they give a consistent story in a few more years.

  30. Frank says:
    October 1, 2013 at 1:52 am
    “Patrick: There is no doubt that atmospheric CO2 has risen since 1950 and 1980. The forcing associated with CO2 increases in proportion with the logarithm of the CO2 concentration. There is some uncertainty associated with the proportionality constant, but such constants do not change with time. Therefore there is no doubt that the radiative forcing from CO2 has risen since 1950 and 1980-.”

    Only IF ALL OTHER THINGS BEING EQUAL ; i.e. when assuming hydrostasis (an atmosphere that does not expand or contract).

    The atmosphere is not hydrostatic (although climate models assume it to be able to apply the barometric equation); this changes the radiative forcing. (TOA height changes; lapse rate changes; density changes)

    Looks like the models got it wrong. What has really happened with radiative forcing? I don’t know. But it’s not warming. The theory is wrong.

  31. Nick Stokes says: October 1, 2013 at 2:43 am
    Nick, Can we accept that effects are not homogenous over the globe?
    Here is a graph from Willis Eschenbach showing SST from Argo in the tropical ocean area –

    Seems to me that the balance of incoming and outgoing might be different in the tropics to the sub-tropics so one value might not fit all.

  32. Geoff Sherrington says: October 1, 2013 at 10:51 pm
    Yes, of course. In fact, forcing is different from TOA balance. Balance has to be maintained, except for temporary variations, whether or not forcing increases.

  33. °Remember the Dying of the Thermometers, 1990? It coincides with a 1.4°C step change in the average. The “Dying”, btw, consisted of ignoring their data. They are still there, almost all of them, but their info never makes it to the HadCRUTs, etc.

Comments are closed.