Skip to content

Science magazine blows the whistle on climate model failure

July 30, 2021

By Paul Homewood


“It’s become clear over the last year or so that we can’t avoid this,” Gavin Schmidt, director of NASA’s Goddard Institute for Space Studies, told Science ­magazine.

The admission is seen as a significant development by scientists who argue that not enough attention has been paid to natural ­cycles in the earth’s climate.

It puts another question mark over the use of the most extreme scenarios generated by models, RCP8.5, to estimate what could be expected in a warming world.


Scientists such as Roy Spencer and John Christy have been saying this for years, but the climate establishment has simply ignored the problem.

Despite Gavin Schmidt’s admission, I suspect the IPCC will still carry on using their overheated computer models.

Given that it these very models upon which governments around the world have based their decisions to destroy their economies, maybe it is time they went back to square one, and had a rethink?

  1. Harry Passfield permalink
    July 30, 2021 6:55 pm

    It’s about TIME!
    Right now the general public are so well aware of the dodgy Covid computer models used by SAGE and, even more-so by Ferguson, they will be really receptive to how climate models (see CMIP5 spaghetti chart) are such a load of cr4p.

  2. Broadlands permalink
    July 30, 2021 7:17 pm

    It may becoming obvious that there are no models that can configure and simulate any of Earth’s natural variability over any meaningful time frame. It seems to have been assumed that the impact of increasing CO2 has exceeded all of those things…the jet streams, the ENSOs, volcanic events. None of those correlates with Mauna Loa CO2. Every time there is a strong El-Nino it grabs the headlines, is acknowledged as being natural, can cause extreme weather… but the implication is that it is AGW CO2 that is the root cause.

  3. David Wojick permalink
    July 30, 2021 7:26 pm

    The issue here is very specific, not about climate models in general. I have been tracking it for a year or so, wondering how it will play out. Now the time has come.

    Every five years or so most of the world’s major modeling centers run a joint exercise, the results of which feed into the latest IPCC report. About 100 models are involved, mostly run by national centers. The latest exercise was called CMIP6, feeding into IPCC AR6, which begins to be released on the 9th of August. This was the 6th CMIP.

    We got an extreme event! About half of the CMIP6 models ran much hotter than they had in CMIPs 1-5. The hot shots show climate carbon sensitivities (CS) above 5 degrees C, while IPCC has always said the range was 1.5-4.5 degrees. Oops!

    The new hotness was traced to greatly increased positive cloud feedbacks, for which there was no big scientific breakthru to justify it.

    What is the IPCC and the modeling community to do? If they accept these hot results it means they were very wrong for 40 years. If they reject the results it calls the modeling into serious question, for the first time by the modeling community itself.

    This ScienceMag article indicates that at least some of the community is rejecting the hot results. What the IPCC does remains to be seen, hopefully shortly but maybe not, as the full 1000 page report is not due out soon, just the summary.

    We might even get a split in the modeling community.

    We skeptics need to make the most of this clear failure.

    Here is URL

    Note “implausibly hot”. Woohoo!

    • David Wojick permalink
      July 30, 2021 7:56 pm

      The modelers dilemma is either (1) all the models have been wrong for 40 years or (2) half are wrong now. Be still my heart!

    • David Wojick permalink
      July 30, 2021 8:59 pm

      Here is the CMIP6 homepage, with a map linked to all the participating modeling centers. Britain looks to have five. (Use two fingers to move and spread the map.)

    • Phoenix44 permalink
      July 31, 2021 8:02 am

      I’m unclear – did they make changes to the treatment of clouds or is the wrong treatment of clouds only now becoming apparent on new runs?

      • David Wojick permalink
        July 31, 2021 11:43 am

        My guess is the former, perhaps a widely adopted new algorithm. It is possible they did it to support the COP 26 call for increased “ambition”.

    • Jordan permalink
      July 31, 2021 3:48 pm

      Thanks for your comments David. Informative and helpful.

      You say: “as the full 1000 page report is not due out soon, just the summary”. That’s fine if we have the assurance that no material changes to the main report will be made. It is not fine if, by publishing a “summary” first, the main report is following a pre-existing script.

      • Gerry, England permalink
        August 2, 2021 11:47 am

        The summary is where the politicians wade in and make claims that the report does not back up. So they need to amend the report to take account of these. After one of the earlier reports was published, the majority of honest scientists demanded their names be removed as the report was not an accurate reflection of what they had said. And with that went any chance of the IPCC reports not being propaganda from an advocacy group.

  4. July 30, 2021 7:41 pm

    Most unfortunately, U-turns are anathema to politicians, whatever their logic and financial benefits might be.

  5. Broadlands permalink
    July 30, 2021 7:58 pm

    Another potential problem for these models is their constant tweaking, not only of the models themselves, but the data that go into them. Constant adjustments to temperatures (often lowering older ones), failure to acknowledge that raw data have been destroyed, incorrect or no adjustments for the UHI effect. All of these problems have been shoved under the rug. Of course, to show some light on any of this now will not help credibility.

    • David Wojick permalink
      July 30, 2021 8:58 pm

      That is not the “implausibility” issue in this case. It is just that about half of the CMIP6 models ran implausibly hot, apparently due to juiced up positive cloud feedback.

      • July 31, 2021 10:07 am

        If clouds make the world hotter in climate models, does lack of them (sun) make it cooler? 🙄

  6. mervhob permalink
    July 30, 2021 11:56 pm

    The truth is, that no algorithmic model based on the approximations of linear algebra can accurately predict the future of a non-linear system. The linear model of the solar system of Claudius Ptolemy, based on epicycles, deferents and equants, was badly out of sync after less than 1000 years, partly because the orbital constants were not well enough known in his day. And this was describing a system that was only weakly non-linear. But both the weather and climate are strongly non-linear – therefore no linear algebraic model can accurately predict their future. The only thing that can be ‘proved’ using the methods of Euler/Lagrange/Laplace etc, are the underlying theorems – they are in effect a consistant ‘software’ whose application to physical reality in a time series is fraught with real difficulty; they are bound to diverge with time.
    As I have written, Euler and Lagrange made no attempt to solve the real mathematical difficulties shown by Newton in his ‘Principia Mathematica’ – instead they replaced it with an algorithmic system in which the worst problems did not exist – linear algebra. If you open your copies of Newton’s Principia, you will find no closed algorithmic expressions – most of the mathematics in the Principia is ‘open’; using synthetic geometry as Newton was well aware that most physical problems were incapable of statement in a closed form with any real rigour.
    So, we should not be surprised that computer programs, however complex the underlying algorithmic structure, diverge with real data over time. And statistical tools can only be accurately applied post priori to data, unless applied to a physically ‘closed’ system. Paul has correctly proceeded in all his analysis and not fallen into the sloppy belief that he can predict the future – hence his data can be trusted.
    The temptation always exists, that having developed a model, to ‘cook’ the previous data to fit the model – as Owen Gingrich shows in his book, ‘The Eye of Heaven’, there is ample evidence that Ptolemy did this, to bolster the acceptance of his system. We should not be surprised therefore, to find that climate computer modellers have been ‘economical with the truth’, removing significance from such historical events as the Roman Climatic Optimum, the Medieval Optimum and the Mini Ice Age, even though the extent and effect of these occurances is well documented and linked to physical events. On that subject, why do most modern analyses start in the mid 1880s? Could it be that that is a convenient ‘low’ due to the explosion of Krakatoa in 1883? The very poor weather after that event and its severe unpredictability is well recorded in cartoons of that period and affected most of the world for the rest of the decade.

    • Phoenix44 permalink
      July 31, 2021 8:01 am

      Very true. Complex non-linear systems cannot be modelled by current computer models.

      As important it is impossible to know starting condition for any model accurately enough so you are never modelling reality anyway. On a chaotic or partly chaotic system, getting your starting conditions wrong can lead to results being totally wrong quite quickly.

      • dave permalink
        July 31, 2021 9:57 am

        Scientists knew well both the uses and the limitations of Science a hundred years ago.

        As one textbook roughly put it (going from my memory)

        “We will NEVER understand the DETAILS of any process or event – such as a particular house catching fire and burning down. But, then, why on earth would one want to? It is enough to know that a house CAN burn down from a knocked over candle.”

        Big shot scientist says “…we can’t avoid this…’

        Why would he want to avoid it? We should always accept joyfully new knowledge*. If I take the wrong turn while driving I will accept that information from my wife – well, perhaps not joyfully but that is because of her tone of voice.

        * Of course, there are the psychological issues of ‘sunk costs.’ and ‘amour propre’ On the first issue, people are always loth to throw away something which it has cost them to obtain. On the second issue, we have all known people who turn very nasty, very fast, when confronted with their mistakes.

      • mervhob permalink
        August 1, 2021 12:43 am

        Phoenix 44 – absolutely correct, starting with the wrong initial conditions rapidly leads to divergence from reality even with small non-linearity. The crude, ‘steady state’ modelling taught in most universities is good for less than 1000 cycles – good enough for an engineer to raise a wet sticky finger in the face of Providence but not to be relied on for something a serious as climate modelling. One of my favourite authors on non-linear modelling, Y.H. Ku states, ‘Nature is non-linear, even the pendulum of Galileo is described by a 2nd order, non-linear differential equation.’ And so it is, as any competent horologist knows!

  7. July 31, 2021 12:41 am

    Greatest marketing champagne ever waged against humanity that a minute degree of planetary warming, and enhanced plant fertilization spells doom for humanity; and that giving loads of tax money to leftists will “fix” the on-coming pleasant climate. And now we don’t even get the warming. Never mind.

    • Tim Leeney permalink
      July 31, 2021 8:41 am

      The marketers still get the champagne. The rest of us do not . . . yet.

  8. Rod Smith permalink
    July 31, 2021 10:21 am

    The big question about Gavin Schmidt’s remark “it’s become clear … that we can”t avoid this” is why anyone would want to avoid what is essentially good news for the world – it is not heating as fast as feared. Clearly the only reason is that it undermines an ideologically motivated agenda.

  9. Alex permalink
    July 31, 2021 7:58 pm

    A more damning failure, one that scuttles the entire fleet:

  10. europeanonion permalink
    August 1, 2021 7:21 pm

    My understanding is that our Solar system orbits in the centre of the Milky Way on a path that takes approximately 225 million years to complete. Added to all the other variables attached to climate issues it seems certain that we are to pass through uncharted territory with who knows what new materials or formulations to be met.

    The longer this charade goes on for the harder it will be for government to back-track. Already we see that those involved on HS2 are making a heap of money from their contracts, unsurprisingly. At every opportunity they refute some killer fact about their doubtful project. Those on the inside deflect any criticism aimed at the their gravy train.

    With car manufacturers forced to tool-up for electric they too will now, probably, defend the project to their last cog, widget and wheel (on hearing vague rumours about the precedence given to walkers and cyclists in the new Highway Code we are not far from having flag bearers proceed us on our journeying ) the declining maintenance of roads the strangulation of the public will is in train and Vehicle Tax looks more like a tithe to the state to be dispensed with as they see fit, rather than a dedicated financing of roads.

    Even if all this blather about climate is scotched the impoverishment of the tax payer will carry-on under other guises. If the tax money was spent properly then issues attached to excess discharges from cars could be solved in a instant. But they have picked a horse even though it is a dumb animal.

    I can only imagine that those who live in what use to be called tyrannies are laughing up their sleeves now when they see how their condition is demonstrably freer than ours which is increasingly more controlling and coercive. Diesel, Covid and climate, three humungous calls made by the state all all destined to demonstrate their ineptitude and disgraceful attitude towards tax payers and democracy.

  11. Gerry, England permalink
    August 2, 2021 12:14 pm

    The main point of this is that a few of the more conscientious climate scientists have been questioning the continued use of the RCP8.5 scenario since it is a fantasy outlook that can never happen. It is true that the other scenarios are not accurate either, but there is concern that RCP8.5 is used to produce scary papers that would collapse with the minimum of scrutiny. Paul has highlighted their use here and when any scary stuff hits the legacy media it is almost certain that buried at the bottom will be RCP8.5.

Comments are closed.

%d bloggers like this: