New Santer Study: 97% Consensus is now 99.99997%

February 27th, 2019 by Roy W. Spencer, Ph. D.

A new paper in Nature Climate Change by Santer et al. (paywalled) claims that the 40 year record of global tropospheric temperatures agrees with climate model simulations of anthropogenic global warming so well that there is less than a 1 in 3.5 million chance (5 sigma, one-tailed test) that the agreement between models and satellites is just by chance.

And, yes, that applies to our (UAH) dataset as well.

While it’s nice that the authors commemorate 40 years of satellite temperature monitoring method (which John Christy and I originally developed), I’m dismayed that this published result could feed a new “one in a million” meme that rivals the “97% of scientists agree” meme, which has been a very successful talking point for politicians, journalists, and liberal arts majors.

John Christy and I examined the study to see just what was done. I will give you the bottom line first, in case you don’t have time to wade through the details:

The new Santer et al. study merely shows that the satellite data have indeed detected warming (not saying how much) that the models can currently only explain with increasing CO2 (since they cannot yet reproduce natural climate variability on multi-decadal time scales).

That’s all.

But we already knew that, didn’t we? So why publish a paper that goes to such great lengths to demonstrate it with an absurdly exaggerated statistic such as 1 in 3.5 million (which corresponds to 99.99997% confidence)? I’ll leave that as a rhetorical question for you to ponder.T

There is so much that should be said, it’s hard to know where to begin.

Current climate models are programmed to only produce human-caused warming

First, you must realize that ANY source of temperature change in the climate system, whether externally forced (e.g. increasing CO2, volcanoes) or internally forced (e.g. weakening ocean vertical circulation, stronger El Ninos) has about the same global temperature signature regionally: more change over land than ocean (yes, even if the ocean is the original source of warming), and as a consequence more warming over the Northern than Southern Hemisphere. In addition, the models tend to warm the tropics more than the extratropics, a pattern which the satellite measurements do not particularly agree with.

Current climate model are adjusted in a rather ad hoc manner to produce no long-term warming (or cooling). This is because the global radiative energy balance that maintains temperatures at a relatively constant level is not known accurately enough from first physical principles (or even from observations), so any unforced trends in the models are considered “spurious” and removed. A handful of weak time-dependent forcings (e.g. ozone depletion, aerosol cooling) are then included in the models which can nudge them somewhat in the warmer or cooler direction temporarily, but only increasing CO2 can cause substantial model warming.

Importantly, we don’t understand natural climate variations, and the models don’t produce it, so CO2 is the only source of warming in today’s state-of-the-art models.

The New Study Methodology

The Santer et al. study address the 40-year period (1979-2018) of tropospheric temperature measurements. They average the models regional pattern of warming during that time, and see how well the satellite data match the models for the geographic pattern.

A few points must be made about this methodology.

  1. As previously mentioned, the models already assume that only CO2 can produce warming, and so their finding of some agreement between model warming and satellite-observed warming is taken to mean proof that the warming is human-caused. It is not. Any natural source of warming (as we will see) would produce about the same kind of agreement, but the models have already been adjusted to exclude that possibility.
  2. Proof of point #1 can be seen in their plot (below) of how the agreement between models and satellite observations increases over time. The fact that the agreement surges during major El Nino warm events is evidence that natural sources of warming can be mis-diagnosed as an anthropogenic signature. What if there is also a multi-decadal source of warming, as has been found to be missing in models compared to observations (e.g. Kravtsov et al., 2018)?
  3. John Christy pointed out that the two major volcanic eruptions (El Chichon and Pinatubo, the latter shown as a blue box in the plot below), which caused temporary cooling, were in the early part of the 40 year record. Even if the model runs did not include increasing CO2, there would still be agreement between warming trends in the models and observations just because of the volcanic cooling early would lead to positive 40-year trends. Obviously, this agreement would not indicate an anthropogenic source, even though the authors methodology would identify it as such.
  4. Their metric for measuring agreement between models and observations basically multiplies the regional warming pattern in the models with the regional warming pattern in the observations. If these patterns were totally uncorrelated, then there would be no diagnosed agreement. But this tells us little about the MAGNITUDE of warming in the observations agreeing with the models. The warming in the observations might only be 1/3 that of the models, or alternatively the warming in the models might be only 1/3 that in the observations. Their metric gives the same value either way. All that is necessary is for the temperature change to be of the same sign, and more warming in either the models or observations will cause an diagnosed increase in the level of agreement metric they use, even if the warming trends are diverging over time.
  5. Their metric of agreement does not even need a geographic “pattern” of warming to reach an absurdly high level of statistical agreement. Warming could be the same everywhere in their 576 gridpoints covering most the Earth, and their metric would sum up the agreement at every gridpoint as independent evidence of a “pattern agreement”, even though no “pattern” of warming exists. This seems like a rather exaggerated statistic.

These are just some of my first impressions of the new study. Ross McKitrick is also examining the paper and will probably have a more elegant explanation of the statistics the paper uses and what those statistics can and cannot show.

Nevertheless, the metric used does demonstrate some level of agreement with high confidence. What exactly is it? As far as I can tell, it’s simply that the satellite observations show some warming in the last 40 years, and so do the models. The expected pattern is fairly uniform globally, which does not tell us much since even El Nino produces fairly uniform warming (and volcanoes produce global cooling). Yet their statistic seems to treat each of the 576 gridpoints as independent, which should have been taken into account (similar to time autocorrelation in time series). It will take more time to examine whether this is indeed the case.

In the end, I believe the study is an attempt to exaggerate the level of agreement between satellite (even UAH) and model warming trends, providing supposed “proof” that the warming is due to increasing CO2, even though natural sources of temperature change (temporary El Nino warming, volcanic cooling early in the record, and who knows what else) can be misinterpreted by their method as human-caused warming.

120 Responses to “New Santer Study: 97% Consensus is now 99.99997%”

Toggle Trackbacks

  1. Ed Mihelich says:

    Thanks, Roy, for an excellent summary. It is too bad that the editors of journals such as this one cannot identify these points prior to publication.

    • Martha says:

      Thank you for pointing out that it is the editors who are to blame for this mess. Many do not understand the process of publication and where responsibility lies in journals.

      • Bob Roberts says:

        Unfortunately the editors take part in the “echo chamber” which reinforces false conclusions regarding the overall “religious” belief that humans have usurped the vastly more powerful forces that actually determine trends in weather, temperature and climate.

        “Attribution” is not science, it’s opinion, but in the end that’s all climate alarmists have.

  2. DMA says:

    If the recorded rise in atmospheric CO2 is not responsive to the recorded changes in emissions (, how can someone detect a fingerprint of the emissions in the temperature record that certainly has not followed the CO2 rise very well.

    • Svante says:

      That is a likely tale.
      We emit twice the CO2 increase in the atmosphere.
      But that has nothing to do with the atmospheric increase.
      Our emissions have no impact, the increase is natural.
      From an engineer that works for the oil and gas industry.
      Yeah right.

      • DMA says:

        Did Mr. Munshi make an error in his analysis that I missed? We emit about 3% of the total annual emissions that enter the atmosphere. There are many natural sources. None of them are constant and 3% is lost in the noise of their variability. Read Harde 2017 and then ( Or start like I did wondering how nature knew how to absorb just enough human CO2 each year so as to keep the rate of change of atmospheric CO2 constant. How did Mother Nature know when human emissions would slow down(2011) or speed up(2003)?
        Let me know if you got the answer. It eluded me.

        • Svante says:

          It’s like the temperature record, you have large natural variation and a small anthropogenic trend.

          These days we have global measurements, see fig. 2 here:

          • Chuck Wiese says:

            Yeah, except those natural variations are two orders of magnitude larger than human emissions and respond to temperature perturbations, and when they do, they drive the atmospheric concentration. It is absurd to think natural emission is static, leaving human emissions responsible for all atmospheric increases back to the industrial revolution.

            The whole CO2 climate change nonsense is a racket and fraud…failed models, manipulated temperature records, incessant propaganda derived from this complete crap.


            Chuck Wiese

          • Svante says:

            Good thing there are daily global measurements down to factory level:

          • Chuck Wiese says:

            Those measurements you tout don’t prove a thing. It looks like model simulation has supplanted reality.

            Perhaps you could take from the article and show where and specifically this “study” proves human emissions are driving atmospheric CO2. I couldn’t find that result trying to wade through all the “definitions’ and assumptions.

            Chuck Wiese

          • Svante says:

            CO2 sources and sinks are not a mystery anymore, we have satellite measurements.

            Like I said, look at fig. 2. You can see that the main sources are China, tropical deforestation, the middle east, europe and the US.

            The resolution is fine, here’s the UMM Said refinery and the Gillette coal field:

          • Chuck Wiese says:

            Did you even bother to watch the video you referenced? You don’t know what you’re talking about. We are nowhere close to knowing any accurate global measurements of CO2 emissive sources.

            For starters, the presenter ,Crisp, seems sure the huge land and ocean source/sink regions have known emissive and ab-sorp-tive properties. He never states how this is so, but is confident land and oceans absorb 10 GtCO2 more each than they emit on an annual basis and then contradicts himself showing a graph of varying ab-sorp-tion and then states he has no clue as to where the CO2 is going from the atmosphere.

            To assume natural emission and ab-sorp-tion from nature is static is just complete bullcrap. It is a fact that temperature perturbs emission and ab-sorp-tion. It’s basic chemistry and biology, but yet none of this instrumentation is anywhere near being able accurately track ab-sorp-tion or emission and has plenty of limitations.

            The calculation to try and determine emissive source depends on comparing the brightness of near infrared light on the oxygen band as a proxy emitting from 765 nanometers and comparing it to brightness from CO2 near infrared ab-sor-bing at 1610 and 2060 nanometers respectively. Then to complicate matters, profile atmospheric slices are taken from climate models and “adjusted” until brightness or emission back at the satellite matches the modified profile. Water vapor variability affects this because the calculation for the CO2 fraction is made for dry air, and water vapor varies continuously with each narrow longitudinal sweep around the earth and it takes a full earth rotation of 24 hours to get one longitudinal sweep. Then any cloud cover and the black ocean reflectivity completely block the ability of the satellite to accurately measure emission back to the satellite.

            Translation: On any given day, this satellite is only able to give an assessment of CO2 concentration or emissive source over 20% of the earth’s surface.

            This system is nowhere near being able to do what you claim which is to isolate sink and source regions of atmospheric CO2 around the earth.

            The pictures of the graphings you provided are as phony as a failed climate model and use those types of farsicle simulations to make a totally ridiculous and baseless claims from.

            Crisp admits they have a long way to go before they have a useful product. So the graphings you submitted in your references have as much credibility as a failed and phony climate model. But leave it up to NASA GISS to present images anyway and pretend the system is giving reliable results.

            Chuck Wiese

          • Svante says:

            It has been verified by ground based measurements.

          • Chuck Wiese says:

            It absolutely cannot do what you claim it can. On any given day, 80% of the earth is not able to be measured by its own limitations, which include water vapor variability, cloud cover and most all water sources.

            It doesn’t matter whether ground based stations verify any part of the small area the satellite can measure.

            You are clueless if you believe this satellite as equipped can map all of the true source and sink regions and the magnitudes of ab-sorp-tion and emission of carbon dioxide.

            Chuck Wiese

          • Svante says:

            It can take thousands of measurements every day.

            A Japanese and a Chinese satellite do the same.
            There are measurements from the ground, sea and air.
            The picture is clear.

            You can feed it into a global circulation model if you want to get a feel for the dynamics of it:

  3. E. Swanson says:

    Dr. Spencer, I haven’t seen the NATURE Climate Change.

    However, your comments regarding the agreement between models and measurements, particularly your UAH data, misses some important facts. Part of the problem is tropospheric aerosols, especially that from the rapid industrialization in Asia as both China and India have increased their coal burning to produce electricity and steel, as well as concrete. The continual emissions of sulfate and black carbon aerosols, which have greatly increased over the pst few decades, would be expected to add a cooling trend to the surface temperature data. Thus, I would not expect that the surface data would show as much warming as that from GCMs which do not include these aerosols.

    Your repeated presentations comparing your UAH products with model results also appear suspect to me. In your work, you must first make some conversion of either the model results or your data to place them on a “level playing field”. For example, you have applied a theoretical model of your MSU/AMSU measurements to the monthly temperature results from the KNMI data archive for computer model results, in order to simulate your TMT or TLT product. One may question how good your conversion is, for example, are the weighting functions used in the conversion a function of season or latitude? Does your conversion weighting adequately model the difference between land and ocean surface microwave emissions? What about the difference between the microwave emissions over sea-ice and open ocean at high latitudes?

    Sorry, I don’t think things are as clear as you claim.

    • Gordon Robertson says:

      swannie…”Your repeated presentations comparing your UAH products with model results also appear suspect to me”.

      Having debated with you and your strange views on science, I find your thought process to be suspect. There is nothing wrong with UAH data and the integrity of the scientists involved, like Roy Spencer and John Christy. There is a whole lot wrong with the pseudo-science programmed into climate models.

      What is really suspect is an out and out alarmist like Santer issuing such garbage.

      • E. Swanson says:

        Gordo, I’ve written 3 papers on the UAH data, 2 of which were peer reviewed. Not that they were exceptional papers, but I’ve at least done a lot of homework more than 20 years studying the mass of other published papers. You’ve yet to show any awareness of my work or that of others who have been critical of Christy & Spencer’s analysis, such as:

        Santer, B. D. et al. Comparing tropospheric warming in climate models and satellite data. J. Clim. 30, 373-392 (2017).

        Po-Chedley, S. & Fu, Q. A bias in the mid-tropospheric channel warm target factor on the NOAA-9 Microwave Sounding Unit. J. Atmos. Ocean. Technol. 29, 646-652 (2012).

    • Bob Roberts says:

      It seems like you’re insisting that Dr. Spencer needs to adjust reality to match admittedly poorly written models, which, as he points out, “cannot yet reproduce natural climate variability on multi-decadal time scales”, most likely in large part due to the fact they greatly over emphasize the role of carbon dioxide in affecting weather, temperature and climate.

      Normally when reality doesn’t match the model, it’s proof the model is wrong. Climate alarmists insist the models are correct and demand that reality be adjusted to match.

      That’s insanity.

      • E. Swanson says:

        Bob Roberts, First, one must verify that “reality” and the models represent the same thing.

        Satellite data from the MSU/AMSU instruments don’t measure surface temperature and the GCM models don’t provide results which match the satellite measurements. Christy and Spencer chose to simulate the MSU/AMSU MT and LT data, manipulating the output of the GCM’s by applying theoretical weighting to the model results for various pressure levels. All these efforts require “adjustments” to the data to compensate for various known problems, such as orbit decay and shift in the time of Local Equator Crossing (LECT).

        The model results are monthly averages, presumably that means averages which begins with gridded daily 24 hour average temperatures. The satellites only cross any latitude twice each day at the same local time in the diurnal cycle, so they can’t capture 24 hour averages. The repeat time over areas of the tropics can be about 2 days, so, again, there’s a question about whether the satellite actually provides a complete 30 average.

        Not to forget, there are 4 groups which have analyzed the MSU/AMSU data to study climate and the UAH v6 product indicates the least trend.

        So, what is your version of “reality”?

    • RealOldOne2 says:

      E.Swanson says: on Feb.27,2019 at 10:01AM “Part of the problem is tropospheric aerosols … The continual emissions of sulfate and black carbon aerosols, which have greatly increased over the past few decades, would be expected to add a cooling trend to the surface temperature data.”
      The only sulfates & aerosols that have demonstrated a cooling effect on global mean temperature are those from major volcanic eruptions that reach the stratosphere and reside there for a year or two. Even the IPCC agrees that tropospheric aerosols are precipitated out within a few days:

      “Aerosols undergo physical and chemical transformations in the atmosphere, especially within clouds, and are removed largely by precipitation. Consequently aerosols in the lower troposphere typically have residence times of a few days.”- IPCC, SAR, WG1, p.20

      E.Swanson says: “from GCMs which do not include these aerosols”
      Tropospheric aerosols are included in GCMs, as shown in this NASA webpage:
      You are just making stuff up to fabricate reasons for the failed predictions of your global warming alarmism belief system.

  4. Ron Clutz says:

    FYI, Lubos Motl weighs in on this, taking exception to claiming hard science standards can be applied to climatology.

    • John F. Hultquist says:

      Thanks Ron.
      I was going to mention it, but checked first.

    • Gordon Robertson says:

      ron…Lubos sums it up well in his closing paragraph:

      “Gavin Schmidt and everyone else who tries to paint hysterical climatology as a hard science analogous to particle physics is simply lying. Particle physics is a hard science and “five sigma proofs” are possible in it, climatology is a soft science and “five sigma proofs” in it are just marketing scams, and cosmology is somewhere in between. We all hope that cosmology will return closer to particle physics but we can’t be sure”.

      Gavin Schmidt is a mathematician who programs climate models. He talks as if he is a nuclear physicist. He was taken to task by engineer Jeffrey Glassman for failing to explain positive feedback correctly. That is scary…very scary. Positive feedback is programmed into climate models under the presumption that an amplification of heat is taking place in the atmosphere due to feedback from CO2.

      That is putting the horse before the cart. Positive feedback depends on an amplifier, feedback itself cannot cause amplification.

      • E. Swanson says:

        Gordo, No, the climate models do not have a positive feedback “programmed in” and there’s no device such as an amplifier. The models are based on the fundamental physical processes as the energy from the Sun flows thru the Earth System and back out to deep space.

        Warming from CO2 (or any other outside source) causes more water to evaporate, increasing the strength of the water vapor greenhouse effect. Rain can fall as water or frozen precipitation and as the Earth warms, there is likely to be more liquid and less frozen precipitation, as well as more rapid melting of the snow which actually falls. In a warmer world, there’s likely to be less sea-ice in the Arctic summer, which also results in less Sunlight reflected by the now darker surface, which then warms the surface. Changes in vegetation will also change albedo, which may go either way. These processes are calculated as effects, that is, they are not directly dependent on the CO2 level, but the temperature changes.

        Gordo, you are stuck with your boneheaded electrical engineering world view which only sees a feedback amplifier circuit, not the other processes which have been called “amplification” or “feedback” as a method to describe model results to the common man. Get over it, won’t you?

        • Bob Roberts says:

          There’s a lot of interesting and often overlooked information in your reply, but also a lot you left out.

          “Warming from CO2 (or any other outside source) causes more water to evaporate, increasing the strength of the water vapor greenhouse effect.”

          So first, let me point out that you also admit it’s water vapor, not carbon dioxide, that’s the actual potent yet misnamed “greenhouse” gas (because the atmosphere and the gasses in it work NOTHING like a greenhouse, which depends on changing convection, where these gasses alter radiation, a completely different process) which is great, since most garden variety (low-information, short attention span, i.e. typical) climate alarmists think that carbon dioxide has enormous warming potential and it simply does not, as your observations hint.

          You do point out some ways more water vapor might contribute to warming but I notice you left out that it also has great potential to prevent warming as well. Water vapor has great potential to be a strong NEGATIVE FEEDBACK source in ways you failed to mention – and it would appear that must be the case, so it’s net effect seems to be near zero. And the other issue you left out is any proof that we’re seeing any sort of increase in global average water vapor – it seems, based on going on 40 years of observations, that claims more carbon dioxide will cause atmospheric water vapor to increase simply haven’t come true, though I do recall a couple weak attempts to claim otherwise.

          You mention the reduction of Arctic ice – but I note that, like typical climate alarmists, you left out the year over year records of greater Antarctic ice that were set at the same time and you leave out that the Arctic ice has also showed some robust recoveries from recently measured low points. It would seem we have not seen a full cycle of Arctic and Antarctic ice trends since we’ve gained the wonderful modern tool with which we may now observe them. Therefore attempting to “attribute” any changes we see versus some randomly assigned “optimum” value have no valid basis.

          And speaking of that, what is the “optimum” average global temperature. There exist reasonable arguments that it is warmer than the current global average. Climate alarmists point out that change, which is a constant, and cannot be stopped, has some negative aspects, but leave out that it also has positive aspects. They tend to ignore the positive and give vastly overrated predictions of coming negative things (which never seem to actually appear), they exactly what Al Gore said was necessary to get people to fall for Catastrophic Anthropogenic Climate Change Alarmism when he insisted on “over representations of factual presentations”

          • E. Swanson says:

            Bob Roberts, We agree that water vapor is another greenhouse gas. Water vapor is already a large component of the global energy balance and the amount of water vapor in the troposphere is a function of temperature. Any increase (decrease) in the Earth’s temperature will result in an increase (decrease) in water vapor’s greenhouse gas effects. As a result, water vapor is called a “positive feedback”, since it tends to amplify what ever change may have caused the overall temperature to change. Furthermore, changes in water vapor will impact clouds in the atmosphere, which results in different influences that are still under debate.

          • E. Swanson says:

            (I’m having trouble posting, sorry)
            Folks who claim a warmer world will not be a problem ignore the fact that humans can not survive in an environment in which the dew point exceeds 35 C because our bodies are cooled by evaporation. There are already occasional instances of such conditions on Earth and further increases in the occurrence of such temperatures will surely lead to major problems. If you think the US immigration problem is critical now, wait until Central America becomes even hotter and more people begin to move North.

          • E. Swanson says:

            (I’m having trouble posting, sorry)
            The Arctic and Antarctic sea-ice situations are vastly different with the Arctic being a high latitude ocean surrounded by large continental land masses and the Antarctic being a large continent with high elevation ice fields surrounded by warm oceans. The Antarctic is also impacted by another man made problem, ozone depletion, and ozone is also a greenhouse gas. I suspect that the Antarctic sea-ice maximum extent may begin to decline as the Ozone Hole continues to heal.

    • Bob Roberts says:

      Yes, I love to point out that while climate alarmists (and the famous Mann “Hockey Schtick”) predict the top curve (worst case scenario), and we’ve actually exceeded the carbon emissions that were supposed to produce it, we’ve actually observed surface (or lower trophosphere) temperature trends which come in UNDER the “zero emissions” scenario presented. Meanwhile, due to multiple “revisions” in the raw data, all done by climate alarmists (Hansen, Jones, etc.) we do see “man-made” warming in global temperature data, but it’s all in the revisions, not in the original raw data.

  5. Gary says:

    I hope you are submitting this critique to Nature Climate Change for publication. Excessive claims need rebuttal, especially when they are wrong.

    • Gordon Robertson says:

      gary…”I hope you are submitting this critique to Nature Climate Change for publication. Excessive claims need rebuttal, especially when they are wrong”.

      You don’t think the alarmists running those journals would pay any attention to a critique by Roy, do you?

  6. Rex says:

    If the study is conceptually rigged, so to speak, why is the confidence level not even higher than 99.99997%?

  7. AZ1971 says:

    What the heck is a “signal to noise ratio” when it’s impossible to ascertain what the level of noise is? Given that ALL of 20th century warming is within statistical noise (based on the starting point of the 1850’s being abnormally cold, not “normal”) this reads as little more than virtue signaling at its finest.

    • Bob Roberts says:

      Indeed – for what is the actual “optimum” average global temperature.

      One can reasonably argue it is not the present one – way too much persistent cold, ice and snow, way too much land that is relatively inhospitable due to same.

      They say warming would expand the deserts, that this has occurred, yet the observations I’ve been reading about say the deserts are greening – coming to life, and I’ve seen proof this is happening in real time. In fact it’s going on right now in the U.S. desert southwest, I’m told and I plan to perhaps go see for myself, again as it’s wonderful to behold.

  8. garyh845 says:

    AXFORD: ‘Unlike GW over the past century. . Medieval warmth was localized’

    Alaska’s Exit & Mendenhall glaciers are local to Greenland?

    Trees recently exposed as the glaciers recede are carbon-dated to be 850-1000 yrs old. Same empirical findings in Alps & S. America.

    MWP was not localized around S Greenland.

    • Eli Mi says:

      garyh845, what exactly is your comment a reply to? It doesn’t seem to be responsive to either the blog post or any of the previous comments. Your comment is still interesting, if off topic, but please explain what it’s responding to, and give references to support your assertions. For example, what is the evidence for the Medieval Warm Period in South America?

      As for the Santer et al. study that is the subject of Dr. Spencer’s blog post, it sounds pretty pathetic that this is what prestigious peer-reviewed climate journals have resorted to, if his analysis is accurate. And all for the purpose of justifying some misleading headlines and stories in the lay press, one has to assume.

    • Bob Roberts says:

      One way “climate science” works is you go out and review past scientific findings and you declare them to be false and you substitute your own false claims for what actual scientists determined when they used real science to determine them.

      Jonathan Overpeck is quoted as saying, I get the sense that Im not the only one who would like to deal a mortal blow to the misuse of supposed warm period terms and myths in the literature”, which has been misquoted as We have to abolish the medieval warm period.

      Indeed there is strong evidence of efforts to “abolish the Little Ice Age AND the Medieval Warm Period” even though he’s misquoted as I pointed out. Still the admission inherent in the quote is that climate alarmists love to claim that current warming is unprecedented, yet it isn’t. And in fact, as admitted by the IPCC (AR5), “current warming” has slowed or perhaps even stopped, despite continued increasing carbon dioxide. But with all the claims of “hottest ever” which are still made despite the increase being less than the admitted margin for error in the estimates used to support such claims, realities such as these get drowned out by the stampeding alarmists who are primed to respond to appeals to authority and appeals to emotion, thus unable to respond to facts, science, logic and reason.

  9. ren says:

    The weakening of the magnetic field in the Western Hemisphere exposes North America to severe winters during periods of low solar wind activity.
    Meridional jetstream prevents the development of El Niño.

  10. ren says:

    Stratospheric intrusion in the northeast of the US.
    The wave of cold air from the central states will shift to the east the US.

  11. Bart says:

    “Yet their statistic seems to treat each of the 576 gridpoints as independent, which should have been taken into account (similar to time autocorrelation in time series).”

    They do that a lot. Like, when they note that N of the last M years are the hottest on record. If those years were independent, it would mean something. Since they’re not, it means very little at all.

  12. Svend Ferdinandsen says:

    Thanks for that: “Importantly, we don’t understand natural climate variations, and the models don’t produce it, so CO2 is the only source of warming in today’s state-of-the-art models.”

    I will try to remember it, because it is in the hart of the models.

  13. Jack Dale says:

    The headline is questionable. The 97% consensus has nothing to do with 5 sigma certainty level.

  14. Mike Flynn says:

    Lubos Motl, a real scientist (a physicist, even) wrote –

    “Gavin Schmidt and everyone else who tries to paint hysterical climatology as a hard science analogous to particle physics is simply lying.”

    Surprise, surprise!


    • Entropic man says:

      He’s right too. Climate science is considerably more complex thanparticle physics.

      • Mike Flynn says:


        Climate is the average of weather – no more, no less.

        No science there at all. That’s why an incompetent mathematician (who claimed a 38% likelihood meant near certainty – what a fool) depends on simpletons such as yourself to believe that he is, in fact, a “climate scientist”.

        You can’t even describe the GHE, can you? Some science. Pseudoscientific fantasy writ large, by a gaggle of bumbling buffoons, in fact.

        Keep worshipping. Who needs facts, when you have faith, eh?


      • Gordon Robertson says:

        entropic…”Hes right too. Climate science is considerably more complex thanparticle physics”.

        I hope you’re kidding.

  15. Edward Caryl says:

    If the models are written to match the increase in CO2 to the warming, and then they succeed in that, where is the surprise? Isn’t this simply circular reasoning with a very small circle?

  16. TomC says:

    Not considering the nearly insignificant amount of CO2 (~.03% or so) in the atmosphere, a more interesting question might be from where & how did the the massive ice sheets ‘arise’ so to speak, in the last IceAge, around 12000 years ago? The ice was basically solid water. That’s a horrendous amount of ice. The preceding amount of water had to come from somewhere, constantly sending moisture south to become snow.
    Recently ancient trees, remnants of mammoths, etc are being found more & more. Where were they during the building of the sheet of ice in Polar regions?

    Some papers from the ’60s were talking about the creation of the Ice sheets, would be accompanied by unusually amounts of rain in the southern areas & much more than usual snow in the areas from the Gulf of Mexico north through the area to the Arctic Sea. Parts of Idaho have reported over 10 feet of snow a couple of days ago.

    Is ‘what’ is trying to tell us something???? Probably need to pay attention to next Winter, the year after that, and the year after that, and……..

  17. Devvon Vienna says:

    I dont support climate change. I dont believe in it. Imhppov, it is like religions and their gods, to scare people into hell if they do not believe. That’s climate change. Second to religion than any terrorist organization.

  18. Stephen P Anderson says:

    Just damn, Dr. Spencer done peed on your party.

  19. pochas94 says:

    It’s what happens in Communism. You are required to believe what “they” tell you. 1984 is getting closer.

  20. Mike Flynn says:

    A quote from Richard Feynman (quite a clever fellow) –

    “However, the theory of meteorology has never been satisfactorily worked out by the physicist.”

    And climate is the average of that which has never been satisfactorily worked out by physicists. Models purporting to state the contrary are a joke, albeit an expensive one on the taxpayers.

    Ben Santer is deluded. He cannot even produce a testable GHE hypothesis. Just another foolish pseudoscientist, disseminating fantasy as fact. Science? Hardly.

    About as stupid as –

    “Atmospheric CO2: Principal Control Knob Governing Earths Temperature” – peer reviewed and all. Obviously, the authors’ peers are as deluded as the authors themselves.

    All part of the rich tapestry of life.


    • Gordon Robertson says:

      mike…”A quote from Richard Feynman (quite a clever fellow)

      However, the theory of meteorology has never been satisfactorily worked out by the physicist.”

      He also claimed that quantum physics works but that no one knows why.

  21. Transport by Zeppelin says:

    Roy; did Santer forget to mention that the models show negative lapse rate feedback, a fundamental result of greenhouse forced warming, while the satellite data shows the opposite?

    This is one of the biggest failings of the models & the theory, as Christy & Ross McKitrick pointed out

    • Randy Bork says:

      Has there been any attempted refutation of that Christy and MicKitrick paper? It seemed rather devastating on its face, and if no one has demonstrated a fatal flaw in its reasoning now 8 months later… Here it would seem Santer and co. are not using an independant model output as a measure of it [quoting from Christy McKitrick July 2018, “Third, the independence criterion means that the target of the prediction must not be an input to the empirical tuning of the model. Once a model has been tuned to match a target, its reproduction of the target is no longer a test of its validity. “

  22. In my current study comparing the CO2 record for Mt Waliguan Observatory on the Tibetan Plateau with the UAH satellite lower troposphere temperature, August 1990 to December 2017, applying first order autoregression analysis gave:
    for CO2 vs Northern Extension zone temperature, a Pearson correlation coefficient of -0.062 with a probability of zero correlation of 27%,
    for CO2 vs Northern Extension zone land temperature, a Pearson correlation coefficient of -0.052 with a probability of zero correlation of 36%,
    for CO2 vs Northern Extension zone ocean temperature, a Pearson correlation coefficient of -0.073 with a probability of zero correlation of 20%,
    that is, either CO2 relates to cooling temperature or there is no causal relationship.

    This is similar to results obtained for other localities, so where is the anthropogenic signal?

    Further, statistical analysis has shown a significant correlation between the annual rate of change of CO2 concentration, at a number of locations, and the UAH satellite lower troposphere Tropics temperature. This implies that temperature is either a direct or indirect cause of the CO2 concentration across the globe. Again no anthropogenic signal. The CO2 is generated in the Tropics, being the area of greatest average temperature, and spreads North and South towards the Poles where it is precipitated in the rain and snow due to its greater solubility at lower temperature.

    For detail see

    The CO2 records all show a distinct 12 month cycle due to the seasons, that is, climate driving CO2. In Spring the CO2 drops due to photosynthesis by new crops but temperature rises. Then in Autumn and Winter, the CO2 rises but temperature falls. The pattern is consistent over years with no sign of an anthropogenic signal and is in the opposite direction to the UN IPCC claim.

    Ice core data has shown that CO2 and temperature rose in unison during inter-glacial periods going back half a million years, before mankind even existed. Aren’t we simply seeing this pattern repeated now regardless of mankind’s presence ?

  23. Aaron S says:

    Two observations are evident: 1. There is a clear anthropogenic signal for global warming, and 2. a discrepancy between predicted amount of warming and observations from satellite data exists and the models are predicting more warming than is observed.

    RSS explains the slow warming rate well:

    Why does this discrepancy exist and what does it mean? One possible explanation is an error in the fundamental physics used by the climate models. In addition to this possibility, there are at least three other plausible explanations for the warming rate differences. There are errors in the forcings used as input to the model simulations (these include forcings due to anthropogenic gases and aerosols, volcanic aerosols, solar input, and changes in ozone), errors in the satellite observations (partially addressed by the use of the uncertainty ensemble), and sequences of internal climate variability in the simulations that are difference from what occurred in the real world. We call to these four explanations model physics errors, model input errors, observational errors, and different variability sequences. They are not mutually exclusive.

  24. Crakar24 says:

    I take it santer has given up on smearing the reputations of operators of millions of weather balloons over the years as incompetent because the balloon data did show his hot spot to exist?

    Obviously his wind sheer has gone bust. Evidence by damned seems to be his new approach to life he has fallen quite far after he represented a zero trend in “fire engine red” and got caught.

  25. Aaron S says:

    Entropic man related to hot spot. Lol… they updated another data set by removing data to see the real trend. Haha. I used to believe this sh!t until the sunspot number was manipulated and can demonstratably be shown false with empirocal data. Credibility is zero.

    “The new dataset was the result of extending an existing data record and then removing artefacts caused by station moves and instrument changes.”

  26. Transport by Zeppelin says:

    RAOBCORE, RICH, and RATPAC, RSS & UAH all say no hot spot.

  27. Stephen Bodner says:

    There are three fundamental problems with the climate models.

    First, the ratio of horizontal to vertical grid size is very large. As a result, the computer equations can not include vertical air flow. A major portion of vertical energy transport is via the upward transfer of moist air. To deal with this problem, the computer modelers use made-up “parametric equations” that have no basic scientific validity.

    Second, the motion of ocean currents can not be calculated for the same basic reason: the horizontal to vertical grid ratio is too large. We also do no have the experimental measurements to calibrate the computer models. The multi-year and multi-decade variation of ocean currents has an important role in modifying atmospheric temperatures and cloud formation.

    Third, clouds play a major role in absorbing and emitting infrared radiation, and in reflecting sunlight. The clouds are too small and too variable to be accurately resolved by the computer grids. In any case, there is insufficient understanding of cloud formation to determine, as the atmosphere continues its slow warming, their future impact on the reflection of sunlight or on infrared transport.

    None of the above three objections are very relevant to weather predictions, because weather predictions are not primarily concerned with accurate calculations of vertical energy flows.

    However, the total worthlessness of their computer modeling of climate does not imply that CO2 will not in the future have a large impact on our weather. Perhaps even the impact could be much worse than any of the climate model predictions. We simply can’t do this science. Do we want to take a chance and assume that the actual warming will continue to be less that what is predicted by these worthless computer models as we dump more and more CO2 into our atmosphere?

    • Entropic man says:

      Stephen Bodner

      “Vertical energy flows”

      I don’t. Know where you got the idea that weather and climate models cannot measure verical energy flows.

      The most recent weather forecasting models use a 10km grid.

      Since the average height of the tropopause is also about 10km that makes each grid square a 10km by 10km by 10km cube.

      CMIP5 uses 50km by 50km grid squares, so the vertical component is 20% of the width.

      There is also the problem of precipitation. You cannot calculate that without vertical transfer of water vapour and latent heat

      • Stephen Bodner says:

        Computer modeling of fluids converts mathematical equations, which have continuous variation, into grids. The equations are then algebraic difference equations. In each grid there is only one value of temperature, humidity, cloud cover, air flow, etc. Adjacent grids must have nearby values, or the solution becomes nonsense. In weather predictions of part of the Earth, in the horizontal direction, the grids have now been reduced to about 10 kilometers, which is great. The grid size is larger for climate modeling. But in the vertical direction the troposphere is of course not uniform in temperature, etc, and is typically divided into 20 or more grids. If the grids were uniformly spaced, each vertical grid would be 1/2 kilometer. That would produce a grid ratio of 20. But actually the grids are not uniformly spaced, to better resolve some altitude regions. So the ratio is greater than 20, and has typically been in the range of 50 to 1000. If the grid ratio is more than two, mathematical errors are introduced when solving the equations. With a grid ratio of 10 or more, the equations effectively split into separate vertical and horizontal flows, with no good coupling between horizontal and vertical directions. Effectively, the climate models are solving a layered set of two-dimensional fluid, with coupling in the vertical direction via made-up parametric equations.

        • Rob says:

          We cannot solve partial differential equations like Navier Stokes, so we model by linearising non linear heat and mass flows across an entire planet and expect to get useful outputs, how?

    • crakar24 says:

      4, They dont include X-rays, GCR, solar wind, CME’s, magnetic effects, electric currents etc, etc , etc and then have the audacity to say they cant explain the extra heat so it must be CO2

  28. Jos says:

    There exist several statistical significance correction methods in case of a Multiple Comparison Problem (Bonferroni correction, Šidák correction, Holm–Bonferroni method).

    The simplest is the Bonferroni correction, which states that for a chosen significance level “alpha”, in case of Multiple Comparison, the appropriate significance level is not “alpha”, but “alpha” divided by the number of comparisons.

    In this case, assuming the authors consider the 5% confidence level as significant (0.05), the significance level for the Multiple Comparison would be 0.05/576 = 0.0000868 (99.991322% or slightly less than 4-sigma).

    That would mean that in reality, if this would have been less than a 4-sigma result, it would have to be considered as a random fluke.

    The claimed 5-sigma (99.99997%) is still better than the 4-sigma, but it is by far not as impressive as 5-sigma suggests.

    The Šidák correction is less conservative, and would put the confidence level around 3-sigma.

    One can debate whether 576 is the correct value for the number of comparisons.

  29. Art Viterito says:

    Excellent article! I was happy to see that you mentioned changes in the thermohaline circulation as a significant driver of global climate. Here is my “2 Sigma” prediction

  30. Hugo says:

    This study is of course nonsense. I tested the latest UAH dataset (global average lower tropospheric temperature ) for a unit root using the Dickey–Fuller test. What you get is a 5-sigma level rejection (1.56*10^-7). So temperature anomalies are likely not a “random walk with drift”. That’s all. I believe this is the statistics they are picking up. This can be reported with a few sentences. The underlying mechanisms though remain unkown.

  31. Adrian says:

    “so any unforced trends in the models are considered “spurious” and removed” So they deny spontaneous climate variations. Lorenz must be spinning in his grave…

  32. ren says:

    The arctic air wave has reached Texas and will shift to the southeast of the US.

  33. Ilkka Ponkanen says:

    The multidecadal signature is AMO:

    1900-30 negative
    1930-63 positive
    1963-95 negative
    1995 – positive

    The turn towards negative AMO will take place within two years time and AMO will be negative around the year 2025. This negative phase will last until approx. the year 2058.

    PDO seems to be cooling and Solar Minimum has started.

    Global temperature anomaly will return to the same level it was in 1979 by the year 2025. This is normal climate variability and has nothing to do with AGW.

    • Gordon Robertson says:

      Ilkka….a study was done years ago by Tsonis et al which studied the relationship of all the oscillations over a century. Tsonis concluded that when the oscillations are in phase the globe warms and when out of phase it cools.

      Tsonis personally suggested we should be studying the oscillations as the source of warming rather than wasting time trying to prove CO2 is the culprit.

  34. John Bills says:

    No scruples in gaining money.
    In 2014 a paper that implicates that all warming from 1993 on is natural:
    In 2018 implicating the way models are used is suboptimal:
    In 2019 almost 100 % sure:

  35. Entropic man says:

    Ikka Ponkanen

    I looked up the GISS smoothed average temperature anomalies for the middle of each AMO phase you defined.

    1915 AMO negative -0.29C
    1946 AMO positive 0.00C
    1979 AMO negative 0.07C
    2011 AMO positive. 0.63C

    Each value is larger than the previous one. This indicates that the warming due to the long term trend is larger than the warming or cooling change due to the AMO.

    The long term warming g has also accelerated. The change from one negative in 1915 to the next negative in 1979 was 0.29+0.07 = 0.36C

    Half a phase later the change from one positive in 1946 to the next positive in 2011 was from 0.00C to 0.63C, an increase of 0.63C.

    On the basis of this data I would expect 2025 to be warmer than the present, not cooler.

  36. Entropic man says:

    John Bills

    For me this, from your second link, is the most interesting.

    It shows that even UAH is detecting a 5 sigma warming trend.

  37. Craig T. says:

    “Current climate models are programmed to only produce human-caused warming”

    Current climate models are programmed with all known relevant forcings. The forcing changing the most is CO2. If any model programmed with that information shows warming with increases in CO2 but not when CO2 levels stay at pre-industrial levels that says a lot about the cause of warming.

    • JDHuffman says:

      Craig, CO2 does not produce any “forcing”. It is NOT an energy source.

      And, you can program a “model” to show any results you want.

      “With four parameters I can fit an elephant, and with five I can make him wiggle his trunk.” –John von Neumann

      • Entropic man says:


        A forcing changes the rate of energy flow through the climate system. It is not itself an energy source.

        A tap controls the flow of water, but does not itself produce water fron nowhere.

        CO2 reduces the outward flow of energy from the climate system to space. The more CO2 the greater the reduction.

        Reducing the rate of outflow below the rate of inflow causes heat to accumulate in the system. More heat, higher temperature.

        • JDHuffman says:

          E-man, that’s not how it works.

          If you want to “model” atmospheric CO2 as a water valve, then adding more water valves would just allow more water flow.

      • Craig T. says:

        It is difficult to create a model that consistently matches past temperatures without a good understanding of the causes of global temperature. And forcing affect temperature. You can’t create a model that matches past temperatures without including CO2 levels.

        If you want to argue that you can program a model to show any result, make a model that matches past temperatures without including CO2. That would be a useful tool for arguing against AGW.

        • JDHuffman says:

          Craig, programming such a model can be very involved, as you know, and require a lot of funding. I can get you the model that matches past temperatures, without any contribution from CO2. And, for the very low price of $2,000,000.

          Let me know when the funding is available. I can provide the model within 24 hours.

          • Craig T. says:

            Ask the Heritage Foundation, that’s more of their thing. Also try EPA head Andrew Wheeler. He’s still close to Murray Energy since Wheeler lobbied for the company up until his appointment. Robert Murray has lots of money he didn’t use to improve mine safety and agrees with your views on climate.

        • Dan Pangburn says:

          The model at matches reported temperatures 98.3% 1895-2017 without considering CO2.

          • Stephen P Anderson says:

            That’s remarkable.

          • Craig T. says:

            Looking at the model at your link I see a problem. Here are two factors listed for the equation:

            T(i) = AGT calculated by adding T(1895) to the reported anomaly, K
            T(1895) = AGT in 1895 = 286.707 K

            If the temperature anomaly for the year is one of the factors in the equation then the model is not predicting the global temperature anomaly.

          • Dan Pangburn says:

            CT,, The real message here is that CO2 has little if any effect on average global temperature and therefore climate. That conclusion used measured data through 2013 and the UAH v6.0 trend 2014-2017. The prediction has nothing to do with it.

            T(i) is used only to calculate the change to the radiation through the atmospheric window. Reported anomalies are used through 2013 resulting in a cumulative effect of the contribution of about -0.0177 K (about 2 % of the total temperature change from 1895) for this factor. I estimated anomalies 2014-2037 for this factor with maximum contribution of -0.026 K in 2037 (about 3 %). The anomaly prediction includes the uncertainty in my estimate of the factor which produces the 3 %.

  38. Joe H says:

    Next step they will publish that 103% of scientists are in consensus on climate change.

    A justification for such a figure could be that any large group of people will have a few with multiple personalities – each of which is entitled to its voice.

    • Craig T. says:

      The paper doesn’t say 99.9% of scientists agree that AGW is happening. It says that the data only has a 1 in 3.5 million chance of not being caused by a warming trend.

      It is fair to say that it doesn’t say anything about the cause of that warming.

  39. Gerry Dail says:

    Dr. Feynman was right.

  40. The new Ben Santer et al. paper, uses the “gold standard”. A standard which is no longer used by anybody.

    The gold standard is a monetary system where a country’s currency or paper money has a value directly linked to gold.

    The gold standard is NOT currently used by any government. Britain stopped using the gold standard in 1931 and the U.S. followed suit in 1933 and abandoned the remnants of the system in 1971.

    The gold standard was completely replaced by fiat temperatures, a term used to describe temperatures that are used because of the IPCC’s order, or fiat, that the temperature must be accepted as proof of global warming.

  41. Roy,

    Thanks for more great information. You continue help your non-science part of your readership respond to the Green Tabernacle Choir.


  42. Anna V says:

    Dr Spencer,

    I have remarked elsewhere that in the popularizations they claim to depend on an analysis method by K.Hasselman, .

    It is behind a paywall , but in the abstract it says the following:

    “A statistically significant climate change signal is regarded as consistent with a given forcing mechanism if the statistical confidence level exceeds a given critical value, but is attributed to that forcing only if all other candidate climate change mechanisms (from a finite set of proposed mechanisms) are rejected at that confidence level. ”

    If their analysis has followed this proposal , there should be more figures than their figure 1, showing that all the other candidate climate change mechanisms they have considered are rejected at 5 sigma.

  43. gallopingcamel says:

    “Climate Science” is almost as squishy as “Behavioral Science” which struggles to reach 2 sigma (~5% chance that the null hypothesis is true):

    Even the “Hard Sciences” have trouble with 3 sigma. The announcement of the mass of the Higgs boson (125.09 ±0.21 GeV/c2) was delayed by ~6 weeks until the 3 sigma standard could be met.

    AFAIK know there is not a single finding in climate science that can pass the 3 sigma test. I challenge the many learned people who visit this site to prove me wrong. You only need one example to do that.

    Ben Santer has always struck me as a bully and a braggart but now he has excited my curiosity. Has he somehow discovered a Unicorn? However I am not curious enough to part with good money to find out.

Leave a Reply