The Nature of the “Scientific Consensus” on Climate Change

My previous post discussed a little-known UN poll that has so far attracted seven million voters from all over the world and which despite being totally unscientific nevertheless provides some interesting insights about public attitudes towards climate change. Here I discuss another little-known climate change poll that also provides some interesting insights but which is otherwise about as different to the UN poll as it’s possible to get. Why? Because there were only 286 respondents, not seven million, and they aren’t just anybody. They’re all climate scientists.

I refer to the Bray & von Storch poll entitled A survey of the perceptions of climate scientists 2013 . It’s quite a poll. It doesn’t ask just a few questions. It asks no fewer than one hundred and thirty-one, some of them highly technical (“The current state of scientific knowledge is developed well enough to allow for a reasonable estimate of the effects on climate of surface albedo?”). Yet 286 of the 4,491 people to whom the poll was sent responded, and most of them to all 131 questions – a respectable level of response to an internet poll considering the time it would have taken to fill it out.

The poll is weighted towards the “consensus” viewpoint. The invitees included “authors of climate related papers in peer reviewed climate related journals …. authors who contributed to Oreskes’ (2004) published conclusions concerning consensus in the climate change issue …. the IPCC list of contributors” and those on “readily available email lists from institute web sites (i.e. NCAR (US National Center for Atmospheric Research), MPI (Max Planck Institute), AMS (American Meteorological Society)) etc”. Almost all the respondents were affiliated with universities or government-funded research organizations and almost half of them had been involved as authors or reviewers of IPCC reports.

But the poll contains a graphic that is arguably the best illustration yet published of what the climate change “scientific consensus” really looks like. We find it on page 59:

268 of 272 climate scientists think that humans have caused at least some of the warming since 1850, representing a 98.5% consensus. But if we use the IPCC’s claim that “most” (i.e more than 50%) of the warming  was human-caused as the criterion the number drops to 81.2%, and if we insist on all of the warming being human-caused, which is essentially what the IPCC’s climate models show, it drops to only 6.3%. Clearly the “scientific consensus” on climate change can be quantified only if we put a hard number on what percentage of observed warming has to be caused by humans before climate change becomes “significant”. (A 12.5% human-caused warming threshold gives a 97.5% consensus among the respondents, incidentally. The oft-quoted 97.5% number seems to have originated in the 2009 Doran poll.)

(What’s your estimate of the percentage of human-caused warming since 1850? Feel free to provide a number. My estimate is either zero or 50% depending on which of my two pet theories sounds more plausible to me at the time, so I give it 25%.)

A full discussion of the results of the Bray & von Storch poll is beyond the scope of a single blog post, so here I will briefly touch on what I consider to be a few of the more revealing responses in something approximating a logical sequence:

First, most respondents regard climate change as a potential threat to humanity but aren’t unanimous as to how serious and dangerous it is:

They are also mostly in favor of taking immediate action to cut global greenhouse gas emissions (“mitigate” means emissions cuts, “adapt to” means higher sea walls etc.) although again there is no unanimity as to how great the need for immediate action is:

They are, however, somewhat less certain as to whether we can predict what the effects of climate change are going to be:

And they can’t agree at all as to whether recent climate-related disasters were caused by climate change or not:

They also have surprisingly little confidence in the predictive skill of climate models only  50 years into the future. In every one of the examples shown below the number of “very poor” votes in fact exceeds the number of “very good” votes. (Note that all the cases except sea level rise apply to regional rather than global climate models because regional models are usually used to evaluate specific climate change impacts. Levels of confidence in global models, however, are not much higher):

I’m going to let these results speak for themselves and leave it up to readers to form their own opinions.

This entry was posted in Climate change, Political commentary and tagged , , , , . Bookmark the permalink.

36 Responses to The Nature of the “Scientific Consensus” on Climate Change

  1. Gavin Doyle says:

    All the more telling when you consider that the livelihoods of these 286 individuals depends on the veracity of the anthropogenic warming models.

  2. bobski2014 says:

    Above, you observed …

    ” First, most respondents regard climate change as a potential threat to humanity but aren’t unanimous as to how serious and dangerous it is:”

    This just seems silly to me, a non-scientist and ex advertising/marketing chap. While a 6.3% cohort response to a >4,000 list is pretty good, it is nevertheless self-selected and therefore inherently suspect.

    It would appear possible that the selection comprises those who see any type of climate change only in a negative light.

    If there is no potential upside to a change in global climate, then at some point in time there was or is an optimum for all types of life on the planet. Any movement away from that is according to this argument, dangerous. This presumes does it not, that ALL types of life optimise with the SAME climate mix. That seems unlikely, if only because plants do rather better with more CO2. So are we only talking about humans? Do we care about anything else?

    What appears to be missing from the arguments of those in the 281 cohort is a definition of the supposed optimum. All their talk, their papers, their expensively funded “research” appears to be founded only on “change”; never on absolutes. Why is this?

    If they’re so damned clever why will they not take pity on this non-scientist ignoramus, and tell me please exactly what the optimum is and when it occurred. Then the UK’s wondrous and hugely expanded team at the risibly entitled Department of Energy and Climate Change could get on re-establish climate perfection.

    I’m not a complete ignoramus by the way, I think I have a fairly good handle on what the game is really about.

    • Leo Smith says:

      Change is always a threat to the status quo.

    • it is nevertheless self-selected and therefore inherently suspect.

      The respondents are both self-selected and pre-selected. Most of the bias was generated in the pre-selection process, which resulted in the poll being sent dominantly to people with ties to the IPCC or to “warmist” scientific institutions. But that’s actually good because the “97.5% consensus” is based on opinions from similar groups of people.

      We have of course no way of knowing how representative the 286 respondents were of the sentiments of the >4,000 total invitees, but the distribution of responses is about what I would expect to see from a group of IPCC climate scientists (right skewed here, left skewed there, higher overall confidence in this, lower overall confidence in that etc.) The fact that the respondents were able to answer complex technical questions related to climate model performance also suggests that they are in fact climate scientists.

      tell me please exactly what the (climate) optimum is and when it occurred.

      That’s easy. Optimum global climatic conditions occurred in pre-industrial times before humans started monkeying with the climate – i.e. around 1750. It’s automatically assumed that any deviation from this pre-industrial optimum will have negative impacts.

  3. Euan Mearns says:

    UK temperatures since 1956 – physical models and interpretation of temperature change

    My wild guess is that 49% of warming is due to CO2 and 51% natural. Very roughly speaking this yields a climate sensitivity of 1.28˚C. But on a bad day I too sometimes wonder if it is not closer to zero.

    One of the main problems with this whole debate is that the warmists do not prescribe different energy policies for different outcomes. Roger, in one of the many threads a few weeks ago you posted a couple of charts. One comparing climate sensitivity to Hadcrut and the other – I can’t remember what the other one was – had lots of lines on it 😉

    • sam Taylor says:


      Regarding their policies I suspect it’s because, much as you aren’t a climate scientist, they are not scholars of energy, and aren’t aware of complex concepts such as EROI, liquid fuels shortages, how inadequate renewables and storage are when integrated into our current grid setup and all the other stuff that goes along with all this. Largely they probably think you can just build a load of wind farms and off you go.

      I rather suspect if they had a more thorough understanding of global energy dynamics they would be tearing their hair out.

  4. Euan Mearns says:

    Roger, thanks for bringing these stats to our attention. I hope that all who have used the 97.5% statistic to try and advance their cause hang their heads in shame.

    You have to start by asking for a definition of what a “Climate Scientist” is. The only conclusion that I have ever been able to reach is that Climate Scientists believe that CO2 is the principle cause of global warming. It is therefore not surprising that 97.5% of climate scientists believe that CO2 is the principle cause of global warming. Its like saying that 97.5% of Christians believe in Jesus (to add a festive flavour to my comment 😉

    So how does the story go. 268 self selected individuals out of total sample of 4491 believe that at least 12.5% of warming is down to Man?

    • Ed says:

      My definition of what a “Climate Scientist” is: someone who has collected real life data, analysed the data and published peer reviewed papers about climate change. If 97.5% of climate scientists believe the principle cause of global warming is man-made CO2 then it’s good enough for me. Will I hang my head in shame? Certainly not.

      Secondly; Jesus probably did exist, the issue is whether he was the son of god. It is not something that can be proved or disproved by collecting and analysing real life data. Therefore it doesn’t matter what Christians, or any other faith group for that matter, believe in. Don’t equate faith with scientific modelling.

      • Euan Mearns says:

        If 97.5% of climate scientists believe the principle cause of global warming is man-made CO2 then it’s good enough for me.

        You seem to be having very serious comprehension problems. 97.5% of a biased unrepresentative group believe that at least 12.5% of warming is down to Man.

      • Ed: Pick a percentage:

        Humans caused 10% of the warming: Consensus 98.5%
        Humans caused 20% of the warming: Consensus 97.1%
        Humans caused 30% of the warming: Consensus 94.1%
        Humans caused 40% of the warming: Consensus 90.1%
        Humans caused 50% of the warming: Consensus 86.8%
        Humans caused 60% of the warming: Consensus 81.6%
        Humans caused 70% of the warming: Consensus 73.2%
        Humans caused 80% of the warming: Consensus 53.3%
        Humans caused 90% of the warming: Consensus 26.6%
        Humans caused 100% of the warming: Consensus 6.3%

    • Its like saying that 97.5% of Christians believe in Jesus

      Good one, Euan.

      One of the more interesting features of the poll is that while most respondents are convinced to a greater or lesser extent that climate change poses a serious and dangerous threat and that immediate action is necessary to combat it, they have limited confidence in the predictive skill of the climate models that the threat predictions are based on. This looks like religion in action too.

      • Ed says:

        I somehow knew that you would like that line of Euan’s in advance, Roger. Oh well. Keep up the good work. You keep me amused at least.

        Climate scientists are like religious fruitcakes. I get it.

        • Ed says:

          Can I retract that last line. It was out of order, offensive and I shouldn’t have written it . I’m sorry.

          (RA note: I don’t look on climate scientists as religious fruitcakes but the comment is in order. I was the one who brought religion up)

      • Roger Andrews says:

        Ed: I see that after giving only minimal consideration to the issue you’ve already lowered the “scientific consensus” percentage from 97.5% to 73.2%.

        And you can lower it a lot farther. According to the IPCC’s climate models humans caused not “most” but all of the warming during the second half of the 20th century:

        And at this level the “consensus” drops to 6.3%.

        97.5% really is a meaningless number.

      • sam Taylor says:

        It looks more like an understanding of the nature of complex systems, if you ask me.

  5. jacobressj says:

    “Climate Scientists believe that CO2 is the principal cause of global warming.”
    And Economists believe in printing money, Doctors in prescribing pills, Lawyers in suing, and Surgeons in operating.

  6. A C Osborn says:

    Whatever Anthropogenic CO2 can possibly contribute to the amount of warmth at the Surface (and I believe it is zero) it is completely negated by the various forms of H2O, convection/conduction etc.
    As to CO2 in total I still believe it to be a miniscule amount compared to the Oceans, H2O and the Adiabatic process. In fact I would still like to know where the “Doubling of CO2” starts on the logarithmic scale, because if it starts at 1ppm then it is totally saturated at about 32ppm.
    See this article by David Archibold which suggests that CO2 added it’s major contribution to warming billions of years ago

  7. Olav says:

    There are different views on this issue and it may be so that man made emissions has negligible effect on climate. I certainly hope so as I do see the difficulties into getting the consensus to do something effectively.
    But hoping that it will play out well is risky. We have fire insurance on our house even as the risk of fire is very low. The risk of serious climate problems is higher… and repairing afterwards is impossible…
    Reductions in CO2 emissions will not be effective unless we do it at a rather drastic scheme. Smaller reductions will not help as burning the remaining easy accessible carbon fuels over 80 years rather than 40 has little effect on CO2 levels in year 2100. Therefore drastic actions are the only way forward.
    Measuring temperature to monitor the course of climate change is difficult to get accurate. Strictly monitoring on sea levels is easier to follow as land level changes due to latest ice age are well understood and easy to adjust for.
    A 2m sea level rise (which will take centuries) is not only a catastrophe for low lying areas as river deltas, islands, coastlines and cities like Miami. It will make every jetty unusable globally. Just as an trivial and not life threatening example: There will not be a single usable beach left on the Northern Shore of the Mediterranean as all beaches there are developed by infrastructure just behind. In Libya and alike places can the ocean create new beaches as the sea level rises but…
    Off cause there is a risk of unnecessary hardship caused by implementing drastic measures which perhaps was not necessary but think about it… Mankind has a 10 000 years of easy documented history before now and hopefully we will have a longer stay on this planet ahead of us. 10 000 years ahead is 333 generations. Letting 10 generations consume all fossil fuels on this planet leaving nothing for the future generations gives a good argument to change course regardless of uncertainties about the dangers from burning the bulk of fossil fuels in the near future.
    I think the world needs a global CO2 tax at a level high enough to make alternatives like nuclear, wind and solar competitive without being subsidized.

    • johndroz says:


      Your comparison is to say that you pay for fire insurance on your house to cover that possible risk. I also have fire insurance — which is about .001 of the value of the property I am protecting ($100± on a $100,000 home).

      Few would object to a .001 true coverage cost for AGW.

      A more accurate comparison is your fire insurance is now $10,000± for a $100,000 home. Would you still think that makes sense?

  8. Paul says:

    Interesting article Roger. I see only 78 questions, not 131. Also there’s quite a lot of questions where the number of answers is not 286 (I saw as low as 230, but I didn’t read thoroughly). The comments at the end are interesting, in parts. Many criticise the number and quality of the questions. There’s even one who says he’s not a climate scientist (out of 286 that is negligible).

    If we think about climate sensitivity probability distributions, it is not really a puzzle that many scientists were uncertain about model predictions yet nevertheless thought that climate change is a danger that should be addressed. I’ve not seen one distribution (even from Lewis etc) that doesn’t have a long tail into significant temperature change. In that context, models are irrelevant. The non-zero probability of a marked change in global temperature should be a worry. Maybe not the only one we face, but I think it explains why the respondents gave an apparently contradictory response.

    • Paul: I never thought anyone would go back and check out my estimate of the number of questions, but now that you’ve done so I am forced to revise it. If we accept that a question comes accompanied by a histogram there are 93 of them. Still a lot.

      Your point about the “fat right tails” observed in Bayesian probability distributions of climate sensitivity estimates is a valid one, but we are unable to judge how much this consideration might have influenced the participants’ attitudes on the dangers of climate change because they weren’t asked to give an opinion on the subject.

      • Paul says:

        I didn’t estimate anything, I just looked at the question numbers 🙂 But, puzzled by your 93, I had another look and now see what you mean – if one counts 9 a,b,c,d,e as 5 questions instead of 1 then there are indeed many more than 78.

        It is striking how many questions refer to judgements about models. It is not advertised how many modellers responded, but I’d expect non-modellers not to have much idea about those questions. Some commented to that effect (in the comments section). That weakens the result. On the other hand, as you say, we have no idea how anyone was influenced by probability distributions.

        I imagine we can all agree that surveys like this are interesting, but, like the Cook et al survey (the famous 97%), ultimately prove nothing.

  9. I imagine we can all agree that surveys like this are interesting, but, like the Cook et al survey (the famous 97%), ultimately prove nothing.

    Hard to disagree with that. Maybe someone should tell the politicians.

  10. Javier says:

    Another very nice entry, Roger. Congratulations.

    First my vote. My estimate is based on the 1,150 years cycle apparent in the last 4000 years of the GISP2 data, that most likely corresponds to the Bond events in the late Holocene. Based on our temporal position in the cycle and the calculated temperature anomaly that the cycle should present in the absence of human fuel fossil contribution, after correcting the Greenland temperature into Global anomaly, we are about 0.2 ºC above where we should be.

    Please observe that this “technique” does not require absolutely any knowledge about climate and therefore it has no bias nor it excludes any relevant factor even if unknown at present.

    Now 0.2 ºC out of 0.7 ºC global warming in the last 100 years, comes out as 30%. So that is my vote.

    I would like to also add my opinion to the poll. It is apparent to me that the graph from question 31 is a bi-modal distribution with a large peak at 80% and a smaller peak at 30%. Since I consider myself skeptic and 30% is my vote, I also consider that peak to represent the skeptics in the poll. That allows us to put a number to the category skeptics. They represent 16% of the universe (45 respondents), while the other 84% can be labelled as believers (241 respondents).

    We can probably trace these 16% of skeptics in most of the answers as the leftmost 16% in each of them.

    We skeptics have two reasons to congratulate. First, almost all great scientific advances had come as a minority of scientists’ thesis was able to conquer a majority’s one. Galileo, Darwin, etc. Second, from 2009 to 2013, just 4 short years, the number of skeptic climatologists has grown from 2.5% to 16%. Now talk about rapid progress. I have seen slower wildfires. With this rate of growth, skeptics will be the majority in 3 years, by 2016.

    • Thank you, Javier

      My 50% human-caused estimate was based on the ~60 year period of the Atlantic Multidecadal Oscillation (which some theorize is caused by orbital forcing from Saturn and Jupiter). There’s a good chance the AMO cycle caused much of the observed warming over the past 40 or so years, and if it did we can expect another 20 or 30 years of no warming.

      I admire your attempts to demonstrate rapid growth in climate change skepticism but I’m afraid I don’t see a bimodal distribution in question 31. Change just one vote and the 30% blip goes away.

      • Javier says:

        The numbers where just a joke, of course, in the line of the 97% meme. I don’t believe much science can be extracted from scientific polls. But I do believe that the number of skeptics is growing, perhaps due to the change in the warming trend since 1998. Scientific journals appear more open towards articles questioning the IPCC conclusions. Some general population polls also show growing skepticism with global warming.

        • The proportion of skeptics in the poll can be estimated from question 5. It’s 7.7%.

          It’s also interesting to note how confidence in the findings of climate science has increased despite another five years of no warming.

  11. Paul says:

    As far as assigning a proportion of warming since 1850 to anthropogenic causes, that is a difficult question. I don’t see a way one can possibly know, but I see some complications with assigning a number.

    Let’s say we think that 100% is anthropogenic. That seems unlikely, but if it were true and if climate sensitivity is low, then most of the warming we can expect has already happened. But if it was all anthropogenic then seeing as we have only seen a rise of 120ppm in CO2 that implies that sensitivity is perhaps relatively high and we can expect more as we keep emitting.

    On the other hand if we say 0% is anthropogenic, again unlikely, then even if sensitivity is low, it has all – maybe 1.5C – yet to come. We only escape that consequence if we fix on very low sensitivities for which there is not much evidence.

    So maybe I’ll go with 30-70% as a good compromise. But there is no science in that.

  12. Pingback: Klein Changes Everything | transitionferndale

  13. Very detailed statistical approach to the anthropogenic impact on environment, form the perception of the interviewed people.

    I only have for certain global crude figures. We are burning fossils approximately one million times faster than Nature took to digest the existing surface biomass in their crustal intestine tectonic folds. And we are throwing this into the atmosphere in the form of gases and heat. I have a hint that this behavior must have some kind of noticeable impact on our global envirnonment. And also that this is not a desirable action for a stable and durable planet. The rest is secondary to me.

  14. Pingback: AWED Energy & Environmental Newsletter: January 5, 2015 - Master Resource

  15. Pingback: Recent Energy And Environmental News – January 5th 2015 | PA Pundits - International

Comments are closed.