Climate sensitivity in CMIP6 GCMs

Anyone who is aware of what’s going on in climate science should have heard that the latest generation of climate models, known as CMIP6, seem to be suggesting a somewhat higher climate sensitivity than suggested by the previous CMIP5 models. The Equilibrium Climate Sensitivity (ECS) from CMIP5 models varied from 2.1K to 4.7K with a mean of 3.3K, while for the CMIP6 models it goes from 1.8K to 5.6K with a mean of 3.9K. In addition, 10 of the 27 CMIP6 GCMs having ECS values above 4.5K.

Thanks to a Twitter thread from Gavin Schmidt (which may have been the result of a tweet from Oliver Bothe) I’ve been made aware of a new paper from Mark Zelinka et al. that diagnoses the causes of higher climate sensitivity in CMIP6 models. According to their analysis, the main reason for the difference between the CMIP5 and CMIP6 ECS values is an enhanced SW low cloud feedback, mostly in the Southern extratropics (latitudes poleward of 30o). The idea being that there is a reduction in low level clouds in these regions which leads to an increase in the absorbed SW solar flux.

Credit: Zelinka et al. (2020)

The figure on the right illustrates the CMIP6 (orange) and CMIP5 (blue) SW low cloud feedback, and their difference (black). It clearly shows an increase poleward of 30S in the CMIP6 models, when compared to the CMIP5 models. As I understand it, this isn’t the only reason for the higher ECS values in the CMIP6 GCMs (they also suggest a sightly larger change in forcing due to a doubling of CO2) but it does seem to be the dominant factor.

Of course, the suggestion of an increased ECS in CMIP6 GCMs does not tell us that the ECS is indeed higher in the real world. The paper seems to suggest that the positive low cloud extratropical feedback is consistent with observations and theory. However, this doesn’t preclude that there could still be an error in some feedbacks that has yet to be established. The paper did, however, conclude with an interesting point. The enhanced low cloud SW feedback occurs in a region with efficient ocean heat uptake. If I understand this (which I may not) this might indicate that the efficient ocean heat uptake could be masking some surface warming while the climate is still changing (i.e., before reaching equilibrium).

I don’t know really know how to conclude this. It seems that there’s a growing understanding of why the CMIP6 GCMs suggest a higher ECS than the CMIP5 GCMs. This could reflect that the real world ECS is also higher than we had expected, but it’s still too early to really tell. I rather hope that it is not.

Links:
An emergent constraint on Transient Climate Response from simulated historical warming in CMIP6 models, by Femke Nijsse, Peter Cox, and Mark Williamson (2020).
Carbon-concentration and carbon-climate feedbacks in CMIP6 models, and their comparison to CMIP5 models, by Vivek Arora et al. (2020).

This entry was posted in Climate sensitivity, Gavin Schmidt, Research, Scientists, The scientific method and tagged , , , , , , , . Bookmark the permalink.

62 Responses to Climate sensitivity in CMIP6 GCMs

  1. Yes, I rather hope real world ECS is not higher than we expected. The ocean heat uptake discussed above seems to fit with this:

    “Because the oceans cover three fifths of the globe, this correction implies that previous estimates of overall global warming have been too low. Moreover it was reported recently that in the one place where it was carefully measured, the underwater melting that is driving disintegration of ice sheets and glaciers is occurring far faster than predicted by theory—as much as two orders of magnitude faster…”

    https://blogs.scientificamerican.com/observations/scientists-have-been-underestimating-the-pace-of-climate-change/

    Also, from that piece:

    “Consistent underestimation is a form of bias—in the literal meaning of a systematic tendency to lean in one direction or another—which raises the question: what is causing this bias in scientific analyses of the climate system?”

    That is a good question. I rather hope that the scientists, physicists and statisticians who fail to recognize their consistent underestimation will recognize and apologize for their bias and the too-rosy expectations that arises from this bias.

    When this happens, when the “too rosy” scientists are convinced and acknowledge that their take on things has been too rosy, it could change the public discussion about the level and urgency of changed needed to respond to this predicament.

    I continue to wonder/hope that the Australia fires might be the tipping point for the scientific and public discussion of global warming. I rather hope that is the case.

    as the chain of command reportedly said in WWII, smoke’m if you’ve got’m.

    Mike

  2. Phil says:

    I saw Gavin’s twitter thread on this and it made me wonder whether this change could be traced back to specific modification(s) to the GCMs from CMIP5 to CMIP6 – assuming, of course, that the models evolve incrementally over time. With my ex-programmers hat on, it seems obvious that one could identify changes to the program code that were likely responsible for such a change in the results (the enhanced low cloud feedback) and then go back to look at the research that induced those changes in the code. But perhaps it isn’t that simple … 🙂

  3. Phil,
    I suspect these codes are so complex that it’s not possible to use changes in the code itself to infer why the ECS has changed between CMIP5 and CMIP6. I think you need to do what was done in this paper; run the codes and analyse the outputs to infer what is different. My understanding, though, is that the changes are made so that the various physical processes better match observations and theory. I also think that the paper is suggesting that some of the changes to the cloud prescriptions do better match observations and theory, which is why it’s not clear if this change to the SW low cloud feedback is indicative of an error in the CMIP6 models, or not.

  4. Everett F Sargent says:

    has suggested that the submission publication deadline (for AR6 WGI (?)) was 2019-12-31 11:59:59 PM. So that this year will be similar to 2013. The acceptance deadline is when?
    http://julesandjames.blogspot.com/2020/01/review-of-blogyear.html

    I’m much more interested in the hindcasts (e. g. the historical record) as an RMSE calculation broken down by latitude and land, ocean and land+ocean.

    My biggest concern is a repeat of CMIP5 near-term post-observational GMST trend lines. Don’t need 0.25-0.3 C/decade circa 2010-2030 again.

  5. Steven Mosher says:

    sbm
    “Recently, the U.K. Met Office announced a revision to the Hadley Center historical analysis of sea surface temperatures (SST), suggesting that the oceans have warmed about 0.1 degree Celsius more than previously thought. ”

    er no.

    Recently, the U.K. Met Office announced a revision to the Hadley Center historical analysis of sea surface temperatures (SST), suggesting that the oceans have warmed about 0.1 degree Celsius more than their previous record indicated”

    There’s more than 1 SST data set. “more than previously thought. ” is ambiguous. More than
    WHO thought. Nobody thought this dataset was the only or the most accurate. sloppy. sometimes
    data sets adjust down, sometimes up. Its a bad skeptic meme that we always adjust up.

  6. Everett F Sargent says:

    sbm sez ,,,

    Oreskes, seriously? I’m so happy that some people aren’t directly involved in the IPCC AR6 processes. But then again a broken watch is correct twice daily, or so I’ve been told.

  7. we have one real model of the climate system. that is the planet where we all live. It is amazingly complex and capturing it with computer models is useful, but the computer models will always need to be reviewed against the real world observations. When the computer models and the real world model drift apart, it is safe to assume that the computer models have failed to capture the complexity of the real world model where we are currently running a gigantic experiment by changing the atmospheric composition in subtle and significant ways.

    Unless, you subscribe to the fables that say the planet is 6,000 years old or that solar output is the source of the planetary warming we are measuring.

    Two words keep coming to mind: precautionary principle.

    Smoke’m if you’ve got’m.

    Mike

  8. Phil says:

    ATTP:

    I was, I think, suggesting something slightly different. This would be to construct a CMIP6-minus GCM where a candidate code change was ”rolled back” to its CMIP5 state and observe whether the black line in Zelinka et al’s diagram was closer to zero. Given what you’ve said, maybe the cause is clear enough already and such an investigation would not shed anymore light on the subject. I guess the length of time required to run these models (and the number of likely code changes to investigate) may prohibit this kind of approach.

  9. Phil,
    I see. I’m not completely sure, but I think many of the CMIP6 GCMs are updated CMIP5 GCMs, so if you rolled back all the changes, you should indeed get the CMIP5 result. However, one could presumably roll back some changes, and not others, to see which effect dominates.

  10. Phil said: “I guess the length of time required to run these models (and the number of likely code changes to investigate) may prohibit this kind of approach.”

    I think that’s right, it would be very expensive and time consuming to check the models, but I think your idea is sound.

    Maybe we should just go with the the results of CMIP6 and adjust changes in emissions (carbon budget) based on the possibility that ECS is higher than CMIP5 indicated? I think the worst that happens in that scenario would be that we might address the problem more quickly than is required. I have it on good authority that addressing the problem of carbon emissions more quickly than is required would also be a mistake.

    I rather hope that we do that, nonetheless. I remain unconvinced that fixing the problem sooner rather than later will actually turn out to be a problem.

    Mike

  11. John Hartz says:

    More insight into how AR6 will differ from AR5…

    As I’ve watched the events of this summer unfolding, I’ve found myself wondering whether the Earth system has now breached a tipping point, an irreversible shift in the stability of the planetary system.

    There may now be so much heat trapped in the system that we may have already triggered a domino effect that could unleash a cascade of abrupt changes that will continue to play out in the years and decades to come.

    Rapid climate change has the potential to reconfigure life on the planet as we know it.

    We know this because the geologic record contains evidence that these events have occurred in the past. The key difference is that we’ve never had 7.5 billion people on the planet, so the human species really is in uncharted territory.

    The scientific community is acknowledging this by including new sections on abrupt climate change throughout key areas of the upcoming IPCC report. We now consider these “low probability, high impact” scenarios an increasingly critical part of our work.

    We are seeing the very worst of our scientific predictions come to pass in these bushfires, Opinion by Joëlle Gergis, Comment is Free, Guardian, Jan 2, 2020

  12. Jon Kirwan says:

    I tend to imagine that, on balance, weaknesses in prior climate models are being improved with some alacrity while strengths in those models are modified only slightly and then only for well-defensible reasons. It’s been my experience, anyway, in almost 50 years of designing (and starting in the early 1980’s, programming them as well) scientific and commercial instrumentation. My work doesn’t get poorer over the years. I learn new things, carefully validate through testing what I think I’ve learned, and my products reflect the newer and better knowledge and perform better. I can honestly say that I’ve never allowed new instrumentation replacing older to in any way under-perform or be less accurate or precise. (At the same price point. But of course, I have used what I’ve learned and other advancements to produce ever cheaper versions with good specifications.)

    Perhaps I’m more respectful of the art and science of climate science than I should be, but I fully expect CMIP6 to have had the same kind of quality workmanship and rigorous attention to important details that I have applied throughout my own life.

  13. John Hartz says:

    Another relevant and poignant section of the Joëlle Gergis opinion piece I cited above…

    As a climate scientist, the thing that really terrifies me is that weather conditions considered extreme by today’s standards will seem sedate in the future. What’s unfolding right now is really just a taste of the new normal.

    At this point I could restate all the lines of scientific evidence that clearly show the links between human-caused climate change and the intensification of extreme weather conditions not just in Australia, but all over the world.

    To avoid sounding like a broken record, instead I will say that as a lead author on the forthcoming Intergovernmental Panel on Climate Change’s (IPCC) Sixth Assessment report of the global climate due out next year, I can assure you that the planetary situation is extremely dire.

    It’s no exaggeration to say my work as scientist now keeps me up at night.

  14. MarkR says:

    Phil:
    “I saw Gavin’s twitter thread on this and it made me wonder whether this change could be traced back to specific modification(s) to the GCMs from CMIP5 to CMIP6”

    Here’s a HadGEM3 paper comparing the CMIP5 and CMIP6 versions with dozens of “part way versions”. The strong Southern low cloud feedback changes come from (i) new calculations of aerosol-cloud microphysics interactions and (ii) new calculations of mixed phase clouds (i.e. clouds with both liquid droplets and ice crystals):
    https://doi.org/10.1029/2019MS001688

    Water droplets are more reflective than ice crystals, so if you melt some of an all-ice cloud, that’s a cooling feedback. Many CMIP5 clouds were totally frozen below -5 C or -10 C, but CALIPSO satellite lidar measurements showed supercooled water is common at those temperatures. So in CMIP6 you make sure you match reality and suddenly you no longer melt as many low level ice clouds because there are fewer of them. This kills off an erroneous cooling feedback that occurred in some models, so your ECS goes up.
    https://agupubs.onlinelibrary.wiley.com/doi/full/10.1029/2018GL081871

    This seems pretty typical to me – model development aims to match observed processes.

  15. Chubbs says:

    Best to wait until all the new information relevant to TCR/ECS is integrated by experts. Another important piece is a recently published review of aerosol forcing (link below) with the following conclusion:

    “These multiple lines of evidence lead to a 68% confidence interval for the total aerosol effective radiative forcing of−1.60 to−0.65 W m−2, or−2.0 to−0.4 W m−2 with a 90% likelihood. Those intervals are of similar width to the last Intergovernmental Panel on Climate Change assessment but shifted towards more negative values”

    https://hal.archives-ouvertes.fr/hal-02322106/file/Bellouin_etal_accepted.pdf

  16. MarkR,
    Thanks, that’s a really interesting paper.

  17. Phil says:

    MarkR – thanks for that post. I would certainly expect liquid water to reflect more IR frequencies than ice (from my Chemical Physics days) – I didn’t know (until now) that CMI5 hadn’t modelled mixed phase clouds very well. And I guess changes to mixed phase clouds would be bound to have a strong latitudinal variation (as shown by Zelinka et al.)

    You, therefore, seem to be of the view that the upwards revision of ECS, or the least the part due to SW low cloud feedback, in CMIP6 is sound ?

  18. David B. Benson says:

    Just from considering what I know about the climate of my mid-Pliocene, the so-called ESCS is about 5 K. A larger value for the climate model CS, larger than 3 K, is consistent with that paleoclimate estimate.

  19. paulski0 says:

    Chubbs,

    Thanks for the link, interesting. Surprised to see Bjorn Stevens on the author list since the paper seems to be disagree with his views, though I’ve noticed a tendency for authors listed as “and” out of alphabetical order turning out to not actually agree with the paper’s conclusions.

  20. RickA says:

    The fact that the mean has jumped from 3.3 to 3.9C in the last 5 years (from CMIP5 to CMIP6) means that we have no more idea of how to model the climate than merely throwing a dart at a dartboard.

    My bet is that in the next five years, the CMPI7 models drop the mean from 3.9C to 3.1C.

    We clearly have not finished our learning process for modeling the climate and have a lot more work to do.

    Plenty more known and unknown unknowns to learn about and incorporate into the models.

    I look forward to learning how the scientists will use the actual temperature increase over the doubling of CO2 from 280 ppm to 560 ppm. That delta T (whatever it turns out to be) has to be useful for double checking the TCR or effective TCR calculations in the models, once it happens. We only have to wait until 2050 or so!

    So 2050 ish (or 560 ppm) will provide a real double check on all the model gyrations, which I assume will continue for the next 30 years.

    But the thrashing around of the models really puts the lie to how well we understand the climate, and how any variable effects it (let alone just CO2).

  21. MarkR says:

    Phil;
    “You, therefore, seem to be of the view that the upwards revision of ECS, or the least the part due to SW low cloud feedback, in CMIP6 is sound ?”

    That’s a great question, I don’t know enough to have a trustworthy opinion yet and I want to learn more.

    The CALIPSO mixed phase cloud data shifts my expected ECS a bit higher. But I suspect we’ll find other deficiencies that were being “hidden” by the southern ocean cloud issues in CMIP5, e.g. maybe it’ll now be obvious that tropical shallow cumulus is a bit off, or something like that.

    I wonder why we’ve seen so much more warming in the western Pacific and less in the east. It’s different than what we saw in CMIP5, and if this *pattern* continues and is caused by CO2-forced warming, then that could bring ECS back down again. But if it’s basically all natural and/or temporary (aerosol? Deforestation?) then we’re looking at unbelievable disaster without rapid CO2 emissions cuts.

  22. Mal Adapted says:

    RickA:

    The fact that the mean has jumped from 3.3 to 3.9C in the last 5 years (from CMIP5 to CMIP6) means that we have no more idea of how to model the climate than merely throwing a dart at a dartboard.

    RickA is apparently here to do here what one “David Young” tried to do on this recent RC thread . Like DY, RickA’s implicit warrant on climate blogs appears to be lukewarmism: in the present example, scoffing at the self-correcting, progressive character of science, and hyping a shift in mean ECS from 3.3 to 3.9 as fatal for the entire modeling enterprise. DY, OTOH, made much of a recent PNAS Perspective by Palmer and Stevens, in which the authors decry the unsuitability of current GCMs for making short-term, detailed regional climate predictions with the confidence desired by policy makers. DY claimed it showed GCMs are “not fit for purpose”, and indeed those are the authors’ very words: the context is the key. Fatal for his lukewarmist case, they explicitly support the consensus of their peers for AGW and rebut the undead “models are unreliable” denialist meme (my bolding):

    For certain, some things are settled. We know that greenhouse gases are accumulating in the atmosphere as a result of human activity and that they are largely responsible for warming of surface temperatures globally. We also are confident in our understanding as to why this warming is expected to be amplified over land masses and the Arctic. Likewise, we are confident in our understanding of how the hydrological cycle amplifies the effects of this warming and how warming amplifies the hydrological cycle. For these and other broad brush strokes of the climate change picture, we are also increasingly confident in our ability to usefully bound the magnitude of the effects. From this certainty stems the conviction that additional warming is best avoided by reducing or reversing emissions of long-lived greenhouse gases.

    As climate scientists, we are rightfully proud of, and eager to talk about, our contribution to settling important and long-standing scientific questions of great societal relevance. What we find more difficult to talk about is our deep dissatisfaction with the ability of our models to inform society about the pace of warming, how this warming plays out regionally, and what it implies for the likelihood of surprises. In our view, the political situation, whereby some influential people and institutions misrepresent doubt about anything to insinuate doubt about everything, certainly contributes to a reluctance to be too openly critical of our models. Unfortunately, circling the wagons leads to false impressions about the source of our confidence and about our ability to meet the scientific challenges posed by a world that we know is warming globally.

    How can we can reconcile our dissatisfaction with the comprehensive models that we use to predict and project global climate with our confidence in the big picture? The answer to this question is actually not so complicated. All one needs to remember is that confidence in the big picture is not primarily derived from the fidelity of comprehensive climate models of the type used to inform national and international assessments of climate change…

    Unlike DY, RickA may have read the whole article. In any event, his comment here places him with those who “misrepresent doubt about anything to insinuate doubt about everything”. Thankfully, he’s no more influential than DY or I!

  23. Rick,

    But the thrashing around of the models really puts the lie to how well we understand the climate, and how any variable effects it (let alone just CO2).

    This is dreadful sophistry. There are many ways in which climate models are skillful (global circulation, warming under increasing forcings, changes to precipitation and evaporation, etc). That they haven’t converged on a precise ECS doesn’t mean that they have no value. If you take away GCMs, what we could conclude? We’d use other evidence (paleo, for example) to conclude that the ECS is probably around 3oC.

  24. dikranmarsupial says:

    Its not as if estimating ECS were the primary purpose in having GCMs.

    Sometimes your knowledge improves by discovering uncertainties that you didn’t previously know about, in which case you should expect the uncertainty (as indicated by the spread of the models) to increase rather than decrease. I agree that it is sophistry to suggest that the disagreement between models increasing over time indicates that we don’t understand the climate.

  25. Chubbs says:

    Most of the uncertainty is on the upper tail of the ECS distribution. We know enough about TCR and the lower tail of the ECS distribution, to motivate strong climate policy. What we don’t know is how much our procrastination is hurting in the long-term. If CMIP6 is right the 1.5C budget becomes roughly the 1.8C budget and so on up the line.

  26. Chubbs,
    Agreed. What was interesting is that at least one of the CMIP6 GCMs suggests an ECS of less than 2K. Would be quite interesting to understand why some produce ECS values that are this low. It’s not completely ruled out, but my understanding was that such low ECS values were regarded as rather implausible.

  27. Chubbs says:

    ATTP – The recent warming has made ECS1.2C in one decade, while maintaining a large heat imbalance.

  28. Chubbs,
    If you’re suggesting that observations themselves make it unlikely that the ECS will be below 2C, then I agree.

  29. Chubbs says:

    Garbled the message. Yes ECS>2C based on recent observations. Don’t need climate models to justify policy.

  30. Without judging a person I don’t know, I can only say that RickA doesn’t sound like a lukewarmer to this lukewarmer.

    Climate models do very well at what they are designed to do–model the broad thrusts of climate trends.

    Although RickA goes way overboard in his assessment, to a certain extent we are all frustrated at the evident fact that climate models stubbornly refuse to do that which they were not designed to do–accurately predict temperature changes on a decadal scale.

  31. JCH says:

    CFMIP – despite Nic Lewis’s involvement, they keep wagging that longer tail.

  32. Tom,

    we are all frustrated at the evident fact that climate models stubbornly refuse to do that which they were not designed to do–accurately predict temperature changes on a decadal scale.

    Well, yes, this isn’t really what climate models are for. There is also a new paper that makes an interesting emergent constraint argument. They argue that the warming since 1970 is a good way to constrain the climate models. If you eliminate the CMIP6 models that don’t match well the warming since 1970, then you can constrain the TCR to be from 1.5K – 2.2K. However, this doesn’t appear to particularly constrain the ECS (there are still some high ECS models that match the warming since 1970).

  33. “climate models stubbornly refuse to do that which they were not designed to do–accurately predict temperature changes on a decadal scale?”

    “All models are wrong, but some are useful”. … The aphorism is generally attributed to the statistician George Box, although the underlying concept predates Box’s writings.

    https://en.wikipedia.org/wiki/All_models_are_wrong

    I think the models might be seen as fit for use if they led our species to avoid catastrophic planetary warming, but that’s based on my goal for their utility. It’s quite different from hoping that the models would accurately predict change on a decadal scale. That time frame is too short for my goals for the models. My concern falls into a time scale of 100 years or more because I want my grandchildren and their children to have good lives.

  34. Chubbs says:

    Tom – If people weren’t discounting climate models so heavily during the hiatus, they would have done a much better job forecasting 2019 temperatures. The same could have been said 20 years ago.

  35. JCH says:

    Among the first decadal prediction models is DM Smith’s, and his initial prediction, for being among the first of its kind, and at an endeavor often described as being impossible, was remarkably accurate. And it went right into the teeth of the deluded “but DaPaws” people.

  36. Mal Adapted says:

    thomaswfuller2:

    Without judging a person I don’t know, I can only say that RickA doesn’t sound like a lukewarmer to this lukewarmer.

    Right, that’s because your definition of lukewarmer is the one you and Mosher promoted on pseudo-skeptical blogs following CRUhack. Nice try, but I’m using the word as defined on SkepticalScience:

    In the context of global warming, the term “lukewarmers” refers to individuals who accept the scientific reality that human greenhouse gas emissions are a primary cause of the current global warming, but who believe that future global warming and the consequences of the associated climate change will not be as bad as the body of scientific evidence indicates. Different “lukewarmers” have different justifications for this belief, but in general, for one reason or another they tend to find the evidence for relatively mild future climate change impacts more compelling than the evidence to the contrary.

    By that definition, we don’t need to know more about RickA than the content of his comments here.

    twf2:

    Climate models do very well at what they are designed to do–model the broad thrusts of climate trends.

    Although RickA goes way overboard in his assessment, to a certain extent we are all frustrated at the evident fact that climate models stubbornly refuse to do that which they were not designed to do–accurately predict temperature changes on a decadal scale.

    Agreed. This is why I don’t consider you a lukewarmer, Tom! It looks to me like you fully acknowledge the expert consensus. You only appear to challenge it, aggressively as you do, because amplifying controversy is your profession ;^).

    Anyway, when Palmer and Stevens declare (my emphasis):

    Whereas present day climate models were fit for the purpose for which they were initially developed, which was to test the basic tenets of our understanding of global climate change, they are inadequate for addressing the needs of society struggling to anticipate the impact of pending changes to weather and climate.

    I doubt any scientifically meta-literate climate realist would disagree! “All models are wrong, but some are useful.” Per dikranmarsupial:

    Its not as if estimating ECS were the primary purpose in having GCMs.

    Yep. Palmer and Stevens again:

    How can we can reconcile our dissatisfaction with the comprehensive models that we use to predict and project global climate with our confidence in the big picture? The answer to this question is actually not so complicated. All one needs to remember is that confidence in the big picture is not primarily derived from the fidelity of comprehensive climate models of the type used to inform national and international assessments of climate change.

    That sounds about right to me.

  37. Mal Adapted, sounds like we agree–well, at least there’s nothing in your comment that I dispute other than your overly facile definition of lukewarmer.

    Just for future reference, a lukewarmer is pragmatically defined as someone who, if offered an over/under bet on ECS of 3C, will take the under. That’s really all there is to it.

  38. Joshua says:

    > Just for future reference, a lukewarmer is pragmatically defined as someone who, if offered an over/under bet on ECS of 3C, will take the under. That’s really all there is to it.

    So sayeth the decider.

  39. Mal Adapted says:

    thomaswfuller2:

    Just for future reference, a lukewarmer is pragmatically defined as someone who, if offered an over/under bet on ECS of 3C, will take the under. That’s really all there is to it.

    Quoting the respected logician CL Dodgson:

    `When I use a word,’ Humpty Dumpty said in rather a scornful tone, `it means just what I choose it to mean — neither more nor less.’

    `The question is,’ said Alice, `whether you can make words mean so many different things.’

    `The question is,’ said Humpty Dumpty, `which is to be master – – that’s all.’

    Sorry Tom, you ain’t the master o’ me. I’m not yours, either, but if you don’t like ‘lukewarmer’ in its more common usage, I suggest you don’t call yourself that here. Like I said, AFAICT it doesn’t actually apply to you anyway.

  40. Mark B says:

    [i]Just for future reference, a lukewarmer is pragmatically defined as someone who, if offered an over/under bet on ECS of 3C, will take the under. That’s really all there is to it.[/i]

    By this definition Dragon Slayers are lukewarmers and I’m pretty sure you don’t mean that.

    A cynical viewer might suggest that picking the arbitrary dividing line at approximately the middle of the consensus distribution while distancing oneself from the nutcases on one’s own side of that line is a tactic to claim as middle ground something that isn’t supportable as such on the evidence.

  41. John Hartz says:

    Seems to me that lukewarmist would be a better label than lukewarmer. 🙂

  42. Or you could just return to luckwarmer, denier, delayer. Whatever floats your boat.

  43. dikranmarsupial says:

    “Just for future reference, a lukewarmer is pragmatically defined as someone who, if offered an over/under bet on ECS of 3C, will take the under. That’s really all there is to it.”

    Ironic that you should call someone else’s definition “facile”. That would make the IPCC FAR “lukewarmer” and the reports since then borderline “lukewarmer”.

    It is also ignoring the key point that the loss function rises faster than linear, so it isn’t much comfort to know that there is a reasonable change that ECS is below 3C if we can’t rule out their being a good chance of it being higher, in which case the risks are still considerable. You might want to take an intellectual evens bet on below 3C, but you would be a fool to think that was a rational basis for climate policy.

  44. Dikran marsupial, I’ll give you a lukewarm reply. I’m not at all certain that we can attribute any damages to date to climate change (although I am almost certain it has had an effect), and that includes fires in Australia and California and monsoon like rains in Houston and other parts of the South. So I consider the idea that loss function is more than linear to be more conjecture than anything else.

    It may prove to be true. But I don’t think there is evidence for it–yet–in the real world. Along with the IPCC, I hold that damages will be real and attributable some time after 2040. And quite a bit of whatever loss we shall see will be down to whether or not we heed the IPCC report on impacts and quit financing construction and reconstruction in vulnerable areas.

  45. Tom,
    Climate change is almost certainly irreversible on relevant timescales (without NETs). Waiting to find out if the damages are linear, or non-linear, seems like a particularly poor strategy, given the system that we’re playing with.

  46. John Hartz says:

    I presume that all of the ECS ratios specified in the OP and in the comments are the products of the ensemble outputs of the GSMs used in both AR5 and AR6. Do we know what the individual GCMs produced?

  47. ATTP, which is why I have advocated mitigation and pre-adaptation measures for over a decade now. Carbon tax? Check. Technology transfer to the developing world? Check. Quickest possible phase-out of coal fired generating plants? Check. Increased support for renewables? Check. Investment in charging infrastructure for EVs? Check. Increased funding for R&D, especially in storage and batteries? Check. X prizes for innovation? Check. Withdrawal of government insurance guarantees for buildings in flood plains and in fire zones? Check. Restoration of mangrove ‘farms’ on tropical islands? Check.

    For a decade, during which… well, never mind.

  48. Chubbs says:

    Just saw this in Femke Nijsse”s twitter:

    “Many of the latest CMIP6 ESMs have larger climate sensitivities, with 6 of 23 models having TCR values above 2.5 K, and an ensemble mean TCR of 2.1±0.4 K. On the face of it, these latest ESM results suggest that the IPCC likely range of TCR may need revising upwards, which would cast further doubt on the feasibility of the Paris targets. Here we show that rather than increasing the uncertainty in climate sensitivity, the CMIP6 models help to further constrain the likely range of TCR to 1.5-2.2 K, with a central estimate of 1.82 K. We reach this conclusion through an emergent constraint approach which relates the value of TCR to the global warming from 1970 onwards.”

    https://www.earth-syst-dynam-discuss.net/esd-2019-86/esd-2019-86.pdf

  49. Pingback: Another CMIP6 climate sensitivity constraint | …and Then There's Physics

  50. verytallguy says:

    “So I consider the idea that loss function is more than linear to be more conjecture than anything else”

    The loss function is self evidently more than linear; consider in extremis the loss at very high dT, say ten degrees fir arguments sake.

  51. vtg,
    Indeed, there must be a temperature change (as you say, about 10K) for which the damage would be essentially infinite.

  52. John Hartz says:

    Although a tad off-topic, here’s an example of a spot-on, short-range climate forecast,,,

    As the nation’s horror bushfire season shows no sign of abating, a landmark 2008 report that warned of these looming conditions is once again in the spotlight.

    Twelve years ago, economist Ross Garnaut led an independent study of the impacts of climate change on the Australian economy.

    The Garnaut Climate Change Review’s final report said projections of fire weather “suggest that fire seasons will start earlier, end slightly later, and generally be more intense”.

    “This effect increases over time, but should be directly observable by 2020.”

    How a climate change study from 12 years ago warned of this horror bushfire season by Nick Baker, Australia, SBS News, Jan 6, 2020

  53. Joshua says:

    Anders –

    > Climate change is almost certainly irreversible on relevant timescales (without NETs). Waiting to find out if the damages are linear, or non-linear, seems like a particularly poor strategy, given the system that we’re playing with.

    It’s fascinating how so many people are so bad at applying basic strategic thinking to low probability high damage risk – even those who have been heavily invested for a decade or more at contextualizing the implications of low probability high damage risk in an issue such as climate change.

    I’ve recently been wondering about evolutionary psych in the context of how bad people are at thinking about risk. I’m not a huge believer in the ability of ev-psych theorists to raise their theories above the grade of Just-So stories, but I guess there’s a logic to thinking that there is an evolutionary basis to our generally shared propensitity towards flawed thinking about risk. There’s probably an evolutionary advantage buried in there somewhere. But damned if I can even guess as to what that is in this case.

  54. Joshua says:

    Oh, what is “NETs” there?

  55. Everett F Sargent says:

    Negative Emissions Technologies (was my guess)

  56. EFS,
    Yes, NETs is Negative Emission Technologies.

  57. dikranmarsupial says:

    ” I’ll give you a lukewarm reply. ”

    yawn

    “I’m not at all certain that we can attribute any damages to date to climate change (although I am almost
    certain it has had an effect),”

    Sophistry. Climate policy should be based on FUTURE damages, not current damages.

    “So I consider the idea that loss function is more than linear to be more conjecture than anything else.”

    It seems like hubris to reject one of the IPCC’s reports so blithely. This is obviously not true. For example, while the Thames barrier is high enough to protect London from storm surges, the damages to London are essentially zero. However as soon as storm surges become high enough to overtop the barrier, then damages start to grow. That is obviously a non-linear damage function. VTG has given you another example.

    And of course, I notice that you have ignored my substantive point, which was the complete lack of utility of your “pragmatic” definition of a “lukewamer”, that would include the first IPCC WG1 report, and is borderline for the subsequent reports.

    The SKS definition is much more useful, a lukewarmer is someone who argues that the effects of climate change are sufficiently low that we shouldn’t take strong action to mitigate against it.

  58. Mal Adapted says:

    Joshua:

    I’ve recently been wondering about evolutionary psych in the context of how bad people are at thinking about risk. I’m not a huge believer in the ability of ev-psych theorists to raise their theories above the grade of Just-So stories, but I guess there’s a logic to thinking that there is an evolutionary basis to our generally shared propensitity towards flawed thinking about risk. There’s probably an evolutionary advantage buried in there somewhere. But damned if I can even guess as to what that is in this case.

    Hmm, speaking as a lifelong student of evolution, I think it more likely that adaptation by natural selection is simply not able to perfect our thinking about risk, or anything else. Random variation and selective retention can create infinitely many adaptive states, but not every imaginable state is equally reachable. Each incremental change is constrained by the laws of physics, and by historical contingency: that is, trade-offs with all our other adaptations. I’m afraid the best we can hope for, as individuals and a species, is to be marginally smarter than our predators and competitors.

  59. Mal Adapted says:

    thomaswfuller2:

    Or you could just return to luckwarmer, denier, delayer. Whatever floats your boat.

    Since I’m not a betting man, “lukewarmer” means all of those things to me. You’re not really any of them, Tom. How about “gadfly“? Come on, admit it: you’re just buzzing around biting our hides for blood and laying eggs 8^D!

  60. angech says:

    Joshua “There’s probably an evolutionary advantage buried in there somewhere. But damned if I can even guess as to what that is in this case.”

    Mal might mean
    “Helmuth von Moltke the Elder – Wikiquote
    No plan of operations extends with any certainty beyond the first contact with the main hostile force”

  61. Dave_Geologist says:

    Although thomaswfuller2 goes way overboard in his assessment, to a certain extent we are all frustrated at the evident fact that thomaswfuller2 stubbornly refuses to accept the facts; that climate models do that which they were not designed to do–accurately predict temperature changes on a decadal scale.

    (I’m sure you’ve had it pointed out to you before Tom, but if you’re puzzled, Google “uncertainty bounds”. If you’re still confused, try precision vs. accuracy.)

  62. Pingback: A couple of highlights | …and Then There's Physics

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.