RealClimate logo

The best case for worst case scenarios

The “end of the world” or “good for you” are the two least likely among the spectrum of potential outcomes.

Stephen Schneider

Scientists have been looking at best, middling and worst case scenarios for anthropogenic climate change for decades. For instance, Stephen Schneider himself took a turn back in 2009. And others have postulated both far more rosy and far more catastrophic possibilities as well (with somewhat variable evidentiary bases).

This question came up last year in the wake of a high profile piece “The Uninhabitable Earth” by David Wallace-Wells in New York magazine. That article was widely read, and heavily discussed on social media – notably by David Roberts, Mike Mann and others, was the subjected to a Climate Feedback audit, a Salon Facebook live show with Kate Marvel and the author, and a Kavli conversation at NYU with Mike Mann this week as well. A book length version is imminent.

In a similar vein, Eric Holthaus wrote “Ice Apocalypse” about worst-case scenarios of Antarctic ice sheet change and the implications for sea level rise. Again, this received a lot of attention and some serious responses (notably one from Tamsin Edwards).

It came up again in discussions about the 4th National Assessment Report which (unsurprisingly) used both high and low end scenarios to bracket plausible trajectories for future climate.

However, I’m not specifically interested in discussing these articles or reports (many others have done so already), but rather why it always so difficult and controversial to write about the worst cases.

There are basically three (somewhat overlapping) reasons:

  1. The credibility problem: What are the plausible worst cases? And how can one tell?
  2. The reticence problem: Are scientists self-censoring to avoid talking about extremely unpleasant outcomes?
  3. The consequentialist problem: Do scientists avoid talking about the most alarming cases to motivate engagement?

These factors all intersect in much of the commentary related to this topic (and in many of the articles linked above), but it’s useful perhaps to tackle them independently.

1. Credibility

It should go without saying that imagination untethered from reality is not a good basis for discussing the future outside of science fiction novels. However, since the worst cases have not yet occurred, some amount of extrapolation, and yes, imagination, is needed to explore what “black swans” or “unknown unknowns” might lurk in our future. But it’s also the case that extrapolations from incorrect or inconsistent premises are less than useful. Unfortunately, this is often hard for even specialists to navigate, let alone journalists.

To be clear, “unknown unknowns” are real. A classic example in environmental science is the Antarctic polar ozone hole which was not predicted ahead of time (see my previous post on that history) and occurred as a result of chemistry that was theoretically known about but not considered salient and thus not implemented in predictions.

Possible candidates for “surprises in the greenhouse”, are shifts in ecosystem functioning because of the climate sensitivity of an under-appreciated key species (think pine bark beetles and the like), under-appreciated sensitivities in the West Antarctic Ice Sheet, or the North Atlantic overturning, and/or carbon feedbacks in the Arctic. Perhaps more important are the potential societal feedbacks to climate events – involving system collapses, refugee crises, health service outages etc. Strictly speaking these are “known unknowns” – we know that we don’t know enough about them. Some truly “unknown unknowns” may emerge as we get closer to Pliocene conditions of course…

But some things can be examined and ruled out. Imminent massive methane releases that are large enough to seriously affect global climate are not going to happen (there isn’t that much methane around, the Arctic was warmer than present both in the early Holocene and last interglacial and nothing similar has occurred). Neither will a massive oxygen depletion event in the ocean release clouds of hydrogen sulfide poisoning all life on land. Insta-freeze conditions driven by a collapse in the North Atlantic circulation (cf. “The Day After Tomorrow”) can be equally easily discounted.

Importantly, not every possibility that ever gets into a peer reviewed paper is equally plausible. Assessments do lag the literature by a few years, but generally (but not always) give much more robust summaries.

2. Reticence

The notion that scientists are so conservative that they hesitate to discuss dire outcomes that their science supports is quite prevalent in many treatments of worst case scenarios. It’s a useful idea, since it allows people to discount any scientists that gainsay a particularly exciting doomsday mechanism (see point #1), but is it actually true?

There have been two papers that really tried to make this point, one by Hansen (2007) (discussing the ‘scientific reticence’ among ice sheet modelers to admit to the possibility of rapid dynamic ice loss), and more recently Brysse et al (2013) who suggest that scientists might be ‘erring on the side of least drama’ (ESLD). Ironically, both papers couch their suggestions in the familiar caveats that they are nominally complaining about.

I am however unconvinced by this thesis. The examples put forward (including ice sheet responses and sea level rise, and a failed 1992 prediction of Arctic ozone depletion, etc) demonstrate biases towards quantitative over qualitative reasoning, and serve as a lesson in better caveating contingent predictions, but as evidence for ESLD they are weak tea.

There are plenty of scientists happy to make dramatic predictions (with varying levels of competence). Wadhams and Mislowski made dramatic predictions of imminent Arctic sea ice loss in the 2010s (based on nothing more than exponential extrapolation of a curve) with much misplaced confidence. Their critics (including me) were not ESLD when they pointed out the lack of physical basis in their claims. Similarly, claims by Keenlyside et al in 2008 of imminent global cooling were dramatic, but again, not strongly based in reality.  But these critiques were not made out of a fear of more drama. Indeed, we also made dramatic predictions about Arctic ozone loss in 2005 (but that was skillful). 

The recent interest in ice shelf calving as a mechanism of rapid ice loss (see Tamsin’s blog) was marked by a dramatic claim based on quantitative modelling, later tempered by better statistical analysis (not by a desire to minimise drama). 

Thus while this notion is quite resistant to being debunked (because of course the reticent scientists aren’t going to admit this!), I’m not convinced that there is any such pattern behind the (undoubted) missteps that have occurred in writing the IPCC reports and the like.

3. Consequentialism

The last point is similar in appearance to the previous, but has a very different basis. Recent social science research (for instance, as discussed by Mann and Hasool (also here)) suggests that fear-based messaging is not effective at building engagement for solving (or mitigating) long-term ‘chronic’ problems (indeed, it’s not clear that panic and/or fear are the best motivators for any constructive solutions to problems). Thus an argument has been made that, yes, scientists are downplaying worst case scenarios, but not because they have a personal or professional aversion to drama (point #2), but because they want to motivate the general public to become engaged in climate change solutions and they feel that this is only possible if there is hope of not only averting catastrophe but also of building a better world. 

Curiously, on this reading, the scientists could find themselves in a reverse double ethical bind – constrained to minimize the consequences of climate change in order to build support for the kind of actions that could avert them.

However, for this to be a real motivation, many things need to be true. It would have to widely accepted that downplaying seriously bad expected consequences would indeed be a greater motivation to action, despite the risk of losses of credibility should the ruse be rumbled. It would also need the communicators who are expressing hope (and/or courage) in the face of alarming findings to be cynically promoting feelings that they do not share. And of course, it would have to be the case that actually telling the truth would be demotivating. The evidence for any of this seems slim. 


To get to the worst cases, two things have to happen – we have to be incredibly stupid and incredibly unlucky. Dismissing plausible worst case scenarios adds to the likelihood of both. Conversely, dwelling on impossible catastrophes is a massive drain of mental energy and focus. But the fundamental question raised by the three points above is who should be listened to and trusted on these questions?

It seems clear to me that attempts to game the communication/action nexus either through deliberate scientific reticence or consequentialism are mostly pointless because none of us know with any certainty what the consequences of our science communication efforts will be. Does the shift in the Overton window from high profile boldness end up being more effective than technical focus on ‘achievable’ incremental progress or does the backlash shut down possibilities? Examples can be found for both cases. Do the millions of extra eyes that see a dramatic climate change story compensate for technical errors or idiosyncratic framings?  Can we get dramatic and widely read stories that don’t have either? These are genuinely difficult questions whose solutions lie far outside the expertise of any individual climate scientist or communicator.

My own view is that scientists generally try to do the right thing, sharing the truth as best they see it, and so, in the main are neither overly reticent nor are they playing a consequentialist game. But it is also clear that with a wickedly complex issue like climate it is easy to go beyond what you know personally to be true and stray into areas where you are less sure-footed. However, if people stick only to their narrow specialties, we are going to miss the issues that arise at their intersections.

Indeed, the true worst case scenario might be one where we don’t venture out from our safe harbors of knowledge to explore the more treacherous shores of uncertainty. As we do, we will need to be careful as well as bold as we map those shoals.


  1. S. Schneider, "The worst-case scenario", Nature, vol. 458, pp. 1104-1105, 2009.
  2. J.E. Hansen, "Scientific reticence and sea level rise", Environmental Research Letters, vol. 2, pp. 024002, 2007.
  3. K. Brysse, N. Oreskes, J. O’Reilly, and M. Oppenheimer, "Climate change prediction: Erring on the side of least drama?", Global Environmental Change, vol. 23, pp. 327-337, 2013.

130 Responses to “The best case for worst case scenarios”

  1. 1
    Hank Roberts says:

    Just adding a cite to the item about

    putting out lots of CO2 and shutting down ocean circulation, which leads to ocean anoxia, and if that gets shallow enough (into the photic zone), you get the proliferation of certain bacteria that produce hydrogen sulfide.

  2. 2
    sidd says:

    Re: “For instance, Stephen Schneider himself took a turn back in 2009.”

    The link in that sentence seems to need login credentials ?


    [Response: Fixed. thanks – gavin]

  3. 3
    Lance Olsen says:

    Hans Joachim Schellnhuber has said the language of global warming “doesn’t capture the scale of destruction<>.”

    Much the same concern about the language we use to describe the rise of heat had already been expressed in an assessment by Harvard’s John A. Paulson School of Engineering and Applied Sciences; “For decades, we called it ‘global warming,’ an innocuous-sounding phrase invoking a gentle increase in worldwide temperatures, like turning up the thermostat in a house.

    “People asked, so the climate is getting warmer. Who cares?” said Michael B. McElroy, the Gilbert Butler Professor of Environmental Studies at Harvard University. “And scientists are partly to blame for that because of how we’ve described climate change<>.”

    Climate science has been struggling with this issue for years.

    For example, in its October 13 2006 issue, Science quoted researcher Brian O’Neill’s concern that the IPCC reports don’t convey the full range of risks; “the extreme scenarios that tend to fall out of the IPCC process may be exactly the ones we should most worry about.”

    In that same issue, Science quoted Michael Schlesinger, a climate scientist at the University of Illinois, Urbana- Champaign. Schlesinger remarked that, “Things are happening right now with the ice sheets that were not predicted to happen until 2100. My worry is that we may have passed the window of opportunity where learning is still useful.”

    Science returned to these concerns in June, 2007, quoting climate researcher Stefan Rahmstorf’s comment that, “The IPCC has been overly cautious in not wanting to give any large number to [future] sea-level rise.”

    Reporting that “Scientists are still trying to strike a balance between their habitual caution and growing concern over uncertain but disastrous greenhouse outcomes,” Science also quoted glaciologist Robert Thomas’ remark that, “ ‘Most scientists don’t want to, but I think we need a way to explore’ the extreme end of the range of possibilities.” Thomas told Science that scientists need “a better way” than IPCC’s consensus approach, “so we can communicate with the public without becoming scaremongers.”

  4. 4
    Dan DaSilva says:

    One thing you can count on is that “unknown unknowns” will always be with us. What can be foreseen about “unknown unknowns”? Well, there will always be yet to be discovered “system collapses, refugee crises, health service outages” that push toward greater alarm. The problem will always be worse than we thought and climate sensitivity always greater than once forecast.

    The result is like the shepherd boy who keeps sensing bigger and bigger wolves that circle the flock just out of sight yet never attack. These wolves grow larger while never eating any sheep. They have found a food source supplied by alarmist desires to mold the future to their political fantasy.

  5. 5
    rhymeswithgoalie says:

    Having grown up in the Cold War, we were faced with the serious possibility of nuclear disaster, but there seemed to be a general feeling of lack of control over the outcome, or the sense that we weren’t actively responsible for bad outcomes.

    Contrast to AGW, where we have the added stress of knowing that we can directly contribute to the problem or need to make the effort to reduce our carbon footprints or pressure politicians to fight big money. With this stressful burden, it’s hard to feel cheered when told the worst-case may be much less likely if we still have to worry about the still-really-bad-case.

    Tangential trivia: The post-apocalyptic premise of the movie “Snowpiercer” was that a geoengineering project overshot and led to a snowball Earth.

  6. 6
    Mike Roberts says:

    I think the last word “shoals” should be “shores” as per the previous sentence.

    “However, if people stick only to their narrow specialties, we are going to miss the issues that arise at their intersections.”

    So what is the answer to getting at what that intersection reveals? Is that what writers and journalists do? But, if so, how do we increase their reach?

  7. 7
    David Beebe says:

    “Indeed the true worst case scenario might be one where we don’t venture out from our safe harbors of knowledge to explore the more treacherous shores of uncertainty. As we do, we will need to be careful as well as bold as we map those shoals.”

    Roger on that.
    I’ve been a commercial fisherman in Southeast Alaska for the last 35 years and prompted to comment given your metaphors. The Gulf of Alaska and Inside Passage waters are already lively enough with natural hazards: rogue waves, microbursts, tsunamis, debris torrents, shoals, uncharted rocks, icebergs, massive logs accompanied and driven by unpredictable weather and menacing tidal currents… We don’t need to add ocean warming, ocean acidification, Anthropogenic Climate Disruption, dead zones, etc. etc. to this mix.

    After all, there’s a very big difference between fear based messaging and sensible caution.

    I’m still alive because I don’t stake my life on “knowledge”-driven prognostication. I’ve lost too many friends who wagered their skill and “knowledge” against known hazards and the unknowable. They not only lost their lives but their boats and hapless crewmen who happened to also be their relatives and friends.

    So, Why has it been so hard, for dead fishermen and “bold” scientists alike, to admit there are no “safe harbors of knowledge,” only consequential risks while dancing with uncertainty, chaos and complexity?

    Given the gravity of historical trends observed with ACD, the existential risks of triggering even more climate forcing feedbacks, and the opportunity costs of forsaking the precautionary principle, I’d like to think scientists could reasonably advocate for the latter.

  8. 8
    Marco says:

    Hank #1 – maybe relevant to note that the “Peter Ward” in the link refers to Peter D. Ward, not Peter L. Ward, who is a bit of a…uhm…special case when it comes to climate change.

  9. 9
    stefan says:

    Gavin, that is a very thoughtful and well-argued article! I mostly agree but am not fully convinced by your dismissal of reticence. Despite Wadhams and Maslowski – I do not think that there are “plenty” climate scientists happy to make dramatic predictions; rather these are the exceptions. First of all, even us as scientists are just people that grew up in the status quo and thus find it hard to believe that the world could become very different. There probably is a subconscious bias there towards believing in a stable world, if you’ve lived all your life in one. Second, there is a price to pay for sticking your neck out – colleagues that ridicule you, public attacks, loss of credibility. It’s much safer for a scientist to stay within the moderate mainstream. And finally, despite the undoubted benefit and importance of assessments, their consensus writing procedure plausibly introduces a bias towards the lowest common denominator which many scientists can agree on. That would tend to curtail even plausible “worst case scenarios”.

  10. 10
    nigelj says:

    “What can be foreseen about “unknown unknowns”?Well, there will always be yet to be discovered “system collapses, refugee crises, health service outages” that push toward greater alarm. The problem will always be worse than we thought and climate sensitivity always greater than once forecast.”

    These are mostly not unknown unknowns, they are highly plausible outcomes and could even be approximately quantified in some cases. They will obviously be worse the higher temperatures get, but will still not be”unknown unknowns”.

    Great article Gavin.

  11. 11
    Dan bloom says:

    Google an important term coined by a psychotherapist in California. “The thought unknown” …his coinage should be part of this discussion.

  12. 12
    Urs Neu says:

    Thanks Gavin for this very interesting discussion of an important issue. I agree with Stefan’s comment on a possible bias and I would add another point: (most) scientists tend to restrict their assessments to systems they can describe and calculate and are very cautious with estimates concerning developments they only roughly understand and cannot simulate accurately with models. This leads to the fact, that developments not skillfully addressed by models might be neglected or left out when discussing worst case scenarios (as e.g. was the case with possible ice sheet instability when calculating the likely range of sea level rise in earlier IPCC reports). Of course, this has some validity, but it still might lead to a biased impression concerning worst case scenarios.

    Another example that I am concerned about is the implications of Arctic warming on circulation changes, where we are still struggling with adequate projections of possible changes and their impacts. For many regions, in the near future (decadal timescale) the corresponding impacts will probably be much more important than the effect of mean global warming. An increase of persistence in rossby wave patterns or changes in their shape distributions (amplitude etc.), might have quite dramatic consequences in regions like California or the Mediterranean, or the mid-latitudes in general. However, since we are – besides some ideas and hints – rather unsure what really could happen and models seem still not very skillfull in this regard, we are very careful to talk and even think about possible “worst case scenarios”, e.g. that the present dryness in California might soon become quasi steady-state, because the rossby wave pattern in this region is in a way pre-determined for dynamical reasons (due to the Rocky mountains), and weakening dynamics of Rossby wave patterns might make a deviation from this pre-determined pattern more and more unlikely. This is just speculating what could happen, yet without a model-based oder statistical background, but not unthinkable in my opinion. And that might be a real worst-case scenario for California… How should scientists act in such cases? At the moment we are looking at models (although not sure about their skills concerning circulation changes) and patterns that are most easy to describe (e.g. blockings) or statistics. Is that enough?

  13. 13
    Dan bloom says:

    Christopher Bollas coined the term “the unthought known.” See Wikipedia page entry. I asked Dr Bollas if this could Appleby climate change discussions and said sure.

  14. 14

    We know these “robust summaries” by the IPCC.

    => How the IPCC Underestimated Climate Change

    => The IPCC May Have Underestimated Future Warming Trends

    Also => Global warming will happen faster than we think

    Last not least a meta study => Ice Melt, Sea Level Rise and Superstorms Video Abstract

    These “robust summaries” aren’t as “robust” as Gavin is stating.

  15. 15

    Thanks for generating this vitally important discussion.

    For comic relief – a Mankoff cartoon favorite reads:
    “While the end-of-the-world scenario will be rife with unimaginable horrors, we believe that the pre-end period will be filled with unprecedented opportunities for profit”×768.jpg

  16. 16
    Paul Damascene says:

    One problem I find with the author’s thesis lies in his focus on worst cases around a single issue. In worst-case modelling in industry, there is now a shift of emphasis away from beginning with a credible worst case to beginning with a worst-imaginable case–precisely to document the scope of issues considered–before identifying a credible worst case within that domain. To borrow from this terminology, then, one could contemplate a whole series of credible worst cases on a number of issues (whether related or unrelated) that in the aggregate result in something worse than any if not all of the worst imaginable cases for any one issue. I would argue that anthropogenic climate disruption is more like that — a cascade of related and unrelated bad outcomes — some imaginable, some credible, some unforeseen, whether in themselves or in their interactions — that, overall, yield an outcome or outcomes worse than credibly expected or even imagined.

  17. 17
    BJ Chippindale says:

    Gavin – VERY nicely written. I do not however agree with the “incredibly stupid and incredibly unlucky”. We merely have to be human and miss one effect for it to be more than civilization can stand.

    I’m afraid that there are some aspects of this that are at the interface with human civilization that we aren’t quite capturing.

    First: The climate is destabilizing. The CO2 hammer has put a step function input into the complex system and the earth’s climate is going to ring like a bell. The speed of the change and the effect of that suddenness is not being well considered.

    It is all well and good to work out the stable end-point. I expect the planet will get there quite handily. We OTOH, may not. I have little patience with those who are married to “linearity” here. The time of linear response is I think, over.

    Second: The thing we capture least well in the climate models is rainfall, snowfall and clouds. We know more moisture will go into the air. We know it will fall out of the sky. Where, when and in what quantities? With the CO2 still rising and the delay in the resulting change imposed by the ocean’s inertia, we can only be sure that we aren’t sure.

    Third: If you don’t think a stable climate is important, ask a farmer.

    It doesn’t take a lot of similar events to wreck the agricultural base that feeds 7.5 billion people. It doesn’t take a lot of hunger to wreck a government. It doesn’t take a lot of wrecked governments of people better armed than fed, to wreck human civilization for an awful lot of (possibly all) humans.

    As an Engineer I learned to look hard at worst cases, and not to indulge in any belief in any deity but Murphy. My God has not disappointed me, punishing hubris and rewarding caution quite reliably. Perhaps that shapes my thinking, but given the risks, I think Science and Scientists need to consider that the worst case is all too likely.

    We are humans, and we are genetically incompetent to deal with problems that take more than a lifetime to appear.

    Your comments about how fraught with difficult communicating about it is are well reasoned and the clear mass media effect I see is one of “it isn’t time to panic yet” which is to say “do nothing drastic” and that is not actually going to work out well for us if we’ve destabilized (and are still hammering) the only climate we have. Worst case is a loss of civilization – tribes of isolated nomads who have forgotten everything we’ve learned in the past 3000 years.

    Worse than that is not possible for any intelligent citizen of the planet.

    regards BJ

  18. 18
    David Young says:

    Well, There is a massive bias in the media to exaggerate every disaster and every danger because it sells papers and generates clicks. The Weather channel will constantly cite some statistic that is “the worst in xxx years.” That’s called cherry picking. On site reporters always seek out the “worst” events for the same reason or ones that fit cultural or ideological biases.

    The public realizes this I believe and has become cynical about the propagandistic uses of fear and loathing. They are also very tired of the constant stream of pseudo-science advice on health and diet. The failures here of the scientific community are obvious and the science often very flawed.

    Scientists who make unfounded predictions will quickly lose all credibility. There is an element of cosmic justice here. Ethical behavior and objectivity actually really has better consequences than attempting to steer outcomes by shading the truth.

  19. 19
    jai mitchell says:

    Anderegg, W.L. et. al.
    American Meteorological Society – September 2014

    CONCLUDING REMARKS. The two case studies analyzed here illustrate the intricacies and complexities in avoiding both type 1 and type 2 errors in scientific assessments. Oppenheimer and colleagues (2007) have noted that searching for consensus in an assessment process such as the IPCC can be counterproductive to risk assessment. We suggest that assessment can further institutionalize the aversion to type 1 errors and attendant risk of committing type 2 errors. Both in paradigm and procedure, the scientific method and culture prioritize type 1 error aversion (Hansson 2013) and “erring on the side of least drama” (O’Reilly et al. 2011) or “scientific reticence” (Hansen 2007), and this can be amplified by both publication bias and scientific assessment(Freudenburg and Muselli 2010; Lemons et al. 1997; O’Reilly et al. 2011). Thus, the high consequence and tails of the distribution of climate impacts, 1448 | SEPTEMBER 2014 where experts may disagree on likelihood or where understanding is still limited, can often be left out or understated in the assessment process (Oppenheimer et al. 2007; Socolow 2011). As participants in the IPCC assessments, we have observed the excessive focus on avoiding type 1 errors at various stages in the assessment process, which may have worsened following the Himalayan glacier event.

  20. 20
    Dan bloom says:

    Anybody home?

  21. 21
    Ron R. says:

    My own view is that scientists generally try to do the right thing, sharing the truth as best they see it…

    My view too. That’s what sets the narrative apart from the one right wingers like to tell, that rather than a group of educated people trying rigorously to find and reveal the truth, they belong to some evil illuminati-like cabal, hoping to get rich off of a giant intentional lie. Right, that’s why there have been so many studies and counter studies done trying to narrow down the facts. They are doing the best they can do with the availabile tools at hand.

    It’s always easier to see the truth in hindsight. At the beginning, and along the way, your trying to use foresight. Much much harder. For one thing, there’s always random factors that pop up out of nowhere. Then there are all of the malignant attacks from those hoping to discredit to work for selfish reasons. Big Dirty Energy comes to mind. That constant attack makes humans draw inward, to circle the wagons, which can lead to inaccuracies. Yet environmental scientists struggle to stay on balanced, to find the truth. Their findings can always be overturned with evidence.

  22. 22
    Ron R. says:

    “Their findings can always be overturned with evidence”. That no one’s been able to (legitimately) so far should, like evolution theory, be enough to convince the still unconvinced.

  23. 23
  24. 24
    Steve Tubb says:

    I am not a scientist, but do understand the basic principles driving climate change and of course the consequences. Generally we now know what will happen and could happen. Are we stupid and unlucky? Possibly, but I think we can explain our (humanity’s) responses so far and into the future in a different way. As I see it, the problems lie in the following: 1) Consequences of CC will not be globally specific and uniform for various reasons including location, time lag and response; 2) Humanity currently seems to possess troubling states of mind including risk aversion, crisis denial (until it hits and even afterward), innovation dependence and attainment security (don’t want to change or give up what we have achieved unless something better comes along); 3) The values (what we value and how we value) we adhere to are all over the map – difficult to get a focused consensus on anything; 4) Only a relatively small part of humanity is truly engaged in the dialogue about and preparation for CC; 5) Opportunities for real change are lacking and some will require enormous amounts of investment and time (decades) to achieve.
    Frankly, I believe we are well into CC and over time incrementally conditions will change and consequences will be real. We need to plan and prepare for those consequences and importantly for the light-bulb moment for most of humanity when it is realized science has been correct all along.
    Rather than talking about mitigation (a fancy word) we need to talk about prevention. The worst is yet to come, whatever the worst might be. To ensure the worst is not catastrophic we need to seriously prevent it from happening.
    When the shit hits the fan, as it will and as we are now seeing, my real concern lies in the relatively hidden consequences and humanity’s response and the inevitable migrations that will occur. People will need to be accommodated in new locations and I do not think we have really begun to have that conversation about logistics and costs ($, environmental, relocations, food security, political cooperation, etc, etc, etc.)
    I referred to values above and respectfully suggest we need to adjust comprehensively away from power, greed and irresponsibility toward respect, trust and compassion. I also think injection of a dose of reality is needed and therein lies a problem as many are still not willing to accept the science and the inevitability of CC. Consensus would be preferred, but at what stage do we realize imposition is the only alternative? I really hope it does not come to that scenario.

  25. 25
    Patrick Mazza says:

    Still, you have non-reticent scientists such as James Hansen and Kevin Anderson telling us we need radical carbon emissions reductions this next decade. Anderson = 10%/yr. to have a 50% chance of holding under 2 deg. C. Hansen saying we need 15%/yr. to stabilize close to current levels and return to 350ppm by 2100. Hansen also asserting if we go to 2 deg. C slow feedbacks will move us to 3 deg. C. I am sure people on this site are well aware. So, considering the inertia of world political and economic systems, it seems unwise not to emphasize that highly negative scenarios veering toward worst case are not low probabilities.

  26. 26
    Hervé says:

    Dear Gavin and all,
    Thanks for the thoughtful and stimulating article. I somehow agree with Stefan who claims that scientists are, like most people, strongly influenced by their “real life” experiences and are thus biased towards “believing in a stable world”. Experience is not necessarily an asset when facing an unprecedented problem.
    Some more useful food for thought can be found in:
    – The Imperative of Responsibility: In Search of Ethics for the Technological Age (translation of Das Prinzip Verantwortung) trans. Hans Jonas and David Herr (1979). ISBN 0-226-40597-4 (University of Chicago Press, 1984) ISBN 0-226-40596-6
    – Pour un catastrophisme éclairé – Quand l’impossible est certain
    Jean-Pierre Dupuy, Eds du Seuil (2002). Not sure it was translated in English.
    Any other reading suggestion is welcome.
    All the best,

  27. 27
    Michael D Sweet. says:

    While for the most I agree with Gavin, I think the design of the IPCC process limits discussion of worst cases. When I look at the last two IPCC reports estimates of sea level rise I see experts saying that they were too conservative, especially on the high end. This directly relates to Gavin’s comments and suggests scientific reticense in the IPCC reports.

    It seems to me that the IPCC puts in the low value that 80% of scientists agree is most likely. This seems like a good amount. Then they put in a value for the highest amount of rise that 80% of scientists agree is the smallest amount of high rise. RealClimate reports that the high end of possible sea level rise that experts expect is greater than the high end estimate of the IPCC! They should put in a high value that 80% of scientists think is the highest possible rise, which is more comparable to using a low value that 80% of experts feel is the lowest possible.

    While I see the point in using the IPCC estimates for discussion, by design they minimize the anticipated risk from all sources. As I understand it, the IPCC takes the lowest risk that 80% of scientists agree on. For risk assessment the IPCC should use at least the average risk scientists estimate but I think the amount that 80% of scientists estimate is the highest possible would be better.

  28. 28
    Tom Sager says:

    This conjecture on worst-case scenarios violates Murphy’s Law. “incredibly stupid and incredibly unlucky.” Indeed! What kind of people would amass enough nuclear weapons to destroy the world many times over and then reason that it makes them secure? — incredibly stupid ones! I’m not sure luck has anything to do with this. “Be not deceived; God is not mocked: for whatsoever a man soweth, that shall he also reap.” There are so many ways in which our overly complex “civilization” could collapse. Perhaps each one has a small probability; but the sum could be significant, perhaps overwhelmingly so.

  29. 29
    Ruth Anthony-Gardner says:

    This was published on the 26th. Report of the Stratocumulus Tipping Point came out the day before.
    As far as I’m concerned, an additional 8°C jump on top of already devastating warming is a game changer, when discussion worst case scenarios.

  30. 30

    Thanks for your very interesting article Gavin.

    Google an expression like “xxx is happening faster than expected” where xxx is climate change, glacier melt, warming, etc. and you get enormous numbers of hits. Then try “xxx is slower than expected” or “xxx is changing as expected” and you get next to nothing.

    Doesn’t this tell us that our climate science is extremely biased towards being way too conservative, and that our projections are almost always way too slow, and that we can expect that rates of change will likely be much higher in future projections?

    Thus, worst case scenarios are at a higher risk of happening than we think.

  31. 31

    DDS, #4–

    These wolves grow larger while never eating any sheep.

    If you look at climate-related catastrophes occurring this century, it doesn’t take very long to get to 100,000 premature deaths and $100 billion in economic loss–and that’s just by a superficial, informal count. Attempts at more rigorous accountings have gone as high as a mean *annual* toll of 400,000 dead.

    So it would seem that there are quite a few missing “sheep”. Of course, being missing, they are unable to make any noise that might serve to draw attention. So you have to be willing to look.

  32. 32
    Martin Manning says:

    To follow up the point made by Michael Sweet, I think the most obvious examples of IPCC assessments adopting an overly conservative approach are in the area of social responses to impacts. For example, while Chapter 12 of the 2014 WG2-AR5 report had a section on climate change as a cause of conflict, there was no mention that the Canadian IISD group had actually predicted the Syrian war could occur two years before it did. They had noted that drought had led to destabilisation and forced migration with 160 villages being abandoned and said “in our opinion, climate change is clearly a potential factor in future conflict in the region for six specific reasons: …” (see

    Current Climate Change Reports is putting out papers on conflict, e.g. Miles-Novelo & Anderson (2019), Climate Change and Psychology: Effects of Rapid Global Warming on Violence and Aggression, online 31 January 2019. This cites 89 other papers relevant to the topic. But the chapter structure for the WG2-AR6 is now set and it is not clear where all this can be covered.

  33. 33
    Hank Roberts says:

    In the past decade ocean oxygen levels have taken a dive—an alarming trend that is linked to climate change, says Andreas Oschlies, an oceanographer at the Helmholtz Center for Ocean Research Kiel in Germany, whose team tracks ocean oxygen levels worldwide. “We were surprised by the intensity of the changes we saw, how rapidly oxygen is going down in the ocean and how large the effects on marine ecosystems are,” he says.

    It is no surprise to scientists that warming oceans are losing oxygen, but the scale of the dip calls for urgent attention, Oschlies says. Oxygen levels in some tropical regions have dropped by a startling 40 percent in the last 50 years, some recent studies reveal. Levels have dropped more subtly elsewhere, with an average loss of 2 percent globally.

  34. 34
    John Atkeison says:

    Thanks for an excellent presentation and great discussion. (My instinct is to bet that things will get somewhat worse, somewhat faster, than formally expected because I think that is what we’ve seen over the recent 20 years. Yes? No?)

    I have a primary rule: tell the truth. Tell it as best you can, as wholly as you can, as realistically as you can, as personally honest as you can.
    In presentations to civic organizations like Kiwanis, Rotary, etc., I have found a sea change in attitude over the past 7 or 8 years. Many more people are knowledgeable about global warming and the climate warping that it causes. There is a wide range of understanding with regard to sorting out the issues that were explored in this post, but 80% of people know some damn thing is going on– even here in Nebraska.
    One more thing- being truthful and accurate means keeping up. The tone and content of The Special Report on 1.5ºC ( is a clear warning. IMHO, the Green New Deal as in the AOC/Markey Resolution ( as introduced in Congress is a directly appropriate response, in timing and in measures generally proposed. It is a terrific start to the discussion.

  35. 35
    nigelj says:

    Well written article that is compelling. I’m not a scientist (I came close to doing a geology or chemistry degree) but I have followed the climate debates. Just tell the plain truth about climate issues including possible high risk scenarios . The IPCC can only base things on the best understanding of risks at some point in time. In public messaging use the high risk scenarious that do have a strong evidence based foundation, like the hothouse earth scenario which has historical evidence, something the public could probably relate to.

    I agree I don’t think it would be wise to downplay risks in some convoluted strategy to motivate people. Too much could go wrong with this and it could damage the credibility of scientists and the authorities. Even worse changing tack now could send the mistaken message that the climate problem is insignificant.

    I would also say avoid exaggerations, and speculation not backed up with any solid evidence, which will also backfire and destroy credibility of scientists. The IPCC reports are quite scary enough if the right parts are emphasised.

    I understand some research appears to show fear is not useful in terms of long term issues like climate change, however fear is a powerful human motivator. For example it would seem strange to not discuss potential refugee problems etc in case it causes a fear reaction. Research shows fear works as a motivator for change in health related campaigns, provided its based on facts and not exaggerated. Perhaps fear could be used, but not over played, and should be combined with positive messaging on solutions. I don’t see that they are mutually exclusive.

    I think its also useful to consider how the risks are communicated to the public. Both scientists and the media and governments obviously have some part to play in this. The IPCC reports are good, except that the summary for policy makers is very nuanced with statements about probabilities of more severe weather etc and not much about spelling out in plain language the risks of severe heatwaves and possible warming above 4 degrees etcetera, and key issues get lost among the vast quantity of information. Media articles take a similar line possibly following on from this lead. I think M Sweet is also right about the way sea level is assessed and presented. As a result the public get vague statements about “more severe weather” that is not really sufficient to get the risks across.

  36. 36
    Al Bundy says:

    David Beebe,
    Excellent comment. I won’t detract by opening my mouth; instead I simply salute.

    Duh, but “duh” is curiously hard for most humans to assimilate.

    Wrong. You are conflating orders of magnitude. A decision that only affects the error bars and is scrubbed by statistical variation of individual cases is irrelevant. Reducing humans to “Do you want me to use blue or yellow diaper pins to strap you in your inevitable diaper isn’t providing a useful choice. It is just placating the powerless.

    I’m not going to read further at this moment. I’ll analyze later. But the core issue, the POINT, is “are estimates trending towards less or more harm over time?

    It doesn’t matter why the answer is what it is. It is friggin critical what the answer is

  37. 37
    Ray Ladbury says:

    What people fail to comprehend is that “worst case” is not necessarily what science does best. The whole enterprise of science is dedicated to understanding the significant influences/causes/drivers in the systems under study. Models are always simplifications of reality–and the simplification may fail as conditions diverge from prior experience. Add to this the consensus nature of the IPCC–or any scientific effort that seeks to provide advice to policy makers, and it is surprising the product is not more conservative than it is.

    Normally, it is the policy makers that must anticipate the worst case–or at least worst-case for a given confidence. However, policy types are afraid to touch climate change.

  38. 38

    Martin Manning, #32–

    One could quibble about the IISD “predicting” the conflict: while the events that occurred in Syria are pretty precise manifestations of their Threats #1-4, the absence of a timeline for conflict, combined with the long time horizon of the progression of the effects of climate change, would be used by denialati as ‘wiggle room’.

    Nevertheless, I find the report strikingly prescient. I’m surprised, given the significance of the Syrian war to global politics, that this analysis isn’t better known. It deserves to be–even if the drought was already begun when the report was written.

  39. 39
    zebra says:

    Excellent discussion. I found comments #12 and #16 particularly compelling.

    I would find it helpful to have some kind of chart– all on one page type– that classifies outcomes (1) in terms of their relative probabilities and (2) relative harms.

    Of course this is a very difficult topic for communication. Let’s remember that there are headlines about potential non-climate related disasters like India-Pakistan getting out of hand because of… oh yeah, luck and stupidity. That’s before they are actually fighting over water and refugees.

    It’s hard to blame “the public” for blanking out one thing or the other.

  40. 40
    Radge Havers says:

    Great posting.

    IMO when discussing worst case scenarios for the general public, probably shouldn’t neglect to at least give a nod to other contributing factors, and to also mention that the current case scenario is already showing signs of lethality.

    If you want to down play the hyperbole, fine and good so long as you provide an alternative narrative.

    Anyway, an oldie but goodie;

    How to screw up. Count the ways:
    — Pollution
    — Urban sprawl, and countless other land-use issues
    — Resource depeltion
    — Population
    — …

    You know, (and why I tend to be less than optimistic) something that continues to be purposely downplayed in media discussions of civil behavior, is the intractable effects of decades of “culture war” and propaganda. The value of reasonable, rational give and take has intentionally been destroyed.

    IOW, good faith has been defenestrated, and to mix a metaphor, all the kings horses and all the kings men…


    “The unthought known” thanks, useful idea.

  41. 41
    Larry Gilman says:

    _Does_ fear demotivate? It’s my impression that people are most likely to act when they experience a blend of urgency and agency. Dire climate predictions induce urgency, a.k.a. fear — and how not? Isn’t fear a reasonable response to the spectrum of likely outcomes, and to the strong tendency of the scientific picture to worsen as more is learned? Do not many climate scientists testify to feeling waves of horror (e.g.,, )? But horror, one doesn’t need a social scientist to know, paralyzes in the absence of agency, perceived non-helplessness. So we need both: urgency and agency, doom plus zoom.

    It’s precisely this combination that is driving climate activism of unprecedented scope around the world right now. The people most prominently active around climate change, including the Sunrise movement, Extinction Rebellion,, the backers of the Green New Deal, and others, universally testify to being motivated at least partly by fear and grief. These very scared people are doing the most, not the least. Three cheers, then, for fearmongering! — if it is science-based and linked to real options for action.

    Re. Mr. Schneider’s main concern: Are climate scientists in fact softpedaling in the belief that scared people don’t act? Mr. Schneider seems to think not, writing that for it this “to be a real motivation” it would “have to be true . . . that actually telling the truth would be demotivating.”

    It wouldn’t have to be true, though: it would only have to be believed. And how many climate scientists believe that tactical softpedaling is the way to go can only known through surveys. We should be agnostic, unless or until such data exist, about how widespread tactical softpedaling is among climate scientists.

  42. 42
    Nemesis says:

    I just have to look out of my window or read the daily news or analyze the economic and political system to know the worst case scenario is real. I started studying the shit some 30 years ago and it’s easy to summarize what I’ve seen during these decades and what I still see:

    Worse and worse and worse adds up to worst.

    No wonder, as you can not expect reasonable and healthy outcomes from a criminal, sick and totally flawed system. No way.

    Uh oh, but what about “optimism”?! Hahaha, calculated optimism without any reason is a primary factor that led us exactly to where we are right now. But they (the funny economy in the first place^^) will preach “optimism” without any reason until they are done. Calculated “optimism” in the cooking pot of Nature is the road to hell.

  43. 43
    Phil Hays says:

    Green New Deal doesn’t have the time scale correct.

    To convert the USA and/or the world to non-carbon energy is likely going to take more than 50 years.

    Start with alternative energy sources. Wind, fission, geothermal, tidal, OTEC, hydro, fusion all have issues that will likely prevent them from becoming the major source of energy. Fission might have a case. Fusion, once we find out how to make it practical, might have a better case.

    That leaves solar. Solar is growing rapidly, and is past 1% of electric power generated. If we solve all the technical problems with a nearly 100% solar based economy, and we keep ramping up solar cell production, someplace around 2070 or so we might be to a 90% solar and other renewables based economy. To get the last 10% might take another 50 years or more.

    We don’t have a realistic hope of avoiding 1.5C warming. Or 2C warming. That was my conclusion back in 1980, and I don’t see anything getting any better.

  44. 44
    Stephen Verchinski says:

    ” Imminent massive methane releases that are large enough to seriously affect global climate are not going to happen (there isn’t that much methane around, the Arctic was warmer than present both in the early Holocene and last interglacial and nothing similar has occurred). ”

    Curious as to which scientific report was used to make this statement. It seems to contradict the IPCC C.1.3. discussion of the release posed by some to release 100 GT

  45. 45
    Tom Sager says:

    For Martin Manning (comment #32): Thanks for link to iisd article. I’ve been looking for something like this for a long time. Have you other similar articles to recommend?

  46. 46
    Al Bundy says:

    BJ Chippindale,
    You impress me. If you’d like to be part of the answer, contact me at

    Steve Tubb, Tom Sager, Paul Beckwith, John Atkeison


    Lololol..use fear of brown refugees to motivate white bigots!

    Brilliant, dude.

    Stuck up people insist that their way is the only proper way. Your description of scientists sounds like the very def, eh?

    You are not just welcome, but core.

    Guys, you’ve read my thoughts. You know I don’t blow smoke.

    It is time.

  47. 47
    John Monro says:

    Hello folks, thought provoking item, and some thoughtful and insightful comments. Thank you all.

    I look at it like this, as an analogy. Many people have one incredibly expensive item that they own or at least are borrowing to own, and that’s their property. The loss of a property would impoverish most of us. So we insure. We spend a lot of hard-earned money for insurance, and many large multinational corporations make a lot of money to help us do this. But when we insure, we never do on the best, or even the medium case scenario. We ensure for the worst imaginable. That’s fire, obviously, but also earthquake and flood, or perhaps a falling aircraft or meteorite, totally overwhelming incidents that would destroy our property, and perhaps our lives. But possibly not a single one more than 1% likely in any one person’s lifetime. We also insure for less serious things, burglary, damage from a car crashing into the house or whatever. Fortunately, if our house is damaged beyond repair, then the insurance company can come along and rebuild it or provide you with enough funds to buy another.

    Humanity lives on what is basically a very large property, a unique piece of real estate in the universe which is completely self contained. It’s pretty self evident that if we seriously damage this property, there’s no galactic insurance company is going to come along to help us rebuild it or relocate us.

    If we lived in an uninsured house, we’d worry most days about this, and equally we’d do everything in our power to prevent some untoward event. We’d spend money on burglar alarms, fire extinguishers, flood protection or whatever, and we’d make damn sure that no activity of ours would put our home at risk. We probably wouldn’t have a brass foundry or oil fractionating facility or chemical laboratory in the basement.

    We live on an uninsured and uninsurable planet. We now know perfectly well that the things we are doing, on a mighty international industrial scale, is seriously and perhaps irrevocably damaging this planet. This isn’t just CO2 or methane emissions, and consequent global warming, unbearable heat, fires, drought and flood, but it’s plastic pollution, mining and waste, ocean depletion and acidification, loss of soil from over-intensive and inappropriate agriculture, it’s the collapse of insect populations and of the early signs of a mass extinction. And it’s nuclear arms and geopolitical dysfunction and rivalry. It’s social breakdown and political extremism.

    The arguments about the severity of global warming in the face of these many existential threats are, I believe, peripheral. Global warming is happening, that’s all we basically need to know. We have to deal with it, now, just as we have to deal with all these other issues, now. We either look after our planet, totally, all the time, or we fail to and suffer the consequences.

    We, humanity that is, and its leadership, urgently have to change. (By urgent, I mean pressing, dire or desperate, it means it’s an emergency, because the warnings are insistent and unrelenting) Realistically, I don’t think we will, as history is replete with failed civilisations. Indeed we’ve now had a generation in which to change for the better, but we are actually doing worse, even worse than the worse predictions of the IPCC. In the news today is that in Shanghai they’re completing this “horizontal skyscraper”. Most people see this as a triumph of technology, something to wonder and glory in. It may be a “triumph of technology (of today), but it’s also totally misapplied technology in the face of all these existential threats – maybe something to wonder about but certainly nothing to glorify, just terrify. When I look at cities like Shanghai, New York or Las Vegas or all the other exponentially expanding skyscraper cities of China, the Middle East and around the world, I just see human hubris. I just see totally thoughtless application of economic and social principles that have long since passed their use by date. I just see us digging this huge entropic hole on our planet, which we have now made so deep we will never be able to climb out of it.

  48. 48
    John Monro says:

    …….and the existential threat that’s really obvious, maybe even the most important, and I omitted to mention, that’s overpopulation.

  49. 49
    Nemesis says:

    @John Monro, #48

    ” …….and the existential threat that’s really obvious, maybe even the most important, and I omitted to mention, that’s overpopulation.”

    Any solution?

    I saw 10 fellow human beings at the table and 10 loafs of bread. Now 1 of these 10 fellow human beings took 9 loafs of bread and said “we have a problem, our house is overpopulated!” Now guess what happened next…

    … boing poum tchak…

    Btw, how many kids do have? I have none and I’m happy.

  50. 50

    Phil Hayes, #43–

    You’re wrong to dismiss wind; it’s an excellent complement to solar, and will continue to grow robustly. IMO, the new energy economy is emerging all around us, and will be here long before 2070. Most likely, it will still have a significant nuclear (fission) component, though a patchy one.