Wednesday 19 July 2017

How to talk about climate doomsday scenarios

David Wallace-Wells wrote a 7000 word cover story in New York Magazine on how unchecked climate change may make the Earth uninhabitable. With 2.5 million readers this longread was the most read article in the magazine's history.

That shows the impact a well written article can have. It also points to a change in the mood in America. Since Trump pulled out of the Paris Agreement, happy to risk climate rapture for the quarterly earning of his donors, attention for climate change in America has spiked.

Americans are living through a nightmare, where you see the danger coming, but cannot convince others to stand up to it. Until you wake up bathing in sweat and pick up your phone from the night stand to read tweets from climate "sceptics" mocking you for facing reality. The same people that make a nightmare out of a perfectly solvable problem.

Sixteen climate scientists of Climate Feedback (including me) reviewed the NY Mag article. This number of reviewers may also be a record.

My summary would be that while the dangers of unfettered climate change are real, we found many inaccuracies, which typically exaggerated the problem. Thus the article was rated as having a "low scientific credibility". Both the NY Mag article and the Climate Feedback rating and earlier criticism by Michael Mann have sparked some controversy.

This post will be about how I would prefer the media to report on worse case scenarios. A second post will be about whether our "nitpicking" was the science police striking again?

I have no idea how 2100 looks like. Put yourself in the place of a well-informed citizen of 1900 and what they may have thought today looks like. Meters of horse shit on the streets due to the growth of traffic? Or imagine how an American thought this time would look like 50 years ago. Flying cars and Mars colonies?

Maybe in 2100 humanity has gone extinct, maybe civilization is gone, maybe humans are enslaved by corporations, maybe currently poor countries are also affluent and corporation can no longer repress us, maybe after another century of development we lead wonderful lives, maybe we are building our first intersolar cruiser, maybe no one cares about intersolar cruisers and people impress each other with poetry and four dimensional chess. Very likely they will be painfully embarrassed for me for the options I gave.

I have no idea how they view climate change in 2100. Do they see it as the biggest historical liability put on them? Are they annoyed at the tax burden for the huge necessary geo-engineering program? Do they wonder why people in 2017 thought it was such a big problem, while it was so easy to solve? Are they happy that due to the geo-engineering program they now have weather satellites and it only rains at night in urban areas?

Even in the best case scenario we are taking the climate system out of known territories. There will be many surprises and to be honest those are what worry me the most. The Uncertainty monster is not our friend and that makes it very hard to say which worst case scenarios are unrealistic.

It is custom to accept smaller risks the bigger the stakes are. Cars and smoking kill many people, but one at a time. A risk someone may be willing to take personally will be larger than the risk one takes with a community, a country or our civilization. The risk of dying in a car accident is 1 in 84 (1.2 %), this would be an unacceptable risk for civilization or humanity. Thus we have to look at the tails.

Finally, we expect the impacts of climate change to accelerate: Because some variability is normal, the first degree of warming makes much less problems than the next. Thus the risks of above average warming are expected to contribute much to the total risk. It is thus good that the article explores what surprises may be in store and talks about scenarios that are not likely, but risky.

Four horsemen of the apocalypse

The article reaches the worst case scenarios in four ways:
  1. The worst case for the emissions of greenhouse gases.
  2. The worst case for how sensitive the climate system responds.
  3. The worst case for the impacts and how humanity responds.
  4. The worst case for the scientific assessment of the evidence.
1. The worst case emission scenario was the [[RCP8.5 scenario]] of the IPCC. These scenarios are really just that: scenarios. No probability is assigned to them.

This is the IPCC report from 2013 and the scenarios were created well before that. My impression is that with the [[Paris climate agreement]] and the fast drop in the prices of renewable energy and storage, the RCP8.5 scenario is no longer very realistic. Another optimistic sign is that industrial emissions have stabilized the last 3 years. However, the US mitigation sceptical movement and their president will keep on fighting to make this dystopia a reality. So it is a legitimate question what kind of a world fossil fuel companies and these people want to create.

2. It is completely legitimate to explore the tails of the probability distribution of the climate sensitivity. Even if it had only 30% probability, Trump did get elected. Even if the chance is just one out of six, you sometimes role a six. And let's not start about Russian roulette. Unfortunately the tail of the distribution is not well constrained and very high sensitivities are hard to exclude.

3. The uncertainty of some impacts can be quantified reasonably well. These are the ones with the most physics in them such as heat waves and large-scale increases in precipitation. Then it is legitimate to go into the tail.

Other impacts are not understood well enough yet (Will ice sheets collapse? How much greenhouse gasses will the soil release due to heating?) or will never be fully predictable because of societal and technological influences (Will The Netherlands evacuate before or after Noah's flood? Will plant breeding keep up with climate change? Will societies be able to cope with climate refugees?). In such cases I would like to hear a balanced spectrum of views, including extreme ones.

Because the broad sweeping article discussed many climate change impacts it could not do justice to complexity of individual impacts. Climate change is typically just one stressor of many. When The Netherlands floods, the climate "sceptics" will not suddenly wake up and say "silly me, I was wrong, now I recognize that climate change is a problem, sorry about that". They will say the storm was to blame and it was really bad luck the storm came from the North West and its maximum coincided with high tides, the dikes were not strong enough, the maintenance not good enough and especially the government is to blame.

Looking at history or at the future only from a climate change perspective brings back bad memories of [[climate determinism]]. The Age recently reported on farm workers in Central America suffering and dying from chronic kidney disease. The regions where this new decease happens is well predicted by warming and changes in insolation. Simultaneously the problem is that these people are so poor that they have to work on hot days and also have a strong work ethic that promotes this. They tend not to drink during work and when they do it are often soft drinks because they are perceived as safer. A large part of the problem is funding for preventative care and people die because they cannot afford dialysis.

This example shows two things. First of all, like the dikes breaking in The Netherlands, the problem has many aspects. Secondly, this was a problem because it was new. There will be many surprises due to climate change. The study of (rare) diseases helps in understand how a healthy body works. It shows what is important for healthy functioning. Medicine can study many bodies, we only have one Earth and will very likely be surprised what the Earth did for us without us realising it.

4. Like for non-physical impacts, where I am hesitant to go into the tail is when it comes to the interpretation of the evidence. That quickly ends in cherry picking experts that say what you want to hear. Those are strategies for mitigation sceptics. Even if those experts do not stray from the evidence and only hold a pessimistic view; I feel this is not for serious science reporting. It is fine to explain the ideas of such experts, but they should be balanced with other views.

Concluding, for the objective part of the problem: if you clearly say you are looking at the worst case feel free to go deep into the tail of the probability distribution. Only looking at mean changes understates the risks. The tail is a big part of the risk and thus very important. Do not forget to talk about many further surprises and that the Uncertainty Monster has an ugly bite.

When it comes to the more subjective parts, please balance pessimistic with optimistic voices. Subjective judgements are unavoidable when it comes to worst case scenarios and the far future where the changes will be largest. People can legitimately have different world views and as a science nerd I would like to hear the full range of legitimate views.

An article needs a focus, but please consider that climate change is one stressor of many. Climate change impacts are complicated, do them justice like a great novelist would and do not make a cartoon out of them.

Related reading

The updated New York Magazine piece By David Wallace-Wells: The Uninhabitable Earth - Famine, economic collapse, a sun that cooks us: What climate change could wreak — sooner than you think. (The reviewed original, with annotations)

The Climate Feedback Feedback: Scientists explain what New York Magazine article on “The Uninhabitable Earth” gets wrong.

New York Magazine now also published extended interviews with the scientists interviewed for the piece: James Hansen, Peter Ward, Walley Broker, Michael Mann, Michael Oppenheimer.

A balanced article in The Atlantic: Are We as Doomed as That New York Magazine Article Says? Why it's so hard to talk about the worst problem in the world.

Michael E. Mann: The ‘Fat Tail’ of Climate Change Risk

Fans of Judith Curry: the uncertainty monster is not your friend

Introduction to Climate Feedback: Climate scientists are now grading climate journalism

Michael E. Mann, Susan Joy Hassol and Tom Toles in the WP: Doomsday scenarios are as harmful as climate change denial. Good people, but I am not buying it: one negative journalistic story in a full media diet does not make people despair, hopeless and paralysed. Plus reality is what it is.

* Painting of the Four Horsemen by Viktor Vasnetsov -, Public Domain, Link


  1. By "Middle America", do you mean Central America (= Costa Rica, Nicaragua, etc.)? Middle American means something different.

  2. Ah, wikipedia disagrees. Hadn't heard of that definition.

  3. Wikipedia's Middle America fits even better than Central America, but I guess like you many will not know it. It is the term we use in Germany. Will update.

  4. "reality is what it is."
    Climate change is a matter of risk.
    Exploring that risk in an honest way must include the full distribution of possibility.

    The potential cost for our future civilization goes from near zero to infinity. Infinite means no more civilization to pay the costs.

    I read the piece and thought it explored the extreme end of consequence and clearly stated that was the aim.
    I also read a range of responses and believe that many in the scientific community over reacted in censuring an honest attempt to talk about the fat tail of possibility.

    In some ways this is a similar capture of argument as the pause nonsense was.
    We have a situation in that we can not talk honestly about such risk because others have poisoned the well with lies and misrepresentation down playing the potential effects.
    We are allowing the fringe to dictate what message is reasonable not the science.

  5. Agree with the first part. We should talk more in terms of risk.

    The Climate Feedback review is my next post. It would be nice to be in a situation where scientists do not have to correct misinformation. If only because that may scare people from participating in the debate. But now that scientists have been forced critique climate "sceptics" because they pollute the public debate so enormously, scientists should be fair and critique all sides.

    I think all sides should stop overusing the term censorship. It is a serious problem and not at all comparable to adults having a serious discussion.

  6. Victor, your faith that feedbacks *not included in RCP 8.5* can't make up the shortfall in fossil fuel emissions that were included is touching. I suppose that explains why you were so interested in arguing with me about that 204 Pg C-eq./2100 figure for permafrost. Hmm, disputing the science details to avoid talking about the big picture. Where have we seen that tactic before?

    But I hope you're right.

  7. I was not interested in arguing with you, I had a hard time getting a clear answer. It will be at least partially accounted for. We expect that the biosphere (and oceans) will take up a smaller fraction of CO2 than it did in the past. Looking forward to your review article.

    The CO2 horseman is not a bigger one than the other 3 horsemen.

    I also hope I am right or even pessimistic. There was just good news that adding sea weed t cow feed can reduce their methane emissions. That sounds like a quite cheep way to reduce an important source of greenhouse gasses. Let's hope some of the surprises will be good ones and not blend them out.

  8. 2100 victor?

    I suspect the EU elite will still be trying to screw as much money out of the UK as possible so that we can escape. Other than that we really can't know. They say the past is another country. As for the future, a century from now is a completely closed book other than we shall still, as always, have extremes of weather. Will they be more extreme? We need to look to the past in order to prepare for the future and there are many periods in most centuries with weather experiences much worse than today, other than around the last century which appears to have been quite a benign period with similarities to the 18th century and the 12th century.

    The least we can do is to make our infrastructure more robust as in recent decades we seem to have lost the notion of extremes. In particular I would like to make our heavily dependent society improve its electrical infrastructure to cope with another non climate related carrington event. To me this is a far more present and likely danger than a widespread catastrophe from man made climate change.
    ( as is cyber warfare)

  9. Censuring .
    express severe disapproval of (someone or something), especially in a formal statement.

  10. Tony, yes I expect that even a century after leaving the EU, part of the UK will still claim that all the bad consequences of their policy preferences are due to the EU. Some things will never change.

    With more energy available, more evaporation, more precipitation, it would be highly surprising if the extremes did not become worse or more frequent, except for frost, ice rain and that kind of things.

    Any other claims of less extreme weather will take a lot of high-quality research before I believe it. A hint is the severe weather in the USA compared to the mind weather in the UK, which is farther from the heat source.

  11. Griff, my apologies for the misunderstanding due to my limited English.

    I have not seen any censuring of this post yet or of my contribution to Climate Feedback where I wrote that the tail is very important. Also Michael Mann wrote about the importance of the fat tail (now in "related reading") while he is now censured (if I use it right) for pointing out errors in the NYM piece.

    One could argue that I am not important enough or that scientists have a tendency to like the science to be accurate. I prefer the last option.

  12. Victor

    I assume you have read the 'US weather review', a monthly and annual digest of weather occurrences in the stats, the earliest is around 1840

    There are numerous catastrophic weather events listed. In what ways are today's worse? Mind you of course there are no records of the extreme times we can see in European records that pre date the existence of the States.

    However, although not directly related to your article, the perils of a carrington event or a concerted cyber attack are far more serious for humanity.


  13. Of course there were extreme events in the past. I never claimed they are new, I only expect them to be bigger and more frequent.

    With much more past to chose from, with worse infrastructure and weather prediction still forbidden witchcraft the consequences used to be very large. That historians thus report on terrible extreme events is really not a valid reason to think there will be no increase.

    Yes, there are many more perils, that was partially what this post was about. I hope there are no people thinking that by creating more vitriol, pitting tribes against each other, and refusing to remove an important stressor we are going to face the other perils faster.

  14. Victor

    you said

    'That historians thus report on terrible extreme events is really not a valid reason to think there will be no increase.' Your point about there being more of the past to chose from is also fair comment.

    However, the claim is often made that there HAS been an increase in these terrible events. In other words we have already far exceeded the limits of natural variability and are doing so ever more frequently.

    I see no proper unequivocal evidence of that. If you believe there has been more such events it would be good to see some authenticated examples, preferably from Europe where records are longer and easier to cross check.

    As an example, there was a debate about Russian heatwaves on ATTP recently. There are many authenticated heatwaves every bit as bad as the one cited from a few years ago. There were around 5 in the 19th century alone.

    Incidentally, I don't want to be accused of a diversion, but it would be my observation that humanity is not very good at dealing with more than one 'crisis' at a time. Therefore are there other crises worse than AGW that need our attention first?

    I can think of 5 off the top of my head that need resolving and would have a direct impact on either health or security. So, in the great scheme of things does AGW warrant being our most important and most costly focus that takes up so much of our effort?


  15. I expect you know the IPCC special report on extremes (SREX) and the evidence it gives for changes in extremes.

    We always work on several problems at once. I am a fan of variability. Some higher risk science project mixed in with studies that are very likely to be useful.

    Climate change is one of the easiest ones to solve. Peace, rule of law, prosperity and end to corruption are much harder. Stopping drug war subsidising of drug cartels and the terror war producing more terrorists would be easier, we would just need the political will to stop doing that. But mostly the same demographic would be against that because it hurts other groups more than their own group.

  16. Victor

    Of course I know that IPCC report. It is very good, as all of them are. A wet day here enabled me to revisit it. :)

    The problem I have with many of these studies is that natural variability, I believe (and perhaps you do to) can not be measured by 'observed changes since 1950' nor anecdotes of the changes a 80 yearold Kenyan farmer has witnessed.

    I gave the example of the Russian heatwaves observed back to the 17th century as an indicator that perhaps modern Russian accounts are not out of the ordinary.

    Accounts of parrots falling out of the sky in Australia due to intense heat a couple of years ago are matched in the Watkins diary of 250 years ago.

    I find this sort of extremes are not out of the ordinary when looking further than a persons life span.

    I tend to like this sort of account from the IPCC report and like to see them followed through for as long a time scale as possible.

    "Before AD 1000, in the low-lying coastal floodplain of the southern North Sea and around the Rhine delta, the area that is now The
    Netherlands, the inhabitants lived on dwelling mounds...'

    Are todays conditions in Holland worse than the extreme storms of the 13th century that this account referenced?

    We can follow these detailed extreme events narrative through a small number of countries back some 250 years and an even smaller number of countries back some 1000 years (when anecdotes, superstition and religion can colour the accounts.)

    This is the sort of stuff I need to see in order to determine the worsening (or not) of extreme events in modern times.

    I have spent quite a few years in examining the British records and can see little sign of increasing extremes. Perhaps the same is true of other countries with long records such as Holland?

    From these we can see that climate is not constant, but can observe it is often highly variable

    How far back can we go to see if these extremes repeat themselves? I am comfortable with 1000 years and for one off events perhaps longer.

    To know their true extent we need to go to the start of the Holocene and I am not aware of any records reliable enough to home in on individual extreme events.

  17. Yes, it is a pity that many studies are on relatively short series. We are working on the digitisation of earlier daily data, but that is expensive. (It is also not helped by a mitigation sceptical movement that wants to punish the messenger when they are in power by cutting climate science budgets. It is also the last thing one would expect from a movement that claims to be interested in science and claims they need more evidence to be convinced.)

    Science does not give up so easily. Science will not say: Well we do not have daily measurements for the last 10.000 years from all over the world. So I guess we will never know whether extremes are changing.

    You do the best you can, do not only use time series analysis on local series, but look at many series, at the full distribution and try to understand the reasons for the changes to gain understanding.

    Do I remember right that there was an English village that was hit by 1000 year floods three times the last years? That could still be a coincidence. Only looking at the most extreme cases and only looking locally without understanding will not given answers before the end of civilisation.

    I would personally prefer more studies looking at changes in the variability, not jumping ahead to changes in extremes. For the variance you have much more data and the data is also more reliable than for extremes.

    What makes the Russian anecdote more reliable than the Kenyan?

  18. Hello Victor

    I don't really think we are in disagreement with each other. I remember a year or so ago you were trying to get funding for a project and hope you managed to obtain it. The more data we have on which to make decisions the better. It is perverse for sceptics to argue otherwise.

    There are many Russian studies from various sources covering several hundred years. Some of the great Russian novelists and composers recorded these events as they travelled around the country, as well as coming from more conventional diary sources. The choking smoke from smouldering peat bogs were a regular fact of life in Moscow. I think they were mentioned in 'War and Peace' but I don't intend to go through that very long and turgid novel to confirm this!

    The 80 year old Kenyan farmer may well be accurate in his recollections but it is one location over a very short time scale and we have no idea whether his experience is unique or commonplace in the historic record.

    As regards the 3 floods. I was on the Environment Agency flood defence committee for some 10 years and think the village you refer to was on my patch.

    It is in a location in western England that was at one time one large inland sea. From it King Alfred took refuge from the Vikings before setting out to defeat them in due course

    The area was drained by pumps for farming, but some 5 or 6 years ago the powers that be decided to let it become a nature reserve and turned off the pumps, some of which were sold, and also reduced the height of the dikes.

    Subsequently the water returned in considerable amounts, flooding large areas except for those slightly higher areas which in anglo saxon were called such names as 'great island' for reasons obvious to everyone except for the powers that be.

    The pumps have now been replaced, dykes reinstated and ditches dug. The area will always have a tendency to flood though, and whilst suitable for farming should not really be used for housing. However the pressures to live 'close to the water' are as strong in Britain as they are in Holland I expect.

    You will drive above this area on the M5 between the Met office and Bristol. In a wet winter it is typified by long expanses of water.

    Incidentally, as you know, it is thought the sea levels were higher in Alfred's time than today, which enabled the Vikings to sail their shallow long boats up rivers that are not navigable today.


  19. Victor

    I had meant to make one final point about our knowledge of extreme events in past times.

    They self evidently would have to occur in the first place

    Secondly; someone must have been in the locality to witness it (increasingly unlikely further back in time as populations were small)

    Thirdly; the event would have had to be recorded. (who by? Most people were illiterate and the average farm worker would not have had the time or inclination to seek out a scribe)

    Fourthly; the record would have to be stored

    Fifth; the record would have had to survive hundreds of years of conflict, fire and floods.

    Sixth; the record would have had to be relocated by a researcher

    Seventh; the piece of research would then have to be accessed by such as ourselves.

    It has been scientifically calculated (by me!) that of the 100% of extreme events that occurred over the centuries, 97% have never reached the attention of the researchers of the modern day for the reasons cited.

    In other words, only 3% of the extreme events that occurred (and were witnessed) have reached our attention.

    All joking aside, far more extreme events must have occurred in the past than we can possibly be aware of.

    Have a good afternoon.


  20. Sorry, tonyb, whatever shred of credibility you had came to an end with your *unsourced* claim of higher sea level during Viking times. You need to base your claims in the scientific literature or after a while you'll just get sniggered at. Being correct about your understanding of that literature would also be good. Sadly that's rather a lot of work.

    Victor, your proposal that I write my own review paper as punishment for instantly fulfilling your request for a citation to one is odd behavior for a scientists. But we all have egos, I suppose.

  21. Hello Steve

    As I was writing specifically about Victors passing comments about the UK 3 times in a 1000 years flooding, to cite in detail other material would have been rather over the top for a sunday morning reply. Of course there is uncertainty, which is why I wrote 'thought'

    However, I am a bit surprised by your comment as you make it appear that there is no knowledge at all about the known sea levels during the period and therefore my comment has come out of left field.

    The sea and river levels during Viking times and later have been written about by such as Hubert Lamb, Jean Groves, Brian Fagan, Professor Duck, Naja Mikkelsen and Antoon Kuijpers and there is a a very thorough 1000 page book called ‘The Viking world’ by Spink, with numerous references and graphics of sea levels through the ages compiled by such as moberg and grinsted

    I have lived close to two towns, one on the South Devon coast and one inland at the Thames Valley (both in the UK) which were raided by the Vikings due (it is thought) to higher water levels allowing greater incursion by shallow draught Viking boats. They similarly sailed up many of Europe’s rivers

    There was a general fall in sea levels during the very early part of the Viking period, following the dark ages cooling, then a substantial rise during the early stages of the Medieval Climate anomaly which affected coastal pasture lands and is thought to be one of the precipitates of their subsequent invasions of other lands and colonisation of Greenland.

    The sea level peak around 1400 coincided with extensive castle building by Edward 1 who, in the late 13th century, built Harlech castle in North Wales, where I once attended an archaeological dig. The castle had a water gate to allow for provisioning by sea. The sea has long receded from here.

    Deposition and land level changes always confuse things of course, but it is believed that sea levels there were higher than today, but the difference between the levels then and now are relatively small, as modern sea levels have continued the rise observed since around 1750 by Gordon Manley who saw this period as the start of general glacial retreat

    I wrote an article with numerous references on sea level change through the Holocene to Roman Times (which again had a high water stand) prior to the dark ages cool period. I have been accumulating material for a part two which would go up to the colder periods around 1450 when sea levels started to fall as ice grew. No doubt you will be looking forward to that.


  22. I'm sure Tony Watts would be happy to post that for you, tonyb. Sadly the linked article doesn't support your point. To make a claim about absolute sea level based on regional data you have to account for everything else that influences it, including PGR, current changes, tide changes and human modification.

    In the meantime try It has a good discussion of various influences on general and local sea level in the UK. The rest of the scientific literature on the subject awaits you.

  23. steve

    I am making a comment about the possible sea levels in a specific area at a specific time. I am not making a global sea level claim.

    I have seen and read a great deal of the scientific papers about sea levels, being fortunate enough to live close to the met office. They have many books and papers in their library and archive on the subject.

    As you know sea level change is a highly complex subject with each of the ocean basins reacting differently. As can be seen in such as Proudman, The British Oceanograhic data or Noaa's tides and current charts many places are seeing sea level falls and others are experiencing rises. In both directions the changes are often greater than average, although generally they are rising slowly, as they have been for some 250 years.

    Why should that not have happened in the past during a known warm period in Europe, with their high waters of a thousand years ago being offset by lower levels elsewhere?

    All highly complicated by land changes relative to the sea.

    The people I cited studied the subjects thoroughly and were renowned scientists and in the books I cited there are numerous authenticated references. I am well aware of Lambeck and have cited the author myself and the paper is also cited by various of the scientists I have mentioned to you, with regards to earlier Holocene sea level changes and glaciation changes. I am not sure of its relevance to the period in question compared to some of the other studies.


  24. "Why should that not have happened in the past during a known warm period in Europe, with their high waters of a thousand years ago being offset by lower levels elsewhere?"


    See That's the big dog in the sea level fight.

    Discuss local sea levels all you like. Just don't make generalized claims about global sea level trends that are at odds with the literature.


Comments are welcome, but comments without arguments may be deleted. Please try to remain on topic. (See also moderation page.)

I read every comment before publishing it. Spam comments are useless.

This comment box can be stretched for more space.